In this, the second part of Dr Will Johnson’s discussion on translating Shakuntala for performance, the Cardiff professor explains how he treats Kalidasa’s famed poetry, using what he notes is reputed to be the pinnacle of Sanskritic let alone Kalidasan dramatic verse. He also explores the role of rasa in the context of this verse as well as more generally, before noting a particular peculiarity about Sanskrit as opposed to Western drama – its ‘emotional directness’:
How did I deal with the translation of the verse or poetry, the third singularity of Shakuntala?
Metre in Sanskrit verse is derived from patterns of light and heavy syllables: in English verse, metre is mostly, but not exclusively, derived from patterns of stress. Sanskrit verse is unrhymed; English verse is rhymed or not, as the form dictates and the poet chooses. Some translators, egged on by proponents of ‘otherness’, demonstrate their virtuosity by reproducing the syllabic pattern of the original verse in English. While this may sometimes (and with a dollop of luck) work for English and other languages, for English and Sanskrit it seems to me the translator’s equivalent of dressing up cats and dogs in coats and hats: bizarre, distracting, and aesthetically dud. Another approach, which crossed my mind, was to use 33 different English verse forms (assuming I could find 33), to mirror the variety in the original. But that, I suspected, would turn it into an anthology as opposed to a performable play. (Much of Kalidasa has in fact been anthologised in India.) I wanted a more unified tone across verses, especially those spoken by the same character. Why? Largely to create an identifiable tone and rhythm for the piece in English (I won’t call it rasa, but maybe something moving in that direction).
Another possibility, of course, was to ignore the verse completely and flatten everything into a uniform prose. Most of the translations that look like verse on the page are actually nevertheless in prose, or in what gets called ‘rhythmic prose’ (shouldn’t all prose be rhythmic?). I rejected that as well. What I did in the end, for better or worse, was to suggest the verse, and its range of differing metres (and to some extent the difference between Sanskrit and Prakrit verses), by using rhyme (sometimes), half-rhyme (frequently), assonance, alliteration, and, at the very least, looking for something more pronounced than a simple prose rhythm. So this, of course, was not meant to match the original on a like for like basis. Again my guideline was what would work in performance: I was looking for a way in which I could retain the verse, and at the same time allow it to be spoken fluently, speedily, and naturally by the actors. To that end, I wanted the verse to flow fairly seamlessly in and out of the prose, so I sometimes began a sentence in prose and moved midway into verse (to a marginal extent the original does that too). An earlier translator, Michael Coulson, lightly compared the difference between the prose and verse passages to that between recitative and aria in opera. In those terms, I was often aiming for something closer to a dramatic aria than simply a beautiful song suspending the action.
I want now to look at one example of what I did with a particular verse: it’s a hard case, and I’m not really satisfied with the result. I also did something pretty heretical here with my translation , although this is atypical in being extreme heresy – I more normally settled for milder varieties. Apart from the treatment of the verse, this example also demonstrates something about how rasa and emotion are evoked through language, and how Sanskrit and English need different means if they are to accomplish similar ends.
The example comes from a group of Sanskrit verses in Act 4. The translator is on a hiding to nothing from the start, since, according to literary legend, this Act is the greatest Act in Sanskrit drama, and these verses the pinnacle of Sanskrit poetry. The verse I want to look at, which happens to be the most famous of these, is spoken by Kanva, Shakuntala’s foster father, as his pregnant daughter prepares to leave the forest to join the king in the court. It’s in a form of Sanskrit metre called shardulavikridita (‘tiger’s play’), which has 19 syllables to the line (or pada), arranged in an entirely regular pattern of heavy and light syllables. My translation, as it happens, is also syllabically regular, but in a different way.
¯ ¯ ¯ ˘ ˘ ¯ ˘ ¯ ˘ ˘ ˘ ¯, ¯ ¯ ˘ ¯ ¯ ˘ ¯
yasyaty adya shakuntaleti hridayam samspristam utkanthaya
¯ ¯ ¯ ˘ ˘ ¯ ˘ ¯ ˘ ˘ ˘ ¯, ¯ ¯ ˘ ¯ ¯ ˘ ¯
*kantha: stambita-baspa-vritti-kalusash cinta-jadam darshanam |
¯ ¯ ¯ ˘ ˘ ¯ ˘ ¯ ˘ ˘ ˘ ¯, ¯ ¯ ˘ ¯ ¯ ˘ ¯
vaiklavyam mama tavad idrisham idam snehad aranyaukasa:
¯ ¯ ¯ ˘ ˘ ¯ ˘ ¯ ˘ ˘ ˘ ¯, ¯ ¯ ˘ ¯ ¯ ˘ ¯
pidyante grihina: katham tu tanaya-vishlesa-du:khair navai: ||
यास्यत्य् अद्य शकुन्तलेति हृदयम् सम्स्पृष्टम् उत्कण्ठया
कण्ठः स्तम्बित-बाष्प-वृत्ति-कलुषश् चिन्ता-जडं दर्शनम्।
वैक्लव्यं मम तावद् ईदृशम् इदम् स्नेहाद् अरण्यौकसः
पीड्यन्ते गृहिणः कथम् तु तनया-विश्लेष-दुःखैर् नवैः
My usual aim in translating verse is compression, which again is predicated on a certain non-traditional style of performance. In this case, the original Sanskrit has 76 syllables; my translation comes in at 56 – and here I’m using syllables as a rough guide to the relative time it takes to speak something, not in connection with metre. In fact, I could have delivered the significant content even more concisely, in 48 syllables, but I’ll explain shortly why I chose not to do that. A quick survey of some other translations shows a couple that exceed the Sanskrit original in length, and others which go in the same direction I took, although I haven’t yet found one more compressed than my own. (My syllable count in English may be slightly fuzzy, because it depends, of course, on pronuciation/enunciation)
Kale’s is probably the most literal of these translations (in some places it’s not grammatical English at all), mine may well be the least literal. There are reasons for this, not all of them to do with my competence or incompetence as a Sanskritist. Here’s Kale’s translation, followed by mine.
At the thought that Shakuntala is to go away to-day, my heart is smitten with grief; my throat is choked owing to the flow of tears suppressed; and my sight is heavy through anxiety. Such (so great), indeed, through affection (for Shakuntala), is this affliction of me that dwells in a forest, how much then must householders be tormented by the fresh (hitherto unexperienced) pangs of separation from their daughters?
(87 syllables – not counting the material in brackets)
Shakuntala must leave today –
My sight grows dark with what may come,
My throat is choked, my heart contracts,
A hard ascetic cracked by love.
Then what must wordly fathers feel,
A child departing in this way?
Shakuntala must leave today.
(56 incl.the 8 syllables of the repeated line)
What I wanted, essentially, was to relocate the emotional centre of gravity. In the original it’s embedded in the speaker’s description of his own sorrowful emotional state, which he then projects onto fathers in general. I didn’t want to change the emotion itself, but I did want a more enacted and characterised expression of it. You may think the achieved difference marginal, but my translation does to some extent deliberately work against some of the principles of Sanskrit aesthetics or rasa theory. For a start, simply in terms of duration, it doesn’t allow for so much tasting of the emotion – it won’t suit the connoisseur. More significantly, as I’ve suggested, I took a pragmatic decision to relocate the emotion to a place where those trained in the Western performance tradition (actor and audience) might have a better hope of finding it. And I attempted to do this by getting as much emotional force as possible into the active realisation that Shakuntala will – must leave today. In the first instance, this involved changing what is a reflective, future tense in the original – yasyati (‘she will leave’) to an ineluctable must. Next, I rendered what Sanskritists will recognise as the ‘iti clause’ – yasyaty adya shakuntaleti (sakuntala iti) – into direct speech: ‘Shakuntala must leave today’, where some others chose the indirect: ‘ My heart is touched with sadness since Shakuntala must go today’, ‘At the thought that Shakuntala is to go away to-day, my heart is smitten with grief’, and so on.
Again, I think their indirection here is quite in tune with the rasic intention of the original: it’s Kanva’s savouring of the sorrowful emotion associated with the karuna rasa, on behalf of the audience, that is at the heart of this. A savouring that takes the form of ‘I am thinking to myself that Shakuntala will leave today, and because of that I am sad’ … The cause of his sadness – Shakuntala leaving – is in some ways the least important element. And it’s taken a step further, since by the end of the verse in the original, the strength of Kanva’s personal emotion, however strongly or mildly expressed, has been further generalised, in good rasa-like fashion, to worldly fathers. This famous verse is therefore not in the play to tell us about Kanva’s character – it’s there to induce a certain, situation-bred aesthetic response in the audience. In a sense, he tells us, the ‘worldly fathers’ in the audience, what to feel. We don’t make the connection directly, so it is generalised for us too. We are invited to savour it.
This aesthetic seemed to me unlikely to have the same effect on a Western audience of non-connoisseurs, who tend to empathise with characters rather than generalised emotions. I therefore wanted, instinctively perhaps, to bring the emotion back to the character. I couldn’t see how to do it to begin with, but then, with the help of a rhyme, I realised that Kanva should repeat his initial assertion, where there is no repeat in the original. What is happening to him finally sinks in at the end of the verse.
To develop that a little … It seems to me that, in most Western drama, the actor finds the emotion in the line, and then enacts it through the language – where the language, and therefore the emotion, belong to the character. The emotion is – precisely – ‘characterised’. In Sanskrit drama, on the other hand, the emotion is generalised: the whole point is to detach it from the character – to depersonalise it. So the chief character (the hero or nayaka) is both the medium of the emotion, and its connoisseur, the rasika. The ultimate aim is to enable the rasa. Hence the assertion by some Western academics that ‘characters’ in Sanskrit dramas are stereotypes – because the important thing is not the development of character, but the rasa generated by the media of language and gesture. The ideal, in fact, is that, in the experience of the performance, the differences between character, actor, and audience should disappear, or be transcended in the aesthetic experience – the rasa which has been created.
I am not arguing for the impossibility of rasa in Western theatre – in fact I believe I have attended performances of Western dramas where something very similar has occurred (although this is necessarily subjective). But my point is that to get to a similar end in a different language and a different cultural context, you may have to employ a dissimilar means – i.e. different performance techniques, and translations at the service of those techniques. It’s not so much poetry that gets lost in translation, it’s rasa. And if you go looking for it in the new language in the same places you identified it in the original, it’s likely to remain lost.
So that in a way is my answer to the rasa singularity. But let me add something else.
It seems to me that from Shakespeare to Pinter, a lot of (the best) Western drama, relies on, and gets its dramatic effect from ambiguity (Jonathan Bate has written about this in relation to Shakespeare). For the audience, it mirrors our experience of life – our uncertainties and ambivalence about situations and people. There is no ambiguity in that sense – in the sense of what we’re supposed to feel – in Shakuntala. In fact, if you compare Kalidasa’s play with the source story in the Mahabharata, you see that he deliberately removes all possible ambiguities. As far as the audience is concerned, there is nothing hidden in the character or situation: everything is in the emotion and the rasa, in the intensity and perfection of the meeting of language and emotion in performance. In short (and here I’m talking about Shakuntala, not necessarily all types of Sanskrit drama ), with the possible exception of the scene with the fisherman, it’s not like life as it’s lived at all – it’s idealised, and like its main linguistic medium, Sanskrit, it’s refined (artificial, even) (one of the meanings of the word samskrita is ‘refined’). There’s never any doubt about what we’re supposed to feel, because we’re ‘told’ in language, gesture and music precisely what it is that we ‘should’ or need to be feeling. The only problem is having sufficient sensibility and sensitivity to understand (that’s to say, experience) this language of rasa and emotion. And that requires training as well as an inherent susceptibility. This lack of ambiguity – this emotional directness – accounts for the feeling of transparency, the feeling that there is nothing hidden, that everything is on the surface of the language, when we encounter such plays in translation. Without the requisite training or acculturation, what we’re left with is the shadow of the poetry, notoriously impossible to translate at a similar level of achievement.
By way of contrast, it seems that Shakespeare, for instance, translates amazingly well, if his worldwide popularity is anything to go by. And again, I think that’s precisely because the ambiguity of his characters and their situations translates well (an ambiguity expressed through language, of course, but to some extent separable from its immediate form). In contrast, Kalidasa, another great poet and playwright, removed from his cultural environment, can seem unexceptional, transparent – pantomimic even. Unless, that is, we can find a match of performance style and translation which releases the emotional charge, which liberates the rasa. And from the translator’s point of view, I think that’s something that can only be discovered in performance. In other words, the translation’s only the half of it. But I believe that the terms of the translation do, as I’ve suggested, configure the possible terms of the performance. Above all the translation and the performance style have to gel.
So when I translated Shakuntala, what counts as ‘authenticity’ in performance, and what original performances may or may not have been like, were not major questions for me. Rather, I was concerned to produce a version that would work in performance in vastly altered cultural circumstances. And, for better or worse, I thought I had produced a version that would work best in a language-focused production, with movement and mime based on stage directions, with music and song where required, but not otherwise with too many added dance-like elements.
But, as a couple of college productions in the US, using my translation, have demonstrated, translators have as little direct control over productions as dead playwrights.
All rights reserved by the author.
Dr Will Johnson is Senior Lecturer in Indian Religions at Cardiff University. Dr Johnson translated Kalidasa’s Shakuntala for Oxford World Classics (The Recognition of Shakuntala: A Play in Seven Acts) in 2001. His other translations include The Sauptikaparvan of the Mahabharata: the Massacre at Night and Mahabharata Book Three: The Forest, Volume Four for the Clay Sanskrit Library.
The new stage production of Shakuntala based on Dr Johnson’s translation will be on for three weeks at the Union Theatre in Waterloo, London, between the 20th January and the 2nd February 2009. http://uniontheatre.org/