From the Archive to the Field: New Research on Albanian Epic Songs

This essay aims primarily to point out some new directions in research on Albanian epic song. [1] Starting from archival materials and the issues they raise for further investigation in the field, it will focus in particular on certain structural elements of the songs related to their performance. The Albanian-language archival materials of the Milman Parry Collection of Oral Literature (henceforth MPCOL), collected in the 1930s by Milman Parry and Albert Lord, provide significant insight into the tradition. The first section offers an overview of the major collections and scholarly research on Albanian epics through the twentieth century, a rich body of work that includes local fieldwork-based studies as well as important foreign contributions, and that documents the geographical scope and contents of the Albanian epic tradition, which constitutes an important part of the oral heritage still alive in Europe today. The second section focuses more narrowly on the materials collected by Milman Parry and Albert Lord in the 1930s, exploring the ways in which those songs as well as the singers’ biographies—which have often been neglected in discussion of the MPCOL—still hold valuable insights for the study of oral poetry more broadly. In particular, this essay explores the implications of the collection process itself, as embedded in recordings and texts from the MPCOL. Finally, the third section summarizes some recent research conducted by the authors focusing on the singer Isa Elezi-Lekëgjekaj, one of the most important living practitioners of this epic tradition. His performances highlight some of the demands placed on the body and the resulting bodily techniques that emerge as part of this performative tradition.

Epic songs are a fundamental part of Albanian cultural identity. Albanian culture includes two linguistic areas: the Tosk dialect area covers the Southern part of Albania, while the Geg dialects are spoken in Northern Albania, Kosova, and the areas inhabited by Albanians in Macedonia and Montenegro. The practice of epic singing persists most actively in the Geg area, in Northern Albania and Kosova, and especially in specific regions such as Rugova, in Kosova, at the border with Albania and Montenegro. The main repertoire of epics songs is called Këngë (or Kânge, in the Geg dialect) Kreshnike, usually translated as ‘songs of the frontier warriors’; it is also called kënge trimash ‘songs of the heroes’ or kënge të moçme ‘ancient songs’. Both textually and in their performance, Këngë Kreshnike have their own particular traits that distinguish them from other poetic repertoires. [2] They are usually performed by a singer who accompanies himself with a one-string bowed instrument, called lahuta in Albanian. The poems feature legendary characters, dealing most often with the deeds of two brothers, Muji and Halil, their friends and relatives, or other heroes like Gjergj Elez Alia, as they struggle against their enemies and encounter the presence of supernatural elements. The songs are performed during holiday gatherings, social visits to people’s homes, and family celebrations.

Research and publications on Këngë Kreshnike date back to the end of the nineteenth century. [3] Besides local collectors, sometimes supported by local institutions, some foreign scholars have also played an important role in the study of the Albanian epics. Indeed, an early 15-line fragment, clearly recognizable as belonging to the Këngë Kreshnike, was published by the Austrian linguist Gustav Meyer in 1897. At the beginning of the twentieth century, complete versions of Albanian epic songs were published thanks to local collectors like Hasan Fejzullahu and Gasper Jakova-Merturi. The establishment of a large corpus of Albanian epic songs is mainly due to the work of the Franciscan friars and Jesuits based in the city of Shkodër. The town served as the old capital of a vilayet during the Ottoman Empire; it was an important cultural center with historical connections to Western culture and Venice and a strong presence of the Catholic church. In Shkodër, thanks to the initiative of the friar Bernardin Palaj, songs identified as Këngë Kreshnike started to appear in the journal Hylli i Dritës (The Star of Light). This journal, founded by Father Gjergj Fishta in 1913, was an important forum for bringing together cultural and political ideas; from the outset, it focused primarily on Albanian history, literature, and folklore, including excerpts of Fishta’s poem Lahuta e Malcís (The Highland Lute). A collection of a substantial number of Këngë Kreshnike, edited by the friars Bernardin Palaj and Donat Kurti, appears in the series Visaret e Kombit (The Treasure of the Nation) published in 1937, the 25th anniversary of Albanian independence from the Ottoman empire; the volume has the title Kânge kreshnikësh dhe legjenda (Songs of the frontier warriors and legends), and includes 34 songs about Muji and Halil. Additionally, important research was carried out by foreign scholars, including (among others) the Austrian scholar Maximilian Lambertz, the Italian friar Fulvio Cordignano, and especially Milman Parry and Albert Lord in the 1930s, whose landmark research is discussed below.

After World War II and the establishment of Enver Hoxha’s regime in Albania, local scholars, based in Tiranë and Prishtinë, carried out research on epics with the support of various institutions, like the Instituti i Shkencavet (Institute of Sciences) and the Instituti i Folklorit (Institute of Folklore), which would later become the Instituti i Kultures Popullore (Institute of Folk Culture). This institute created a framework for the study of traditional Albanian folklore, with the aim of building a national culture. The research focused primarily on the collection of texts, published in volumes edited by Qemal Haxhihasani in 1955 and 1966. Later, scholars began focusing on the study of musical melodies too, with a book edited in 1983 by Ferial Daja. In parallel to these local developments, foreign scholars, such as Stavro Skendi and Agniya Desnickaya, also made authoritative contributions that deal in turn with the controversial issue of the relationship between the repertoire of Këngë Kreshnike and that of neighboring Slavic traditions. [4]

Meanwhile, research conducted in Kosova is of particular importance in the field of epic. Inhabited by a large majority of Albanians, yet remaining outside the national borders of the state of Albania when it was formed, Kosova was for decades—both before and after the breakup of Yugoslavia—an autonomous province of Serbia, with moments of tension with Belgrade, until its unilateral declaration of independence in 2008. Despite this sensitive situation (which has sometimes hindered the study of aspects of Albanian culture), scholars in Kosova have been able to promote and publish important work. While the Këngë Kreshnike in Albania represents only a portion of the local oral traditional corpus, in Kosova—where the Geg dialect is widely present—the epic is at the center of scholarly research. The institutional base for research on oral heritage in Kosova, including the epic tradition, has been the Instituti Albanologjik i Prishtinës (The Institute of Albanology, Prishtina), which played an important role for decades in the cultural and political consciousness of the Albanians of Kosova. After a pioneering period in the 1950s, the institute began its full range of activities in 1967. Led for many years by the distinguished scholar Anton Çetta, the institute supported the publication of multiple volumes on epic, with research carried out by scholars like Rrustem Berisha, Anton Nikë Berisha, and Zymer Ujkan Neziri. As early as 1953, they began publishing texts of epic songs in the volume of folk songs, Këngë popullore shqiptare të Kosovës dhe Metohisë, with subsequent contributions made by several volumes of Këngë Kreshnike published in 1974, 1991, and 1993. More recently, Neziri has been editing the most ambitious project supported by the Institute, a multivolume study of the epics of the area of Rugova, which remains the principal repository of living tradition to the present. Three volumes have already been published, with two of those devoted to the repertories of individual singers. [5]

For decades following World War II, the international isolation of Tirana and Prishtina hindered the circulation of these materials and scholarly discussion internationally. In Albania more recently, important changes have started to emerge in the realm of cultural politics and the activities of local scholars, while scholarly contact with Western countries has also increased. In Kosova, too, from the fall of Yugoslavia to its recent declaration of independence in 2008, an important process of opening and increasing contact with foreign scholars is underway. Local research projects of note, in addition to Neziri’s long-term project, include the publication of other collections of epics and reprints of older collections that are no longer available: in 2008, Anton Nikë Berisha reprinted 51 texts from older collections, and in 2009 Nimon Alimusai republished 20 canonical texts. In addition, books devoted to folk songs and oral tradition (e.g. Krasniqi and Krasniqi 2007) or textbooks for school (e.g. Neziri 1999) include significant parts of epics. The reprinting of the work of the friars of Shkodër, especially the landmark volume of Visaret e Kombit, unavailable for decades, is especially important. Published in 2007 to mark the 70th anniversary of the first edition, the reprint grew out of an initiative of the Franciscan community, still present today in Shkodër, with their important library.

A key step in fostering deeper contacts and communication between scholars in Albania and abroad has been the translation of important studies and collections into and from Albanian, such as Stavro Skendi’s book (1954; transl. 2007), which includes in appendix an especially important article by Russian scholar Agniya Desnickaya. Translations of key collections and texts from Albanian into Russian, Serbo-Croatian, and French date back to the mid-twentieth century. And since the first English publication of Këngë Kreshnike (Elsie 1994), the work of Robert Elsie has been of particular importance in translating, disseminating, and popularizing Albanian language and literary culture. For instance, his 2004 book of translations with Janice Mathie-Heck, Songs of the Frontier Warriors: Këngë Kreshnikësh, is a bilingual edition of a selection of epic songs with texts taken from the old collection of Visaret e Kombit, republished in a modernized orthography. The book marks an important milestone in making these texts accessible to a much wider audience.

A recurring theme in recent scholarly discussion on Albanian epic is the UNESCO Intangible Heritage List, especially since 2005, when UNESCO declared polyphonic vocal music from southern Albania a Masterpiece of the Oral and Intangible Heritage of Humanity. [6] A subsequent discussion arose about the possibility of petitioning UNESCO to include the Kënge Kreshnike on the Intangible Heritage List as well. On the one hand, there is clear consensus that some areas, especially the region of Rugova in Kosova, have great relevance for oral epic tradition; on the other hand, the complex political situation in Kosova makes a formal nomination for the Intangible Heritage List very unlikely, since it requires the support of a country with direct interest in preserving that tradition. The topic has been very popular in recent years at conferences and festivals, such as the festival of Gjirokastër, Albania, in September 2009, where musical performances and a scholarly conference were held jointly. [7] The issue was raised again at an August 2010 conference organized in Prishtina by the two main cultural institutions of Albania and Kosova working on epic tradition, the Institute of Albanology, Prishtina, and the Center for Albanian Studies, Tirana (Qendra e Studimeve Albanologjike, Tiranë), with an international academic board chaired by Zymer Neziri. [8] The conference brought together a sizable number of Albanian scholars from Tirana, Prishtina, and Skopje, along with international scholars from across Europe, the United States, and Japan, fostering a transnational conversation about the importance of Albanian epic and the current state of scholarly research on the topic.

Another key corpus for the documentation of and research on Albanian oral poetry is the Albanian-language holdings of the Milman Parry Collection of Oral Literature (MPCOL) at Harvard University. [9] Despite being collected nearly 80 years ago, these materials still point to new directions in the study of oral literature. There are two main groups of materials in the MPCOL collected by Parry and Lord as part of their broader research on South Slavic epics. The first group comprises recordings of epics, performed by bilingual singers from Novi Pazar—especially Salih Ugljanin in November 1934—as well as lyric songs and other fragments, most notably featuring the polyglot singer Fatima Biberović, originally from Prizren but living in Gacko when Parry and Lord recorded her in April and May 1935. The second group includes transcribed texts, collected by Albert Lord in a set of notebooks during his 1937 trip to northern Albania. [10]

This Albanian corpus from the MPCOL is revealing, both in terms of the study of the Këngë Kreshnike and more generally in relation to epic repertoire and performance. The use of recording technology, especially the special apparatus used by the researchers, allowed them to document these Albanian performances and to produce the first uninterrupted recording of a complete epic in Albanian. Lord’s notebooks from Albania include a body of texts exceeding 22,000 lines, collected from about 40 singers. Its texts are comparable to other collections made in Albania and Kosova, while also featuring the longest Albanian oral poem ever collected. [11]

The collection of these materials by Lord coincides with some of the most relevant publications of Këngë Kreshnike by local collectors. In 1937, the year of Lord’s Albanian trip, the landmark volume Visaret e Kombit was also published. While Lord was in Shkodër, the point of departure for his collecting journey, he was in touch with Franciscan friars and other local collectors who helped him in his research. In a notebook in the MPCOL entitled Adresser, which Lord used as a diary for his journey, he mentions the Franciscan press of Shkodër (p. 8), as well as the friar Bernardin Palaj (p. 12), present in some of Lord’s photos. He also collaborated with Murat Paci, a teacher at the local gymnasium high school, who sent Lord a notebook with transcribed texts after his return to Cambridge, which is also preserved in the MPCOL.

While in Albania, Lord worked with some of the same singers as local collectors, who deemed certain of these singers the best of their generation. Following Parry’s methodology from a few years earlier, Lord provides essential details about the singers and their lives, in addition to the song transcriptions. He typically notes biographical data for the singers (including age, religion, writing ability, training, languages spoken, travels, and contact with other singers), song titles for the singer’s entire repertoire, and the complete transcription of the text of one or more representative songs. We rarely find this kind of information in other epic collections of the time. A singular case is that of Gjergj Pllumbi, a famous singer from Shala, whose poetic repertoire was documented by other leading scholars like Ernest Koliqi, Fulvio Cordignano, and Anton Çetta. In Lord’s manuscript, we find arguably the most reliable information about Pllumbi’s activity and background, otherwise depicted in a romantic and often unrealistic manner by other scholars at the time. For example, Koliqi drafted a kind of heroic portrait of Pllumbi (like the heroes of his songs), stressing his role as a local chief, while highlighting the complete isolation that surrounded him as a rhapsodist living in the mountains. In contrast, Lord notes key details like his teacher’s name (Mark Kolesh from Theth), the trips he took to many parts of Yugoslavia, and the complete list of songs in his repertoire. [12]

All these elements are certainly important in terms of documentation, ethnographic data, the cultural history of the collection, and the expansion of the songs of the repertoire of the Këngë Kreshnike itself. However, even more significantly, these archival materials return our attention to the field, encouraging us to pay particular attention to aspects of performance. Such a shift in focus, made possible by the existence of audio recordings, is especially crucial for the performance of Albanian epics, with consequences for the question of their versification.

The Albanian recordings by Ugljanin, which include both sung lines and rhythmically recited songs, offer a useful point of departure. The presence of recited, rather than sung, texts—a phenomenon occurring in the Slavic materials too—is due to the period of mourning after the assassination of the Yugoslav King Alexander I in Marseille on October 9, 1934. For several weeks after his death, singing was restricted. In the Albanian epics, the method of performing lines—whether sung or recited—results in corresponding changes in the meter of versification. This shift can be seen in one of the most unique recordings of Ugljanin, the song Mark e Musa (PN 659, 154 lines long). Here, Ugljanin begins the song by singing the first six lines. Then, without any interruption, he shifts to rhythmic recitation of the texts.

In the performance of this song, the presence or absence of sung melody corresponds to changes in the metrical structure. The verses presented in the sung performance (1–6) are similar to the decasyllabic (with two long stresses on the final syllables), while those of the recited texts (7–154) are similar to the octosyllabic, with two isochronous stresses (usually on the 3rd and penultimate or final syllables). This change in the verse results from the switch from singing to recitation, all within the performance of a single song. Here is the transcription of the text performed by Ugljanin with stressed syllables highlighted (numbers enclosed in parentheses indicate stressed syllables in the recited version):

sung:

recited:

sung:

recited:

In fact, this example suggests a kind of expansion/compression system of the line, depending on the presence or absence of repetitions and filler words, which allows a shift from a sung, decasyllabic mode of performance to a recited, octosyllabic one. Such changes appear in many other examples of songs from Albania and Arbëresh oral tradition. [13] However, in the case of epics, this point has particular significance because it is also bound up with the very method of collecting songs.

Recording a song while it is sung, recording its recited text, or transcribing it in a recited version under dictation are all different ways of collecting a song with different performance conditions. All three forms were used in collecting the MPCOL materials, as explained by Lord. [14] Indeed, Parry began collecting texts by dictation, much as earlier collectors had done for their published collections of epics. However, even when he was using sound-recording devices, Parry did not stop collecting dictated texts. His conceptualization of a Homeric scholar (or scholar of epic, more generally) mirrors the hypothesis of the Homeric text as a written transcription of a dictated performance. [15]

With regard to the difference between dictating and singing, Lord’s essay, “Homer’s Originality: Oral Dictated Texts,” goes further still, articulating a qualitative difference, referred to by singers themselves, between a text dictated with the aim of making a written transcription and a sung performance. [16] The singers seem perfectly aware of this difference between a sung performance of a text and a dictation for transcription, as evidenced by a conversation between Nikola Vujnović and the bilingual singer Sulejman Makić:

For Parry and Lord, the moment of transcribing a recited text, dictated line by line, becomes a relevant form of performance, even if it is not the usual one for singers. [18] In the case of the Albanian epics, the difference between sung and recited performance goes even further because, as we have seen, it entails a change in meter. This change raises new questions about Albanian epics, which are commonly thought to be characterized by a decasyllabic line. Although the most common Albanian folk verse is the octosyllabic line, in all the major collections, legendary epics mainly employ decasyllabic lines, and have been consequently defined in this way by the most authoritative scholars. [19] Moreover, the case of Ugljanin is particularly striking because this change occurs during a single performance. Performance emerges as a kind of Procrustean bed in which lines can be stretched or shortened with repetition and filler words as they are adapted to the structure created by the presence of melody or the rhythm of recitation. [20]

The song of Mark e Musa serves as the starting point for a research project that shifts from the geographies documented in the MPCOL to the border area of Rugova, mentioned above as the center of Albanian-language epic performance today, focusing on different ways of performing epic songs. Ugljanin’s method of singing a song in decasyllabic lines while reciting it in octosyllabic lines is confirmed by research conducted with many other singers, especially one of the most acclaimed singers active today in Rugova, Isa Elezi-Lekëgjekaj. [21] Isa Elezi is perfectly aware, like other singers, that singing and reciting are different modes of performing a text. For example, a group of researchers asked Isa to perform the same song, Kanga e marteses se Halilit (The Song of the Wedding of Halil), on different occasions. These occasions were filmed in order to document variations of wording between one performance and another. [22] In the first performance (Version A, singing mode), Isa sang the text while accompanying himself on the lahuta, according to the “normal, sung” performance method. The lines are about 11–12 syllables (with repetitions and filler words) and the rhythm is set by the melody as well as by the movements of the singer’s hands while bowing the instrument:

In the second performance (Version B, reciting mode), Isa rhythmically recited the text without singing. The results reveal precisely the same performance practice as that of Ugljanin’s recitation: the rhythmically-recited verse has eight syllables (plus or minus one) and the rhythm is set by the regular presence of two isochronous accents (highlighted):

At this point, Isa was asked to do a third performance: to dictate the text line-by-line in order to allow a listener to transcribe it without the help of a recording. The aim was to recreate the mode of collection of many texts before the advent of sound recording, a mode that remained in use even after the introduction of recording technologies, including in Lord’s 1937 collection of Albanian materials. In this third performance (Version C, dictating mode), the timing of Isa’s recitation depended on the speed of the scribe taking dictation. Even though the number of syllables resembles the previous recited performance (Version B), now the rhythm is broken by the pauses required by the dictation—indicated approximately in seconds in square brackets—and the the regularity of isochronous accents is totally lost:

The poem Kanga e marteses se Halilit is performed by Isa Elezi differently each time, each with a rhythmic specificity that affects the meter of the verse. These results problematize the presumed ideal of a decasyllabic line. Moreover, even the categories of octosyllabic and decasyllabic verses, as most scholarship describes them, prove to be inadequate, since the rhythm of the isochronous stresses (especially in the recited version) appears to be much more important than the number and length of syllables, which in these examples is subject to slight variation.

Ultimately, the variations in “performances” of the same song suggest a disconnect between the reality of performance and the fixity of lines typically found in the published corpus of epic texts. At a minimum, this raises the question of the role scholars play in collecting—or more precisely in some cases, in establishing, or even co-creating—the texts of epic songs. Scholars as early as Skendi have acknowledged this point, highlighting how collectors often justified their interventions in editing and re-arranging texts according to romantic ideals in order to mold them according to certain artistic and literary values. [23] Such methods made it possible to coherently analyze printed verses of oral poetry by existing literary and metrical standards. At the same time, it meant that these texts were not necessarily representative of real performance practice, especially with regards to rhythm and meter. This gap makes recordings like Ugljanin’s Mark e Musa even more precious, since we find in it a real performance of a singer—from a period when epic performances still took place—without such intensive mediation by scholars.

As so often happens with fieldwork, surprises also lead to new research questions beyond those being addressed. Such was the case with our research with Isa Elezi, which went beyond simply verifying the existence of different ways to perform a text with regard to rhythm. During a recording session in August 2008, Isa Elezi seemed frustrated by our requests for long sessions of recitations without singing or playing the lahuta for accompaniment as he sat in front of listeners armed with a video camera and an audio recorder. Suddenly he arose and started to recite while pacing around the room. He was reciting the text as he had in the second version (Version B, reciting mode), but now he was marking the two stresses of each recited line with his footsteps. He explained his rationale clearly:

As pointed out above, singers find themselves in an unusual situation when called on to dictate a text without singing or to perform without an instrument that is otherwise always used. Such discomfort has been documented with singers from many musical cultures around the world. However, the element most sorely missed by singers in the recited version would seem not to be melody per se, but rather body movement, which helps the singer maintain rhythm. For example, other scholars have noted how certain singers, when asked to perform without lahuta, use sticks of wood to imitate its bowing movement. A striking case took place during the last Gjirokastër festival in September 2009. A singer was performing a newly composed song with his grandchildren: while the grandfather played the lahuta, his grandson sang, moving his hands rhythmically to simulate the actions of playing, imitating the rhythmic bowing movements of his grandfather. [24] As many theories of embodied performance would suggest, corporeal movement plays a crucial role in rhythmically framing the performance and in reinforcing the technique of memorizing and composing using formulas. As one such example, a well-known passage in Eric Havelock’s Preface to Plato connects mnemonic mechanisms with bodily performance by way of rhythmic movement. Moreover, for Havelock, this aspect is critical for constructing the Greek concept of the Mousikē, a multisensory, multiaesthetic phenomenon that would later be split into different performing arts: poetry, music and dance. [25]

Demonstrating a certain continuity of performance methods and concepts across time and space, the living epic forms of the Western Balkans today show traces of the importance of body movement in epic performance. The “walking performance” of Isa Elezi is relevant, not only because it demonstrates how the body becomes involved in rhythmic movement, but also because of the singer’s conscious use of footsteps as a rhythmic marker in versification. In the case of this walking performance, the regular movement of his rhythmic footsteps establishes a rhythmic frame that “pulls” the words of each line or formula. This approach suggests that we might think of the footstep—a recurring trope in poetry since ancient time—not only in a theoretical and metaphorical way, but perhaps also more concretely in connection with actual physical performance. More broadly, it reminds us once again of the salient connection between poetry, song, and the rhythmic movement of the body, present since the performing arts of antiquity.