Translating Kareva

by Miriam McIlfatrick-Ksenofontov

Translating Kareva

Miriam McIlfatrick-Ksenofontov

I first encountered Doris Kareva's poetry as a student of the Estonian language when I was given a copy of Armuaeg (Time of Grace) for pronunciation practice. I clearly remember the pleasure I felt in reading her poems aloud. I understood little more than isolated words and phrases, fragments of meaning, but the sounds and the rhythms had already captivated me. Such is the power of poetry. Years later, Doris asked me to edit some translations for a bilingual anthology of modern Estonian poetry; it included a number of her own poems, and what struck me immediately was the almost total absence of the rhythm and rhyme that I had been so acutely aware of in my early readings of her poetry. Doris Kareva's voice, as I had heard it, had disappeared. Then, Doris herself made two memorable remarks about her own poems. The first was after the premiere of the stage adaptation of her Mandragora collection in 2003, when, in considerable consternation, she said that hearing her poems in a male voice ‘changed everything’. The second remark came when she was sending some poems to a UK publisher; I suggested including a cover letter, but Doris declined saying that the poems should speak for themselves. Viewed together, these incidents indicate the significance of the voice of the poem for the poet, for the reader/listener and for the translator.

This was not something new to me – my experience of poetry began in my childhood in Ireland, when it was still commonplace in our home for friends or relatives to turn up of an evening, uninvited, but not unexpected. Ceilidhers my mother called them – an anglicised version of the Irish ceilidh meaning a social gathering. As the house filled and refreshments were consumed, news and views on local matters were interspersed with stories and jokes old and new, and with outbursts of piano or fiddle music, song or poetry. I think particularly of a maternal great-uncle who never came without his fiddle, and a paternal cousin who made poems about his farmland, including one about his pride and joy, an old steam engine called Wee Mary Anne. So from the outset, poetry for me had to do with voice and performance, and was part and parcel of daily life. Even poetry ‘readings’ in Ireland tended to be very vocal affairs. Our best-known poets did not start out reading their poems to silent, respectful audiences in halls; audiences could be responsive and noisy – listening attentively, engrossed in the event and chipping in with asides and jibes along the way.

My concept of poetry then is that it is, above all, voice. Even when it is consigned to paper, the desire for live expression still lies at the heart of all poetry; even when we read a poem we still ‘hear’ its cadences and feel its rhythm. While the poem may be visible as a verbal object with irreplaceable and immovable characters in print, its voice is not captured and held in them. Language remains free to speak in each reading, rereading, and to each and every reader. There is no passive transmission of a message determined by the poet and fixed in the poem to be unearthed by the reader. This is vital interaction; it is the event that makes the poem, the event is the poem.

The challenge for the translator then is to avoid seeing the poem as fixed language in a solid object and the reader as the receiver or seeker of a single invariant message. I find it helpful to view the poem not as the final text but as the last draft in the writing process. Between the poem and the translator a situation of dialogue can be established. This involves asking questions of the poem in order to find what meanings it insists on – not what we want to extract or what we think the poet might have intended. The questions need to be open and precise; for example, a question such as: "what does this word (image, rhyme, comma, etc.) do in/to the poem?" will not generate as much as: "how would the poem be different if this word (rhyme, etc.) were replaced by another or removed altogether?" and "if this word were replaced by x, what would the effect be?" The guiding principle is that any choice made by the poet, consciously or not, inevitably involved the rejection of alternatives. In this way, we can view the elements on the page not only as ‘original’ but also as what is left when others have left the scene (of the creation of the text). The translator teases the poem open and cultivates a sense of curiosity and wonder at the interplay of all its elements. This is a valuable first step towards understanding how poem and language are one and the same.

The poetry of Doris Kareva offers a considerable challenge to any approach to translation. She is a striking example of a poet in whose work sounds and meanings of words overlap and interact; indeed, at times, the whole momentum of a poem seems to depend solely on the sound and physicality of the words. The poem Concerto strumenti e voce would seem to overtly state its musical and vocal intentions, and when I first read the poem I noted both from the title and the sound the resemblance to a musical composition. The words appeared to be somewhat subservient to the musicality of the poem. In the spirit of exploration, I attempted a literal translation simply to see what it would produce. In the first sequence (I use the term sequence, rather than verse, since music is sequential) I found that only one word was exactly the same in English (mandragora); this was a satisfying discovery since it is the title and a key word in the work as a whole. Then there were a number of words which were foreign but known in English and which retained the same pronunciation (mandala, rosa mundi, gloria, grandiosa), or which retained the same pronunciation with only minor spelling changes (rhododendron, Cassiopeia, Christiania, Alexandria). And finally, words which could be translated literally and retain some of the rhythm, stress or sound value of the original (laurel, coriander, salamander), or retained the same spelling but with a shift of stress (Andromeda, mania), or changed both spelling and pronunciation (androgyne). This left only ‘undruk’ and ‘ja’. The latter, as ‘and’ fitted in with the rhythm of the line in which it appeared. ‘Undruk’ was the first point at which the literal translation procedure had to face a more complex choice. The literal translations: ‘underskirt’ and ‘petticoat’ communicated something of the original meaning and kept some of the alliteration of the line, but the rhythm of the line would have been substantially altered. If I held to my idea that the words were subservient to the musicality of the poem, I could simply have opted for any word that offered a corresponding alliterative or assonant value. I toyed with words and word parts as disparate as: ruff, ruche, dandy, dandruff, etc. but felt they did not fit – or at least not with what I had produced so far.

At this point I began another line of questioning with the poem: why should I feel that these words were inappropriate? What did ‘undruk’ do in the original that ‘dandruff’, for example, did not do in my version? I had perceived lexical ‘echoes’ or ‘strands’ that ran through the poem. Some were obvious: names from Greek mythology; musical instruments; Latin terms; plants and shrubs; words connected with various religious and mystical practices; words that were rooted in the Estonian language, culture, cultural heritage, etc.; Estonian words that were distorted into ambiguity but still carried recognisable elements and associations for an Estonian-speaker; the list could go on for as long as the mind is able to find associations, but I have listed what struck me as the primary and most immediately obvious ones. For example, I found echoes of ‘undruk’ in those words that were rooted in Estonian culture such as ‘untsantsakas’, ‘mats mis mats’, ‘kärakas’, and ‘krõks’ , and this led me to search for a word in English that would function in a similar way. I finally settled on ‘cambric’, defined in the dictionary as "a finely woven linen or cotton fabric, usually white" but I had not found it by consulting a dictionary. The main reason for my choice was that I remembered it from other texts, novels from the beginning of the 20th century in which cambric was a commonly used fabric in women's undergarments, and also from the Simon and Garfunkel song "Are you going to Scarborough fair" – the line "Tell her to make me a cambric shirt". My associations were with old-fashioned undergarments and song. Thus, I had found an answer to my questions; I had created connections. I found no echo for ‘dandruff’ in the poem at that time (although who knows if on another reading or dialogue with the poem I might not), whereas ‘cambric’ had a resonance for me in the first sequence, as described above, that would reappear later the poem, in relation to ‘striptiis?’ in the third sequence.

It was when I started to work closely with the second sequence that I realised that my purely literal approach would only take me so far. It was clear that by keeping to it, I would lose much of the sound effect of the poem, e.g. ‘Nile’ for ‘Niilus’ in line 2 would lose in assonant value, whereas ‘Indus’ offers both assonant and alliterative links and is also a river. Likewise, ‘kriit’ and ‘igrek’ in line 5 would literally disappear if translated as ‘chalk’ and ‘y’. Skimming down the poem again, I saw that the cumulative effect of literal translation would simply silence the poem. Therefore I moved on still listening for echoes but now looking more actively for other links as well.

As I explored further, I felt that words, sounds and ideas reverberated throughout the whole space of the poem, I discovered that many words reverberated in more than one place. Consequently, I aimed to emulate this potential for movement in my choice of words in the translated text. The knock-on effect of this was a constant to-ing and fro-ing through the original for associations and then the same process in the English. I found myself in what Lawrence Venuti refers to as "a site of uncontrollable polysemy" . Whatever word I chose for the translation in some sense reduced the meaning of the original word, but it also released new potentialities, and these in turn echoed back strangely to the original. The tranquillity of the original was disturbed, but inspired by Derrida's suggestion that "a good translation should always abuse", I continued in the same vein. And certainly some of my choices may constitute an ‘abuse’, a ‘going beyond’ what Kareva might have had in mind, and then again some of them will undoubtedly exceed or even conflict with what I myself intended when they are read by another reader. These were thoughts that crossed my mind, but not concerns. For example, in the third sequence, ‘metseen – mis seen?’  becoming ‘maitresse – Mae West’, was inspired by that mistress of the double-entendre herself. And, in the fourth sequence, ‘Udmurdi-Burundi / urjukk, burundukk’  becoming ‘tandoori patchouli / Udmurt-Buriat’ had its roots in a play of associations for the English speaker: between what is familiar and known yet still foreign in source, name and essence; and also between places and peoples (I derived some satisfaction in remaining faithful to a Finno-Ugric connection by retaining the Udmurt, but they had moved and found new neighbours!); and there was also a hint of the exotic (‘urjukk, burundukk’ becoming ‘tandoori, patchouli’). Admittedly, by this stage I had been driven to dictionaries and other reference sources in order to check literal meanings and the root and/or full extent of many words. I drew inspiration from George Steiner's words that "a true reader is a dictionary addict (...) grasping the life of words" . It was in this way that I rediscovered another poem I had long forgotten: ‘Ulalume’ written in 1847 by Edgar Allan Poe, originally as an elocution piece and known for its focus on sound. I do not think that it is by chance that it makes its appearance in Kareva's poem.

It is impossible to remember or accurately recount the entire process of reading and creating involved in the making of this new poem. It is more useful to sum up what I was aware of doing and the effect it was having on the poem. Throughout the whole process, I was intensely aware of the difference that my translation was making. I realised that I was entering that movement of difference which is a fundamental property of languages. I felt that my mind had become the site of passage of the poem. Consciously or not, I was identifying those features of the original text which represent something that is essential to the poem. Where they did not lend themselves to literal translation, I attempted to be ‘faithful’ to them (and thus to the original poem) by questioning what it is that they do to and in the poem and matching it in the new poem, while still allowing my own language to shape the new contours of the new poem. In the end, I see that there was a great deal of Derrida's notion of a certain economy of translation in that I was attempting to relate same to same and other to other. The poem initially seemed to lend itself to insightful exploration through literal translation due to the general looseness of syntactic features. However, once it became clear that this alone would not account for the coherence of the poem, I began to incorporate other strategies at the same time. Viewed in isolation, the words often appeared to be semantically disparate, essentially different, but there were a number of identifiable ways in which they were linking and giving the poem its coherence. There is a musicality and a mantra-like intensity about the whole, largely due to the sound (assonance, alliteration, end-rhyme) and the rhythm. These along with the lexical threads that weave through it are what give the poem its voice and they are what I found in my process of dismantling the poem into draft texts.

On a final note, I would add that I did have another version of the last two lines of the final sequence of the poem, the place where the voice almost disappears and it subsides into pure rhythmic sound. Having attempted to emulate the riddling juxtapositions of the third line of this sequence, where recognisable words or word parts are repositioned in combinations whose sole reason of existence seems to be their sound quality and effect, mixed with their semi-articulate magical, mystical and religious tones (‘unelumm, lumeruum, jumalumm’ – ‘runic loom, lunar swoon, jubilum’). I tried to continue this sound effect in the ending, hence ‘tumm, tumm, tumm / trumm’ became ‘dumb, dumb, dumb / drum’. In terms of an association with sound and music this was promising, but a primary connotation of ‘dumb’ – in the sense of ‘stupid’ – repeated three times began to have a jarring effect that distracted and detracted from the pure sound quality that was demanded. Instead, I opted for the decidedly more sombre ‘doom’ and ‘tomb’ version for its sound quality and its finality at the end of the poem. Whether or not it is a gloomy ending to the poem is for the new reader to feel.

Miriam M-K was born and grew up in Northern Ireland and currently lives in Tallinn where she teaches English language and Irish literature at Tallinn University. She studied Romance languages at Nottingham University and has worked as a teacher and teacher trainer in Spain and Portugal. She is a freelance translator of poetry, and the creative process in translating poetry is the subject of her ongoing academic research; her MA thesis was a case study on the English translations of Doris Kareva's Mandragora poems.

‘Concerto Strumenti e Voce’ (Mandragora, pp.48-49)

Andromeda, mandragora,
rododendron, mandala,
undruk, undruk, Kassiopeia,
mania grandiosa, rosa mundi,
loorber, loorber, koriander,
androgüün ja salamander –
Sirius, viiul, vioola, tequila,
Aiolos, Niilus, iood, iood,
Oibibio, Iphigeneia,
Kriit, igrek, kriteerium, kärakas, krõks,
katekismus katki, karkääks!
Karakul, curriculum vitae,
kuradi kurat –
Klara kinkis Karlile klarneti, Karl!
Karl varastas Klara korallid,
küll küll kiriküüt!

Untsantsakas, netsuke, mats mis mats, müts,
metseen – mis seen? Miss Universum –
mis siis? Striptiis? Repriis?
Tantra, mantra, yantra.
Tantra, mantra, yantra.
Mater, mater dolorosa,
hosianna, roosamanna,
krutsifix, Asterix,
aksakall, parallel-Ariel,

Vintväät, Viruvärav, efendi,
urjukk, burundukk,
purpur, jaspis ja pits.
Andalusia lits.

Adenoid, asteroid, oikumeeni hüpnoid,
Dalai-Lloyd, Ululuum,
unelumm, lumeruum, jumalumm,
tumm, tumm, tumm

Translation of ‘Concerto Strumenti e Voce’ (by the author of the thesis)

Andromeda, mandragora,
rhododendron, mandala
cambric, cambric, Cassiopeia,
mania grandiosa, rosa mundi,
laurel, laurel, coriander
androgyne and salamander –

Sirius, zither, viola, tequila,
Aeolus, Indus, iota, iota,
Oibibio, Iphigeneia,
Crete, éclair, criterion, cardinal, crook,
Catechism cut in two, cuckoo,
Karakul, curriculum vitae,
Christ caramba –
Clara criticised Karl’s clarinet, Karl!
Karl snatched Clara's corals,
Cucumber cool!

Know-it-all, netsuke, Midsummer mad,
maitresse – Mae West, Miss Universe,
misdeed? Striptease? Reprise?
Tantra, mantra, yantra.
Tantra, mantra, yantra.
Mater, mater dolorosa,
hosannah, sweet manna,
Crucifix, Asterix,
aquarelle, parallel-Ariel,

Vice versa, Vichy water, efendi,
tandoori, patchouli,
Udmurt, Buriat,
purple, jasper and pitch.
Andalusian bitch.

Adenoid, asteroid, ecumenic hypnoid.
Dalai-Lloyd, Ulalume,
runic loom, lunar swoon, jubilum,
doom, doom, doom,