Dr Will Johnson, Senior Lecturer in Indian Religions at Cardiff University, translated Kalidasa’s Shakuntala for Oxford World Classics in 2001 with the intention of producing a ‘performable’ translation of the Sanskrit drama.
His translation has formed the basis for two American productions and an upcoming London one directed by Tarek Iskander of Community 20 theatre company.
In this, the first of his two part discussion, Dr Johnson describes how he dealt with the play’s multilingual text and the ‘two other languages’ of Sanskrit drama – music and movement – bearing in mind his modern (Western) audience:
Translating for Performance
What follows is an informal reflection on my experience, seven or eight years ago, of translating the best-known classical Sanskrit drama, Kalidasa’s The Recognition of Shakuntala for Oxford World’s Classics. In so far as it is an analysis of my ‘method’ (although aspiration might be a better word), then it is definitely an analysis after the event – along the lines of ‘how might I make a different set of mistakes next time’.
I was, of course, faced with the same challenges that confront any translator of a Sanskrit drama. But rather than give a generalised description of these, I shall attempt to account for the specific decisions I made about how to deal with (or swerve around) a number of those challenges in relation to my imagined audience (my perceived target audience, to use the jargon of translation studies).
Readers of this website may well be familiar with the conventions of Sanskrit drama, so I shall merely list them here, with a brief commentary for future reference.
1. ‘Sanskrit’ drama, it was in fact composed not just in Sanskrit but also in a number of related languages called Prakrits. By Kalidasa’s time (4th to 5th Century CE) these were all literary rather than natural languages.
The manner in which Sanskrit and the Prakrits are used in Shakuntala is standardised in this way: Sanskrit is spoken by high-caste, educated males – so by the king and his higher-grade officials, by brahmins and by male ascetics. Females (of whatever rank), young boys, and lower-caste men speak different kinds of Prakrit. There are certain exceptions to this, notably the stock comic brahmin who is the king’s companion, the Vidusaka. In spite of his high-caste status he speaks a type of Prakrit, partly for comic effect and partly to illustrate his dubious character. Women of sufficient education (ascetics and courtesans) are also occasionally permitted to speak Sanskrit.
2. According to the developed tradition, two other languages are employed in the performance of a Sanskrit drama, although you barely glimpse them in the written text. These are the languages of music and movement. Just how large a part they played in early performances of Shakuntala seems to me open to question. But, whatever Kalidasa’s original intentions, later performance tradition, and many modern revivals, certainly treat Shakuntala as a dance-drama, and tend to lay on the full repertoire of hand, face, and bodily movements, in addition to the spoken language.
3. Like other plays of this type, Shakuntala is in a mixture of verse and prose.
4. According to Bharata’s Natyashastra the means by which a play achieves its emotional and aesthetic effect on the audience are explained by a principle which is usually labelled ‘rasa theory’. Rasa literally means ‘juice’ or ‘flavour’– so the essential metaphor is one of tasting. Who has the experience of tasting? The audience. Who provides it? The playwright in combination with the performers. How is it done? According to the Natyashastra the precise purpose of all worthwhile drama is the creation, out of the principal emotions evoked within the play, of a harmonious and complementary whole; and this in turn engenders a related, but impersonal, and universalised, experience of joy and bliss in the audience. It is the particular combination of poetry, action, plot, movement, sound, and gesture which brings about this rasa. But it doesn’t happen spontaneously: the poetry has to be good enough, the actors have to be good enough, and the audience has to be good enough – it needs to be trained or acculturated for the experience to occur. In other words, an audience of cognoscenti, educated in the subtleties of this aesthetic, have their common response conditioned, both by the specifics of the performance and by the dominant emotion or emotions that persist throughout it.
The nature of Shakuntala – of Sanskrit drama in general – therefore implies a certain sophistication in the audience – that it should be cosmopolitan, literate, courtly, urban and urbane, and educated in the convention; and it’s certainly not clear to me whether such demands were ameliorated to any degree at all by movement and music, or whether these simply meant that even more was expected of the audience. Almost certainly the latter, I think. It’s sometimes claimed that movement and music are ‘universal languages’. That seems to me no more significant than saying language itself is universal. You still have to learn how to look and to listen if you’re to understand – you still need to acquire a particular vocabulary and syntax. At some level you still have to translate.
The Natyashastra says drama is for all, but it can hardly have had dramas such as Shakuntala in mind. Unless its definition of ‘all’ is in fact restricted to the significant ‘all’ of high class, educated persons. (And what does the fact that the rasika is male tell us about the gender constitution of the ideal audience? Or is it assumed that the rasa experience ultimately transcends gender – the assumption made by some later devotional movements?)
If the Natyashastra‘s claim – that drama is a fifth Veda accessible to all castes or classes (sarva varnika) – refers to all those involved in the production (smiths, carpenters, painters and so on), then it may well be true (and is perhaps analogous to their involvement in the great Vedic rituals). But if it refers to the audience, I’m not sure how you square the apparent democratic instincts of Natyashastra with the actual form of Sanskrit drama.
So now I come to my own translation. How did I deal with these conventions or singularities of Sanskrit drama?
My first decision effectively dictated all those that followed. I decided that I had to produce a performable version. Partly this was because I didn’t think that any of the other translations I had looked at were strikingly performable, and mine might at least differ in that respect, but also because I have a long-standing interest in theatre, and so I know that plays come to life in performance in ways that can’t be imagined when you simply read them on the page. In other words, I would have chosen the performable, over the solely literary, whatever the circumstances. Which was not to say I was abandoning the literary altogether – I’ll come on to that presently. I say ‘performable’, which is not quite the same as ‘performing’, since I would take a performing version to be one tailored to the practical requirements of a specific production. A performable version is therefore one that has the potential to be configured in various ways – a performing version in waiting, as it were.
Once you’ve decided that you’re translating for performance, you have to be totally pragmatic – the only thing that counts is what works (i.e. what communicates the desired effect to the audience). There is a catch to this of course, in that you don’t always know what works until you see it in peformance – which is why new (and sometimes not so new) plays are often changed quite radically in rehearsal, which is where performing versions are usually made. I only had that luxury in a limited way: I tried out a version of the prologue and Act 1 with some graduate students in the early stages of my translation. But most of the time, for better of worse, I had to rely on my own experience of performance – mostly as the member of an audience – to imagine what might and might not work. Effectively, I was directing a kind of performance in my head throughout the translation process. (Incidentally, this probably applies in a different way to all translation. It’s significant that one of the very few readable books on the nature and problems of literary translation – by Robert Wechsler – is called Performing Without A Stage, although it barely mentions the theatre as such.)
As you might guess, my perception of my audience, and how it would differ from Kalidasa’s, turned out to be the crucial factor here. (This isn’t the same, of course, as giving the audience what it wants, or what it’s comfortable with, but it is about the limits of what’s effectively communicable.) We’ve established that in terms of a particular performance style and its associated conventions, Kalidasa was writing for an audience of cognoscenti or connoisseurs. The audience for my translation would be completely different – and what they wouldn’t be is cognoscenti. Nor was I going to translate for those second order cognoscenti – academic specialists in the field. Why would they need another translation of this particular text anyway, other than to purr over, or, more likely, shred? I wasn’t in any particular sense aiming for their students either – although I think Oxford University Press may have had other ideas about that. Rather, I was aiming for an imagined mixed audience sitting in a theatre somewhere in metropolitan, or even provincial, Britain on a dull Tuesday evening in November. In other words, I could simply not afford to assume what Kalidasa could assume. And if I translated him without reimagining and transmuting his principal assumptions (including those derived from the style of performance and the conventions of a drama in Sanskrit in the 5th century CE), what I translated would not be performable – that’s to say it would not communicate anything significant to a modern audience.
This of course raises plenty of other problems, not least the apparent over-transparency of the play in English. Without benefit of footnotes (usually unperformable), and particularly if cut from its original length, it may seem like just another fairy tale, or even, in performance, an exotic pantomime. Once you’ve gone down this line, then your only defence may be in attempting to convey, however inadequately, some of the play’s literary quality, and so its flavour or rasa, in a shifted, transformed sense. And in general I tried to move the play from the highly stylised and artificial (in linguistic and dramatic terms) towards something more naturalistic and in tune with current performance styles. Of course, hardly any translations outlast their times. Classic texts get retranslated – so if you’re aiming for something performable, it should, in my view, be within the limits of current style – i.e. what works for this audience now. And I would say that my translation is, in a way, close to the limits of what is performable now– to make it more performable (and make it perhaps a ‘performing’ version), I would have had to step even further from the original, and probably out of the jacket of a ‘World’s Classic’ altogether. I would have had to destroy completely the reader’s (and editor’s) illusion that they were getting ‘Shakuntala as it is’ (to borrow a phrase).
So this brings me to how I dealt with the first singularity of Shakuntala, its mixture of languages. And this was easy. Like every other translator, I simply ignored it and treated the play as monolingual. Because it’s insoluble, it’s no problem at all. The intelligent bridle at the idea something’s insoluble, of course. You think – and I thought too – that there must be some ingenious solution. All the women speak in Welsh or French, all the comic characters have Geordie accents. But this would be the equivalent of a postwar American translation of Ivan Denisovitch, where the inhabitants of the gulag were all G.I.s. Think of the historical, social and political overtones. Think of the unintended comedy. There are modern examples of multilingual plays – some Canadian dramas for instance (Robert Lepage works across languages to some extent). And there have been attempts to translate them. But again, think of the specific associations.
Obviously I hoped to convey some of the differences in the way that characters speak through light changes in idiom, and so on. It’s possible to an extent for the comic characters, but how do you express the differences between men and women when they’re in conversation with each other in two different languages, let alone the differences between one Prakrit and another? This particular distinction – this verbal polyphony – is not so much lost in translation, as lost to it altogether. But since we’re talking about performance, and engagement with a modern audience, there are not so many opportunities for it to make sense in contemporary terms anyway. Multilingual plays need multilingual audiences. (Maybe someone operating with, say, Hindi and English in a British Asian context could fashion something. But more likely they couldn’t.) I believe that in the South Indian Kootiyattam performance tradition (which is a form of Sanskrit theatre), the Prakrits are sometimes replaced by the local vernacular (Malayålam), but these are not full translations – you (the audience) still have the Sanskrit to deal with. Dance and gesture have probably become the principal language of Kootiyattam. And modern Indian productions of Shakuntala which don’t take the heritage option tend to be mostly in one language (Hindi or another of India’s many contemporary languages) as far as I can discover.
And in fact, from a relatively early date the Indian tradition itself had became monolingual in respect of classical Sanskrit drama. By the 10th century, probably all Sanskrit plays, including Shakuntala, were indeed performed entirely in Sanskrit, commentators having supplied a Sanskrit version of the Prakrits, known as a chåyå or ‘shadow’, for the benefit of the Sanskrit speaking elite. While Sanskrit continued to be understood as the lingua franca, the Prakrits had given way way to other vernaculars and, perhaps more pertinently, other literary languages – i.e. vernacular literary languages.
What about the second singularity of Shakuntala, and of Sanskrit drama in general, its assumption of a multiform performance style, including highly coded gestures and music? You could say that this isn’t the translator’s problem, but if you’re translating for performance it may certainly become one. Sanskrit dramas are leisurely, they take their time – after all you need to savour their rasa – and so they take time to perform. If you add in dance, and continual hand, face and body gestures, they take even longer. Indeed, originally they may have often been performed in installments, or at least over a whole night, like Kathakali dance dramas, and particular acts would certainly have been performed on their own at the request of patrons. One of the longer versions of Shakuntala probably had material added by Kalidasa, or someone else, for precisely that purpose.
Clearly audiences were expected to see them time and time again, deepening their experience at each performance, relishing or tasting slight and subtle variations on a well-known theme. Most modern, Western audiences, on the other hand, will only see a play once and are not made up of connoisseurs in that sense. To expand this a little: the one thing needed by connoisseurs (and here I mean nothing more than an acculturated audience) is time – repeated close contact with a story that, in one sense, is soon already well-known. But what can be derived from such contact is not instantly realised: it requires ‘training’, in the sense of habituation and concentrated exposure – a point made by the great Kashmiri Shaiva theologian Abhinavagupta (drawing on and developing rasa theory in the 10th century C.E.). Describing the most sensitive readers of poetry (kavya) – the sahridayas or connoisseurs, he says
येषाम् काव्यानुशीलनवशाद् विशदीकृते मनो–मुकुरे वर्णनीयतन्मयी–भवना–योग्यता इति
The sensitive readers are those whose mirror-like minds are made crystal clear by their constant practice (anushIlana) of the reading of poetry in such a way that their minds become identified with whatever is described in poetry. (my italics)
Dhvanyaloka-Locana 1.1 (trans. Matilal).
Matilal The Word and the World (‘Meaning in Literary Criticism’) p. 177:
AnushIlana – the word translated by ‘constant practice’ – also has the sense of ‘constant repetition’ – there is something mantra-like about it. And in fact, according to Abhinavagupta, the true connoisseur is more than half-way to spiritual enlightenment.
This is certainly a long distance from the conditions that apply for most modern (Western) audiences. The economics of modern theatre and ticket prices do not encourage connoisseurship in this sense. Time in this case is all too obviously money for everyone involved. The translator can hardly be blamed, therefore, if she or he settles for a performance of the imagination – one which takes place in the mind of a reader – the novelisation, if you like, of the drama, and avoids the performance of these plays altogether. Rashly, perhaps, I took a different line and tried to recreate something that might be actually performable within the conventions of Western theatre. But to do that I had to counter, to some extent, the connoisseurs’ approach.
One of my strategies was to speed up the delivery of what was already, for a Western audience, a long, seven act play. Not by cutting the text (it seems to me that you need most of the poetry to have any chance of getting some of the rasic effect – and it’s always the poetry that gets cut by directors), but by driving the action forward through the language. My assumption was that, apart from the usual gestures, and maybe some naturalistic mime, the actor would not require extra time to do other things as he or she spoke the line– i.e. they wouldn’t need to employ the full panoply of coded gestures you get in traditional performances of this material.
This article continues in ‘Translating for Performance – Dr Will Johnson, Part 2’