TRANSLATING TURKIC ORAL EPICS INTO ENGLISH AND GERMAN: PROBLEMS AND INSIGHTS

It is a well-known fact that the Turkic-speaking peoples of Eurasia and Siberia possess a rich heritage of oral epic poetry. Much has been written down, and in some areas the oral tradition of epic poetry is still flourishing today. While a few of the early texts, written down in the 19th century, are available in older German translations (A. Schiefner, W. Radloff), the majority of these epics can only be accessed either in their original language or (in some cases) in Russian translation. Translations of Turkic oral epics into European languages such as English, German or French are urgently needed in order to familiarize epic scholars outside the Turkic-or Russian-speaking world with this important corpus. Translating Turkic epics into European languages poses, however, a number of problems. In the following some of these problems are identified and discussed with reference to my translations of the Uzbek oral epics Ravshan and Alpamysh into German, the Karakalpak epic Edige and the Kirghiz epic Manas both into English. ´The latter translation is still ongoing; two volumes have so far been completed. The translation problems concern basic methodological questions such as the choice between a literary and a literal translation and the strategies available to overcome differences in linguistic structure between the source languages and the target languages. An important element in translation is not only the linguistic, but also the stylistic level of the text. In addition, a translation has to pay attention to paralinguistic aspects (e.g., music and performance modes) and to the cultural world of the original. The translator is not only a mediator between languages, but also between cultures.


Introduction
In literary criticism an epic is generally defined as a long narrative in verse, which is centred on heroic deeds and war-like exploits and often retells the early history of a tribe, ethnic group or people.This understanding is based on the Homeric epics, the Iliad and the Odyssey.These epics have been continually read and translated from Antiquity to the present day and occupy hence an important place in world literature.While the original language of these epics, Ancient Greek, is known today almost only by scholars in the classics, there are innumerable translations into all the major languages on our globe, which guarantee a universal reception.The Homeric epics demonstrate the importance of translation for world literature and culture.
There is a rich oral epic tradition among the Turkic peoples of Central Asia and Siberia, which can rival any other epic tradition in world literature.Despite its richness, however, these epics are little known to readers who cannot read them in their original languages.It was in the nineteenth century that the Turkic oral traditions were first brought to the attention of the scholarly community in both Europe and the Russian Empire.The Finnish philologist and ethnologist M. Alexander Castrén, whose interest was focused on the Uralic languages, also studied Turkic languages during his travels in Siberia in the 1840s and collected a number of oral epics from the Koibal (who speak a dialect of Khakas) and Karagas (whose language is today better known as Tofalar) at the foot of the Sayan mountain range in Central Asia.After Castrén's death Anton Schiefner published both his travel books Вестник Северо-Восточного федерального университета имени М. К. Аммосова: Серия Эпосоведение, № 1 (01) 2016 and his "Essay of Koibal and Karagas Grammar".In this grammar, some epic songs together with their German translation are found in the appendix [1].Fifteen years earlier, the Polish poet, scholar and diplomat Alexander Chodzko had published an English translation of the Azerbaijanian version of the Köroghlu cycle [2].In 1866 the first volume of Wilhelm Radloff's monumental collection of folklore texts of the Turkic peoples appeared.For the first six volumes Radloff also provided a German translation; of special importance for epic studies is the fifth volume, devoted to the oral epics of the Kirghiz, which also contains an important theoretical preface [3].Despite these early translations, the epic poetry of Central Asia and Siberia has remained widely unknown.Schiefner's and Radloff's translations are today difficult to read and difficult to understand.Unless one takes the trouble to look at the original text, much remains unclear.Chodzko's translation still reads well, but the book is very difficult to find, even in good university libraries.Of course, there are more modern translations of Turkic epics into German, English, or French and occasionally also other languages, but the list is comparatively short.To mention just one contemporary translation: A. T. Hatto re-edited Radloff's Kirghiz volume and translated the text into English.About his translation Hatto writes: "The translations opposite the text have no pretensions either as literary creations or as 'exact' renderings of the remote Kirghiz such as would please a grammarian.They are made 1) in order to show those familiar with Kirghiz or at least a not too distant Turkic language, how the writer understands the text; 2) to serve as a first commentary on the original, to be completed by reference to the Commentary proper; 3) to enable those who have to rely on a translation, safely to assimilate the narrative content for their own scholarly purposes or pleasures" [4].
As Hatto points out, literal translation can be of different degrees: literal in the sense of a linguistic gloss or literal in the sense of adhering to the sense of the original as much as possible and yet producing an idiomatic and grammatically correct text in the translation language.Of course, the problem lies in the qualification "as much as possible".It is on this point that a voluminous scholarly literature on the principles and methodology of translation, both from a linguistic and a literary point of view, has developed over the last half century [5].
In the following I will illustrate some of the problems from my own translations into German and English.Before doing so, however, it has to be mentioned that there is an important translation literature in Russian.A great number of oral epics of the Turkic-speaking peoples are available in highly polished literary translations.Lev Pen'kovskiy (1894-1971) translated the Uzbek epic Alpamysh (1949), the poet Arseniy Tarkovskiy    (1975).Any reader of these translations will be impressed by their high poetic quality.As is to be expected, these literary translations are not always faithful to the wording of the originals.As V. M. Gatsak notes, translations of oral poetry tend to move further and further away from the original the more they focus on rendering the poem in poetic form [6]. Apart from these literary translations, there are also a great number of scholarly translations of Turkic oral epics into Russian.Two excellent series need to be mentioned here, the series "Epics of the Peoples of Eurasia" (Эпос народов Евразии), formerly called "Epics of the peoples of the USSR" (Эпос народов СССР), and the series "Monuments of the Folklore of the Peoples of Siberia and the Far East" (Памятники фольклора народов Сибири и Дальнего Востока).In the volumes of these series both the texts and their Russian translations are given; in the latter series a small record or CD generally accompanies the volumes.The translations are intended to be as accurate as possible, without however impairing their readability.Verse is translated in verse-lines, though not in rhythmic lines; in this way, the verse structure of the epics is graphically symbolized and readers are encouraged to read the texts as verse rather than as prose [7].Unfortunately, Russian is little known to scholars in Europe and North America (apart from the Slavic-speaking areas of Eastern and South-Eastern Europe), which means that an important source for the better understanding of the epics of Siberia and Central Asia remains untapped.It is therefore of special momentum that the Yakut oral epic tradition is now available in English in the magnificent translation of Oyunsky's Nyurgun Bootur [8].

Rhyme and style
Despite some very basic similarities between the various Turkic epic traditions, which in my opinion go back to a common historical and cultural background, there are also a great number of differences between, for instance, the olonkho of the Yakuts and the dastan of the Uzbeks.These differences demand different translation strategies.I have translated two Uzbek dastans into German: Alpomish and Ravshan, an epic from the Köroghlu (or Goroghli) cycle [9].These epics are in a mixture of verse and prose, and while the prose can be translated in a straightforward fashion, the verse poses special demands.The metre of the verse passages is a 7/8-syllable or an 11/12-syllable line; the verse lines are rhymed, often in regular, but often also in irregular patterns.For the reader of the translation it should be clear that verse passages are in verse, i. e. the lines have to be printed like verse lines.But what about metre?Two reasons speak against rhyme in a German or English translation.(1) Since the 19th century, rhyme is not any more the dominant form of verse; in fact a rhymed poem (unless very well composed) will often sound antiquated and amateurish.
(2) In order to find rhymes in the target language (German or English), the literal translation has to be sacrificed.The translation is then in danger of being only loosely related to the original.The remarks Arthur Waley made about the absence of rhyme in his translations from the Chinese are also true of translations from the Turkic languages: "I have not used rhyme because it is impossible to produce in English rhyme-effects at all similar to those of the original, where the same rhyme sometimes runs through a whole poem.Also, because the restrictions of rhyme necessarily injure either the vigour of one's language or the literalness of one's version.I do not, at any rate, know of any example to the contrary" [10].
Nevertheless, it seems important to me to convey the sense of poetry to the reader.This can be done in two ways.(1) Even if the lines have a differing number of syllables and no rhymes, one can try to translate rhythmically, i. e. with a regular succession of accentuated and non-accentuated syllables.This has been the principle of my Alpomish translation.(2) The Uzbek dastans have generally a flowery, ornamental style in their verse passages, influenced by Persian poetry.This is especially noticeable in Ravshan.Here the ornamental style can be imitated in translation by a choice of poetic vocabulary and the use of a non-colloquial, "refined" register.

Paralinguistic features: music and gestures
As is well known, oral epics are generally sung rather than just recited.In epics composed in a mixture of verse and prose, only the verse passages are sung, in most Turkic traditions to the accompaniment of a bowed or plucked instrument.The music is an integral part of performance and a text edition or translation without music ignores an important aspect of the oral epic.Therefore the volumes in the series "Monuments of the Folklore of the Peoples of Siberia and the Far East" contain audio samples as well as chapters on the music.In my edition and translation of the Karakalpak epic of Edige, which I recorded from the singer Jumabay Bazarov  in 1993, I included a CD with audio and video examples.As I wanted the reader to be aware that poetic passages were sung to a particular tune (to the accompaniment of the Karakalpak qobyz, a bowed lute), I have added the type of melody used for a particular passage and specified at which line the singer changed from singing into declaiming.The various types of melodies are discussed and illustrated in an extra chapter [11].
As is well-known, the oral epics or dastans of the Uzbeks and Karakalpaks (as well as those of other Turkic peoples, for instance the Uighurs and Turkmens) are as a rule composed in a mixture of verse and prose.In the prose parts, which are spoken or recited, the narrators' performance can become fairly lively and dramatic: the epic singer gesticulates, adopts different facial expressions to signify different moods, and modulates his voice both with regard to loudness and pitch.If these elements are incorporated into the original text edition, then they should also be part of the translation.In my edition and translation of Edige I have added photos illustrating the singer's gestures and have indicated in the text and translation when he interrupted his performance to sip his tea.

Lexical complexity in Manas
The greatest challenge is presented by the translation of the Kirghiz epic of Manas.One problem is the length of the epic.A translator needs time and leisure to devote him or herself to the task of translation.I am presently translating the version of Manas written down from (and partly by) the singer Jüsüp Mamay   lines; the first part, Manas proper, contains somewhat over 50,000 verse lines; it is to this first part that I will limit my translation work.So far, I have translated 20,000 lines in two volumes [12].Apart from its length, Manas bristles with linguistic and stylistic problems.An interesting discovery in the translation process was that, although Manas has formulas like other oral epics, it is on the whole surprisingly unformulaic.New words come up continuously and repetitiveness is kept to a minimum.The traditional singer commands a huge vocabulary, and not all of the vocabulary has been recorded by lexicographers.K. K. Yudakhin's Kirghiz-Russian dictionary is an excellent tool, but it does let down the translator frequently [13].Passages are often structured by what one might call "rhyme-strings" or "rhyme-chains".These are words in rhyme-position, which in the form of chains of different length occur repeatedly in the epic.One such chain or string is composed of the verbs dürküröp, kürküröp, dünggüröp, küüldöp in the following passage from Mamay's Manas: Көпчүлүктүн баарысы, 7890 Өрт алгандай дүркүрөп, Чагылган күндөй күркүрөп, Жер козголуп дүңгүрөп, Ызы-чуу түшүп күүлдөп...
All the many people gathered 7890 Shouted as if a fire had broken out, Roared as if there were thunder and lightning, Boomed and made the ground shake, Clamoured and made a lot of noise... [14] This is just one instance of this rhyme-string; in the first part of Mamay's Manas cycle, 21 instances of this string occur.In this string not only the verbs in the passage quoted above are found, but also other verbs such as küngürlöp (groaning), büjürlöp (romping) or bürküröp (splashing).When looking also at the Manas versions of Sayaqbay Qaralaev and Saghymbay Orozbaqov, we can see that this chain is composed of a total of 19 verbs, which occur in combinations of two, three, four, five, six and eight rhyme-words.
The difficulty for the translator is that the Kirghiz words are mostly onomatopoetic.Their morphological structure contains derivational syllables such as kür, gür, kir, qyr, ghyr, which characterize these words as imitative, mostly of sound, but also of movement.English has only a limited number of verbs of a comparable type; examples are to drizzle, to frizzle, to dribble or to nibble.The same is true of German.Apart from the rhyme-chain illustrated above, there are many other rhyme-strings in Kirghiz epics.Among the verbs found in these strings a large group is formed by verbs of movement.These verbs describe different types of movement, for many of which there are no correspondences in English, and the verb in question has therefore to be further qualified; but even with extra qualifications the various nuances cannot properly be captured in English; here are some examples, many of them belonging to rhyme-chains: "to move heavily": dalbangda-(далбаңда-), dardangda-(дардаңда-) "to move clumsily": barbangda- "to move in a dignified manner": baqjangda-(бакжаңда-) "to move smoothly": qalqylda-(калкылда-) "to move faster and faster": küülön-(күүлөн-) "to move like a stump (maimed limb)": moltongdo-(молтоңдо-), cholongdo-(чолоңдо-), choltongdo-(чолтоңдо-) "to move energetically": sabala-(сабала-), tüyül-(түйүл-) "to move gracefully": suyqay-(суйкай-) "to move slowly": chalqy-(чалкы-).Both the rhyme-position and the semantic content of these verbs present problems for the translator.There is no way these words can be translated as rhyme-words in English (or German), and even their semantic shades cannot always be rendered by lexemes in the target language.The translator can do no more than indicate the lexical richness of the original in his or her translation by as wide a choice of synonyms and semantically related lexemes as possible in the translation language.The special effect and ingenuity of rhyme-strings, however, will have to remain a feature of the original which the translation cannot adequately capture.Fortunately, the translation can nevertheless transmit a flavour of the original, by attempting to strike a balance between a literal translation, which pays close attention to the source language, and a literary translation, which exploits the lexical and stylistic possibilities of the target language to give adequate renderings of the original.Striking this balance is an ever new endeavour.There are no manuals for translators of Turkic oral epics and a solution must be sought for every tradition and for every text.

Problems of syntax
While the decision between a literary and a literal translation is a fundamental one every translator has to face, other translation problems are more specific and demand different solutions depending on the target language and the intended readership.One problem concerns language structure and in particular syntax.The Turkic languages abound in participial and gerundival constructions, which cannot be imitated in all languages.The Turkic languages have pre-modification with a participleconstruction positioned before the modified noun; this has to become a relative clause in English or German, positioned after the noun: Uzbek kecha kelgan kishilar is rendered as "the people who came yesterday" in English (and "the Leute, die gestern gekommen sind" in German).This gives a different weight to the nominal modification.Sometimes such participle structures can be taken over in the target language (Uzbek keladigan mihmanlar "the arriving guests"), but as a general principle pre-modification becomes post-modification in English and German.Another problem is caused by gerundival constructions, which normally have to be translated as clauses (generally adverbial clauses) and in this way made more definite in meaning than in the original.The following lines from Jüsüp Mamay's version of Manas can serve as an example: Кепти угуп күүлөнүп, Бузулуп өңү түрлөнүп, Манас турду күркүрөп, Жан алчудай сүрдөнүп.(20586-20589) Here we have six gerunds (ughup from uq-'to hear', küülönüp from küülön-here 'to become excited', buzulup from buzul-'to break apart', türlönüp from türlön-'to look awe-inspiring', kürküröp from kürkürö-'to boom', sürdönüp from sürdön-'to get confused') and one finite verb (turdu).In Kirghiz (as in other Turkic languages) the gerund in -p expresses an action previous to that of the finite verb; but it can also express a simultaneous action and have additional functions as a complement to the finite verb [15].The verb tur-in our example can be interpreted as a full verb (here: 'stood up') or as a descriptive modal verb, modifying the preceding gerund in -p, with a durative meaning [16].
In the original all these meanings are, as it were, "dormant"; in a translation into German or English (and the same is true of other Germanic languages and generally also of the Romance languages) they have to be made specific.One interpretation is to take the first gerund (ughup) as expressing a previous event ('after he had heard the speech'), while taking the others as simultaneous with the finite verb.As to turdu, it could be interpreted as modifying the following gerund, if the unusual position of the gerund is considered a poetic licence ('Manas continued to boom'), but it seems more natural in the context of this passage to translate turdu as a full verb.A literal translation, taking these specifications into account, would then read: 'after he had heard the speech, he became excited, his face darkened and took on an awe-inspiring appearance, Manas stood up and shouted with a booming voice, out of his mind (confused) as if he were to take life.'Having made these choices for a literal translation, the translator will, however, in the next stage want to vary the verb and tense sequence somewhat.Here is one suggestion for these lines: The words he had just heard fired him on: His face assumed a threatening and dark expression; Manas stood up, thundering, And in a murderous frenzy.

The translation of culture-specific words
A different problem is the translation of culture-specific words.Encyclopedic dictionaries of the English language such as the Oxford English Dictionary or Webster's New World Dictionary of the American Language list the word kumiss (also spelled koumis(s) and koumyss) and gloss the word as "mare's or camel's milk fermented (or distilled) and used as a drink by Tatar nomads of Asia" (Webster).Clearly this word, so frequent in Kirghiz and Kazakh epics, can be used in its Anglicized form in a translation with perhaps only a short annotation as to its meaning.Other culture-specific words might not have entered the English or German lexicon as loan-words and might not have exact equivalents in the target language.There are two alternatives for dealing with these words: (1) to leave them untranslated and explain their meaning in a footnote or a glossary or (2) to find a rough equivalent or use a paraphrase.Words for clothes, weapons, food, the yurt, the nomad's habitation, and its parts, manners and customs are some of the lexical areas where the retention of a word in the original language in the translation makes sense.Depending on the readership rough equivalents or paraphrases might be preferable to a profusion of words taken from the original language.'Bride-price' for kalym or 'present for bringing good news' for süyünchü are acceptable translations.Sometimes, however, rough equivalents or paraphrases might be misleading.'Cuirass' is an acceptable translation of Kirghiz charayna (чарайна).However, the translator has to be careful to avoid suggestions of a knight in armour as typical of the European Middle Ages.The armour of the Central Asian nomadic warrior is much lighter than that of the European knight, especially of the late Middle Ages [17].
A special problem concerns proper names, of people, horses and weapons.Sometimes the whole name has a meaning, as is generally the case with the names of horses (e. g.Manas's horse Aqqula, 'light-grey'); sometimes a name has a meaningful attribute: Tazbaymart, 'scabby Baymart' or Qïrghïlchal 'the old man Qïrghïl' in Manas.Arguments can be found for translating the names or for leaving them (with perhaps a note as to their meaning), and in practice both positions are found.An example of the strict adherence to the principle of translating all names is the French translation of the Secret History of the Mongols by M.-D.Even and R. Pop [18].

Conclusion
The problems posed by the translation of culture-specific words and of names underline one of the most basic goals of translation: to make the reader familiar with a poetic text in its cultural context.While there is a sizable literature on translation problems from the point of view of linguistics and stylistics, this aspect is often ignored.The translator, however, is not only a translator of words, but also a mediator between cultures.The less familiar the world of the original, the more cultural glosses will be necessary.In the end, however, the translation will have to be more than a guide to another culture.It will have to give the readers not only a taste of the original, but a reliable alternative to the original, an alternative that will become sufficiently familiar for them to consider the translation a part of their own literary universe.