Albert B. Lord, The Singer of Tales
Introduction to the Second Edition
Part I. The Theory
1. Introduction 2. Singers: Performance and Training 3. The Formula 4. The Theme 5. Songs and the Song 6. Writing and Oral Tradition Part II. The Application
7. Homer 8. The Odyssey 9. The Iliad 10. Some Notes on Medieval Epic Appendix I Appendix II Appendix III Appendix IV Appendix V Appendix VI
Chapter 5. Songs and the Song
As long as one thought of the oral poet as a singer who carried in his head a song in more or less the exact form in which he had learned it from another singer, as long as one used for investigation ballads and comparatively short epics, the question of what an oral song is could not arise. It was, we assumed, essentially like any other poem; its text was more or less fixed. But when we look more closely at the process of oral composition and come to appreciate more fully the creative role of the individual singer in carrying forward the tradition, we must begin to query our concept of a song.
When the singer of tales, equipped with a store of formulas and themes and a technique of composition, takes his place before an audience and tells his story, he follows the plan which he has learned along with the other elements of his profession.  Whereas the singer thinks of his song in terms of a flexible plan of themes, some of which are essential and some of which are not, we think of it as a given text which undergoes change from one singing to another. We are more aware of change than the singer is, because we have a concept of the fixity of a performance or of its recording on wire or tape or plastic or in writing. We think of change in content and in wording; for, to us, at some moment both wording and content have been established. To the singer the song, which cannot be changed (since to change it would, in his mind, be to tell an untrue story or to falsify history), is the essence of the story itself. His idea of stability, to which he is deeply devoted, does not include the wording, which to him has never been fixed, nor the unessential parts of the story. He builds his performance, or song in our sense, on the stable skeleton of narrative, which is the song in his sense.
When one asks a singer what songs he knows, he will begin by saying that he knows the song, for example, about Marko Kraljević when he fought with Musa, or he will identify it by its first lines.  In other words, the song is the story of what someone did or what happened to some hero, but it is also the song itself expressed in verse. It is not just a story; it is not merely a tale divorced from its telling. Sulejman Makić said that he could repeat a song that he had heard only once, provided that he heard it to the gusle (I, p. 266). This is a most significant clue. The story in the poet-singer's mind is a story in song. Were it not for remarks like that of Makić, we  might be led to think that the singer needs only "a story," which he then retells in the language of verse. But now we know that the story itself must have the particular form which it has only when it is told in verse.
Any particular song is different in the mouth of each of its singers. If we consider it in the thought of a single singer during the years in which he sings it, we find that it is different at different stages in his career. Its clearness of outline will depend upon how many times he sings it; whether it is an established part of his repertory or merely a song which he sings occasionally. The length of the song is also important, because a short song will naturally tend to become more stable the more it is sung,
In some respects the larger themes and the song are alike. Their outward form and their specific content are ever changing. Yet there is a basic idea or combination of ideas that is fairly stable. We can say, then, that a song is the story about a given hero, but its expressed forms are multiple, and each of these expressed forms or tellings of the story is itself a separate song, in its own right, authentic and valid as a song unto itself. We must distinguish then two concepts of song in oral poetry. One is the general idea of the story, which we use when we speak in larger terms, for example, of the song of the wedding of Smailagić Meho, which actually includes all singings of it. The other concept of song is that of a particular performance or text, such as Avdo Međedović's song, "The Wedding of Smailagić Meho," dictated during the month of July, 1935.
Our real difficulty arises from the fact that, unlike the oral poet, we are not accustomed to thinking in terms of fluidity. We find it difficult to grasp something that is multiform. It seems to us necessary to construct an ideal text or to seek an original, and we remain dissatisfied with an ever-changing phenomenon. I believe that once we know the facts of oral composition we must cease trying to find an original of any traditional song. From one point of view each performance is an original. From another point of view it is impossible to retrace the work of generations of singers to that moment when some singer first sang a particular song.
We are occasionally fortunate enough to be present at a first singing, and we are then disappointed, because the singer has not perfected the song with much practice and by the test of repeated performance.  Even after he has—and it may change much as he works it over—it must be accepted and sung by other singers in order to become a part of the tradition, and in their hands it will go through other changes, and so the process continues from generation to generation. We cannot retrace these steps in any particular song. There was an original, of course, but we must be content with the texts that we have and not endeavor to "correct" or "perfect" them in accordance with a purely arbitrary guess at what the original might have been.
Indeed, we should be fully aware that even had we this "original," let us say, of the wedding of Smailagić Meho, we would not have the original of  the basic story, that is, the song of the young man who goes forth into the world to win his spurs. We would have only the application of this story to the hero Meho. Each performance is the specific song, and at the same time it is the generic song. The song we are listening to is "the song"; for each performance is more than a performance; it is a re-creation. Following this line of thinking, we might term a singer's first singing of a song as a creation of the song in his experience. Both synchronically and historically there would be numerous creations and re-creations of the song. This concept of the relationship between "songs" (performances of the same specific or generic song) is closer to the truth than the concept of an "original" and "variants." In a sense each performance is "an" original, if not "the" original.
The truth of the matter is that our concept of "the original," of "the song," simply makes no sense in oral tradition. To us it seems so basic, so logical, since we are brought up in a society in which writing has fixed the norm of a stable first creation in art, that we feel there must be an "original" for everything. The first singing in oral tradition does not coincide with this concept of the "original." We might as well be prepared to face the fact that we are in a different world of thought, the patterns of which do not always fit our cherished terms. In oral tradition the idea of an original is illogical.
It follows, then, that we cannot correctly speak of a "variant," since there is no "original" to be varied! Yet songs are related to one another in varying degrees; not, however, in the relationship of variant to original, in spite of the recourse so often made to an erroneous concept of "oral transmission"; for "oral transmission," "oral composition," "oral creation," and "oral performance" are all one and the same thing. Our greatest error is to attempt to make "scientifically" rigid a phenomenon that is fluid.
But if we are pursuing a will-o'-the-wisp when we seek an original, we are deluded by a mirage when we try to construct an ideal form of any given song. If we take all the extant texts of the song of Smailagić Meho and from them extract all the common elements, we have constructed something that never existed in reality or even in the mind of any of the singers of that song. We have simply then the common elements in this restricted number of texts, nothing more, nothing less.
It seems to me highly significant that the words "author" and "original" have either no meaning at all in oral tradition or a meaning quite different from the one usually assigned to them. The anonymity of folk epic is a fiction, because the singer has a name. We have created for ourselves in regard to both these terms problems that are not of any major importance.
It should be clear from the foregoing that the author of an oral epic, that is, the text of a performance, is the performer, the singer before us. Given normal eyesight on the part of the spectator, he is not multiple, but single. The. author of any of our texts, unless an editor has tampered with it, is the man who dictated, sang, chanted, or otherwise gave expression to  it. A performance is unique; it is a creation, not a reproduction, and it can therefore have only one author.
Actually, only the man with writing seems to worry about this, just as only he looks for the nonexistent, illogical, and irrelevant "original." Singers deny that they are the creators of the song. They learned it from other singers. We know now that both are right, each according to his meaning of "song." To attempt to find the first singer of a song is as futile as to try to discover the first singing. And yet, just as the first singing could not be called the "original," so the first man to sing a song cannot be considered its "author," because of the peculiar relationship, already discussed, between his singing and all subsequent singings. From that point of view a song has no "author" but a multiplicity of authors, each singing being a creation, each singing having its own single "author." This is, however, a very different concept of multiple authorship from that, or more properly those, in general use among Homerists.
Change and stability—these are the two elements of the traditional process that we must seek to comprehend. What is it that changes and why and how? What remains stable and why? In order to answer these questions, we should consider three groups of thematic analyses of songs, reports, as it were, on three groups of experiments. The first contains experiments on transmission of a song from one singer to another, the second illustrates the differences in a single singer's performances of a given song at brief intervals of time, and the third shows what happens to a song in a singer's repertory over a longer period, namely sixteen or seventeen years.
The cases of direct transmission without any intervening period of time (Salih Ugljanin from Nikola Vujnović, Avdo Međedović from Mumin Vlahovljak) demonstrate that the learner follows the text that he hears fairly closely in terms of basic story. The first of this pair of experiments ("The Wedding of Relja of Pazar," Parry 656) shows Salih Ugljanin omitting a number of details, but expanding two speeches and the marriage theme at the end.  Expansion is what one would normally expect from an old singer repeating the song of one much younger; in fact it is surprising that Ugljanin did not expand more. The omissions are interesting because the majority of them are concerned with the mountain spirits (vile). It would seem that these creatures of the imagination do not live in groups with a leader in Ugljanin's world. In his songs they appear alone and hence without a leader. His own picture of these spirits has, therefore, prevailed over that of Vujnović except for the necessary action of one of them attacking Relja and marrying him. The changes of detail, however, do not change the essence of the story. They follow the principle we have seen at work in the previous chapter; namely, that in learning a song the singer depends for details on his already habitual presentation of themes.
These changes in detail in Ugljanin's singing of the wedding of Relja  of Pazar are minor, and perhaps of no great significance; but the changes that Međedović made in Vlahovljak's song of Bećiragić Meho attract our attention (Parry 12471, 12468; thematic analysis in Appendix I). The most striking difference between the two is, of course, the degree and quality of expansion in Avdo's performance, from 2294 lines to 6313 lines, nearly three times the length of Mumin's song. Naturally such expansion involves the addition of many details, and we have already seen in the previous chapter the kind of changes that are normal in transmission on the level of the individual theme. The learner usually employs his own form of any given theme rather than the form that he has heard from the other singer. Sometimes the singer who is learning has to restrain himself from doing this, that is, from following his own theme rather than another's. If he did not take care, he would fall into self-contradiction later. Avdo is trapped in this way briefly, and not very significantly, at the beginning of theme 5 (see Appendix I). According to his habit, when a letter is delivered, the recipient opens it and reads it and the head of the assembly asks about the letter. This Avdo causes to happen in our story, forgetting that the messenger is waiting for a reward; momentarily Avdo is carried on by habit and for a few lines neglects the theme of paying the messenger, a really important theme in this song, a distinctive part of it. Later Avdo has to repeat the theme of reading the letter, thus causing a minor inconsistency.
There was something also in the story of Meho's capture of Nikola Vodogazović (theme 7) that surprised Avdo. He tells us that he was surprised; that is to say, he underlines the fact that Meho went to Janok, not in disguise, but in the clothes of a Turkish border warrior. Avdo emphasizes this at the time of telling, and even makes it the lever for introducing the repetition of this part of the tale-within-a-tale when Meho stands before the vizier in Budim. Avdo underlines this lack of disguise, because expeditions to Janok are generally for rescue, not for capture, and they are accomplished in disguise. The tavern maid usually recognizes the hero as a blood brother by tokens of some kind, and aids in the rescue. Everything in Avdo's experience of the tradition at this point indicates that the hero should be in disguise, and Avdo must restrain himself from disguising him. He indicates his problem and his feeling about it; yet he follows his model.
Indeed, after the story-within-a-story, Avdo is still worried about the question of disguise; he still feels that he must "correct" Mumin's telling. When assistance is offered to Meho from others in the assembly (theme 8), Mumin is content with Meho's wearing the clothes and riding the horse of his uncle, or of Halil, or of Bećir; when one wears another's clothes, one is that man for the moment and is thus disguised. Avdo did not accept this method, and on Meho's second journey to Janok, Avdo is determined that he will be in disguise. The disguise is proper now, anyway, because he is going on a rescue mission, not one of capture. Therefore, in Avdo's singing,  when Meho's uncle offers help to the youth, he states specifically that Meho must this time disguise himself and his horse. In the case of Halil's оffеr, Halil says merely that his sister will dress Meho as she does him when he goes on such raids. Later the dress is specified as that of a Viennese standard-bearer (theme 9).
Disguise must eventually involve recognition. The sense of "disguise" which must at one time have belonged to the borrowed horse and armor as we find it in Mumin's song has disappeared; so that when Meho in Halil's clothes and riding Halil's famous strawberry roan goes to Jela's tavern in Janok, she immediately recognizes him. There is no need for a recognition scene. In fact, all feel that Meho is not in disguise. But Avdo has restored to the theme a consciousness of disguise by adding a second disguising, the clothing and armor of the enemy country, in addition to borrowed horse and armor. The consequent recognition by Jela, the tavern maid, is not by elaborate signs; Meho simply tells her who he is. Yet one of the elements of recognition, one of those artifacts that constitute stage properties for such scenes, the musical instrument, together with a song, is present in Avdo's tale. Oddly enough the instrument is in the hands of the maid, not of the hero; the significant fact is that both it and the song are present at this moment in the tale. They were not in Avdo's model but have been added by him to the song as he heard it.
Another kind of change illustrated in this same experiment with Mumin and Avdo is concerned with a shift in the order of events. We have seen that Mumin listed three оffеrs of assistance from members of the assembly; one from each of the following: uncle Ahmet, Halil, and Bećir, son of Mustajbey. Avdo lists only the first two. What is involved here is the boundary between themes. In Mumin's story the theme of assembly can be said to end at line 1320, when Meho departed for Kladuša. The scene in Kladuša follows, after which the story shifts back again to Udbina, Bećir, and Mustajbey; the shift is skillfully accomplished by having Meho pass through Udbina as he leaves for Janok, and by describing the aghas as watching him depart. The theme of assembly could also be said not to end until after the conversation between Bećir and his father; it might be thought of rather as being interrupted by the scene in Kladuša. Such a framing is one of the common methods of indicating concurrent actions. Avdo's technique is somewhat different at this point. With him the assembly theme does not end when Meho departs for Kladuša but continues through the conversation between Bećir and Mustajbey and the summoning of the Border. Then the scene is transferred to Kladuša, whence the hero departs directly for Janok without going by way of Udbina, or at least without any mention of that town.
A similar change is found toward the end of the song (theme 13). In this case, however, it is Avdo who interrupts the final scene before the great battle (the scene in which Meho leaves Jela and goes to join Anđelica at  the crossroads) by returning to Udbina and to the Turks preparing to go to the assistance of Meho. Mumin had included the raising of the border army at the end of the conversation of Bećir with his father, Mustajbey (theme 10), but its scope was not so great as that of Avdo.
We have seen in this song changes stemming from addition of details and description, expansion by ornamentation, changes in action (such as those concerned with the disguise) that seem to stem from the tension of essentials preserving certain conglomerates or configurations of themes, changes in the order of appearance of the dramatis personae, shifting of themes from one place to another, forming new balances and patterns. Yet the story has remained essentially the same; the changes have not been of the kind that distort the· tale. If anything, they have enhanced it.
In the two experiments so far discussed the "pupil" sang immediately after hearing the song from the "teacher." We have the exact text of both performances. In other instances in the Parry material, the "pupil's" text was taken down or recorded some period, usually many years, after the learning, and in only one case do we have, by chance, the exact text from which the learning was done. This is the case of Avdo Međedović and the twelve thousand-line dictated text of "The Wedding of Smailagić Meho" (Parry 6840). This remarkable song merits special comment. Expansion by ornamentation is obvious in this case throughout the song; the first theme, of which we have written in the previous chapter, illustrates this ornamentation extremely well. We have already observed that the description or the hero is not purely decoration, but is rather especially meaningful. Here Meho is the hero born of old age, the darling of family and empire, owner of special gifts, horse, sword, armor, and clothing. Avdo has emphasized, indeed brought into significant prominence, these characteristics of Meho by his expansion of this part of the theme.
There is one change that may have been brought about simply by Avdo's own personal sensitivity to human relations in a heroic society. In the song book the question as to why Meho is sad was posed by the uncle without any prompting by the head of the assembly. Avdo represents the leader as calling the uncle and suggesting that he interrogate the youth, with the caution that it not be done immediately but only after a short period, lest the boy be embarrassed.
As in "Béciragić Meho," so in "Smailagić Meho," there are several cases of a change in the order of events. Thus, in the song book Meho talks first with the maiden in the coach at the Glina, hears her story, and then he and Osman fight with the guards. In Avdo's telling, Meho and Osman first fight with the guards and then Meho talks with Fata and hears her story. Moreover, Avdo has Fata ask for Meho's identity, and then has Meho ask Fata for her hand in marriage. She replies that her mother has told her she was destined for him anyway. There is none of this in the song book version, not at this point in the tale at any rate. In the song book, Meho  asks Fata's mother for her daughter's hand when he first meets her in Budim bewailing the fate of her husband and child. He offers to give her good news of Fata, if she will give her to him. Avdo's version of the meeting between Meho and Fata's mother is somewhat more dramatic—even melodramatic! The mother is about to leap from the window to her death in the courtyard when Meho shouts to her that her daughter has returned with him.
Avdo has also added a detail that is worth noting, because he makes use of it again later in the story, namely the letter that Osman found on the body of the captain of the guard who was taking Fata by coach to General Petar. A letter or some sign proving betrayal is a not uncommon theme (cf. the cross falling from the turban of the pontiff in the story of the capture of Bagdad, I, p. 85). In this case the letter is used later when Meho and Osman are back with Smail in Kanidža to prove that they are not lying about the vizier's treachery.
Surely one of Avdo's most attractive qualities as a singer of tales—and he has many—is his sense of heroic ethic. In his elaboration of the theme of Meho's departure from Budim, he shows Meho presented by Fata's mother with two proposals, each of which has something to recommend it. First, Fata's mother suggests that he take Fata back with him immediately lest the vizier capture her during his absence. This is sensible, but he refuses on grounds that he does not wish the vizier to think him afraid. The second suggestion is that he stay with Fata a month to enjoy her love lest he be killed in the battle that is sure to come. This suggestion was taken from the song book, which merely stated, however, that Meho refused to stay overnight. This too Meho refuses as dishonorable.
Less attractive to us as expansion, but typical of epic, especially of dictated oral epic, is the extended catalogue. Avdo pulls out all the stops! Those who have maintained that the catalogue of ships in the Iliad must be a historical document of considerable accuracy and antiquity should pay close attention, I believe, to the "accuracy" with which Avdo has reproduced the song book! Had his sense of historicity been at all strong, he would at least have tried to follow the book with care. The book and his friend who could read were available, and he most certainly could have had this part memorized "cold." Obviously this did not concern him. His catalogue here is essentially the same catalogue as in his other songs! Note that the list of chieftains who arrive in response to the invitations contains names of some individuals who were not mentioned in the invitations themselves.
Avdo has felt it necessary to relate the final battle, or battles, at considerable length, giving special care to the attack on Budim. It is noteworthy that he has rejected the facts even of the song book story here. Avdo rejects Mustajbey as the captor of the vizier and as the new pasha. Moreover, after the battle has been described, Avdo has chosen to draw together with completeness many of the threads of the story otherwise left hanging. The  sultan is informed of what has happened, new administrators and rulers are appointed, Fata's father and others are brought back from exile, and the wedding is finally held. The song book version left it to be taken for granted that all these details would have been worked out; Avdo has simply told in full what the other left to the imagination.
Making the imagination articulate is one of Homer's secrets also, and the following passage from Avdo's song, presented here as further illustration of his technique of expansion and elaboration, is reminiscent of Telemachus' journey in the Odyssey and not unhomeric in quality. On the second night of their journey to Budim, Meho and Osman stop at Knez Vujadin's house.
The corresponding section in the song book, Smailagić Meho, Mostar, 1925, tells of the stop at the house of Toroman Vuk and covers eleven lines (194–204):
They drove on their battle steeds and travelled that second day. They made their lodging far off in the small village of Veselica with the Vlah, Toroman Vuk, There they will spend the second night, and they will be lodged graciously, and graciously received, and graciously served. Then in the morning they will arise early. They pressed forward through the border country.
Here is Avdo's account:
They flew over the green plain even as a star in summer across the sky. They passed villages and crossed mountains. They travelled a whole day until nightfall, and they covered as much of Bosnia as they had planned to that day. They had come to the dwelling of Knez Vujadin. The Knez was at home with his wife and his two sons. They were looking from the window when the two imperial dragons came in sight, all glittering with gold and glorious in their plumes. The horses beneath them were in full panoply. Both heroes were like unto imperial pashas, and much better riders were they. Their array was much better than a pasha's or even a vizier's, or even, indeed, a great imperial general's. The Knez's two sons flew to the window and pressed their foreheads against it. When they saw, they wondered, and they cried out to their father: "O father, here is a wonderful sight that we have never seen before! Here are two heroes on two golden horses! They must be either pashas or viziers."
When the Knez looked and saw the son of Smail the Pilgrim with his plumed cap and the feather of an alajbey on it, and beneath him his winged horse, and when he saw the standard-bearer, that mighty hero, on his fine steed, then did Vujadin feel distressed, for here was the son of the Pilgrim in person. He recognized him because of Osman and his white horse, for all the Border knew Osman, and all the Kingdoms too. Then said Vujadin to his sons: "Run quickly to open the courtyard gate; open both portals wide before the two imperial dragons! Give them greeting and stand at attention, as if they were pashas or viziers, for to you they are indeed pashas and viziers. Tonight you shall neither sleep nor sit in their presence, but you must cross your arms upon your breast and speak no word, but serve these heroes in silence. Show that you honor them highly, both for my sake and for the prestige of your house, that in the years when I am no more you may bring it good repute!"
The Knez's two sons watched, and then they ran even as two mountain wolves and opened the courtyard gate. The two heroes drove in their horses. They said: "Good evening!" and the youths in the courtyard replied, and bowed low before them. Then they embraced the youths. Old Vujadin came flying from the house to the bottom of  the stairs. He shouted greeting to the heroes and took them by their lordly hands. His two sons seized the horses by the bridle and walked them up and down. Then Lady Vujadin took the two spears from the saddlehorn and carried them to the upper chamber of the house to the master's room, where but few guests are admitted. That room was kept for such heroes as these. It was strewn with Venetian cloth, and round about were silk couches and fine pillows covered with white silk and embroidered in the center with gold. They parted the curtained doorway of the room and entered. Then came his two dear daughters-in-law, like unto two white mountain spirits. They took the men's boots and socks and the swords from their waists. When the two youths were seated, they gazed at the ornaments in the room, at the cushions on the couch, all silk and embroidered with gold. In the middle of the room was a table spread with Venetian cloth and on it a metal platter heaped with all sorts of food; and on the table were coffee-urns with golden handles, and cups of crystal. Next to this was a mother-of-pearl table with a six-winged cask holding forty stone, two pitchers adorned with mother-of-pearl, and four three-liter glasses covered with a silk napkin. Around the table were four chairs…
[Glasses are filled, and talk begins. Vujadin says to Meho:]
"What is happening on the Austrian border? How are the lords of the Border? Do you still lead raiding bands over the hills, raiding bands and larger armies? Do you reach even as far as the Austrian Empire, broadening the borders of Sulejman's kingdom? Have the young men become better than their elders? What think you, Mehmed; are the old men better than the young?" And Mehmed answered: "Thoughts differ, but mine shall ever be that the old men are the better."…
In the meantime the boys had brought back the tired horses from their walk, had taken off the golden saddles and the girths and all the trappings. They sponged the horses and dried their manes with a cloth. Then they covered them with blankets, gave them barley, and waited for the beasts to eat it. They put hay in the mangers, closed the door of the stable, and went into the house to continue to do service. Their hats they left on the pegs, and they stood bareheaded before Mehmed and his standard-bearer Osman...
Finally the couches were spread for steep, and the youths settled comfortably. All night the boys watched over their lords, lest, tired from drink, they should be disturbed and seek either wine or water.
When dawn broke, Osman called to Mehmed: "O Mehmed, we have slept too long." Vujadin and his sons tried as best they could to persuade their guests to stay longer, but it was of no avail. The Knez's sons prepared the horses. Meantime the youths were ready and descended to the courtyard. The maidens brought their spears, the boys led out their horses, and the youths mounted. The night had passed and now the day-star shone and dawn unfolded its wings.
Mehmed put his hand into his pocket and gave each of the maidens five gold pieces. But the Knez's children would not accept them: "No, Mehmed, you shall not pay for your lodgings. This is not an inn or a tavern, but a dwelling for men of breeding." But Mehmed would not listen: "This is not pay, my children, but a gift of love. Let the girls buy combs and powder!"
Then he rode to the courtyard gate, and behind him Osman on his white stallion, even as a star across a clear sky. Dawn spread its wings and soon the two youths were riding by the cool Klim near Budim, four hours away.
Avdo has gone to great lengths to elaborate this theme and the theme of the other overnight stop on Meho's journey. There is some evidence to indicate that singers do not ornament unimportant points in their stories.  Halting places on journeys, scenes of hospitality, both here and in Homer, may deserve the emphasis given them neither because they are realistic pictures of heroic life nor because they are artistically useful in showing passage of time, but because the archetypal journey in epic was of a ceremonial nature and its stages were marked by significant events and meaningful encounters. Perhaps in such changes Avdo is following a tribal and traditional sense of what is important, although he himself would merely claim that he knew how to "ornament" a song.
Certainly not all singers would make the changes that Avdo has. In the Parry Collection there are songs from singers who learned them from printed texts. Their songs, however, are very close to the printed versions, and one realizes that the singer was attempting consciously to memorize or at least to follow closely what was printed. Singers like Avdo, in whom the feeling of the traditional is still strong, make no attempt to memorize, as we know, even when a song is read to them, but singers imbued with the idea that the written text is the proper one strive to keep to it even verbally if possible. With them the tradition is dead or dying. It could be truly said, I believe, that the only way in which they can compensate for their lack of awareness of the tradition, that awareness that we are beginning to see as deeply conservative, religiously maintaining the meaning of a song, is to memorize or attempt to memorize. The true representative of the tradition has other methods of learning, unfamiliar to the nontraditional.
One might expect that when a son learns from his father, whom he hears at an early age and frequently during the most formative years, the song will not vary much if at all in the process of transmission. Although we do not have any texts from direct experiment of son and father at the period of learning, we do have texts at a later period, when the son is grown up and has become a singer in his own right. A study of these texts seems to indicate that the changes in transmission that we found characteristic of the experiments with Ugljanin and Međedović, both mature singers learning from mature singers, are present here as well. Two examples from Kolašin in Montenegro will illustrate these changes.
We have the song of "Čevljanin Rade and the Captain of Spuž" from Antonije Ćetković (Parry 6718), the seventy-year-old father, and also from Milan Ćetković (Parry 6714), the son, aged twenty-two. Both were literate. The year of recording was 1935. The son's song is shorter than the father's, 249 lines as compared to 445 lines. The father ornaments more than the son does. We note this from the very beginning as we read the two songs parallel to one another.
|Čevljanin Rade is drinking wine||Čevljanin Rade is drinking wine|
|In the midst of Cevo in the white tower.||In broad Cevo on the border.|
|An adorned mountain woman is serving the wine,||An adorned mountain woman is serving the wine, |
|In her right hand a beaker and a golden cup.|
|When Rade had his fill of wine,||When they had their fill of wine,|
|He began to talk of many things,|
|How many Turks he had cut down|
|Around Spuž the bloody town;|
|And the adorned mountain woman listened to him,|
|Listened and then said to him:||The adorned mountain woman asked him:|
|"My lord, Čevljanin Rade,||"My lord, Čevljanin Rade,|
|I know well that you are a good hero.|
|So, my dear lord,||So, by the true God,|
|Are you afraid of any hero?||Are you afraid of any hero?|
|If he call you tomorrow to combat,||If he call you tomorrow to combat,"|
|Would you dare to go out to fight?"|
He answers that he is afraid only of the Captain of Spuž. At this moment a letter arrives from the Captain of Spuž. In the manner of its arrival we note the difference between the two singers; this is another case of the same theme being told in two different ways. Antonije, the father, simply tells of the letter arriving at the place where they were sitting and falling into Rade's lap; Rade takes it, breaks the seal, and reads it. Milan, the son, at this point expands the telling; Rade hears the knocker and asks the woman to go to see who it is; for it might be a letter-carrier. The woman obeys and goes down stairs to the courtyard, lets in the letter-carrier, who goes up to Rade's room, puts the letter on Rade's lap, and stands back to serve him. Rade reads the letter and is troubled. The woman asks him the reason, and he tells her the contents of the letter. At this point the son's song is slightly longer than the father's, 52 lines as compared with 34 lines.
Since the letter is important in starting the action of the song, Milan has paid closer attention to it, and the verbal resemblances between his text and that of his father are great, especially in the beginning. The captain says that he has heard of the worth of Rade's wife; he too has a worthy wife. He challenges Rade to single combat at Spuž in the presence of both their wives. The way in which the prize is stated differs noticeably in the two songs. The son says simply: "Let the winner take both women!" (line 71). The father elaborates: "If you cut off my head, take my white-throated lady to Čevo on the border, and love her whenever you awake. If I cut off your head, I shall take your adorned mountain woman to Spuž on the border, and love her whenever I awake" (lines 49–56). Here the father is distinguished by the kind of ornamentation which he favors. One might say that the moral qualities of the singers are herein reflected.
The next section of the song tells of the preparation of Rade and his wife to go to Spuž and of their journey there. Milan, the son, is not so skilled in description as his father, Antonije. Rade arrives at Spuž in line 94 in Milan's song, and at line 124 in Antonije's. The father relates how the  captain is there already drinking wine, and how Rade too sits down to drink as his wife waits upon him. But Milan says nothing of Rade's drinking and omits the striking picture of the captain's horse as it stands impatiently waiting for its master: "Before the tent he had tied his bedouin mare; and what a horse she was—may the wolves devour her! She struck with her hoofs and scraped with her ears, and with her teeth she bit the black earth. She was awaiting her master. The mare was thirsting for blood, eager to drink the blood of heroes" (lines 129–135). Milan missed this magnificent and frightening creature.
Now the two opponents send their wives to the enemy to discover whether either man is wearing steel breastplate under his shirt. Rade's wife finds that the captain has on a breastplate, but the captain tells her that if she informs Rade, the captain will be killed and his wife will go to Rade, and Rade's wife will become her slave. Rade's wife, therefore, reports to him that the captain is not wearing steel. The captain's wife then makes trial of Rade and discovers that he has no breastplate and reports this truly to the captain. The two singers' versions of this theme are very close, although the father uses more ornamentation than the son (the theme ends at line 254 in the former and at line 154 in the latter). Antonije presents a conversation between Rade and the captain's wife in which Rade asks her why she seems sad when she discovers he has no breastplate. "Is the captain wearing one?" asks Rade. She tells him to ask his own wife; for she herself cannot be untrue to the captain. In Antonije's song this conversation balances that between Rade's wife and the captain; the balance is lacking in Milan's song.
The combat itself is not only much shorter in Milan's version (lines 155–213) than in his father's (lines 255–385) but it is also rather different. Antonije tells how they first fire their pistols and miss; then they fight with swords and Rade's sword strikes fire from the captain's breastplate, whereas the captain draws blood from Rade. Rade breaks the captain's sword, and they begin to wrestle. The foam from Rade's lips is bloody; he sees he will perish and he calls his wife to aid him (line 321). In Milan's song the two first fight with swords until Rade's foam is bloody and the swords are broken. Then they wrestle and Rade throws the captain and is about to cut off his head, when the captain shouts to Rade's wife to help him, saying that if she does not, Rade will win the captain's wife and she will become that woman's slave (line 181). The breastplate, about which there was so much time spent earlier in the song, is not even mentioned again by Milan.
In Antonije's tale Rade's wife takes a sword from the ground and wounds him, whereupon he asks the captain's wife for aid. She comes and kills the mountain woman and then helps Rade against the captain. Finally she cuts his belt and his trousers fall about his legs and trip him. Rade falls upon him, the captain's wife brings Rade a sword, and he kills the captain. He  takes horse and wife home. Milan tells how Rade's wife comes to the aid of the captain and attacks Rade, who then appeals to the captain's wife, on the grounds that if he perishes the captain will make her a slave of the new wife. The captain's wife kills the woman with her teeth and gives a sword to Rade, who cuts off the captain's head. Milan avoids the rather cheap trick with the trousers. 
It would seem that the differences between the same song performed by a father and also by a son who learned it from him are the same in kind and even in number as in transmission between other individuals. This may be surprising at first blush. We should remember, however, that the social groups involved are small, and it often happens that the boy will hear other singers as frequently as he hears his father. Moreover—and this is the real point—the differences are inherent in the very process of transmission and composition.
At the end of the previous chapter we saw that some themes have a tendency to cling together, held by a kind of tension, and to form recurrent patterns of groups of themes. They adhere to one another so tenaciously that their use transcends the boundaries of any one song or of any group of songs. They are found in return songs, weddings, rescues, captures, and taking of cities. If, for one reason or another one of these themes is omitted or expressed only by implication or inadequately in the singing of one man, it will reappear in full bloom in the singing of another who has learned the song from him, but is already aware from other songs of the tensions binding the themes.
We can illustrate this phenomenon from the Novi Pazar material in Volume I. In Makić's song of the rescue of the Alibey's children (No. 24) he has omitted the theme of the disguise, possibly taking it for granted that the disguise is to be understood. The other elements of the thematic complex, long captivity, recognition, rescue, are present. In Zogić's version, the theme of disguise is made explicit and elaborated. Makić, of course, may have made this element more explicit in other singings than the single one which we have from him, but in a real sense Zogić's addition or elaboration here has not been a radical change in the story as he heard it, whether the disguise was expressed or implicit in Makić's performance. The point is that the disguise is there. Zogìc has also used a different means of recognition, namely by the breastplate, than that employed by Makić, who simply has the hero declare his identity. These are two of the multi-forms of this theme. The change is not basic.
While the foregoing example explains why a theme was added or elaborated, that is, in order to fill a gap which could not be left, it does not explain the particular form of the new material. Zogić could very well have had Alija go home, put on a disguise, and depart for enemy country. This form of the disguise theme is common and even usual in the group of themes in question. There is, however, another group of themes in which  an orphan hero goes forth on adventure with borrowed armor and weapons, and the adventure involved is seeking to rescue someone (often a father) or to win a bride. This group is closely akin to the return group. Its hero is generally Sirotan (Orphan) Alija, and he and Bojičić Alija are frequently and understandably confused. Zogić has modulated, as it were, into the orphan group, whence he obtains the material for this theme. By doing so, however, he makes his narrative come into conflict with the form of the recognition theme that he has chosen. Alija is recognized by the breastplate which he always wears and which he won in single combat! Yet earlier he had no weapons or armor and had to borrow them from his uncle. The two themes do not go together because they belong in two different, though related, groups of themes. The strength of the association of Alija with the orphan group has brought into the return group a theme inconsistent with one of the multiforme of the recognition theme belonging in the return group.
The examples of transmission that have been given leave no doubt that it would be a fruitless task to attempt to reconstruct the text of a song purporting to be the model for any other given text. When we turn to the study of different performances of a song by a given singer, we usually find here also no small divergences in text, yet a conservativeness in regard to story. Several examples are to be found in the published volumes of the Parry Collection,  but I include here an example from the Christian tradition of Marko Kraljević songs in central Hercegovina in the district of Stolac.
Petar Vidić is no more than an average singer, and for that reason a comparison of his texts is not without significance. He is the type of singer who must carry the brunt of the transmission of the art. From the thematic analysis of his four versions of the song of Marko and Nina (Appendix II) we can see the changes in the content of a song that take place between performances. There are more Petar Vidić's in any tradition than there are Homers!
Petar's version of this song seems to have changed in the year intervening between Parry 6 (1933) and Parry 804 (1934). He did not know the song very well in 1933, but the encounter with the American collector had revived interest in it, and when Parry returned the next year Petar was prepared. He now sings a song of 279 lines rather than 154. Actually the expansion that has brought this length about occurs in the first part of the story, up to the entrance of Marko into Nina's tower. The expressed fear of Marko that Nina will attack his home while he is in the army and the instructions to his mother are additions, as are also the account of the harrowing of Marko's castle, the actual capture of his wife and sister, and the sending of the falcon with the letter to Marko. These elements are additions, but there is nothing especially new in them. Expanded also is the scene in Stambol of Marko's obtaining permission from the sultan and men to accompany him. 
Petar had handled the disguising of Marko and his men in monks' clothing in a very summary manner in No. 6, but he made up for this the next year in No. 804. In so doing, however, he omitted the important scene of Marko's meeting at the spring Zloglav with his wife, who does not recognize him, although she does recognize his horse. In No. 804, instead of encountering her at the spring, Marko is questioned by her from a window when he arrives in the courtyard of Nina's castle. Marko's deceptive tale is given at the spring in No. 6 and in the courtyard in No. 804. Not that the spring is missing in the later text. It is there, but introduces not the scene with the wife but the scene at the nearby church where the monks' clothing is captured, the theme to which Petar had given short shrift the preceding year. In concentrating on this, he has forgotten about the women at Zloglav until it is too late. It is important, however, for Marko to tell his deceptive story to his wife, and Petar substitutes her for Nina in the courtyard scene.
Such are the main differences in the first part of these two texts from Petar Vidić. In 1933 his knowledge of the song was imperfect, and he was having great difficulty in dictating. In 1934 he was more at home singing No. 804 and seems possibly to have brushed up on the song in the meantime. But the expansion led him to forget an important scene; it may also be that the recording apparatus, which he saw for the first time in 1934, had excited him.
Parry 805 was dictated immediately after No. 804. Its comparative brevity stems undoubtedly from Petar's inability to dictate and the great feeling of discomposure that arose therefrom. He nevertheless straightened out the difficulty of the relationship between the scene of the church and the disguise, the deceptive story told to the wife, and the story told to Nina. When Petar sang the song again two days later (Parry 846), the early part of the story, the arrival of the letter from the sultan, Marko's experiences in Stambol, and the activity of Nina, are elaborated at greater length than in any of the other texts. And he has preserved the same order of disguise and deceptive stories as in No. 805.
There is a certain amount of confusion in the endings of all Petar's versions: the name of Marko's chief companion is different in each text; he and the other companions are killed in No. 6 but not in No. 804. Because the story of Marko and Nina is a form of the tale of the return of the hero to find his wife about to remarry, believing that he is dead, the question of whether the chief companion (and other companions) are killed is not without significance to the comparative study of this story. Odysseus' crew perishes, but his son and friends in Ithaca, who aid him in slaying the suitors, are untouched.
Petar's texts do not answer the question as to whether Marko's companions are to be equated with Odysseus' crew or with his friends, but they indicate that the question is a real one; for sometimes the companions are crew and sometimes friends. They are killed in No. 6 (crew), survive in No. 804 (friends), and in Nos. 805 and 846 they all survive except for the chief  companion, who is missing (a compromise). From this score we can see that it is a point on which Petar was not very clear, but which he felt to be significant enough to be kept, even when he had forgotten it in one performance (a performance in which he also forgot about the scene at the spring with Marko's wife and the other women). It is possible that Petar's compromise was necessary. Return tales seem to involve the death of someone close to the returning hero; in Yugoslav tradition this person is generally the hero's mother. I believe that it is highly likely that the death of the chief companion here is a substitute for the death of the mother.
The experiment with Petar Vidić and his song of Marko and Nina is very helpful, because it illustrates the way in which a singer, even when faced with unfamiliar circumstances of performance such as dictating and singing for a microphone, struggles with the phenomena of stability and change. It also demonstrates the value of having not merely two, but even three or more texts of the same song from the same singer. Numbers 805 and 846, in spite of differences in amount of elaboration and hence in length, differences resulting from the dictating technique, show stability from performance to performance, not of text, but of thematic structure. On the other hand, at first glance, Numbers 6 and 804 seem to indicate change. It is to be noted, however, that the differences between them are of the same order as the differences discussed in the section on transmission; namely, elaboration or lack of it, shifts in sequence, substitution. Actually the singer is attempting to regain a lost stability of story. He regains it in No. 805. Once a singer learns a song it attains a kind of thematic stability as long as he keeps singing it; but when he sings it infrequently, it begins to suffer from reduced ornamentation, and lapses of memory of the story. We have seen that it can be restored to active duty from the moth-ball fleet.
Not all parts of the song appear to be equally shadowy when a song is inactive. It may well be that the elements that remain, no matter how lacking in elaboration, are the most significant in the story. Nothing vital to the tale is missing from Vidić's No. 6 in 1933. The essentials are all there.
We have observed the thematic changes that take place between versions of the same song sung by a single singer over varying intervals of time. The longest interval, however, was only a year. From the material that I collected in 1950 and 1951 we can examine the differences found in versions of the same song by the same singer separated by a period of fifteen or sixteen years. The translation below presents thematic analysis in parallel columns of such an experiment. The song analyzed is the same song of Marko and Nina analyzed above in Vidić's versions and thus incidentally offers opportunity to observe the thematic content of versions of the same song by different singers. The singer of the versions below is a Moslem, Halil Bajgorić, from Dabrica near Stolac. He is, in fact, a close neighbor of Vidić. In Dabrica are the ruins of a fortress called Koštun, and the inhabitants of this valley associate the song of Marko and Nina with the fortress which is ever in their sight. Parry 6695 was sung for the records in  the spring of 1935; it contains 464 lines. Lord 84 was sung for the wire recorder June 7, 1950 and contains only 209 lines. The singer was in a hurry to finish and depart, since he had been called by the authorities from his work in the fields. The song must be considered as incomplete; otherwise the ending is unsatisfactory.
|Parry 6695 (1935)||Lord 84 (1950)|
|Twelve days after his wedding, Marko receives a letter||Marko goes to fight the Arabs for the sultan.|
|from the sultan telling him that the Arabs|
|have attacked and asking Marko to come to|
|his assistance. Marko prepares his horse|
|and himself, tells his mother that he is|
|going, says farewell to her and to his wife,|
|and sets out for Stambol.|
|In Stambol the sultan greets Marko, explains the|
|siluation to him, and then Marko goes to|
|Arabia where he fights with and overcomes|
|The sultan recalls Marko to Stambol and gives him|
|gifts. Again the sultan sends him out to|
|fight with the Arabs.|
|During the battle Marko receives a letter from his||While Marko is away Nina captures his tower, steals|
|mother telling him that Nina has captured||his wife, and treads upon his old mother.|
|his tower, stolen his wife, and trodden||His mother writes a letter to Marko telling|
|upon his mother. She asks him to come to||him what has happened.|
|Marko goes to Stambol, tells the sultan about Nina,||Same|
|and asks for help.|
|The sultan tells Marko that it is useless and that he||Same|
|will not give him help, but Marko asks only|
|for Alilagha and 30 heroes. This the sultan grants.|
|Marko disguises himself and his men as monks and||Same, except that they go first to Prilip where Marko|
|they proceed to Dabrica.||talks with his mother. Then they all go on|
|At the spring of Zloglav they find 30 women washing||Same, except that there are 100 women instead of 30.|
|clothes, among them Marko's wife. When|||
|she sees Marko's horse, she asks the monk|
|where he got the horse. He explains that|
|Marko has died in a fight with the Turks (!)|
|and that Marko gave him his horse in|
|return for burying him. Marko's wife weeps.|
|They proceed to Koštim. Marko tells the guards that||Arrived at Koştun, Marko leaves his companions at the|
|he is a monk who has come to marry Nina,||gate while he goes to talk with Nina. He|
|and that he has 30 companions who will||tells Nina that Marko is dead. Nina asks him|
|prune the vines. He is welcomed by Nina,||if he will marry him to Marko's wife. Marko agrees.|
|who asks him if he has ever been in the|
|Turkish army. Marko says that he has.|
|Nina asks Marko if he will sing and dance. Marko||At the wedding feast Marko asks Nina for permission|
|dances and the tower trembles. Nina says||to dance a little. He dances and the tower|
|that the monk must have learned to dance||trembles. Nina says that the monk must|
|from Marko. As a gift he gives him Marko's||have learned to dance from Marko.|
|from its scabbard.|
|Marko reveals himself, draws the sword and cuts off||Marko reveals himself, draws the sword, and cuts off|
|Nina's head. A fight ensues. Marko sends||Nina's head. Marko drives his enemies to|
|Alilagha to the gate to prevent any of the||the gate where they are met by his companions.|
|enemy from escaping.|
|Nina's three brothers escape and Marko pursues them.|
|He kills Vidoje at Vidovo Polje; Stephen at|
|Stephen's Cross; and Jasen at Jasena.|
|Marko gives Koštun to Alilagha and makes him a bey.|
|Then Marko returns with his wife to his|
|mother in Prilip.|
As we might, indeed, have expected, the story has remained essentially the same in both versions. The main difference between the two is to be found at the beginning and at the end. In the earlier version the beginning was greatly elaborated and the end was given full treatment. The second version has reduced the beginning of the song to a minimum and the end has been clipped. It is very likely that the clipped ending is caused not by the period of time between the two versions but simply by the singer's eagerness to depart. Certainly the discrepancies between the two versions are not necessarily greater than the discrepancies possible between two singings under different circumstances. There is no evidence that time has much changed Bajgorić's version of this song. But we can observe the opposite of what we saw with the first two texts of Vidić's performances of the same song; namely the regression from an elaborated text to a reduced text. One could infer, I believe correctly, that Bajgorić had not sung this song for a considerable period of time, and the circumstances of performance were not ideal.
It will be useful to see the results of a similar experiment with another singer and another song. The singer is Sulejman Fortić of Novi Pazar, and the song is the taking of Bagdad. Parry 676 (II, No. 22) was sung for the  phonograph records November 24, 1934, and Lord 10 (II, No. 23) was sung for the wires May 17, 1950. The former has 875 lines and the latter 812. The difference in length is negligible when compared with the difference in our first example.
With Fortić, also, the story is essentially the same in both the earlier and the later versions. It is instructive to note that sometimes the later version is fuller than the earlier and sometimes the opposite is true. One must bear in mind that the performer in this case was only twenty-nine years old at the time of the first version and that he was not yet a fully trained singer. Fifteen years later he was a more accomplished singer, although he is by no means a very good one even at his best. There are, however, two major differences between the two versions that cannot be explained as elaboration, or lack of it, but illustrate the substitution of one incident for another. In 1934 Fortić told how the messenger from the sultan went to Kajnidža, did not find Alija at home, and was directed by his mother to the mosque garden where Alija was assembled with the other men. In this he follows his master's, Ugljanin's, singing of the story faithfully. In 1951, possibly because he felt that as president of the National Front in Novi Pazar the mention of religious institutions such as mosques was not wise or fitting, he has omitted this incident, thus avoiding forbidden gatherings of Moslems at their churches. However, the feeling that Alija could be reached only through an intermediary was also very strong, and Fortić substituted another incident for the one with the mosque. The messenger must go first to Mujo and Halil in Kladuša and they will take him to Alija. Again the change is only apparently basic. The significant idea has been kept, and only the form of it has changed. The idea itself was felt to be so important that the simple solution of having the messenger proceed directly to Alija was avoided.
The second major difference between the two versions is in the ending. The way in which the earlier version ends is unorthodox. None of the other versions of this song has an unhappy and unsatisfactory ending. This ending may have been something that Fortie had himself improvised earlier at a time when he had not heard the song many times all the way through or when he had forgotten it. In other words, this unhappy ending may itself have been a change introduced into the song by Fortić himself. The later version of Fortić with its happy ending would seem to indicate that the singer had heard the more orthodox versions of the song since the earlier singing and had brought his own version more into conformity with them. The later version would then be an example of the corrective influence of the tradition. When a singer deviates too greatly from the traditional version of a song in regard to an essential theme, he is brought back into line, not by the audience but by the songs and singers of the tradition itself. 
For confirmation of the changes that take place over a period of time in  oral narrative song, from another though closely related language district, I am deeply indebted to Professor and Mrs. St. Stojkov of Sofia, Bulgaria. In July 1958 the Bulgarian Committee for Friendship and Cultural Relations with Foreign Countries arranged for the Stojkovs to take me to two of the villages near Sofia to listen to and record epic singing. We went to hear singers whom the Bulgarians had recorded on tape seven years earlier and whose songs, transcribed from these tapes, I had read in the Ethnographic Museum in Sofia. Mrs. Stojkov brought carefully typed copies of these texts into the field, and we asked the singers to sing these songs while she and I followed the texts. She was surprised to find that the singers changed the texts even very considerably. In one case the changes were so great that it was impossible to follow the written text. There was no doubt that in the person of the aged Vasilenka a pure oral tradition was still alive and operative. 
If we cease to expect verbal identity between different performances of the same song, whether they be by different singers or by the same one, whether they be over a shorter or a longer period of time, we are bound to notice that there are a few simple types of differences between them; (1) elaboration or simplification; the same thing told with more or less detail; (2) different order in a series; usually the reverse order, but sometimes merely a different order. In respect to the first of these types, the elaboration is usually significant, whereas the simplification indicates either a limited scope on the part of the singer, a restriction of time, or lack of practice in the song. In any case, the elaboration, in spite of what the singer himself may think or say about it, is not "pure" ornamentation; it has meaning in terms of the tradition from which it stems. In regard to the second type, one might conjecture as to why the change of order is often to the reverse. It would seem to be a sort of "chiasmus." Singers often use a series of questions followed by the answers in reverse order. Such a shift of order is regular practice. We should not be too surprised, therefore, to find it in transmission from one singer to another, since it is common in the singing of a given individual. Neither of these two types is a change in the "essence" of a song. If the tradition moved from singing to singing, from singer to singer, only in these two ways, one would not arrive at the diversity of "versions" and "variants" of a single song which is so characteristic of oral traditional material. They account for some of the differences but not for all of them, and certainly not for the most radical.
But there are other types of change. The substitution of one multiform of a theme for another, one kind of recognition scene for another kind, for example, one kind of disguise for another, is not uncommon, we have seen, as songs pass from one singer to another. The endings of songs are less stable, more open to variation, than their beginnings. Here the tension between themes that arises from habitual association comes into operation. It may help to provide an ending when either there was none in the  singer's experience or a given song or what there was seemed vague and hazy in his mind. The process may involve more than a mere ending to a song and actually lead the singer to mix songs, passing from one song pattern to another at a point at which the two patterns coincide. Singers recognize the fact that this kind of thing happens, because they criticize other singers for "mixing" songs. It is well for us to understand how this comes about. It is not haphazard, but the result of perfectly understandable and knowable forces. To the superficial observer, changes in oral tradition may seem chaotic and arbitrary. In reality this is not so. It cannot be said that "anything goes." Nor are these changes due in the ordinary sense to failure of memory of a fixed text, first, of course, because there is no fixed text, second, because there is no concept among singers of memorization as we know it, and third, because at a number of points in any song there are forces leading in several directions, any one of which the singer may take. If his experience of the particular song is weak, either as a whole or at any part, the force in a direction divergent from the one he has heard may be the strongest.
It is worth pointing out again that the changes of which we have been speaking have been brought about, not by forces seeking change for its own sake, nor by pure chance, but by an insistent, conservative urge for preservation of an essential idea as expressed either in a single theme or in a group of themes. Multiformity is essentially conservative in traditional lore, all outward appearances to the contrary.
The result of this multitudinous pattern of stresses and strains for the maintenance of stability is the typical multiformity of songs in oral tradition. The only way to make this multiformity graphic is to compare a number of texts of return songs from the Parry Collection. (See Appendices III and IV.)
The fact that the same song occurs attached to different heroes would seem to indicate that the story is more important than the historical hero to which it is attached. There is a close relationship between hero and tale, but with some tales at least the type of hero is more significant than the specific hero. It is convenient to group songs according to their story content, or thematic configurations, because songs seem to continue in spite of the particular historical hero; they are not connected irrevocably to any single hero.
If we classify the songs by their content we find a number of well-defined categories: weddings, rescues, returns, and captures of cities. They are well defined, and yet they overlap. We can illustrate this by isolating some of these groups and studying their basic patterns.
Parry was especially interested in the particular groups I have just mentioned and his collection is filled with examples of them. The reason for his concern for these groups lies in their obvious similarity to the Homeric poems and also, insofar as we have information about them, to the epics  of the ancient Greeks called "cyclic." Because of the parallel with the Odyssey, let us begin with the return songs in Yugoslav tradition.
The focal point of the return song is the return itself, and this is always surrounded by (a) disguise, (b) deceptive story, and (c) recognition. Almost invariably in this group the return is preceded by (a) shouting of the hero in prison after a long period of years, and (b) release with stipulation of return to prison. The section of the story that precedes the return home usually contains within it, by flashback, or else it is introduced by, the tale of how the hero was summoned from home on his wedding night to go to war, and of how he was captured. On the other hand, almost invariably the return home is followed by (a) the return of the hero to the enemy prison, and (b) the rescue of someone else from that prison. These four essential elements, (1) tale of capture, (2) shouting and release, (3) return home, and (4) sequel, are roughly parallel to the story of Odysseus.
It is, I believe, very significant that, whereas the songs that contain the return home are preceded by a tale of capture, shouting, and release and are followed by a sequel in the form of a rescue, there is a group of songs beginning with a tale of capture and shouting that does not lead to release and return, but to refusal of release and to rescue of the hero by someone else. In other words, songs beginning with the first two elements can lead in two directions, either to release or to rescue. And we cannot help but be struck by the similarity here with the twofold plan of Athena at the beginning of the Odyssey, (a) to send Telemachus to seek news of his father, who is weeping on the shores of Ogygia, and (b) to send Hermes to release Odysseus.
We can see that the return songs proper are in reality part of a larger group that entails captivity, shouting, and rescue, for even the return group contains a sequel to the return home that relates a rescue. Those containing the return theme end with the rescue (a) of the other prisoners who had been with the hero, or (b) of the hero's son who had been captured during his absence, or (c) of the hero himself. In the first two instances the hero does the rescuing; in the last case he is himself rescued by his wife's suitor, by his friend, or by his son. Although the rescue is accomplished in several ways, the commonest is by capturing the enemy's son and negotiating an exchange of prisoners. In the subgroup which tells of refusal of release followed by rescue, the rescue of the hero is done by his wife or by the Turks of the Border in battle!
There is a group of rescue songs that begins with the arrival of a message from a prisoner or prisoners who have been in captivity for a long period of time, often announcing that they are soon to be executed and asking for help. This opening thematic group coincides with the situation of the return group in which the release is refused; only the shouting scene is missing. In the rescue group a hero undertakes the task of rescuing the long lost hero.  He first disguises himself and when he arrives, after a journey, in the enemy country, there is a recognition scene with someone who is friendly, usually a woman; there are games or tests of some kind, and then flight. Very often not only rescue but also a wedding is involved.
Although it would seem that we have changed heroes in midstream, from the imprisoned to the rescuer, the pattern of disguise, journey, deceptive story, recognition, games, is like that of the return songs proper. In this case the disguise, journey, recognition complex is associated with the trip to the enemy country, and in the return songs it is connected with the trip from that country. There is a change in personnel and in geographical direction but not in basic thematic material. One might say that this group of rescue songs is a group of return songs in reverse.
There is also a group of wedding songs which follow the same pattern as the group of rescue songs. Bride stealing, especially when the bride is willing, as she is in these cases, but hindered by the enemy, is basically the same as rescue. We have seen, moreover, that even the rescue songs in this category (and they seem to include all those in which the imprisoned has been absent for a long time) also involve a wedding.
We find this same complex (disguise, journey, deceptive story, recognition, games or tests, wedding) in one other group of songs, namely, in that concerned with capture of cities. This is particularly clear in the version of the "Song of Bagdad" (Parry 20, Lord 79, by Osman Mekić of Stolac) in which Dadić Omer, who has accepted the challenge of Kajtaz to meet him in single combat in place of the sultan, is accosted by some magic birds on his way to Bagdad. They advise him to change clothes with a beggar whom he will meet. In this disguise he meets the queen of Bagdad in the castle garden, there is a recognition scene, and this is followed by the combat, and finally by the abduction of the queen and the capture of the city. Characteristic of the group of songs concerned with the capture of a city is the fact that the siege has been going on unsuccessfully for a long period of time.
It would seem, then, that all these groups, return, rescue, wedding, capture of city, are the same song in respect to the pattern of captivity and freedom (release or rescue) that we have been studying, in many disguises! This is, moreover, not a subsidiary pattern, not a minor part of the story, but actually the central action.
It is clear now, therefore, that the songs of this whole group are basically one song, at any rate from the disguise and departure to the final rescue. But how about the beginnings of the songs in this category? The beginnings are centered in the captivity itself. It is a long period of captivity, usually in a far distant enemy country. And in one way or another the story of the capture, of how the hero happened to find himself in this predicament, is related. Sometimes this is told in direct narrative, but more often, especially in the return tales proper, it is in a flashback, that is to say, the hero tells a newly-arrived prisoner the story of his capture. In the other groups, especially  the rescue group proper, the tale of capture is related in the message received from the prisoner, or sometimes the person who has received the message relates it to the group or individual who will undertake the rescue. In the case of wedding songs, the hero often gives an account of how he happened to meet or to know of the girl in the far country to whom he is now betrothed and whom he now wishes, venturing forth, to win back.
Any single performance that we may choose of a song in this group of interrelated families must be understood in terms of its brothers and sisters, and even its cousins of several removes. While recognizing the fact that the singer knows the whole song before he starts to sing (not textually, of course, but thematically), nevertheless, at some time when he reaches key points in the performance of the song he finds that he is drawn in one direction or another by the similarities with related groups at those points. The intensity of that pull may differ from performance to performance, but it is always there and the singer always relives that tense moment. Even though the pattern of the song he intends to sing is set early in the performance, forces moving in other directions will still be felt at critical junctures, simply because the theme involved can lead in more than one path.
When we look back over these examples of transmission, we are, I believe, struck by the conservativeness of the tradition. The basic story is carefully preserved. Moreover, the changes fall into certain clear categories, of which the following emerge: (1) saying the same thing in fewer or more lines, because of singers' methods of line composition and of linking lines together, (2) expansion of ornamentation, adding of details of description (that may not be without significance), (3) changes of order in a sequence (this may arise from a different sense of balance on the part of the learner, or even from what might be called a chiastic arrangement where one singer reverses the order given by the other), (4) addition of material not in a given text of the teacher, but found in texts of other singers in the district, (5) omission of material, and (6) substitution of one theme for another, in a story configuration held together by inner tensions.
In a variety of ways a song in tradition is separate, yet inseparable from other songs. 
[ back ] 1. See Gerhard Gesemann, "Kompositionsschema und heroisch-epische Stilisierung" in his Studien zur südslavischen Volksepik (Reichenberg, 1926), pp. 65–96. Of interest also in connection with song structure is A. Schmaus, Studije o krajinskoj epici (Zagreb, 1953), passim.
[ back ] 2. Cf. Makić's statements in Parry and Lord, Serbocroatian Heroic Songs, I, 266–267:
N: What's the song which you sang just now? S: That's the song about when the two pashas spent the winter [the first line of the song]. N: Where? S: At Temišvar. N: All right. If anyone were to say to you: "Sing me such and such a song." What would he say to you? Or, if you were to say to someone: "I sang such and such a song," meaning the one you sang just now? What is it called? What name shall we put on it? ... all right, "two pashas spent the winter," but—S: In Sement. "Sing me the song about Temišvar." That's it, the song of Temišvar ... that's what people say to me ... N: What happened at Temišvar? S: What happened? You've already heard what happened. When the pashas spent the winter and all seven kings surrounded them. N: I know, but there are many songs about Temišvar ... Have you ever heard that there are more songs about Temišvar than just that one? S: I haven't heard any songs about Temišvar except this one. N: Well, I've heard that there are other songs about Temišvar. S: Well, what isn't there in the world? Every singer has his own songs. N: Then what shall we put that song down as? What's its name? S: Its name is "At Muhač."
[ back ] 3. As an example of a song that has not been perfected by much singing and is close in frequency of performance, if not in time, to its first singing, see Salih Ugljanin's song of the Greek War (Parry and Lord, I and II, No. 10).
The only actual text I know of made up in our presence in 1934 is a song, coaxed out of Salih Ugljanin, about Parry and Nikola and the collecting. It is in Parry Text 655, records 965–966. Actually the songs made up about collectors are not very good examples because collectors and collecting are not inspiring nor proper subjects of epic! I give the conversation which preceded Salih's song:
The only actual text I know of made up in our presence in 1934 is a song, coaxed out of Salih Ugljanin, about Parry and Nikola and the collecting. It is in Parry Text 655, records 965–966. Actually the songs made up about collectors are not very good examples because collectors and collecting are not inspiring nor proper subjects of epic! I give the conversation which preceded Salih's song:
Nikola: Would you be able to make up a song about how we came here and found you to sing songs for us? Salih: Oho! Then I would be able. N: You would? In Bosnian? S: In Bosnian. Parry: Right away? S: Right away. N: Now? S: Now, while you are here. Since you have come now, right? N: Yes. S: I wouldn't be able to do it in that time. I wouldn't, by God! N: Would you be able, beginning now, sitting here, to tell it straight off, as in a song? S: Well, I don't know. Let me see now. I don't know what … N: No, no, but how we came here and how we found you. S: Since you came here. N: Yes. Parry: How you have worked for us. How you dictated songs. S: Ha! Parry: How Nikola wrote them. S: Yes, Yes. N: Could you do that? S: I could. Now, this one which I told you, I could tell it to you. N: Again you have not understood me! S: I can't do anything. N: Could you make up a song? S: Ha! N: How you recited to me and I wrote the songs. Could you create a song about that? How we gave you fifty dinars a day and sometimes sixty, and cigarettes and tea and other things, and... ? S: By Allah, that I could. And in Bosnian? N: Yes. How will it be? Come on, begin so that we may see. S: What's the name of the boss? N: Milman. S: Milman? N: Yes. S: And you're Nikola? N: Yes. S: As for the other let him... (Lord is referred to here in the next room at the recording machine.) N: What? What did you say? S: We won't include him, you know, but only you two. Parry: All right, as you like. N: Come on, Salja, let's hear it! S: Let me see, on what day did we begin? Parry: But nicely, you know! As if to the gusle. S: Yes, by Allah! N: We began to work here on Monday, and today is Saturday. S: Yes, indeed, and we worked all day until it was night. N: Yes. S: Every day until it was night we worked. N: Yes, S: And so we shall sing a song. I'll speak more loudly, right now.
|Od kratkoga vakta i zemana,||Since a short time ago,|
|Ima puno, tamo šes' da ... taman šes' dana,||All of six da ... exactly six days,|
|Pa od dana ponedijonika,||Well, since Monday,|
|Kako ođe smo na skupnicu.||We have been gathered here.|
|5||Ja i Nikola pesme iskazali,||5||Nikola and I have recited songs,|
|Ja kazao, Nikola pisao,||I dictated and Nikola wrote,|
|Od ob jutra do do noći ravne.||From morning until level night.|
|Sve smo redom pesme ispisalji,||We have written out all songs one after another,|
|I mene su pošteno platilji.||And they paid me honorably.|
|10||I ja sam svaki dan dolazijo,||10||And I came every day,|
|Sve od jutra do većere ravno,||Ever from morning until evening straight,|
|I sve po jednu smo pesmu ispje...ispisalji,||And ever a song each day we sa...wrote,|
|I aćik smo je dokazalji,||And frankly we declared,|
|Od koga smo pesmu taku ćulji,||From whom we heard such and such a song,|
|15||I svakome ime upisalji,||15||And we wrote the name of each,|
|I Nikola redom upisao,||And Nikola recorded them one after another,|
|I u grohot smo se osmijali,||And we laughed heartily,|
|I ćajeve poćesto smo pilji,||And rather frequently we drank tea,|
|A cigara preko hesapa gorjeli.||And we burned cigarettes beyond counting.|
|20||Tako je bilo za šes' dana ravno.||20||So it was for six days straight.|
|Danas šesti što smo uradilji.||Today is the sixth we have been working.|
|Nekoljiko ćuda poprićalji.||We have related several wonders.|
|Veljiki smo smijeh otvorilji.||We have started great laughter.|
|Nekoljiko jada poprićalji,||We have related several sad tales,|
|25||Što su stari prićalji iftijari,||25||What the aged elders have related,|
|To smo danas mi ponovilji.||That we have repeated today.|
N: Go on, keep on talking! P: Is there any more? S: There isn't. N: There isn't any more, is that it? S: Yes. For six days, I counted out the days for you, that we drank tea and wrote.
For a song about Parry or in his honor, written by Milovan Vojičić and given to him, see Appendix VI.
For a song about Parry or in his honor, written by Milovan Vojičić and given to him, see Appendix VI.
[ back ] 4. The two songs are of about the same length, Nikola's version having 156 lines, and Salih's 142. The differences between them are: (1) Nikola has Marko suggest that they go to Golješ to steal the vile's horses: Salih mentions only that they are to look at them; (2) in requesting Relja to sing, Nikola has Marko say that he is sleepy and wants Relja to sing to keep him awake (lines 33–36): Salih omits; (3) Nikola has Relja reply to Marko that he is afraid to sing because of the wolves and bandits (lines 37–42): Salih omits; (4) on the other hand, Salih has expanded Marko's speech to Relja, in which he tells him not to be afraid, from Nikola's nine lines to fifteen, but the general contents are the same (N: lines 44–52, S: lines 33–47); (5) in Nikola's version Jevrosima offers her primacy to the vila who will pluck out Relja's eyes (lines 70–71): Salih omits; (6) Salih does not mention the name of the second vila (N: line 75); (7) Nikola has Anđelija say that she will bring to Jevrosima Relja's eyes (line 79): Salih omits; (8) Nikola's picture of Marko riding cross-legged on his horse and swearing like a drunkard (lines 86–87) is not repeated by Salih; (9) the effect of the whipping on the vila's flesh (lines 118–119) is omitted by Salih; (10) Nikola's detail of releasing the falcons to supervise the gathering of herbs by the vila (line 127) is omitted in Salih's version; (11) on the other hand, Salih has expanded Marko's speech to Relja in which he stops the beating and suggests the marriage (N: lines 137–141; S: 117–127); and (12) Salih has expanded the theme of the marriage itself from Nikola's six lines (151–156) to eleven (132–142).
[ back ] 5. For other examples see Kolašin texts, Parry 6771 and 489; 6780 and 6736, and Appendix V.
[ back ] 6. There are approximately 70 such experiments in the Parry Collection, most of them from the districts of Stolac (30) and Gacko (20). A number of instances (actually 8) will be found in vols. I and II in the material from Novi Pazar. From other regions there are a total of 12 experiments: 3 from Kolašin, 1 from Glamoč, 9 from Bihać, and 7 from Bijelo Polje. Including the Marko and Nina song in Appendix II, which we have analyzed (see Chart V), the experiment was applied 6 times to Petar Vidić. The song Marko and Nina was the only case in this group of Vidić's containing texts at an interval of a year, and using more than two texts; the other instances are of only two texts not separated by much time, only a few days. In each of these cases one text was sung for the records and the other dictated for Nikola Vujnović to write down; this fact must also be taken into consideration in comparing the texts.
[ back ] 7. As a final example of what may happen to a song over many years in the hands of a single singer, one may take Avdo Međedović's "Osmanbey Delibegović and Pavičević Luka" (Parry 12389; 12441). This second of the two longest songs in the collection is in one way even more remarkable than the "Wedding of Smailagić Meho." It was sung for the records in 1935 and not taken down from dictation. The text contains 13,331 lines and fills ninety-seven twelve-inch phonograph records. I have estimated that actual singing time was between sixteen and seventeen hours. The singer sang for three or four days, lost his voice, was given a ten-day vacation, and finished in another three or four days, the whole process taking about three weeks.
[ back ] 8. The Russian folklorist A. M. Astakhova has written: "A variant becomes interesting and significant not only as one of the indispensable links for the clarification of the bases of the plot, but also as a manifestation of the living creative energy of the masses" (Russkij bylinnyj epos na Severe, Petrozavodsk, 1948, p. 5, quoted from R. M. Volkov, "K probleme varianta v izučenii bylin," Russkij fol'klor, 2:98 (1957).