Patterns of Transmission: Mothers and Daughters in the Milman Parry Collection of Oral Literature

Aida Vidan, Harvard University

From the time of the Homeric epics down to the period when Milman Parry and Albert Lord discovered “new Homers” such as Avdo Međedović in the Slavic part of the Balkans, the epic tradition has captivated the attention of both scholars and audiences. By contrast, shorter lyric songs and ballads from the same region traditionally performed by women (and sometimes simply called “women’s songs,” a problematic term to which I will return), while collected and published along with the epics, have rarely received the same scope of analysis. And yet, owing to the wealth of sources preserved in the collections of Vuk S. Karadžić, the Matica hrvatska archives and publications (Zagreb), and the manuscripts and recordings in the Milman Parry Collection of Oral Literature (Harvard University), it is exactly the shorter verse genres that offer possibilities for investigation unavailable with other types of oral materials. In this article I will focus on one such example, namely the transmission of oral songs within the circle of an extended family.

The new research practices introduced by Parry and Lord during their trips to the former Yugoslavia in the 1930s, 1950s and 1960s (the last two undertaken by Lord alone) are now seen to have set a standard and have been treated in multiple scholarly contributions. For this reason I will not analyze them in detail here. Instead, I wish to underscore only those elements directly relevant for this article. As is well known, Parry and Lord returned from the Balkans in 1935 with a large body of not only epic materials and interviews with singers, but also a trove of ballads and lyric songs that served as the source for Bartók and Lord’s volume Serbo-Croatian Folk Songs (1951). Ever since his first engagement with the shorter verse genres Lord pointed to the need for a continual investigation of this part of the tradition: “Further study of the texts of women’s songs and a comparison of dictated texts with songs recorded in performance, with an investigation of how the two relate, could allow for a reconstruction of the total repertory of formulas, verses, and even whole songs that constitute the oral tradition of lyric songs and ballads of Muslims in Gacko. It would also allow for a comparison of texts collected from younger people with texts of the same songs obtained from older individuals. This would help the reader gain an insight into transformations which appear from one generation of singers to the next.” (Lord, 1953: xxiii; my transl.). Lord returned to this subject in two chapters of his last book, The Singer Resumes the Tale, signaling a plan to undertake a further study, which sadly he did not have a chance to complete.

Even some of the earliest collectors, such as Vuk Karadžić, had a strong sense that lyric songs and ballads held great potential for research and that in some dimensions they preserved the tradition more faithfully than the epics. In 1824, in the Leipzig edition of his Narodne srpske pjesme, he made the following observation: “Concerning the age of our songs, I would say that we have women’s songs that are older than heroic songs because we have few heroic songs that are older than Kosovo and there is not a single one that comes from before the Nemanjić period, while among women’s songs some may be even a thousand years old.” (Karadžić, 1969: vol. 1: 539; my translation). [1] Similarly, Vatroslav Jagić, an important Croatian linguist active in Leipzig and Vienna, in his discussion of bugarštice [2] points out that the older epic songs, which often deal with the tragic fates of heroes, reflect the solemn, tragic tone of the bugarštice from which they likely originated. This tone, he finds, is lost in the more recent epic poetry (which is far more heroic), but is better preserved in the lyric poetry of the South Slavs. He therefore concludes that it is the lyric poetry that reflects more faithfully the nature of the oldest layer of traditional poetry (Jagić, 1880: 229-30). [3] Nikola Andrić, one of the principal editors of the Croatian series Hrvatske narodne pjesme containing oral traditional songs (published by Matica hrvatska publishing house), in his discussion of the lengthy ballads by one of the best Croatian female singers, Kate Murat née Palunko, speculates that these songs are “the best proof of how our heroic men’s songs evolved from women’s songs” (Andrić, 1996: 623; my translation). [4] Not surprisingly, for scholars of the older generation the age and historical background of the songs were central preoccupations. Today, however, the strategies employed and questions posed are of a broader scope. Owing to their modern collection practices, Lord and Parry in fact opened up additional research avenues that were unavailable in earlier times. In the quote provided above, Lord notes the possibility of studying songs supplied by different generations of singers, thereby allowing an insight into various transformations that occur in the process of transmission. In order to investigate the existence of various patterns, it is a prerequisite to have a large body of material from a diverse population of singers who performed on multiple occasions. It is also necessary to know who these singers are and how they are related to one another. Despite the wealth of oral resources from the Balkans from the nineteenth and early twentieth centuries, when the traditional singing was still at its prime, most of the collections do not allow one to carry out this type of analysis.

While the nineteenth-century collectors often disregarded the identity of the singer or, at best, noted down the name and the locality, Parry, Lord and their assistants systematically collected the data related to the singer and, when possible, conducted interviews with the singers as well. Although from our vantage point we may wish for more substantial information, the fact remains that the Parry Collection is one of the rare repositories that allows one to reconstruct several family trees of singers from the same community of Gacko (Herzegovina) and its neighboring villages in the year 1935, when the collecting took place. Of the nearly 13,000 texts contained in the Parry Collection, 11,396 are so-called “women’s songs” performed by 257 female singers from the Gacko district and 52 male singers from Gacko and other areas of Bosnia and Herzegovina. As will be evident from the data presented in this article, the term “women’s songs” is relative, since lyric songs and ballads could be and were (albeit more rarely) performed by male singers as well. What follows here is a discussion of the extended family tree of one of the central female singers from the Parry Collection, within which we will encounter a mother-daughter pair, as well as a mother and son, both performing shorter verse genres.

The importance of ethnological data in this type of research has been emphasized for the last few decades, especially in the context of the performance-centered approach (Bauman & Briggs, 1990). Although this orientation is considerably younger than Parry and Lord’s field trips, their methodology fits well even within the newer theoretical models. As Bauman and Briggs point out, “A crucial move in the establishment of performance approaches was a shift from the study of texts to the analysis of the emergence of texts in contexts.” (Bauman & Briggs, 1990: 66). It was Parry and Lord’s awareness of the importance of the context that prompted them to note down biographical data and to persuade singers to be interviewed. These materials, including the interviews with their principal local assistants (Nikola Vujnović for epic songs, and Ibrahim Hrustanović and Hamdija Šaković for lyric songs and ballads) provide invaluable ethnographic data, but also constitute the first steps toward contextualization. [5] As deftly outlined in the same article by Bauman and Briggs, the move from context to contextualization and entextualization [6] has followed on the heels of performance-centered analysis and has served as an attempt to deal with a Pandora’s box of questions that grew from the term “context.” The central preoccupation appears to be the position of the text – and here we should tread carefully since “text” in an oral performance is only the verbal content of a given performance that is subsequently codified as text – in relation to the cohort of contextual components which could be roughly divided as micro-sociological (situational context pertaining to a specific performance) and macro-sociological (broader “cultural reality,” Bronisław Malinowski’s term; Bauman & Briggs, 1990: 68). It appears that the latter category in particular emerges as too all-encompassing and that “[t]he seemingly simple task of describing ‘the context’ of a performance can accordingly become an infinite regress” (Bauman & Briggs, 1990: 68). It can include subcategories, such as the moral and aesthetic values of the specific community, socio-historical context, context of communicative system, and others as underscored by both Malinowski and Bauman. This cognitive process in which context becomes reified leads to a sense of false objectivity, claim Bauman and Briggs (1990: 68). They argue that the positivism of such an approach lies in its superimposition of “context” as a category existing independently of the discourse of a given performance. Attempts to redress this problem have resulted in the term “contextualization,” which reflects the positing of “context” as a category that is negotiated between participants in social interaction. Rather than having to deal with a potentially infinite set of factors that constitute “the context,” performance-based analysis takes into consideration cues generated in the process of a specific performance, which in turn then signal to researchers how to form meaningful and functional patterns, as well as which contextual components to favor. In the case of the present study, while many other relevant components could be taken into consideration, the pattern of investigation emerges on the basis of familial relationships identified by the singers themselves and their underscoring of the relevance of the family unit and the community as a key element in the transmission of traditional songs. The performance-centered analysis has called simultaneously for both a more contextual and a more textual approach. As paradoxical as it may seem, the two actually go hand in hand: “In order to avoid reifying ‘the context’ it is necessary to study the textual details that illuminate the manner in which participants are collectively constructing the world around them” (Bauman & Briggs, 1990: 69). In our example, a study of familial relationships, biographical data, and the circumstances in which the songs were performed and collected, completes only one segment of the task at hand and establishes the functional pattern that organizes investigation of “the text,” or to use Lord’s term, which better suits this purpose, the multiform. [7]

A large portion of the ballads and lyric songs (more exactly: 1728) in the Parry collection were written down by Parry and Lord’s local assistants, Ibrahim Hrustanović and Hamdija Šaković, who took down these songs from both dictation and singing. They collected materials primarily from the female family members (first and foremost their mothers and an aunt), but also from neighbors and other (mostly female) members of the community. Another collector closely related to this group is Hamdija Šaković’s cousin, Halid Dizdarević, who collected 141 songs, of which 140 are from his mother, Đula Dizdarević, the central figure of this article. This approach, partly dictated by a complex set of circumstances, resulted in an outstanding quality of songs and a minimal interference from the collectors, who happened to be close family/community members, a factor of extreme importance in a traditional Muslim society. [8] The remainder of the songs were written down by other singers’ children or literate community members, but unfortunately we do not have interviews with these other collectors or their informants, as we have with Hamdija, Ibrahim and some of their family members. Even these other collectors, however, noted down the singers’ relevant biographical data and often indicated their own relationship to the singer, especially in those cases when a daughter wrote down songs from her mother, grandmother, or an aunt. Parry and Lord themselves recorded approximately 250 songs and segments of songs (along with interviews with five singers) onto aluminum disks.

Some of the best ballads and lyric songs come from three singers — Hasnija Hrustanović, Emina Šaković, and Đula Dizdarević — who were all illiterate. Their age is rounded up since there were no birth records for female children (they remembered things in relation to significant historical events in the region). My estimate is that Hasnija was about 65, Đula about 67, and Emina 54 at the time the songs were collected, meaning that Hasnija was born approximately in 1870, Đula in 1868, and Emina in 1881. Hasnija provided more songs than any other singer in the Parry Collection: 466 in total, comprising approximately 10,000 lines of oral poetry. Emina supplied 330 songs, and Đula 338. Of particular interest to us is the fact that all the principal singers’ sons were involved in collecting and, in addition, that Emina and Đula also happened to be sisters. The largest number of songs was from Hasnija Hrustanović, written down by her son Ibrahim Hrustanović; similarly Hamdija Šaković, Emina’s son, wrote down most of his mother’s songs in the Parry Collection, and as mentioned above, Halid Dizdarević collected almost exclusively from his mother, Đula Dizdarević. Some of the collectors were female (in the present study, Sifa Kurtović, the granddaughter of Đula), but since in Herzegovina and generally in the Balkan region in the 1930s male children were more likely to receive an education, most of the collectors were male, while in the nineteenth century they were exclusively male. In general, it was difficult for female collectors to get involved in these processes because of the need to travel and communicate with unknown informants, a major obstacle for women raised in a patriarchal environment. For this reason those few who participated in Parry and Lord collecting efforts wrote down songs only within the closest circle of other female family members. We thus have an interesting situation with a tendency for women to convey the knowledge of singing to their daughters, and at a point when literacy was on the rise, for the sons to serve as a link for transferring an oral heritage into a written one. As previously noted, their interference in this process was minimal since they belonged to the same community as their informants. In fact, it is hard to imagine a more seamless transmission, since the presence of an external researcher, as evidenced on the recorded materials in the Parry Collection, actually introduced disturbance or noise in the communication channel (Vidan, 2003:10). Ibrahim Hrustanović, Hamdija Šaković, and Halid Dizdarević’s knowledge of the local circumstances was critical in accessing the most productive informants, as was their familiarity with the cultural landscape and conduct codes. Moreover, in their own interviews with Parry and Lord they volunteered data on their informants and the community, and in this way have contributed significantly to our understanding of the songs. It is necessary therefore to credit the local collectors not only for their contribution in preserving this heritage, but also in securing the contextual information which is at the very core of our efforts in systematizing and interpreting the songs.

In the interview with Parry and Lord, Hamdija Šaković explains that his mother (as well as his aunt Đula) comes from the Pašić family. In other words, Pašić is Đula’s and Emina’s maiden name. The family was so well off that the female children did not have to work in the field and could stay inside, embroider, and sing (Vidan, 2003: 83; PN 6314, phonograph records 2993-2994). These favorable circumstances most certainly affected both the singers’ acquisition of the formulaic language and the subject matter of the traditional songs, and made an impact on their ability to compose in performance (Lord’s term; Lord, 1959:1). Hamdija states that his mother Emina – and we can presume his aunt Đula as well – learned the songs from their mother, whose maiden name was Tanović. In the Parry Collection there are several singers by the name of Tanović, including one of the oldest singers, Kana Tanović, who was ninety at the time of collecting and who provided several superb ballads. It remains to be established whether Đula and Emina were related to this group of singers. Furthermore, we know that Đula and Emina’s grandmother was from the renowned family of Ljubović from Odžak, Nevesinje. We can thus reconstruct the female family lineage, which is of particular relevance since in the interview Hamdija quotes his mother Emina, who explicitly states that the lyric songs and ballads were transmitted exclusively through the female members of the family, from mother to daughter.

The focus of this article is Emina’s older sister Đula Dizdarević, and we are fortunate to have an interview with her conducted by Nikola Vujnović in the presence of Parry and Lord in Gacko in 1935 (PN 6520, 6523, phonograph records 3743-3757, 3828). Đula confirms that she comes from the Pašić family and that she was born in the village of Mulji. Although in the interview she claims to be seventy-five, the age of her son indicates that she must have been a few years younger, likely in her late sixties, as already noted. She married Murataga Dizdarević’s son from Korita. Đula had twelve sons and one daughter. Seven of her sons died, three lived in Gacko along with her daughter, one son lived in Dubrovnik, and one emigrated to America. She married into a family which had once had plenty of land and about a hundred serfs, but most of the wealth was lost in clashes with the neighboring Montenegrins and in various periods of unrest. In the part of the interview where Đula discusses her singing she reminisces that her father was a renowned guslar (epic singer) and that she still remembers one of his songs (preserved in the Parry Collection, PN 11688,entitled “Četovanje Save Pejovića” and collected by Halid Dizdarević), a detail that testifies to her exposure to and ability to compose in the epic genre. She boasts that there was no festivity in the village where she would not be invited because of her voice, but admits that her singing abilities have declined with age, an observation also made by Béla Bartók (Bartók & Lord, 1951:251). In terms of her knowledge of traditional songs, Đula estimates that all of her songs, written down and recorded combined, constitute perhaps only, in her own words “one ninth” of all the songs she knew, since when she went home, she would remember even more songs. Of course this estimate may be somewhat inflated, but it is undeniable that this singer, her sister Emina, and Hasnija Hrustanović all had an enviable knowledge of the tradition, as well as a talent to create new songs on the pattern of the old ones. Đula liked the old custom of singing with the baking pan (“tevsija”), which is revolved in the proximity of one’s mouth, thereby enhancing the singer’s voice. She learned most of her songs in her native village, but she heard others from the Gacko region and she thought that the songs in the whole Gacko area were the same. Needless to say, the contemporary reader and the oral singer would have quite different notions of “sameness” – Đula here refers to the existence of multiforms or variants, not songs that overlap verbatim. It is quite curious that in one part of the interview the singer claims that she could, and many times did, compose new songs, but that later on in the discussion, likely owing to confusion and/or embarrassment, denies such ability. A close examination of her songs in the Parry Collection indicates, however, that Đula was able to compose new songs in performance, could craft multiforms, and that she never used rote memorization.

Đula had one daughter, Arifa Kurtović, who at the time of collecting was in her mid to late thirties. She was also illiterate and was married to Šaćir Kurtović, who was in his forties. The Parry Collection contains materials by several singers by the name of Arifa Kurtović (not an unusual occurrence in small communities with only a handful of last names), and caution is needed not to confuse the identity of these singers. Songs were collected both from Arifa and her husband Šaćir; some were noted down by Hamdija Šaković (her first cousin) and some by Arifa and Šaćir’s daughter Sifa. Another important member of this extended family is represented in the materials of the Parry Collection, Šaćir’s mother Umija Kurtović née Fazlagić, who claimed to be eighty-five at the time and was born in Kula Fazlagića near Gacko. She too was illiterate. In the Parry Collection there are 84 lyric songs and ballads by Arifa, 57 by Šaćir’s mother Umija, and 3 lyric songs by Šaćir.

Let us now look more closely at the textual data provided by the broader family of Đula Dizdarević and the transformations that occurred in the process of conveying the traditional songs from one generation to another. I will not focus as much on the lateral influences between the two sisters, Đula and Emina, since elsewhere I have published a comparison of some of their songs in an analysis of textual stability in the South Slavic traditional ballad (Vidan, 2003: 45-63). It is sufficient to recall here the conclusion that Đula and Emina were more prone to choosing both similar wording and narrative patterns than the other singers. This finding was established on the basis of a close examination of the formulaic language and the principal themes occurring in a cycle of multiforms dealing with the subject of the forced marriage of a girl who was subsequently abducted by her beloved. However, even Đula’s and Emina’s songs have many points of differentiation. In addition, two multiforms of this song provided by Emina are quite different, especially in their final segment, thereby attesting to her ability to compose in performance and to modify a song “on the go.” This is further supported by a large body of songs in the Parry Collection by both singers, examination of which confirms that the singers recalled formulas and themes but never memorized songs word for word.

If we now turn our attention to the songs of Đula’s daughter, Arifa Kurtović, we recognize several noteworthy tendencies: first, a vast majority of Arifa’s songs exists in Đula’s repertoire, and those that are not found there appear in Emina’s repertoire. Sometimes they exist in both sisters’ repertoires (for an exact breakdown, see appendix 1). One needs to keep in mind that Đula’s and Emina’s repertoires overlapped because of their kinship and exposure to the same sources. This indeed confirms that Arifa learned her songs largely from her mother and possibly her aunt. Secondly, Arifa has far fewer songs that are creative in the sense of producing new themes or even stringing together familiar themes into new narrative patterns. In other words, her combinatory abilities, which are essential for building the song both on the verse and narrative levels, show a significant decline in relation to the older generation of singers, including her mother. Third, Arifa’s songs are considerably shorter and less ornate, and her repertoire is markedly smaller. She has a broader selection of lyric songs than of ballads, and quite often her songs are truncated or contaminated. Furthermore, hyper- and hyposyllabism are both more frequent in her songs. Finally, a tendency towards an increased textual stability is evident, although we can assert that Arifa still composed in performance rather than followed a matrix of memorized texts. We need to recall here Lord’s notion of textuality: “a sense that a song has a recognizable text and that the singers recognize that fact” (Lord, 1995: 177), as well as Gregory Nagy’s important observation that the words “text” and “textuality” do not imply writing, but rather “[a] degree of a composition’s invariability from performance to performance” (Nagy, 1996: 19). While it is generally accepted that literacy decreases the ability of traditional singers to compose in performance, here we have a mother-daughter pair who are both illiterate, but the daughter reveals a striking decline in performing ability. It is logical to ask which other factors may have caused such a disturbance in the transmission pattern. One could argue, of course, that an element of individual talent needs to be factored in – which is certainly a valid point, but when a larger sample of singers and songs from the Parry Collection is taken into consideration, regardless of the question of literacy, we can identify the same declining trend. Bringing into the picture more contextual elements would likely provide some clues. Thus before we continue with investigation of several of Arifa’s songs, it may be in order to recall the timeline once again: if Đula was born in approximately 1868, and Arifa was in her mid to late thirties at the time of the collection, we can estimate that Arifa was born around the turn of the century. When she was in her early teens, World War I was raging in her homeland, disrupting her youth and the daily patterns of living, of which traditional singing was an important aspect. We have seen a similar decline in the oral tradition of the region two other times in the twentieth century: during World War II, right after Parry and Lord collected the bulk of the materials held in the Parry Collection, and during the Yugoslav wars of the 1990s, which formed ethnically pure enclaves and caused large displacements of the local populations. In other words, every time a traditional community is subjected to disorder (of which wars are the most egregious category), we are likely to see a disruption in the preservation of its cultural heritage. Could it be that Arifa, in the midst of the unrests and exactly at the stage when a singer absorbs the special language of oral songs, simply could not appropriate the tradition in the same way her mother had under the secure roof of her wealthy family and without the need for hard labor? Even a superficial glance at Đula’s immediate family reveals that the close community started disintegrating, with one of her sons moving to an urban center and another even emigrating to America. This brings us once again to the conclusion that the texts of oral songs constitute a body of empirical evidence which can be fully understood only if we take into consideration the contextual elements. In this case, the process of transmission between the two generations was affected by a cohort of factors, many of which directly point to the rupture in the social texture, which in turn affected the oral heritage in both qualitative and quantitative dimensions. Examining Arifa Kurtović’s songs outside this broader context, or even worse, taking a random song text and analyzing it outside the context of traditional referentiality, to use John Miles Foley’s term, means to dislocate the tradition to such a degree that the text becomes reified in much the same way the context had, as noted by Bauman and Briggs (Bauman & Briggs, 1990:68).

Foley’s insistence on the presence of tradition repeats in a way the quest for contextualiztion, but although it has a performative dimension, in its essence it is more text-centered. In order to understand a given song, especially when it reaches the point of entextualization, Foley holds that it is crucial to be aware of the context of tradition which is “enormously larger and more echoic than the text or work itself,” and which “brings the lifeblood of generations of poems and performances to the individual performance or text. Each element in the phraseology or narrative thematics stands not simply for that singular instance but for the plurality and multiformity that are beyond the reach of textualization” (Foley, 1991: 7). If one considers the importance of both diachronic and synchronic aspects of a tradition when reading the text of an oral traditional song, one quickly realizes the relevance of an anthropological dimension in this type of research, and the need to systematize and study these materials not in isolation, but following the patterns that the community imprinted onto them. In other words, contextualization needs to be performed on the basis of anthropological/historical/ linguistic and other data but also grounded in the other extant texts/recordings from the given tradition. A text by Arifa Kurtović is read differently on its own, differently in conjunction to multiforms rendered by her mother, and yet differently again in relation to all such multiforms provided by a group of singers from her community of Gacko, or, for that matter, in the context of the whole tradition of the region.

In order to be able to demonstrate more closely some of these arguments and to illustrate symptomatic occurrences in songs belonging to two different generations of singers, it is relevant to examine several examples in detail. The first song to be examined is PN 11843 by Arifa Kurtović, which is among her best songs. It consists of 56 lines and in the narrative sense it is the most elaborate of her materials. Many of Arifa’s songs are brief and contaminated in terms of thematic elements, but this is one of the very few with a developed exposition, complication, and resolution. It also has a fair amount of detail and dramatic buildup. This particular song is frequently found in the Parry Collection corpus, but is also widespread in the Slavic part of the Balkans, with versions of it both in the Matica hrvatska series Hrvatske narodne pjesme (vol. 5: texts 161/162) and in Karadžić’s Srpske narodne pjesme (vol. 5: text 691). There are no fewer than three multiforms by Arifa’s mother Đula and one by her aunt Emina, thereby providing a solid basis for comparison. It is indicative that Arifa’s multiform, like the remainder of her songs, has many more metrical errors than those by the older singers. Even if we factor in possible slips on the scribe’s part, this problem persists, since among the compared texts we have two, one by Arifa and the other by Đula, that were both written by Sifa Kurtović (Arifa’s daughter and Đula’s granddaughter). Đula’s multiform as taken down by Sifa is metrically far more stable than Arifa’s. In this song Arifa’s transitions from one narrative segment to another are performed quite fluently, although they do not surpass the quality of those found in Đula’s or Emina’s multiforms. Emina in particular supplied quite a dramatic and emotional version, which happens to be also the longest, most nuanced, and metrically impeccable. When all five songs are read side by side, the combination of the themes contained in them yields the following plot:

1. widowship of lady Pirinbegovica, many nobles seek her in marriage but she will accept only the one who will also receive her son from the first marriage
2. one of the suitors (called Hasanaga in some songs, bey Ljubović in others) makes a promise to take in both her and her son
3. Pirinbegovica sends a message to the suitor to prepare many riches
4. the suitor arrives and orders Pirinbegovica to take along her late husband’s arms
5. on the road to the suitor’s house, he stops and orders Pirinbegovica to throw her son on the grass
6. Pirinbegovica hesitates, the suitor makes an aliitional threat
7. Pirinbegovica has no choice and lowers the child on the grass
8. the suitor slaughters the child
9. Pirinbegovica’s lament/lullaby in which she consoles her son that instead of her
   1. the wind will rock him
   2. the dew will wash him/nurse him
   3. the fairies will nurse him
   4. the leaves will kiss him
10. Pirinbegovica’s message to the fairies
11. the fairies take care of the child
12. arrival to the suitor’s house and celebration
13. illness and death of the suitor
14. the soil, the waters, and the mountain refuse to accept the sinner’s body because he killed/abandoned the child

While all five versions contain most of the themes, not a single multiform has all of them. Also, in some of the multiforms the child is not killed but abandoned in the mountain, however the punishment is the same. The themes from Arifa’s song are found in her mother’s and aunt’s versions, with the exception of the detail concerning the suitor’s order to lady Pirinbegovica to take along her late husband’s arms – presumably so he could make use of it. This element of the suitor’s self-interest is essential since it is a driving force behind his marital proposal. In Đula’s and Emina’s multiforms the arms are not mentioned, however all of them state the fact that the suitor loaded nine or ten horses of the child’s inheritance when he came for the bride. The detail pertaining to the arms, as well as the fact that Arifa’s multiform, despite relying on the same narrative pattern and formulaic expressions, still differs from her mother’s and her aunt’s versions in its verbal realization is proof enough that the singer was able to compose in performance. However, it is noticeable even in this song, which is in its quality considerably better than others in her repertoire, that her ability to utilize the language of traditional songs in a creative and independent manner was on a lower level than that of the older members of her family.

The importance of synchronic comparison comes through particularly clearly in the example of this song since there is another singer of Đula’s generation from the Gacko area who had it in her repertoire. Moreover, her maiden name indicates that she, too, may have been related to the same family. The singer in question is Đula Hrustanović née Pašić (thus she shares Đula’s and Emina’s maiden name), a resident of Međuljice who was seventy-two and illiterate. She provided a superb multiform of the ballad analyzed here, 150 lines in length, in which the greediness of the suitor is underscored on all levels. In this version, which follows the already established narrative line with greater ornamentation and nuance, the principal female character (called Alibegovica) leaves a hundred gold pieces (ducats) by her infant’s side in hope that someone will find him and help him. Soon thereafter, the hunters arrive and save the baby, but Alibegovica, desperate to make sure wild animals did not attack him, sneaks out to the mountain during the first wedding night to discover that both the child and the ducats are gone, but also that the hunters have left a letter for her. In the morning her new husband tries to console her by promising her new offspring, but she questions his good intentions: if a child who brought him so many riches was left behind, she can only imagine how he would treat her and her future children now that she is poor. The suitor shows some remorse, but the damage is irreparable. The singer skillfully skips through a timeline and in the closing part of the ballad offers a development not found in any other multiforms from this cycle. In broad strokes she describes the abandoned child’s path to adulthood and his coming of age, at which point he confronts his mother’s husband, demanding that he return the riches appropriated from him so slyly through the wedding arrangements. If this is not honored, the young man insists on a duel. Knowing the force and skill he would encounter, and facing his own demons, the suitor/husband returns the wealth to Alibegovica’s son. The element of greediness, which is central for understanding the motivation of the suitor, is thus fully elaborated in this multiform. Arifa glosses over it in her song, as do her mother and aunt. In the Gacko circle there existed different renderings of this ballad (as we see in the example of Đula Hrustanović), [9] and on the basis of the proximity of the extant multiforms by the daughter-mother-aunt group, it can be stated with a fair amount of certainty that Arifa learned this song from her mother, possibly from her aunt.

Arifa was clearly able to reproduce with minimal modification what she learned, but was unable to generate new thematic combinations. There are several other songs by her where we have a similar situation, such as PN 12022, in which a young man, after having lived in Istanbul (Stambol) chooses Egypt (Misir) as a place to settle and has a correspondence with his bride-to-be about the type of house she desires. Here again Arifa closely follows the narrative pattern and ornamentation (especially when it comes to the description of the new house) found in her mother’s multiform. However, there are songs from this cycle that are considerably more elaborate, including the one by Arifa’s mother-in-law Umija Kurtović (PN 11848a), which is twice as long. The two songs discussed here are among Arifa’s more successful in her modest repertoire of ballads (only four in the Parry Collection are sung to their conclusions), and in both she shows a strong inclination to follow her mother’s matrix. The other two songs display some signs of contamination (in the sense that a theme is imported from another cycle but is not adapted fully to the new semantic environment). As one might expect, lyric songs, which comprise most of Arifa’s repertoire, show even greater textual stability. Although Arifa is not an exception in her tendency to employ minimal variation in the lyric repertoire, her metric command is also often unsatisfactory, and she fails to connect various thematic components in a sensible narrative. Even though there are some instances where this happens to older singers, including her mother, with Arifa these problems persist through most of her repertoire.

Finally, it might be relevant to compare the sample of songs by Šaćir Kurtović (Arifa’s husband) and his mother Umija Kurtović. Šaćir has no heroic songs in the Parry Collection, thus the three short lyric songs written down by his daughter are the only materials by him. Out of the three songs, two are found in his mother’s repertoire (PN 5009 Omilje mi milje moj[e], omilje mi Jagoda corresponds to Umija’s multiform PN 11886; PN 5012 Poljem se vija Hajdar delija corresponds to Umija’s PN 11892), and the third was sung both by his wife Arifa and her mother Đula (PN 4993 corresponds to Arifa’s PN 2453 and Đula’s 2308). In all three songs textual stability is extremely high and it is likely that Šaćir learned these songs from the female members of his family. Unfortunately, there is no other evidence of his singing abilities, so these examples are ancillary to the central argument in this article.

Recalling in this concluding part once again Bauman’s and Briggs’ theory, it can be said that the texts by the Dizdarević-Kurtović-Šaković group of singers have become entextualized in the process of collecting and archiving, and above all, by the sheer fact that they were extracted from the performative context and captured in written format. Bauman and Briggs state that efforts towards contextualization emphasized that “performance is anchored in and inseparable from its context of use” (Bauman and Briggs, 1990:73), but they raise a red flag by pointing to the counterforces that are at work when discourse is turned into a text. They hold that performance is especially susceptible to entextualization since it is a marked way of expression and in essence an enactment of the poetic function (as defined by Roman Jakobson; Bauman and Briggs, 1990:73; Jakobson, 1990: 69-79 ). While Bauman and Briggs offer a range of approaches to entextualization, the concern here is what happens to the text once it becomes decontextualized. They claim, and rightly so, that the processes of “decontextualization and recontextualization of performed discourse bear upon the political economy of the texts” (Bauman and Briggs, 1990:76). In other words, once the text is decontextualized, we can access it from different positions, evaluate it with diverse tools, interpret it with varying levels of understanding and competence, and attach to it a range of meanings. If these concepts are applied to the materials examined in this article, what are the conclusions? How has the process of entextualization affected our perception and interpretation of these songs? First and foremost, entextualization has pushed the notion of community to a secondary level, thereby diminishing our ability to assess every performance and every multiform in the context of tradition. The texts are extracted from their natural environment and considerable effort is needed to supplant the surrounding data that enables us to interpret the text in a meaningful way. If an isolated text by Arifa Kurtović or Đula Dizdarević (or any other singer for that matter) is read or listened to, can the reader/listener be aware of the full ramifications of the text? This brings us back to Foley’s notion of traditional referentiality. None of the songs analyzed in this article should be read, nor can they be fully appreciated, without the context of other songs from the tradition, a cognizance of the relationships between the singers, and a basic familiarity with the community from which the songs stemmed. Entextualization, when exercised with these points in mind, preserves the relations and texture of the community from which materials were recorded, employs tradition as an essential and indispensable frame of reference, and presents each text as part of a broader but closely entwined system. Existence of such knowledge structure is a prerequisite and often a key for understanding oral traditional materials.

All materials collected in Gacko, 1935

All materials collected in Gacko, 1935

All materials collected in Gacko, 1935

All materials collected in Gacko, 1935

All materials collected in Gacko, 1935

Note: In cases when Arifa has more than one multiform, the songs are cross-referenced. Multiforms by her mother Đula Dizdarević and her aunt Emina Šaković are indicated. When there is a remote relation between the songs (correspondence only in one segment of the song), it is marked that songs are “related.” If only PN (Parry number indicating the text signature in the Parry Collection) is listed, the songs are multiforms.

2241 Siđela sam preksinoć do sedam sahati – contaminated

2242 Đevojka momku prsten povratila

2264 Dvoje mlado sitno smilje bralo – Đula PN 10548

2265 Kićeno nebo zvezdama

2266 Ružu brala dilber Umihana – related to Emina PN 2320

2267 Rasti bolje moj zeleni bore – Arifa PN 12012, Arifa PN 11878, Đula PN 7055, Emina PN 1343

2452 Oj travice bona, što si tako rosna – Đula PN 11753

2453 Na sred polja bijeli se kula – Đula PN 2308

2454 Ovce čuva Omer čobanine – related to Đula PN 2449

2455 Pođe cura na vodu

4983 Igrali se vrani konji – Arifa PN 11849 , Đula PN 11705

4984 Vrani se konji vadaju – Arifa PN 11854, Emina PN 2323

4985 Što Morava mutna teče – Đula PN 7091

4986 Zmaj preleće s mora na Dunavo – Đula PN 2183, Đula PN 2313, Đula PN 11728, Đula PN 6329

4987 Je l’ svanulo, je li sunce granulo – Đula PN 11739

4992 Šta se ono iznad grada đido đido more đidijo – Emina PN 2240

5000 Kraj Saraj’va jedna bašča zelena – Đula PN 2617, Emina PN 2002

5001 Drimaš ‘i šćerko, ne drimam majko – Đula PN 2306

5004 Hajrija, bona, raspleti kose

5008 San zaspala dilber Naza u bašči – Đula PN 7052, Đula PN 11718, Emina PN 1006

5010 Haj moja ružo rumena – Đula PN 11763

5011 Gorom jašu gospoda svatovi – Arifa PN11869; related to Emina PN 2327, Umija Kurtović

(Arifa’s mother-in-law) PN 11908

11812 Kosovo more, Kosovo polje – Đula PN 2217

11815 Zavikala b’jela vila – Đula PN 11698

11821 Zeleni se šaša, mene prosi paša – contaminated

11822 Polećela, nalećela dva goluba bela – related to Đula PN 7073; related to Emina PN 977

and Emina PN1034

11824 Kolo igra, ja ne vid’la – Arifa PN 12020, Đula PN 2212

11825 U đul bašči kraj šimšira

11826 Sarjoš Aljo drume zatvaraše – related to Emina PN 2271

11828 Ljuto kune stara majka Mehmeda

11830 Gledala sam ispod obrvica

11831 Ko se ono bregom šeće

11834 Gledala sam preko bela sveta

11835 Ustala sam jutros rano

11836 Pomračina, selo spava

11837 Ne plač’, ne plač’ draga – contaminated, related to Đula PN 11748 and to the cycle

including Emina PN 1980, Đula PN 2220, Đula PN 11741, Đula PN 11758

11839 U Trebinju gradu – Đula PN 7090

11840 Razbolje se Đerđelez Alija – Emina PN 9690

11841 Falijo se žuti limun kraj mora – Arifa PN 12017, Đula PN 7078

11843 Obudovlje Pirimbegovica – Đula PN 1317, Đula PN 2618, Đula PN 5016, Emina PN 2315

11844 Sitna trava zelena – contaminated; related to Đula PN 2262

11849 Igrali se vrani konji – Arifa PN 4983, Đula PN 11705

11850 Pomuti se Sava i Morava – Emina PN 1048

11854 Vrani se konji vadaju – Arifa PN 4984, Emina PN 2323

11855 Slavuj pile, ne poj’ rano – Emina PN 2312

11857 Koliko je Dabar polje ravno – related to Đula PN 2251

11861 Stalo sunce čuda gledajući – Emina PN 2111

11862 Je l’ svanulo, je li sunce granulo – Đula PN 11739

11864 Viš višnjica rod rodila – Arifa PN 12015, Emina PN 1017

11865 Oj livado, rosna travo – Emina PN 2243

11868 Moja majko, dugo jadna bila – Đula PN 7046, Emina PN 2278

11869 Gorom jašu kićeni svatovi – Arifa PN 5011, Umija Kurtović PN 11908; related to

Emina PN 2327

11872 Ovce čuva Raza i Ismija – Arifa PN 2454, Đula PN 2449

11874 Zaprosijo Zajim Osmanaga – truncated; Emina PN 2274

11878 Rasti bolje moj zeleni bore – Arifa PN 2267, Arifa PN 12012, Đula PN 7055, Emina PN 1343

11882 Pusta kršla jelo Sarajevo

11883 Kasno pođoh iz Morića hana – Đula PN 7069, Emina PN 2340

11884 U kokota sumbulova – Arifa PN 12014, Đula PN 4995

11888 Kroz bašču mi potok teče

11890 San zaspala dilber Naza u bašči – Đula PN 11718, Emina PN 1006

11891 Pošla je devojka za goru na vodu – related to Đula PN 2217¸ Đula PN 7056, Đula PN

11690, Emina PN 1996 , Emina PN 2385, Emina PN 1030, Emina PN 1042

11895 Crven fesić mamo – Emina PN 2107

11896 Prođe mi momče kroz sokak – Emina PN 1007

12001 Sve ptičice lastavice – Emina PN 2001

12002 Vjetar ružu niz polje teraše – Emina PN 1996

12003 Poletijo, dušo, vran gavran

12004 Iz polja ruža procvala – contaminated; Emina PN 996

12005 Šajke lađe svukud po moru

12006 U Mostaru gradu – related to Đula PN 2158

12007 Potrvena po ledini trava – Arifa PN 12023

12008 Zaspala Džeha kraj mora – related to Đula PN 7052

12011 Konja kuje Meho momče mlado – Emina PN 2118

12012 Rasti bolje moj zeleni bore – Arifa PN 11878, Arifa PN 2267, Đula PN 7055, Emina PN 1343

12013 Sitna trava zelena

12014 U kokota sumbulova – Arifa PN 11884, Đula PN 4995

12015 Viš višnjica rod rodila – Arifa PN 11864, Emina PN 1017

12016 Momak mami đevojku – contaminated

12017 Falijo se žuti limun kraj mora – Arifa PN 11841, Đula PN 7078

12018 Poigralo kolo na ravnoj ravnini

12019 Ko j’ za kolo, hajde u kolo – Emina 1000, Emina 2029

12020 Koliko igra, ja ne vid’la – Arifa PN 11824, Đula PN 2212

12021 Puhni mi, puhni hladane – Đula PN 10528, Đula PN 11752, Emina PN 996

12022 Alibegu Misir omilijo – Đula PN 7041, Đula PN 11695, Umija Kurtović PN 11848a,

Umija Kurtović PN 11920

12023 Potrvena trava djetelina – Arifa PN 12007