Since the different tribes of the human Babel began to study each other's literatures, the problem of poetical translation has constantly defied the earnest experimenter. There have been brilliant versions, successful falsifications, honest renderings, but some few lyrics apart a successful translation there has not been. Yet it cannot be that a form of effort so earnestly & persistently pursued and so necessary to the perfection of culture and advance of civilisation, is the vain pursuit of a chimera. Nothing which mankind earnestly attempts is impossible, not even the conversion of copper into gold or the discovery of the elixir of life or the power of aerial motion; but so long as experiment proceeds on mistaken lines, based on a mistaken conception of the very elements of the problem, it must necessarily fail. Man may go on fashioning wings for himself for ever but they will never lift him into the empyrean: the essence of the problem is to conquer the attraction of the earth which cannot be done by any material means. Poetical translation was long dominated by the superstition that the visible word is the chief factor in language and the unit which must be seized on as a basis in rendering; the result is seen in so-called translations which reproduce the sense of the original faultlessly & yet put us into an atmosphere which we at once recognize to be quite alien to the atmosphere of the original; we say then that the rendering is a faithful one or a success of esteem or a makeshift or a caput mortuum according to the nature of our predilections and the measure of our urbanity. The nineteenth century has been the first to recognize generally that there is a spirit behind the word & dominating the word which eludes the "faithful" translator and that it is more important to get at the spirit of a poet than his exact sense. But after its manner it has contented itself with the generalisation and not attempted to discover the lines on which the generalisation must

Page – 239

be crystallised into practice, its extent & its limitations. Every translator has been a law to himself; and the result is anarchic confusion. As the sole tangible benefit there has been discovered a new art not yet perfected of translation into prose poetry. Such translation has many advantages; it allows the translator to avail himself of manifold delicacies of rhythm without undergoing the labour of verse formation and to compromise with the orthodox superstition by rendering the word unit yet with some show of preserving the original flavour. But even in the best of these translations it is little more than a beautiful show. Poetry can only be translated by poetry and verse forms by verse forms. It remains to approach the task of translation in a less haphazard spirit, to realise our essential aim, to define exactly what elements in poetry demand rendering, how far & by what law of equivalent values each may be rendered and if all cannot be reproduced, which of them may in each particular case be sacrificed without injuring the essential worth of the translation. Most of the translations of Kalidasa here offered to the public have been written after the translator had arrived at such a definite account with himself and in conscientious conformity to its results. Others done while he yet saw his goal no more than dimly and was blindly working his way to the final solution, may not be so satisfactory. I do not pretend that I have myself arrived at the right method; but I am certain that reasoned & thoughtful attempts of this sort can alone lead to it. Now that nations are turning away from the study of the great classical languages to physical & practical science and resorting even to modern languages, if for literature at all then for contemporary literature, it is imperative that the ennobling influences spiritual, romantic & imaginative of the old tongues should be popularised in modern speech; otherwise the modern world, vain of its fancied superiority & limiting itself more & more to its own type of ideas with no opportunity of saving immersions in the past & recreative destructions of the present, will soon petrify & perish in the mould of a rigid realism & materialism. Among their influences the beauty & power of their secular & religious poetry is perhaps the most potent & formative.

Page – 240

The choice of meter is the first & most pregnant question that meets a translator. With the growth of Alexandrianism and the diffusion of undigested learning, more & more frequent attempts are being made to reproduce in poetical versions the formal metre of the original. Such attempts rest on a fundamental misconception of the bases of poetry. In poetry as in all other phenomena it is spirit that is at work and form is merely the outward expression & instrument of the spirit. So far is this true that form itself only exists as a manifestation of spirit and has no independent being. When we speak of the Homeric hexameter, we are speaking of a certain balance [of] spiritual force called by us Homer working through emotion into the material shape of a fixed mould of rhythmical sound which obeys both in its limiting sameness & in its variations the law of the spirit within.

The mere quantities are but the most mechanical & outward part of metre. A fanciful mind might draw a parallel between the elements of man & the elements of metre. Just as in man there is the outward food-plasm and within it the vital or sensational man conditioned by & conditioning the food-plasm & within the vital man the emotional or impressional man similarly related and again within that the intellectual man governing the others and again within that the delight of the spirit in its reasoning existence & within that delight like the moon within its halo the Spirit who is Lord of all these, the sitter in the chariot & the master of its driving, so in metre there is the quantitative or accentual arrangement which is its body, & within that body conditioning & conditioned by it the arrangement of pauses & sounds, such as assonance, alliteration, composition of related & varying letters, and again within it conditioning & conditioned by this sensational element & through it the mechanical element is the pure emotional movement of the verse and again within these understanding and guiding all three, bringing the element of restraint, management, subordination to a superior law of harmony, is the intellectual element, the driver of the chariot of sound; within this again is the poetic delight in the creation of

Page – 241

harmonious sound, the august & disinterested pleasure of the really great poet which has nothing in it of frenzy or rather has the exultation & increased strength of frenzy without its loss of self-control; and within this even is the spirit, that unanalysable thing behind metre, style & diction which makes us feel "This is Homer, this is Shakespeare, this is Dante."

[All these are essential before really great verse can be produced; everyone knows that verse may scan well enough & yet be very poor verse; there may beyond this be skilful placings of pause & combinations of sound as in Tennyson's blank verse, but the result is merely artificially elegant & skilful technique; if emotion movement is super-added, the result is melody, lyric sweetness or elegiac grace or flowing & sensuous beauty, as in Shelley, Keats, Gray, but the poet is not yet a master of great harmonies; for this intellect is necessary, a great mind seizing, manipulating & moulding all these by some higher law of harmony, the law of its own spirit. But such management is not possible without the august poetical delight of which I have spoken, and that again is but the outflow of the mighty spirit within, its sense of life & power & its pleasure in the use of that power with no ulterior motive beyond its own delight.]¹

But just as the body of a man is also soul, has in each of its cells a separate portion of spirit, so it is with the mechanical form of a verse. The importance of metre arises from the fact that different arrangements of sound have different spiritual and emotional values, tend to produce that is to say by virtue of the fixed succession of sounds a fixed spiritual atmosphere & a given type of emotional exaltation & the mere creative power of sound though a material thing is yet near to spirit, is very great; great on the material & ascending in force through the moral & intellectual, culminating on the emotional plane. It is a factor of the first importance in music & poetry. In these different

1 Paragraph bracketed in the manuscript. Written at the end of the piece, it was apparently intended for insertion here. — Ed.

Page – 242

arrangements of syllabic sound metre forms the most important, at least the most tangible element. Every poet who has sounded his own consciousness must be aware that management of metre is the gate of his inspiration and the law of his success. There is a double process, his state of mind and spirit suggesting its own syllabic measure, and the metre again confirming, prolonging and recreating the original state of mind and spirit. Inspiration itself seems hardly so much a matter of ideas or feeling as of rhythm. Even when the ideas or the feelings are active, they will not usually run into the right form, the words will not take their right places, the syllables will not fall into a natural harmony. But if one has or succeeds in awaking the right metrical mood, if the metrical form instead of being deliberately created, creates itself or becomes, a magical felicity of thought, diction & harmony attends it & seems even to be created by it. Ideas & words come rapidly & almost as rapidly take their right places as in a well ordered assembly where everyone knows his seat. When the metre comes right, everything else comes right; when the metre has to be created with effort, everything else has to be done with effort, and the result has to be worked on over & over again before it satisfies.

This supreme importance of the metrical form might seem at first sight to justify the transplanters of metre. For if it be the aim of good translation to reproduce not merely the mechanical meanings of words, the corresponding verbal counters used in the rough & ready business of interlingual commerce, but to create the same spiritual, emotional & aesthetic effect as the original, the first condition is obviously to identify our spiritual condition, as far as may be, with that of the poet at the time when he wrote & then to embody the emotion in verse. This cannot be done without finding a metre which shall have the same spiritual and emotional value as the metre of the original. Even when one has been found, there will naturally be no success unless the mind of the translator has sufficient kinship, sufficient points of spiritual & emotional contact and a sufficient basis of common poetical powers not only to enter into but to render the spiritual temperament & the mood of that temperament,

Page – 243

of which his text was the expression; hence a good poetical translation is the rarest thing in the world. Conversely even if all these requisites exist, they will not succeed to the full without the discovery of the right metre. Is the right metre then the metre of the original? Must an adequate version of Homer, a real translation, be couched in the hexameter? At first sight it would seem so. But the issue is here complicated by the hard fact that the same arrangement of quantities or of accents has very seldom the same spiritual & emotional value in two different languages. The hexameter in English, however skilfully managed, has not the same value as the Homeric, the English alexandrine does not render the French; terza rima in Latinised Saxon sounds entirely different from the noble movement of the Divina Commedia, the stiff German blank verse of Goethe & Schiller is not the golden Shakespearian harmony. It is not only that there are mechanical differences, a strongly accentuated language hopelessly varying from those which distribute accent evenly, or a language of ultimate accent like French from one of penultimate accent like Italian or initial accent like English, or one which courts elision from one which shuns it, a million grammatical & syllabic details besides, lead to fundamental differences of sound-notation. Beyond & beneath these outward differences is the essential soul of the language from which they arise, and which in its turn depends mainly upon the ethnological type always different in different countries because the mixture of different root races in two types even when they seem nearly related is never the same. The Swedish type for instance which is largely the same as the Norwegian is yet largely different, while the Danish generally classed in the same Scandinavian group differs radically from both. This is that curse of Babel, after all quite as much a blessing as a curse, which weighs upon no one so heavily as on the conscientious translator of poetry; for the prose translator being more concerned to render the precise idea than emotional effects and the subtle spiritual aura of poetry, treads an immeasurably smoother & more straightforward path. For some metres at least it seems impossible to find adequate equivalents in other languages. Why has there never been a real rendering of Homer

Page – 244

in English? It is not the whole truth to say that no modern can put himself back imaginatively into the half-savage Homeric period; a mind with a sufficient basis of primitive sympathies & sufficient power of imaginative self-control to subdue for a time the modern in him may conceivably be found. But the main, the insuperable obstacle is that no one has ever found or been able to create an English metre with the same spiritual & emotional equivalent as Homer's marvellous hexameters.

That transmetrisation is a false method, is therefore clear. The translator's only resource is to steep himself in the original, quelling that in him which conflicts with its spirit, and remain on the watch for the proper metrical mood in himself. Sometimes the right metre will come to him, sometimes it will not. In the latter case effort in this direction will not have been entirely wasted; for spirit, when one gives it a chance, is always stronger than matter & he will be able to impose something of the desired spiritual atmosphere even upon an unsuitable metrical form. But if he seize on the right metre, he has every chance, supposing him poetically empowered, of creating a translation which shall not only be classical, but shall be the translation. Wilful choice of metre is always fatal. William Morris' Homeric translation failed hopelessly partly because of his affected "Anglosaxon" diction, but still more because he chose to apply a metre good enough possibly for the Volsungasaga to the rendering of a far more mighty & complex spirit. On the other hand Fitzgerald might have produced a very beautiful version in English had he chosen for his Rubaiyat some ordinary English metre, but his unique success was his reward for discovering the true equivalent of the quatrain in English. One need only imagine to oneself the difference if Fitzgerald had chosen the ordinary English quatrain instead of the rhyme system of his original. His Rubaiyat in spite of the serious defect of unfaithfulness will remain the final version of Omar in English, not to be superseded by more faithful renderings, excluding therefore the contingency of a superior poetical genius employing the same metre for a fuller & closer translation.

In Kalidasa another very serious difficulty over & beyond

Page – 245

the usual pitfalls meets the unhappy translator. Few great Sanscrit poems employ the same metre throughout. In the dramas where metrical form is only used when the thought, image or emotion rises above the ordinary level, the poet employs whatever metre he thinks suitable to the mood he is in. In English however such a method would result in opera rather than in drama. I have therefore thought it best, taking into consideration the poetical feeling & harmonious flow of Kalidasa's prose, to use blank verse throughout varying its pitch according as the original form is metrical or prose & the emotion or imagery more or less exalted. In epic work the licence of metrical variation is not quite so great; yet there are several metres considered apt to epic narrative & Kalidasa varies them without scruple in different cantos, sometimes even in the same canto. If blank verse be, as I believe it is, a fair equivalent for the anustubh, the ordinary epic metre, how shall one find others which shall correspond as well to the "Indra's thunderbolt sloka", the "lesser Indra's thunderbolt sloka", the "gambolling of the tiger sloka" and all those other wonderful & grandiose rhythmic structures with fascinating names of which Kalidasa is so mighty a master? Nor would such variation be tolerated by English canons of taste. In the epic & drama the translator is driven to a compromise and therefore to that extent a failure; he may infuse good poems or plays reproducing the architecture & idea-sense of Kalidasa with something of his spirit; but it is a version & not a translation. It is only when he comes to the Cloud Messenger that he is free of this difficulty; for the Cloud Messenger is written throughout in a single & consistent stanza. This Mandakranta or "gently stepping" stanza is entirely quantitative and too complicated to be rendered into any corresponding accentual form. The arrangement of metrical divisions is as follows: spondee-long, dactyl, tribrach, two spondee-shorts, spondee; four lines of this build make up the stanza. Thus

Page – 246

In casting about for a metre I was only certain of one thing that neither blank verse nor the royal quatrain would serve my purpose; the one has not the necessary basis of recurring harmonies; in the other the recurrence is too rigid, sharply defined & unvarying to represent the eternal swell and surge of Kalidasa's stanza. Fortunately by an inspiration, & without deliberate choice, Kalidasa's lines as I began turning them flowed or slipped into the form of triple rhyme and that necessarily suggested the terza rima. This metre, as I have treated it, seems to me to reproduce with as much accuracy as the difference between the languages allows, the spiritual & emotional atmosphere of the Cloud Messenger. The terza rima in English lends itself naturally to the principle of variation in recurrence, which imparts so singular a charm to this poem, recurrence in especial of certain words, images, assonances, harmonies, but recurrence always with a difference so as to keep one note sounding through the whole performance underneath its various harmony. In terza rima the triple rhyme immensely helps this effect, for it allows of the same common rhymes recurring but usually with a difference in one or more of their company.

It is a common opinion that terza rima does not suit the English language and cannot therefore be naturalised, that it must always remain an exotic. This seems to me a fallacy. Any metre capable of accentual representation in harmony with the accentual law of the English language, can be naturalised in English. If it has not yet been done, we must attribute it to some initial error of conception. Byron & Shelley failed because they wanted to create the same effect with this instrument as Dante had done; but terza rima in English can never have the same effect as in Italian. In the one it is a metre of woven harmonies suitable to noble & intellectual narrative; in the other it can only be a metre of woven melodies suitable to beautiful description or elegiac sweetness. To occasional magnificences or sublimities it lends itself admirably, but I should doubt whether it could even

Page – 247

in the strongest hands sustain the burden of a long & noble epic of the soul & mind like the Divina Commedia. But it is not true that it cannot be made in English a perfectly natural, effective & musical form. It is certainly surprising that Shelley with his instinct for melody, did not perceive the conditions of the problem. His lyric metres & within certain limitations his blank verse are always fine, so fine that if the matter & manner were equal to the melody, he would have been one of the few great poets instead of one of the many who have just missed being great. But his Triumph of Life is a metrical failure. We feel that the poet is aiming at a metrical effect which he has not accomplished.

The second question, but a far simpler one, is the use of rhyme. It may be objected that as in the Sanscrit there is no rhyme, the introduction of this element into the English version would disturb the closeness of the spiritual equivalent by the intrusion of a foreign ornament. But this is to argue from a quantitative to an accentual language, which is always a mistake. There are certain effects easily created within the rich quantitative variety of ancient languages, of which an equivalent in English can only be found by the aid of rhyme. No competent critic would declare Tennyson's absurd experiment in Boadicea an equivalent to the rushing, stumbling & leaping metre of the Attis with its singular & rare effects. A proper equivalent would only be found in some rhymed system and preferably I should fancy in some system of unusually related but intricate & closely recurring rhymes. Swinburne might have done it; for Swinburne's work, though with few exceptions poor work as poetry, is a marvellous repertory of successful metrical experiments. I have already indicated the appropriateness of the triple rhyme system of the terza rima to the Cloud Messenger. English is certainly not a language of easy rhyming like the southern tongues of Europe; but given in the poet a copious command of words and a natural swing and felicity, laeta rather than curiosa, it is amply enough provided for any ordinary call upon its resources. There are however two critical superstitions which seriously interfere with the naturalness & ease rhymed poetry

Page – 248

demands, the superstition of the perfect rhyme and the superstition of the original rhyme. It is no objection to a rhyme that it is imperfect. There is nothing occult or cryptic in rhyme, no divine law compelling us to assimilate two rhymed endings to the very ´ letter such as the law of the Vedic chant by which a single letter mispronounced sterilizes the mantra. Rhyme is a convenience and an ornament intended to serve certain artistic purposes, to create certain sound-effects, and if the effect of a perfect rhyme is beautiful, melodious and satisfying, an imperfect rhyme has sometimes its own finer effect far more subtle, haunting and suggestive; by limiting the satisfaction of the ear, it sets a new chord vibrating in the soul. A poem with an excessive proportion of imperfect rhymes is unsatisfactory, because it would not satisfy the natural human craving for regularity & order; but the slavish use of perfect rhymes only would be still more inartistic because it would not satisfy the natural human craving for liberty & variety. In this respect and in a hundred others the disabilities of the English language have been its blessings; the artistic labour & the opportunity for calling a subtler harmony out of discord have given its best poetical literature a force & power quite out of proportion to the natural abilities of the race. There are of course limits to every departure from rigidity but the degree of imperfection admissible in a rhyme is very great so long as it does not evolve harshness or vulgarism. Mrs Browning's rhymes are .. bad in this respect, but why? Because "tyrants" & "silence" is no rhyme at all, while "candles" & "angels" involves a hideous vulgarism; and in less glaring instances the law of double rhymes generally requiring closer correspondence than single is totally disregarded. The right use of imperfect rhymes is not to be forbidden because of occasional abuse. It is also no objection to a rhyme that it is "hackneyed". A hackneyed thought, a hackneyed phrase there may be, but a hackneyed rhyme seems to me a contradiction in terms. Rhyme is no part of the intellectual warp & woof of a poem, but a pure ornament the only object of which is to assist the soul with beauty; it appeals to the soul not through the intellect or imagination but through the ear. Now the oldest & most often used rhymes are generally the most

Page – 249

beautiful and we ought not to sacrifice that beauty merely out of an unreasoning impatience of what is old. Common rhymes have a wonderful charm of their own and come to us laden with a thousand beautiful associations. The pursuit of mere originality can only lead us to such unpardonable extravagances as "haunches stir" & "Manchester". Such rhymes any poet can multiply who chooses to prostitute his genius to the amusement of the gallery, or is sufficiently unpoetic to prefer the freedom of barbarous uncouthness to that self-denial which is the secret of grace & beauty. On the other hand if we pursue originality & beauty together, we end in preciosity or an artificial grace, and what are these but the spirit of Poetry lifting her wings to abandon that land and that literature for a long season or sometimes for ever? Unusual & peculiar rhymes demand to be sparingly used & always for the definite object of setting in relief common rhymes rather than for the sake of their own strangeness.

The question of metre and rhymes being satisfactorily settled there comes the crucial question of fidelity, on which every translator has to make his own choice at his own peril. On one side is the danger of sacrificing the spirit to the letter, on the other the charge of writing a paraphrase or a poem of one's own under the cloak of translation. Here as elsewhere it seems to me that rigid rules are out of place. What we have to keep in mind is not any rigid law, but the object with which we are translating. If we merely want to render, to acquaint foreign peoples with the ideas & subject matter of the writer, as literal a rendering as idiom will allow, will do our business. If we wish to give a poetical version, to clothe the general sense & spirit of the writer in our own words, paraphrase & unfaithfulness become permissible; the writer has not intended to translate and it is idle to criticise him with reference to an ideal he never entertained. But the ideal of a translation is something different from either of these. The translator seeks first to place the mind of the reader in the same spiritual atmosphere as the original; he seeks next to produce in him the same emotions & the same kind of poetical delight and aesthetic gratification, and lastly he seeks to convey to him the thought of the poet & substance in such words as will create,

Page – 250

as far as may be, the same or a similar train of associations, the same pictures or the same sensuous impressions. This is an ideal to which one can never do more than approximate; but the nearer one approximates to it, the better the translation. How it shall be done, depends upon the judgment, the sympathetic instinct of the poet, the extent to which he is imbued with the associations of both languages and can render not merely word by word but shade by shade, not only signification by signification, but suggestion by suggestion. There is one initial stumbling block which can never be quite got over; the mythology, fauna & flora of Indian literature are absolutely alien to Europe. (We are in a different world; this is no peaceful English world of field or garden & woodland with the cheerful song of the thrush or the redbreast, the nightingale warbling in the night by some small & quiet river, the lark soaring in the morning to the pale blue skies; no country of deep snows & light suns & homely toil without spiritual presences save the borrowed fancies of the Greeks or shadowy metaphysical imaginations of the poet's brain that haunt thought's aery wildernesses, no people homely [and] matter of fact, never rising far above earth or sinking far below it. We have instead a mother of gigantic rivers, huge sombre forests and mountains whose lower slopes climb above the clouds;² the roar of the wild beasts fills those forests & the cry of innumerable birds peoples those rivers; & in their midst lives a people who have soared into the highest heavens of the spirit, experienced the grandest & most illimitable thoughts possible to the intellect & sounded the utmost depths of sensuous indulgence; so fierce is the pulse of life that even trees & inanimate things seem to have life, emotions, a real & passionate history and over all move mighty presences of gods & spirits who are still real to the consciousness of the people.)

The life & surroundings in which Indian poetry moves cannot be rendered in the terms of English poetry. Yet to give up

2 An alternative version reads, after "forests": under a burning sun or a magical moonlight;

Page – 251

the problem and content oneself with tumbling out the warm, throbbing Indian word to shiver & starve in the inclement atmosphere of the English language seems to me not only an act of literary inhumanity & a poor-spirited confession of failure, but a piece of laziness likely to defeat its own object. An English reader can gather no picture from & associate no idea of beauty with these outlandish terms. What can he understand when he is told that the atimukta creeper is flowering in the grove of késara trees and the mullica or the [ ] is sending out its fragrance into the night and the chocrovaque is complaining to his mate amid the still ripples of the river that flows through the jambous? Or how does it help him to know that the scarlet mouth of a woman is like the red bimba fruit or the crimson bandhoul flower? People who know Sanscrit seem to imagine that because these words have colour & meaning & beauty to them, they must also convey the same associations to their reader. This is a natural but deplorable mistake; this jargon is merely a disfigurement in English poetry. The cultured may read their work in spite of the jargon out of the unlimited intellectual curiosity natural to culture; the half-cultured may read it because of the jargon out of the ingrained tendency of the half-cultured mind to delight in what is at once unintelligible & inartistic. But their work can neither be a thing of permanent beauty nor serve a really useful object; & work which is neither immortal nor useful what self-respecting man would knowingly go out of his way to do? Difficulties are after all given us in order that we may brace our sinews by surmounting them; the greater the difficulty, the greater our chance of the very highest success. I can only point out rather sketchily how I have myself thought it best to meet the difficulty; a detailed discussion would require a separate volume. In the first place a certain concession may be made but within very narrow & guarded limits to the need for local colour; a few names of trees, flowers, birds etc. may be transliterated into English, but only when they do not look hopelessly outlandish in that form or else have a liquid or haunting beauty of sound; a similar indulgence may be yet more freely permitted in the transliteration of mythological names. But here

Page – 252

the licence ends; a too liberal use of it would entirely destroy the ideal of translation; what is perfectly familiar in the original language must not seem entirely alien to the foreign audience; there must be a certain toning down of strangeness, an attempt to bring home the association to the foreign intelligence, to give at least some idea to a cultured but not Orientally erudite mind. This may be done in many ways & I have availed myself of all. A word may be rendered by some neologism which will help to convey any prominent characteristic or idea associated with the thing it expresses; blossom o' ruby may, for instance, render bandhoula, a flower which is always mentioned for its redness. Or else the word itself may be dropped & the characteristic brought into prominence; for instance instead of saying that a woman is lipped like a ripe bimba, it is, I think, a fair translation to write "Her scarlet mouth is a ripe fruit & red." This device of expressingly declaring the characteristics which the original only mentions, I have frequently employed in the Cloud Messenger, even when equivalent words exist in English, because many objects known in both countries are yet familiar & full of common associations to the Indian mind while to the English they are rare, exotic and slightly associated or only with one particular & often accidental characteristic.3 A kindred method especially with mythological allusions is to explain fully what in the original is implicit; Kalidasa for instance compares a huge dark cloud striding northwards from Crouncharundhra to "the dark foot of Vishnu lifted in impetuous act to quell Bali", śyāmah pādo baliniyamanābhyudyatasyeva visnoh. This I have .. . translated

3 The following passage was written in the top margins of these pages of the manuscript. Its place of insertion was not indicated:

It is an unfortunate tendency of the English mind to seize on what seems to it grotesque or ungainly in an unfamiliar object; thus the elephant & peacock have become almost impossible in English poetry, because the one is associated with lumbering heaviness & the other with absurd strutting. The tendency of the Hindu mind on the other hand is to seize on what is pleasing & beautiful in all things & even to see a charm where the English mind sees a deformity & to extract poetry & grace out of the ugly. The classical instances are the immortal verses in which Valmekie by a storm of beautiful & costly images & epithets has immortalised the hump of Manthara & the still more immortal passage in which he has made the tail of a monkey epic.

Page – 253

"Dark like the cloudy foot of highest God

When starting from the dwarf-shape world-immense

With Titan-quelling step through heaven he strode."

It will be at once objected that this is not translation, but the most licentious paraphrase. This is not so if my original contention be granted that the business of poetical translation is to reproduce not the exact words but the exact image, associations & poetical beauty & flavour of the original. There is not a single word in the translation I have instanced which does not represent something at once suggested to the Indian reader by the words of the text. Vishnu is nothing to the English reader but some monstrous & bizarre Hindu idol; to the Hindu He is God Himself; the word is therefore more correctly represented in English by "highest God" than by Vishnu; śyāmaḥ pādaḥ is closely represented by "dark like the cloudy foot", the word cloudy being necessary both to point the simile which is not so apparent & natural to the English reader as to the Indian and to define the precise sort of darkness indicated by the term śyāmaḥ; Bali has no meaning or association in English, but in the Sanscrit it represents the same idea as "Titan"; only the particular name recalls a certain theosophic legend which is a household word to the Hindu, that of the dwarf-Vishnu who obtained from the Titan Bali as much land as he could cover with three steps, then filling the whole world with himself with one stride measured the earth, with another the heavens and with the third placing his foot on the head of Bali thrust him down into bottomless Hell. All this immediately arises before the mental eye of the Hindu as he reads Kalidasa's finely chosen words. The impetuous & vigorous term abhyudyatasya both in sound & sense suggests the sudden starting up of the world-pervading deity from the dwarf shape he had assumed while the comparison to the cloud reminds him that the second step of the three is referred to, that of Vishnu striding "through heaven." But to the English reader the words of Kalidasa literally transliterated would be a mere artificial conceit devoid of the original sublimity. It is the inability to seize the associations & precise poetical force of Sanscrit words that

Page – 254

has led so many European Sanscritists to describe the poetry of Kalidasa which is hardly surpassed for truth, bold directness & native beauty & grandeur as the artificial poetry of an artificial period. A literal translation would only spread this erroneous impression to the general reader. It must be admitted that in the opposite method one of Kalidasa's finest characteristics is, it is true, entirely lost, his power of expressing by a single simple direct & sufficient word ideas & pictures of the utmost grandeur or shaded complexity; but this is a characteristic which could in no case be possible in any language but the classical Sanscrit which Kalidasa did more than any man to create or at least to perfect. Even the utmost literalness could not transfer this characteristic into English. This method of eliciting all the idea-values of the original of which I have given a rather extreme instance, I have applied with great frequency where a pregnant mythological allusion or a strong or subtle picture or image calls for adequate representation; more especially perhaps in pictures or images connected with birds & animals unfamiliar or but slightly familiar to the English reader. (At the same time I must plead guilty to occasional excesses, to reading into Kalidasa perhaps in a dozen instances what is not there. I can only plead in apology that translators are always incorrigible sinners in this respect and that I have sinned less than others; moreover except in one or two instances these additions have always been suggested either by the sound or substance of the original. I may instance the line

A flickering line of fireflies seen in sleep,

Kalidasa says nothing equivalent to or suggesting "seen in sleep", but I had to render somehow the impression of night & dim unreality created by the dreamy movement & whispering assonances of the lines

alpālpabhāsaṁ

khadyotālīvilasitanibhāṁ vidyudunmeṣadṛṣṭim

with their soft dentals & their wavering & gliding liquids and sibilants. Unable to do this by sound I sought to do it by verbal

Page – 255

expression; and in so far made a confession of incompetence, but in a way that may perhaps carry its own pardon.)

There is yet another method which has to be applied far more cautiously, but is sometimes indispensable. Occasionally it is necessary or at least advisable to discard the original image altogether and replace it by a more intelligible English image. There is no commoner subject of allusion in Sanscrit poetry than the passionate monotoned threnody of the forlorn bird who is divided at night by some mysterious law from his mate, divided if by a single lotus leaf, yet fatally divided. Such at least was the belief suggested by its cry at night to the imaginative Aryans. Nothing can exceed the beauty, pathos & power with which this allusion is employed by Kalidasa. Hear for instance Pururavus as he seeks for his lost Urvasie

Thou wild drake when thy love,

Her body hidden by a lotus-leaf,

Lurks near thee in the pool, deemest her far

And wailest musically to the flowers

A wild deep dirge. Such is thy conjugal

Yearning, thy terror such of even a little

Division from her nearness. Me thus afflicted,

Me so forlorn thou art averse to bless

With just a little tidings of my love.

And again in the Shacountala, the lovers are thus gracefully warned

O Chocrovaque, sob farewell to thy mate.

The night, the night comes down to part you.

Fable as it is, one who has steeped himself in Hindu poetry can never bring himself wholly to disbelieve it. For him the melancholy call of the bird will sound for ever across the chill dividing stream & make musical with pity the huge and solemn night. But when the Yaksha says to the cloud that he will recognize her who is his second life by her sweet rare speech and her loneliness in that city of happy lovers "sole like a lonely Chocrovaque with me her comrade far away", the simile has no

Page – 256

pathos to an English mind and even when explained would only seem "an artificiality common to the court-poetry of the Sanscrit age". I have therefore thought myself justified by the slightness of the allusion in translating "Sole like a widowed bird when all the nests Are making", which translates the idea & the emotion while suggesting a slightly different but related image.

I have indicated above the main principles by which I have guided myself in the task of translation. But there still remains the question, whether while preserving the ideals one may not still adhere more or less closely to the text. The answer to this is that such closeness is imperative, but it must be a closeness of word-value, not merely of word-meaning; into this word-value there enter the elements of association, sound and aesthetic beauty. If these are not translated, the word is not translated, however correct the rendering may be. For instance the words salila, āpaḥ and jala in Sanscrit all mean water, but if jala may be fairly represented by the common English word & the more poetic āpaḥ by "waters" or "ocean" according to the context, what will represent the beautiful suggestions of grace, brightness, softness & clearness which accompany salila? Here it is obvious that we have to seek refuge in sound suggestions & verse-subtleties to do what is not feasible by verbal rendering. Everything therefore depends on the skill & felicity of the translator and he must be judged rather by the accuracy with which he renders the emotional & aesthetic value of each expression than brought to a rigid [accounting] for each word in the original. Moreover the idiom of Sanscrit, especially of classical Sanscrit, is too far divided from the idiom of English. Literal translation from the Greek is possible though sometimes disastrous, but literal translation from the Sanscrit is impossible. There is indeed a school endowed with more valour than discretion and more metaphor than sense who condemn the dressing up of the Aryan beauty in English clothes and therefore demand that not only should the exact words be kept, but the exact idiom. For instance they would perpetrate the following: "Covering with lashes water-heavy from anguish, her eye gone to meet from former pleasantness the nectar-cool lattice-path-entered feet of

Page – 257

the moon and then at once turned away, like a land-lotus-plant on a cloudy day not awake, not sleeping." Now quite apart from the execrable English & the want of rhythm, the succession of the actions and the connexions of thought which are made admirably clear in the Sanscrit by the mere order of the words, is here entirely obscured & lost; moreover the poetic significance of the words prityā (pleasantness) and sābhre, implying here rain as well as cloud and the beautiful force of salilagurubhiḥ (water. heavy) are not even hinted at; while the meaning & application of the simile quite apparent in the original needs bringing out in the English. For the purpose of immediate comparison I give here my own version. "The moon-beams."⁴ This I maintain though not literal is almost as close and meets without overstepping all the requirements of good translation. For the better illustration of the method, I prefer however to quote a more typical stanza.

Śabdāyante madhuram anilaiḥ kīcakāḥ pūryamāṇāḥ,

Saṁsaktābhis tripuravijayo gīyate kinnarībhiḥ;

Nirhrādas te muraja iva cet kandareṣu dhvaniḥ syāt,

Saṅgītartho nanu paśupates tatra bhāvī samagraḥ. .

Rendered into [literal English] this is "The bamboos filling with the winds are noising sweetly, the Tripour-conquest is being sung by the glued-together Kinnaries; if thy thunder should be in the glens like the sound on a drum the material of the concert of the Beast-Lord is to be complete there, eh?" My own translation runs

Of Tripour slain in lovely dances joined

And linked troops the Oreads of the hill

Are singing and inspired with rushing wind

Sweet is the noise of bamboos fluting shrill;

Thou thundering in the mountain-glens with cry

Of drums shouldst the sublime orchestra fill.

4 Sri Aurobindo apparently intended to transcribe a passage from his now-lost translation here. -Ed.

Page – 258

"Of Tripour slain are singing" (tripuravijayo gīyate) requires little comment. The word tripura means the "three cities" [and] refers to the three material qualities of rajas, sattva & tamas, light, passion & darkness, which have to be slain by Sheva the emancipator before the soul can rejoin God; but there is no reference here to the theosophic basis of the legend, but purely to the legend itself, the conquest of the demon Tripoura by Mahadeva. There was no means of avoiding the mythological allusion & its ˙ unfamiliarity had simply to be accepted. Saṁsaktābhiḥ, meaning "linked close together in an uninterrupted chain", is here ` rendered by "joined in linked troops"; but this hardly satisfies the requirement of poetic translation, for the term suggests to an Indian a very common practice which does not, I think, exist in Europe, women taking each other's hands and dancing as they sing, generally in a circle; to express this in English, so as to create the same picture as the Sanscrit conveys, it was necessary to add "in lovely dances". The word Kinnaries presents a serious initial difficulty. The Purana mythologising partly from false etymology has turned these Kinnars into men & women with horse-faces & this description has been copied down into all Sanscrit dictionaries, but the Kinnaries of Valmekie had little resemblance with these Puranic grotesques; they are beings of superhuman beauty, unearthly sweetness of voice & wild freedom who seldom appear on the earth, their home is in the mountains & in the skies; he speaks of a young Kinnar snared & bound by men & the mother wailing over her offspring; and Kekayie lying on the ground in her passion of grief & anger is compared to a Kinnarie fallen from the skies. In all probability they were at first a fugitive image of the strange wild voices of the wind galloping and crying in the mountaintops. The idea of speed would then suggest the idea of galloping horses and by the usual principle of Puranic allegory, which was intellectual rather than artistic, the head, the most prominent & essential member of the human body, would be chosen as the seat of the symbol. Kalidasa had in this as in many other instances to take the Puranic allegorisation of the old poetic figure and new-subject it to the law of artistic beauty. In no case does he depart from the Puranic conception, but his

Page – 259

method is to suppress the ungainly elements of the idea, often preserving it only in an epithet, and bring into prominence all the elements of beauty. Here the horse-faces are entirely suppressed & the picture offered is that of women singing with unearthly voices on the mountain-tops. The use of the word Kinnarie here would have no poetic propriety; to the uninstructed it would mean nothing and to the instructed would suggest only the ungainly horse-face which Kalidasa here ignores and conflict with the idea of wild & divine melody which is emphasized. I have therefore translated "the Oreads of the hills"; these spirits of the mountains are the only image in English which can at all render the idea of beauty & vague strangeness here implied; at the same time I have used the apparently tautologous enlargement "of the hills" because it was necessary to give some idea of the distant, wild & mystic which the Greek Oreads does not in itself quite bring out. I have moreover transposed the two lines in translation for very obvious reasons. The first line demands still more careful translation. The word śabdāyante means literally "sound, make a noise," but unlike its English rendering it is a rare word used by Kalidasa for the sake of a certain effect of sound and a certain shade of signification; while therefore rendering by "noise" I have added the epithet "shrill" to bring it up to the required value. Again the force & sound of pūryamānāḥ cannot be rendered by its literal rendering "filled" and anila, one of the many beautiful & significant Sanscrit words for wind, -vāyu, anila, pavana, samīra, samīrana, vāta, prabhañjana, marut, sadāgati, -suggests powerfully the breath and flowing of wind & is in the Upanishad used as equivalent to prana, the breath or emotional soul; to render adequately the word "inspired" has been preferred to "filled" and the epithet "rushing" added to "wind" Kīcakāh pūryamānāh anilaih in the original suggests at once . the sound of the flute, because the flute is in India made of the . hollow bamboo & the shrillness of the word kīcakāh assists the suggestion; in English it was necessary to define the metaphor. The last two lines of the stanza have been rendered with great closeness except for the omission of nanu and the substitution of

Page – 260

the epithet "sublime" for paśupates. Nanu is a Sanscrit particle which sometimes asks a rhetorical question but more often suggests one answered; the delicate shades suggested by the Sanscrit particles cannot be represented in English or only by gross effects which would be intolerably excessive & rhetorical. The omission of Pasupati, the name of Sheva as the Lord of Wildlife, though not necessary, is I think justified. He is sufficiently suggested by the last stanza & to those who understand the allusion, by the reference to Tripoura; the object of suggesting the wild & sublime which is served in Sanscrit by introducing this name, is equally served in English by the general atmosphere of wild remoteness & the insertion of the epithet "sublime".

This analysis of a single stanza, ex uno disce omnes, will be enough to show the essential fidelity which underlies the apparent freedom of my translation. At the same time it would be disingenuous to deny that in at least a dozen places of each poem, -more perhaps in the longer ones -I have slipped into words & touches which have no justification in the original. This is a literary offence which is always condemnable and always committed. In mitigation of judgment I can only say that it has been done rarely and that the superfluous word or touch is never out of harmony with or unsuggested by the original; it has sprung out of the text and not been foisted upon it. I may instance the line⁵

The remarks I have made apply to all the translations but more especially to the Cloud Messenger. In the drama except in highly poetical passages I have more often than not sacrificed subtlety in order to preserve the directness & incisiveness of the Sanscrit, qualities of great importance to dramatic writing, and in the epic to the dread of diffuseness which would ruin the noble harmony of the original. But the Cloud Messenger demands rather than shuns the careful & subtle rendering of every effect of phrase, sound & association. The Meghaduta of Kalidasa is the most marvellously perfect descriptive and elegiac poem

5 Sri Aurobindo did not write the line he intended to "instance" in his manuscript. -Ed.

Page – 261

in the world's literature. Every possible beauty of phrase, every possible beauty of sound, every grace of literary association, every source of imaginative & sensuous beauty has been woven together into an harmony which is without rival & without fault; for amidst all its wealth of colour, delicacy & sweetness, there is not a word too much or too little, no false note, no excessive or defective touch; the colouring is just & subdued in its richness, the verse movement regular in its variety, the diction simple in its suggestiveness, the emotion convincing & fervent behind a certain high restraint, the imagery precise, right & helpful, not overdone as in the Raghuvansa & yet quite as full of beauty & power. The Shacountala and the Cloud Messenger are the ne plus ultra of Hindu poetic art. Such a poem asks for & repays the utmost pains a translator can give it; it demands all the wealth of word & sound effect, all the power of literary beauty, of imaginative & sensuous charm he has the capacity to extract from the English language. At the same time its qualities of diction & verse cannot be rendered. The diffuseness of English will neither lend itself to the brief suggestiveness of the Sanscrit without being too high-strung, nervous & bare in its strength & so falsifying its flowing harmony & sweetness; nor to its easy harmony without losing close-knit precision & so falsifying its brevity, gravity & majesty. We must be content to lose something in order that we may not lose all.

The prose of Kalidasa's dialogue is the most unpretentious & admirable prose in Sanscrit literature; it is perfectly simple, easy in pitch & natural in tone with a shining, smiling, rippling lucidity, a soft, carolling gait like a little girl running along in a meadow & smiling back at you as she goes. There is the true image of it; a quiet English meadow with wild flowers on a bright summer morning, breezes abroad, the smell of hay in the neighbourhood, honeysuckle on the bank, hedges full of convolvuli or wild roses, a ditch on one side with cress & forget-me-nots & nothing pronounced or poignant except perhaps a stray whiff of meadowsweet from a distance. This admirable unobtrusive charm and just observed music (Coleridge) makes it run easily

Page – 262

into verse in English. In translating one has at first some vague idea of reproducing the form as well as the spirit of the Sanscrit, rendering verse stanza by verse stanza & prose movement by prose movement. But it will soon be discovered that except in the talk of the buffoon & not always then Kalidasa's prose never evokes its just echo, never finds its answering pitch, tone or quality in English prose. The impression it creates is in no way different from Shakespeare's verse taken anywhere at its easiest & sweetest

Your lord does know my mind: I cannot love him:

Yet I suppose him virtuous, know him noble,

Of great estate, of fresh and stainless youth;

In voices well divulged, free, learned and valiant;

And in dimension in the shape of nature

A gracious person; but yet I cannot love him.

He might have took his answer long ago.

Or again still more close in its subtle & telling simplicity

Ol. What is your parentage?

Vi. Above my fortunes, yet my state is well.

I am a gentleman.

Ol. Get you to your lord;

I cannot love him: let him send no more;

Unless perchance you come to me again

To tell me how he takes it.

There is absolutely no difference between this & the prose of Kalidasa, since even the absence of metre is compensated by the natural majesty, grace & rhythmic euphony of the Sanscrit language & the sweet seriousness & lucid effectiveness it naturally wears when it is not tortured for effects.

Page – 263