Bryn Mawr Classical Review

Bryn Mawr Classical Review 2002.02.12

Lauri Honko, Textualising the Siri Epic (Folklore Fellows' Communications 264).   Helsinki:  Suomalainen Tiedeakatemia, 1998.  Pp. 695.  ISBN 951-41-0812-4.  

Lauri Honko, , Chinnapa Gowda, Viveka Rai , Anneli Honko (ed.), The Siri Epic as performed by Gopala Naika I-II (Folklore Fellows' Communications 265-6).   Helsinki:  Suomalainen Tiedeakatemia, 1998.  Pp. lxix + 893.  ISBN 951-41-0814-0 + 951-41-0816-7.  

Reviewed by Minna Skafte Jensen, University of Southern Denmark (
Word count: 4922 words

During a period of seven days in December 1990, a Finnish-Indian scholarly team recorded a great oral epic in Karnataka, India. The poem consists of invocations (563 verses) and the epic proper (15,683 verses). In the volumes here reviewed the text is published with introduction and English translation and the recording process described and discussed.

Let it be stated right from the beginning: this is pioneer work of the greatest importance for the study of Indian literature and folklore, as well as of comparative epic in general. As such it is highly relevant to readers of archaic Greek poetry, and Lauri Honko is well aware of this fact: Homer is mentioned in the first sentence of the monograph (p. 7) and on the first page of the edition (the Siri epic is "only five lines shorter than the Iliad", p. ix).

The poem tells the story of the heroine Siri and her family. She is born miraculously from the bud of an areca-flower, grows up, is betrothed to a prince from a neighbouring area, marries him and becomes pregnant. Her husband, however, visits a harlot and Siri is offended, returns to her father's house and gives birth there to a son, Kumara. Her father dies, and the village men decide to take over her belongings against her protests. Unable to defend herself, she sets her inherited home on fire and departs together with her baby and her trusty maid, Daaru. They have to cross a river and can find no boat, but the river halts its waters so that they can pass on foot. The young Kumara foresees that his mother will remarry and asks her to send him into "maaya", the divine sphere, and she accepts, giving him the maid as company. Alone in a foreign land, she is now courted by two kings, but recognizes in them her two brothers, also born from flowers. They take her into their care and marry her off, with her own consent, to a new husband who already has a wife, Saamu. In this way a new custom was introduced, it says, that had not been known before, that of a second marriage. The two wives, Siri and Saamu, have problems accepting each other but are finally reconciled. Siri has a new child, a girl called Sonne; like her mother, she is born from an areca-flower. Sonne is brought up by an old man, Ajjeru, together with another small girl, Gindye, who like Sonne has had mysterious experiences in her earliest childhood. As they grow older, they both marry, but problems of jealousy come up between them. Sonne curses Gindye, who goes to maaya, where she is united with Kumara. Sonne is punished but afterwards reconciled with her husband. But the couple has children only after having made a vow to Shiva that they will present the god with a golden child in a silver cradle. Sonne then bears girl twins, Abbaya and Daaraya, and they all live happily together, but the parents do not fulfill their promise. When the small girls grow up, they become expert players of a game called "cenne". In due course, they find wooers, but at a cenne game the sisters get angry, and one kills the other, hitting her in the head with the game board. A holy Brahmin, who has been present all along and actually provoked the sisters to fight, now explains that both twins must go to maaya. The surviving sister having thrown the corpse into a well afterwards drowns herself, and the two corpses are floating on the water like areca flower-buds. Their parents lament, and the girls are allowed to return to the human world for three days. They are on the point of being married, but at the last moment find two mortal girls to take over their place as brides and return to maaya to their uncle Kumara. The poem ends with descriptions of Kumara's rituals and his power to protect his worshippers.

Textualising the Siri Epic has four parts: A: The Enigma of Long Epic. B: Textualisation of Oral Epics: Antecedents. C: Textualisation of Oral Epics: The Present Case. D: The Siri Epic: A Synopsis. Illustrations, a full bibliography and an index are added.

Part A consists mainly of a presentation of the material and discussions of methodology. The singer Gopala Naika lives in the village of Machar in South Kanara, a district of Karnataka in Southwestern India. His language is Tulu, which is spoken by c. 2 million persons, mainly Hindus. It is known from a few inscriptions and other early texts but was given a systematic written form only towards the end of the 19th century, when Protestant missionaries began study of its grammar. So far it has a limited written literature. In schools children are taught in Kannada, the official language of Karnataka. Naika is a farmer by profession, but he also has an important religious function as the leader of a cult group of c. 80 women worshipping the heroine Siri. During rituals the devotees are possessed by divine spirits, the priest by Siri's son Kumara and the women by Siri herself or one of the six other female protagonists. Naika's singing activity is closely related to the cult since the epic tells what happened to these divinities while they were still human.

Accordingly, epic singing has an important place in ritual performance. But it is also in everyday use, since it is performed as a work song in the paddy-field. While groups of women collect the small rice plants and transfer them into new fields, their work is accompanied by a singer, often so that the women sing refrains or repeat the lead singer's line in chorus. During the singing the women's hands are rapidly collecting or planting. The humble context of the paddy-field is perhaps a more important factor in developing the epic than the religious festivals, since the paddy-field offers ample time for undisturbed singing, whereas on cult occasions the epic competes with other events. As a young man, Gopala Naika had learned his first songs in the paddy-field, and Honko even speaks of it as a "high school" of epics (pp. 548, 563).

In 1985-6 Naika dictated the Siri epic to Chinnapa Gowda, a version that ran into 8,538 verses. It was his own idea to have the epic written, because he wanted it to be read in schools (p. 13-14).

Methodologically, Honko starts out with problems of definition. It is not easy to define the epic genre in such a way that the definition is valid everywhere. He discusses the proposals made by C.M. Bowra, Ruth Finnegan, Paul Zumthor, John William Johnson and Brenda Beck, and offers as his own solution: "Epics are great narratives about exemplars, originally performed by specialised singers as superstories which excel in length, power of expression and significance of content over other narratives and function as a source of identity representations in the traditional community or group receiving the epic" (p. 28). He divides the world's epics into three main categories: literary, tradition-oriented or "semiliterary", and oral (p. 37), and refers to Virgil's Aeneid, Lönnrot's Kalevala and Gopala Naika's Siri epic respectively as examples.

In his approach, Honko is inspired by Parry and Lord and the widespread further discussions and adaptations of their theory. He is critical towards their idea of the formula, but adds a whole series of new terms to the analytic apparatus. Of central importance is his idea of "multiforms", defined as "repeatable and artistic expressions of variable length which are constitutive for narration and function as generic markers" (100). He discusses this term in relation to Parry and Lord's "themes", and he clarifies his idea by an example: in the recorded epic there are three passages in which a baby is put into a cradle spread with silk, cared for and given a name; this happens to the small Siri, Sonne and Gindye, and takes place in different wording, but following the same mental text (pp. 106-12).

"Mental text" is another new term. Different performances of the "same" passage in Gopala Naika's repertoire varied considerably, even between performances separated by only a few days; on the other hand, the basic story line changed little over the four years during which he was intermittently recorded. Honko's hypothesis is that songs exist in the singer's mind as individual entities which he "edits" before and during performances; "empirical evidence shows that singers take their time in preparing their performances and that they also outline conscious changes in storylines and episodic patterns" (p. 94). Therefore, to Honko "textualisation" does not necessarily mean recording in writing, but may describe a singer's activity when he composes a long, coherent narrative, such as Naika did on the demand of the scholars. The process is surveyed in schematic form on p. 167.

Honko also speaks of intertexts, pools of tradition, performance modes, performance styles, sub-epics, descriptions, episodes, steps, segmentation, emplotment, etic and emic units, to mention only the most frequently recurring terms.

Part B surveys 11 select cases of recorded oral epics: Elias Lönnrot and the Kalevala, Wilhelm Radloff and the Manas epic, Alfred Kroeber and the Mohave epic, Milman Parry, Albert Lord and Serbo-Croatian epics, Gordon Innes and the Sunjata epic, Nigel Phillips and the Anggun Nan Tungga epic, Brenda Beck and the Annanmaar epic, Gene Roghair and the Palnaadu epic, Charles Bird, John Johnson and the Son-Jara epic, Susan Slyomovics and the Siirat Banii Hilaal epic, and finally John Smith and the Paabuujii epic.

In part C the fieldwork that led to the publication of the Siri epic is described and discussed. Even though it takes up almost 400 closely printed pages, Honko refers to it as an introduction (p. 593), which it is both in the sense that its main purpose is to offer the kind of information the reader may need in reading the epic, and that what it describes is the beginning of research into this epic tradition rather than the final outcome of such work.

Honko's team consisted mainly of 5 persons: the singer, two "outsiders" and two "aliens" (p. 502): Gopala Naika, the Indian scholars Chinnapa Gowda and Viveka Rai, as well as Finnish Lauri and Anneli Honko. Among them, they were able to cover most aspects of the research. The overall plan for the project, the decisions about what to collect and how to treat it, were in the hands of Lauri Honko. Gopala Naika did the singing, but he was also important in establishing relations with local people such that they accepted having foreign researchers with video cameras ond other technical devices attend private and public cult events. He was also ready to discuss with the scholars all kinds of matters related to Tulu epics. Both Indian scholars were native speakers of Tulu and experienced researchers in the field; they worked as interpreters in the widest sense of the word, necessarily first and foremost as transcribers and translators of the text, but also as cultural interpreters and links to Indian scholarship. The folklorist Lauri Honko describes his own scholarly background for the project as being mainly experience in the fields of comparative religion and Finnish-Baltic traditions of epic and laments; both he and his wife started out as philologists. During the recording they were handling the machinery, and in the subsequent processing Anneli Honko analysed and close-read the text.

The ambition was to make a documentation which was good, full and thick: 'Good' referred to the technical quality of the picture and sound, 'full' to the coverage of the entire ritual process from its preparations to the departure of the last participants, and 'thick' to the multiple documentation ... (as well as) the collection of data about performers, context and collateral activities" (pp. 225-6).

Honko first attended a performance by Gopala Naika at a "house daliya" in February 1989 and had a long interview with him some days afterwards. This was the basis for asking him to perform a full version for the scholars in 1990. The recording took place in the garden of a private house, furnished with two tape recorders and a video camera. Naika performs without an instrument, so he was simply sitting on a bench, singing and singing. He performed the invocations on December 20th, and the narrative from the 21st to the 28th, with a two days' pause in the middle. The process is described in chapters 8-9 in wonderful and precise detail. We are told what the scholars asked for from the singer, who was present at the different performances, where the singer made his breaks, what kinds of errors he corrected, how long each segment was, and how much time the singer spent on each segment. Net singing time for the whole epic was 26 hours. Except for the dictated version, Naika had not performed his epic as a coherent narrative before.

Chapters 12-13 describe the annual Siri festival in Nidgal as it was celebrated on 28-30 of March, 1991. The scholars had been allowed to record the whole ritual as it took place on the third day of the festival, focusing on Gopala Naika and his group of Siri women. They began in Gopala Naika's house with 14-15 Siris present and Naika as the priest, and the singing consisted mostly in invocations and prayers. From there they went in procession to the temple area in Nidgal; along the route more Siris and their relatives joined the group. In Nidgal there were two to three thousand devotees present, among them c. 10 other groups of Siris and Kumaras, and in contrast to the ritual in Naika's home this part of the festival was noisy and somewhat chaotic; it included dramatic performance such as representation of the tragic event of the two sisters playing cenne. When the time came for epic singing, Naika began with invocations and the first part of the narrative, up to the point when Kumara left for maaya, just before Siri's second marriage. The performance then changed into dialogue, Naika and his two assistants being possessed by Kumara and the women by Siri or one of the six other female protagonists of the epic. The priests asked every Siri individually to tell her "birth story", and the women answered in the voice of the divine spirit by whom they were possessed. Some of the answers developed into considerable epic narratives. Areca flowers were being carried by those possessed and handled in various significant ways during the ritual. This part lasted most of the night, from c. 10.15 p.m. to 4.50 a.m., but did not involve anything like a full, continuous narrative of the epic as it had been sung to the scholars some months before. Oddly enough, the most coherent epic singing at the festival was not Gopala Naika's of the Siri epic but took place as an introduction to pantomime performance in which a low caste impersonator acted the stories of various heroes. While he was being masked and dressed up, his wife and daughter sang to him the narratives that he was going to perform, as if to rehearse with him the roles he was going to play. As for the Siri epic, Honko sums up its relations to the festival as follows: it "has the function of a mythical charter, a kind of 'holy text' in an oral setting" (p. 447, cf. 463).

Besides the daliya witnessed in February 1989, the special performance of the full version in December 1990, and the public Siri festival recorded in March 1991, the Finnish-Indian team undertook important supplementary research. They travelled to various of the localities in which the Siri epic unfolds in search of "epic archaeology", and even though some of these places were not far from Gopala Naika's home, he had never visited them before. But of special importance is that they made lots of further recordings. They interviewed some of the women of his Siri group and attended their singing; they also had Gopala Naika perform in the paddy-field. They recorded another version of the introductory invocations, still in December 1990; four other epics in Gopala Naika's repertoire in January 1993; another performance of the Siri epic at a house daliya in December 1993; and many sessions of interviews with him. In this way, the collected material contains ample evidence for comparing different epics sung by the same singer and different versions of the 'same' story performed by him in different contexts and for studying the opinions he has of his own tradition. This last element is clearly of special importance to Honko; he quotes a critical remark by Dennis Tedlock, "In the classic ethnography, other anthropologists may be quoted at length, but no native ever utters a complete sentence, whether in text or in translation" (p. 513), and explicitly has as one of his main ambitions to let the singer himself come through, not only as an informant of an ancient tradition, but as an individual artist in his own right.

Part D offers a prose summary of the sung version of the epic recorded December 1990. Among other things this means that the monograph makes sense even unaccompanied by the edition.

The Siri Epic as performed by Gopala Naika. I-II. The volumes consist of preface, introduction, text, illustrations, and finally a glossary of names, untranslated words and rare English words in the translation. The introduction is a concise form of the monograph, necessary in order to make the two text volumes readable as an independent publication.

The poem is written in its original Tulu language on the left pages, accompanied by an English translation on the right, more or less like a volume of the Loeb collection. The editors have chosen to transcribe what can be heard in their tapes as precisely as possible without editorial corrections other than those the singer himself asked for. The text is divided into verses according to the singer's breath breaks but not otherwise analysed; it is written without upper cases and punctuation. The translation, however, is printed according to normal European standards of writing. First and foremost it follows the original closely; the scholars' ambition has been that it should be "as verbatim as possible and as readable as necessary" (Textualising the Siri Epic p. 586, p. lxi).

Honko is careful to underline how closely the four scholars cooperated also in the laborious phase of transforming audio- and videotaped song into a written edition. He estimates that this part of the project consumed 20-25 times as much work as the recording. Nevertheless, there was also a degree of specialisation: Chinnapa Gowda was mainly in charge of transcribing and translating interviews, Viveka Rai of establishing the Tulu transcription, Rai and Anneli Honko of making the basic translation, Anneli Honko of analysing the content of the poem and dividing it into "cantos", and finally Lauri Honko gave the translation its definitive form.

The result of their efforts is a marvellous publication in which the patient reader may come close to understanding and enjoying an otherwise inaccessible poem. Guided by the translation it is possible to find not only answers to scholarly questions, but also aesthetic pleasure. The narrative is straightforward and quiet. Even without knowing the sound of the verse you feel that you get a sense of its soft rhythm, interspersed with recurring, refrain-like invocations of the god. Repeated verses and half-verses are further characteristics, and often repetitions are in themselves doubled or even tripled with tiny variations, as for instance "on the threshold she put her foot / on the lintel she put her hand" (782-3) or "with a thousand decorations, with weapons /with a drum tied to a coconut tree / with a drum tied to a mango tree" (1560-62). The adding style that Milman Parry underlined in Homeric poetry is obvious here too: each verse tends to be a meaningful whole. Also more comprehensive units are repeated, although not necessarily verbatim, the "multiforms" of Lauri Honko that resemble Lord's themes. They may consist of shorter passages such as for example vv. 13.741-86 as compared to 13.827-62, but also longer stretches of narrative follow the same patterns; thus Siri and Sonne begin their lives very much in the same way. Now and then narrative inconsistencies seem to result from a confusion of multiforms in a way parallel to what Lord described as typical of composition by theme. The poem is harmoniously rounded with a ring composition, the areca-flower of Siri's birth recurring as a simile at the twin sisters' death.

The editors state that the best commentary is to be found by careful reading of the poem. Even though this principle is both ancient and attractive, I must admit that I would have found a few explanatory notes helpful. The summary at the end of the monograph (Textualising, part D) solves some of the problems, but many remain. All too often the reader feels uncertain about the meaning of what is being said. For instance I found the description of how Siri sends her son and maid to maaya scary: she presses them down into the earth with her foot, one after another. After all going to maaya means leaving this world, however much the son asked for it. Are we to understand that our heroine is killing her own son and his nurse before marrying her new husband? A similar instance of a narrative difficult for the foreigner to understand is the role the Brahmin has in the events of the disastrous cenne game.

In general, however, the poem gets through to its reader, and it is striking how definitely female-centered it is, almost feminist. Except for the important role of Kumara with his cult-initiating functions, the epic is dominated by women engaged in typical women's affairs such as pregnancy and childbirth. The minor male characters take a vivid interest in these matters, too, and are more impressive as fathers than as warriors. In scenes where men and women are in conflict, the narrator most often takes the women's side; a striking example of this is the scene in which Siri joins a male assembly convened to decide upon the question of her inheritance. The narrator describes the "men of justice" through their speeches and the fact that they accept money from Siri's adversaries in such a way that the reader feels in full sympathy with the heroine when "a fury of seven fourteen worlds filled her" (v. 4468). It is easy to link this aspect of the poem with the fact that it is meant for a female audience. However, this seems to be a question that did not interest Honko. The three volumes as a whole constitute an almost overwhelmingly rich work full of experience and thoughts, and to Homeric scholars hoping to learn from living oral traditions about the functions and context of Greek epic it is a treasure of information.

Of special importance is the precise, careful publication of the text. It has been stated often enough that oral performance does not in itself exclude poetry of similar length to the Iliad and the Odyssey; C.M. Bowra said this fifty years ago in his famous survey of the world's epic poetry.1 But documentation is rare, and the present work is way beyond other editions in its care for making the original text accessible to non-native speakers. This quality of the edition adds significantly to the usefulness of the monograph, since it enables readers to control Honko's more personal statements and interpretations.

I am not very happy with Honko's definition of the epic genre cited above. It is cumbersome in its detail and would hardly include the Aeneid as an epic, although this poem is classified as such in his book. Perhaps it is meant as a definition only of oral epic; if so, I miss the element that this genre belongs among those considered to convey true information. As a matter of fact, Honko often underlines that Gopala Naika considers it his job to perform true stories, and that this is important for him in many questions of detail.

Part B of the monograph, the survey of 11 epic traditions and the history of their recording, is of special interest for Homeric studies. Few classicists know how dramatically research in oral epic has expanded during the last decades. The immense growth of fieldwork has made it possible to evaluate the conclusions that Lord made in The Singer of Tales (1960) on the basis of a large spectrum of field experience, and Honko's selection offers an easy introduction to this complicated area.

His own study is full of criticism of the founding fathers of the oral-formulaic theory, Milman Parry and A.B. Lord. But in his overall approach, the questions he puts, and the way he structures his fieldwork, he is manifestly influenced by their theory. The interest in variation in oral traditions and accordingly the practice of recording and publishing a single, specific performance rather than an ideal text, entered fieldwork as a result of Lord's Singer of Tales, and this is, of course, what Honko does; the interviews with the singer also have their model in Parry and Lord's work. Honko's intense interest in all aspects of methodology makes his book one long discussion of the concept of oral epic, and as such, a discussion with Parry and Lord and their adherents. He rejects Parry's definition of the formula, stating that "the life of a formula is variation, not fixity" (p. 113); but if Parry exaggerated the general validity of his definition, it seems to me that Honko commits the same error, generalising from the epic tradition he has under scrutiny to oral poetry as such.

A main point in the oral-formulaic theory is the flexibility of tradition, and this is abundantly exemplified and discussed in Honko's work. His discussions of this question are detailed and interesting. For instance, both of the two "full" versions of the story follow the same storyline, and although one is double the size of the other both have passages the other one does not have. The idea of the mental text that the singer is all the time "editing", appears convincing and is confirmed by Lord's experiences in Montenegro when in some cases singers performed songs in very much the same form as they had done 17 years before.2

"The enigma of long epic" is emphasised as the title of part A and is repeatedly treated. Honko makes no secret of the fact that Gopala Naika's normal performances were different from what he did in the "induced" contexts, answering scholars' requests. The suspicion that long epic is the ambition of collectors rather than of singers is touched upon, and Honko is well aware of the influence scholars' expectations have on the recordings made. After all, the idea that he and his colleagues have of the genre is basically formed not only by Homer but also by Kalevala in the case of the Finns and by Mahabharata and Ramayana for the Indians. His discussions lead to no clear conclusions on this question, which is satisfactory since the existence of very long performance sessions in natural contexts is after all well attested by other fieldworkers.

The question of the epic's function as a charter for ritual is fascinating. The connection between poem and cult is evident and close. Nevertheless it is not quite satisfactory to see the poem as a kind of textbook for cult practice, among other things because things take place in ritual without having parallels in the epic. To classicists it is thought-provoking to compare with ancient Greek religion and poetry, and here the parallel would perhaps be the Homeric hymns rather than the Iliad and the Odyssey. But in both cases I should prefer to see the complex of myths as a collective mental text of which both epics and other kinds of oral literature as well as cult events were performances. Or in another terminology, as the langue of which specific poems or practices were the paroles.

Chapter C9 with the almost pedantically detailed report of what took place during the recording is of special interest to me, and Honko does me the honour of stating that it was provoked by a paper I read at a conference in Turku in 1996. Since I have for decades been speculating on the question of how precisely the Iliad and the Odyssey achieved the written form in which we know them, I am interested in all kinds of information about recording processes. Our knowledge of the context of the Homeric epics is certainly of the kind that Honko would classify as "thin", but in my efforts to learn from "thick" documentation I am regularly frustrated even by modern editions. They say far too little about what exactly took place when a given epic was being dictated or mechanically recorded by a scholar and also about the process taking place between the recording of the text and the publication of the book. Here for the first time in history we are offered solid fact. In relation to the division into 24 books of each poem, it is fascinating to compare Anneli Honko's 56 "cantos" with the 36 segments of varying length established by the pauses Gopala Naika made in his performance. It is my hope that Honko's book will inspire other fieldworkers to include such information too.

For the comparative approach to Homer this extraordinary work is perhaps the most important study since The Singer of Tales.


1.   C.M. Bowra: Heroic Poetry. London: Macmillan 1952, pp. 330-67.
2.   Milman Parry and Albert B. Lord (edd.): Serbocroatian Heroic Songs I. Cambridge, Mass.: Harvard University Press, 1954, pp. 409-16.

Read Latest
Index for 2002
Change Greek Display
Books Available for Review

HTML generated at 13:27:44, Friday, 03 April 2009