The Mahabharata 
translated by John Smith.
Penguin, 834 pp., £16.99, May 2009, 978 0 14 044681 4
Show More
Show More

Many people in India believe that, because the Mahabharata – the ancient epic poem, in Sanskrit, about a disastrous fratricidal war – is such a tragic, violent book, it is dangerous to keep the whole text in your house; most people who have it stow one part of it somewhere else, just to be on the safe side. The Mahabharata, in any case, takes up quite a lot of shelf space: it contains about 75,000 verses – sometimes rounded off to 100,000 – or three million words, some 15 times the combined length of the Hebrew Bible and the New Testament, or seven times the Iliad and the Odyssey combined; and a hundred times more interesting.

It has remained central to Hindu culture since it was first composed, during the period from before 300 BCE to after 300 CE. A.K. Ramanujan used to say that no Indian ever hears the Mahabharata for the first time. For centuries, Indians heard it in the form of public recitations, or performances of dramatised episodes, or in the explanations of scenes depicted in stone or paint on the sides of temples. More recently, they read it in India’s version of Classic Comics (the Amar Chitra Katha series) or saw it in the hugely successful televised version, 94 episodes, based largely on the comic book; the streets of India were empty (or as empty as anything ever is in India) during the broadcast hours on Sunday mornings, from 1988 to 1990. Or they saw various Bollywood versions, or the six-hour film version (1989) of Peter Brook’s nine-hour theatrical adaptation (1985). And now they can read Chindu Sreedharan’s ‘Epicretold’, posted on Twitter, one 140-character tweet at a time.

The bare bones of the central story are these: the five sons of King Pandu, called the Pandavas, were fathered by gods: Yudhishthira by Dharma (the moral law incarnate), Bhima by the Wind, Arjuna by Indra (king of the gods) and the twins by the Ashvins (the Dioskuroi). All five of them married Draupadi. When Yudhishthira lost the kingdom to his cousins in a game of dice, the Pandavas and Draupadi went into exile for 12 years, at the end of which – with the help of their cousin, the incarnate god Krishna, who befriended the Pandavas and whose counsel to Arjuna on the battlefield is the Bhagavad Gita – they regained their kingdom through a cataclysmic battle in which almost everyone on both sides was killed.

But the story of the Pyrrhic victory of the Pandava princes constitutes just a fifth of the epic, its skeleton. Much of the flesh is supplied by myths, folk tales, rituals, histories, theology, philosophy, science, legal debates, poetry and just about any other form of cultural knowledge that anyone wanted to preserve in ancient India. Most of these episodes, many of them about women, are fairly securely hooked to the fabric of the plot: a question about the ancestors of the Pandavas inspires the narrator to tell the story of the birth of their ancestor, Bharata, from Shakuntala, the innocent maiden whom King Dushyanta seduced and abandoned (a story that captivated Goethe); Yudhishthira is consoled, after his own gambling disaster, by the tale of Nala, whose compulsive gambling lost him his kingdom and his wife Damayanti, until she managed to reunite them. Other stories are told as moral lessons to the human heroes and heroines, such as the tale of King Sibi, who chopped off his own flesh to feed a dove fleeing from a hawk (both birds turned out to be gods disguised in order to test him); and Savitri, whose steadfastness persuaded the god of death to spare her doomed husband. Philosophical and legal questions also arise out of the aporias of the plot and are answered in discourses that often go on for thousands of verses.

Above all, the Mahabharata is an exposition of dharma, including the proper conduct of a king, of a warrior, of an individual living in times of calamity, and of a person seeking to attain freedom from rebirth. Time and again, when a character finds that every available moral choice is the wrong choice, or when one of the good guys does something obviously very wrong, he will mutter, or be told, ‘Dharma is subtle (sukshma),’ thin and slippery as a fine silk sari, internally inconsistent as well as disguised, hidden, masked. People try again and again to do the right thing, and fail and fail, until they no longer know what the right thing is.

Whenever the Mahabharata is told or retold, the ethical and religious questions that it raises are given new, contemporary meanings. In 1989, the diplomat Shashi Tharoor retold the Mahabharata as The Great Indian Novel, in which the heroes are recast as thinly veiled forms of Gandhi, Nehru, Indira Gandhi and others. (The hero Karna, who, in the Sanskrit version, slices off the armour that grows on his body and fights against his brothers, appears as Jinnah, who, when he goes over from the Hindu to the Muslim side, seizes a knife and circumcises himself.)

In the Sanskrit Mahabharata, Ekalavya, a low-caste tribal, becomes such a superb archer that he threatens Prince Arjuna, famed as the greatest of all archers, who insists that Ekalavya cut off his right thumb; Ekalavya humbly obeys. Contemporary low-caste Dalits make Ekalavya do for them what in the Sanskrit text he didn’t do for himself: rebel. A poem about the movement to gain water rights for Dalits on the Ganges begins, ‘If you had kept your thumb/History would have happened somehow differently,’ and concludes: ‘My thumb/will never be broken.’ Such poems stand as a Dalit (or Dalit Buddhist) critique of Hinduism, rejecting both Ekalavya and Hinduism. There are Ekalavya education foundations in Ahmedabad and Hyderabad. The Ekalavya Ashram in Andhra Pradesh is a non-profit, tribal welfare facility established in 1990; run by people from the local business community, it serves underprivileged tribal people who can’t afford to educate their children. The televised Mahabharata made a big point of the Ekalavya story, playing it out at great length. The Lords Cricket Ground in India has been renamed the Ekalavya Krida Mandal.

Reinterpretations of this sort have been going on from the moment the Mahabharata began to be composed. It grew and changed in numerous parallel lines spread over the entire subcontinent, constantly retold and rewritten, both in Sanskrit and in vernacular dialects. It flickers back and forth between Sanskrit manuscripts and village storytellers, each adding new gemstones to the old mosaic. It was like an ancient Wikipedia, to which anyone who knew Sanskrit, or who knew someone who knew Sanskrit, could add a bit here, a bit there. But the powerful intertextuality of Hinduism ensured that anyone who added anything to the Mahabharata was well aware of the tradition behind it and fitted his or her own insight, or story, thoughtfully into the ongoing conversation. However diverse its sources, for several thousand years the tradition has regarded it as a conversation among people who knew one another’s views and argued with silent partners.

The curse of the Mahabharata has waylaid several scholars who aspired to English all of its 18 books. J.A.B. van Buitenen set out, in 1973, to translate the whole thing and died, in 1978, after completing barely a third of it. (A group of other scholars have continued the project, but it remains unfinished.) The Clay Sanskrit Library proposed to publish a translation of all the books (together with a number of other Sanskrit texts), but this valuable series has now been abruptly discontinued, after publishing (among a number of other volumes) only about half of the Mahabharata. Kisari Mohan Ganguli, the author of the best-known complete English translation, had a different sort of bad luck: he finished the translation (in 1896) but was cheated of the credit for it. Until quite recently, it was always called ‘the P.C. Roy translation’, attributed to the publisher, Pratap Chandra Roy (who was made a CBE), and Ganguli’s name was unknown. The curse is working (as Anna Russell used to say of Wagner’s Ring).

John Smith, earlier the author of a fine translation and study of the medieval Hindi Epic of Pabuji, is the latest to take up the challenge. He may escape the curse because his version is incomplete: though it spans the entire text, he has translated only 11 per cent of the Sanskrit, bridging the translated passages with detailed summaries. Half the book is translation, half summary. He covers the waterfront, however, in that you can find the whole Mahabharata there, in either condensed or fully expanded form, and the careful index helps you to locate whatever you are looking for. This alternation of condensed and expanded passages is true to the spirit of the original Sanskrit text, in which, time and again, a storyteller will casually remark something of the order of, ‘Oh, this was the pond where Shakuntala met Dushyanta,’ and his companion will say, ‘Now that you have told it in brief, please tell it in full,’ and the storyteller gladly obliges. Punning on the name of the mythical author of the text, Vyasa (which means ‘divider’), the text tells us: ‘After he had produced this great knowledge in detail, Vyasa made a summary of it, since the wise in this world want it both in its entirety (samasa) and in parts (vyasa).’

Smith has given it to us in parts, a manageable fragment, by far the best single-volume English translation extant. It is certainly the easiest to read, and the sparing use of notes (less than one to a page, generally cross-references) keeps it moving along, though occasionally necessitating a simplification. Flying in the face of the sexist old Italian adage that a translation, like a woman, cannot be both beautiful and faithful, this one is both meticulously accurate in its rendition of the Sanskrit and fluent in clear, graceful English sentences. The tone is appropriately formal and a bit old-fashioned, elegant rather than conversational, avoiding contractions (‘cannot’ instead of ‘can’t’) and retaining most of the many epithets (though generally using one basic term for each character even when the text uses several, a useful simplification and a way to avoid cumbersome notes). Wisely, Smith left in Sanskrit just a few central, untranslatable words (such as ‘dharma’, ‘brahman’ and ‘mantra’), which he defines in the glossary. In the awkward matter of ‘begats’ (‘awkward’ because English, unlike Sanskrit, lacks precise sexual terms that are neither obscene nor medical), he generally uses ‘lay with’, a bit fusty but better than, say, ‘had sex with’. (‘Lay with’ may, however, strike some younger readers as archaic to the point of incomprehension; some years ago when I wrote in a journal article that Hera suspected that Zeus ‘would lie with’ some woman he was in love with, a young editor corrected it to read ‘would lie to’ the woman.) Every sentence in Smith’s translation is a pleasure to read, and if you check it against the Sanskrit, it’s there.

This, then, is the book to read if you want to know what is in the Mahabharata without devoting several years to it. Smith expresses his admiration and gratitude for the accuracy of Ganguli’s translation, and tactfully omits to mention that, since Ganguli wrote in what passed for English among the 19th-century Bengali bhadralok, his Victorian translation is rather tedious to read nowadays. Van Buitenen, too, is heavy sledding, with his stilted prose and his idiosyncrasies (such as using ‘barons’ to designate members of the class of kings and warriors), though he also got most of it right, and his notes are terrific. But there is a price to pay for three decisions that Smith made: he translated the ‘critical edition’; he left out the poetry; and he left out most of the religion.

The soi-disant critical edition was created in Poona (now Pune) between 1933 and 1972 by a group of scholars labouring under the misguided assumption that you could organise Sanskrit editions as you could Greek editions, by collating a large number of manuscripts. There is a deeply patriarchal metaphor at the root of the whole idea of tracing text stemmata, as if mimicking the way that secondary stories grow out of the trunk of the main plot; the biblical genealogical tree of x begat y (who begat z …) ignores the mothers – the storytellers – who just tangle the branches (to continue in dryadic mode). For there is no trunk to this textual tree, no pristine Ur-Mahabharata for scholars to isolate from a mass of corrupted texts; if it is a tree at all, it is a banyan, which must have an original root but sends down so many subsequent roots from its branches (other variants) that one can no longer tell which was the original. There are hundreds of Mahabharatas, hundreds of different manuscripts and innumerable oral versions (one reason it is impossible to make an accurate calculation of the number of its verses), one just as ur as the next. The Pune edition is critical, all right, but in the sense that it left the patient (the Mahabharata) in critical condition.

There are several recensions of the Mahabharata, each preserved and cherished by a particular community. The critical edition, by contrast, is like Frankenstein’s monster, pieced together from various scraps of different bodies; its only community is that of the Pune scholars, the Frankensteins. Moreover, it left out a great deal of material that the Indian literary and religious traditions have continued to draw on, such as the passage in which Vyasa dictates the entire text to the elephant-headed god Ganesha. Ganesha made Vyasa promise not to keep him waiting, and Vyasa made Ganesha promise, in return, not to write down anything he did not understand; when Vyasa began to fall behind, he quickly threw in the ‘knots’ that Ganesha had to stop to untangle and that have troubled readers ever since. This passage also demonstrates the tradition’s subtle understanding of the problematic relationship between oral and written versions of the text.

When, therefore, Smith casually remarks, ‘The text translated here is, of course, that of the critical edition,’ there’s no ‘of course’ about it. Smith’s translation, like van Buitenen’s (he also used the critical edition), leaves out all that the Pune edition left out; he could at least have summarised some of the most important amputations, just as he summarises the bits of the critical edition that he does not translate.

Even within the critical edition, hard choices had to be made to cram the text into a single volume, even an 834-page volume. The literary embellishments went the way of all flesh. For instance: near the end of the story, the hero Arjuna sees that the great city of Dvaraka on the west coast of Gujarat has been deserted. Smith’s summary says, ‘He finds Dvaraka appearing like a widowed woman,’ but he doesn’t mention the longer extended simile that follows the simple image of the widow:

Arjuna saw Dvaraka as a river, with its water consisting of the people, horses for fish, chariots for boats, the sound of musical instruments and chariots for the sound of its waves, and the houses’ bathing ghats for crocodiles; with jewels for its clumps of aquatic moss, adamantine ramparts for its garlands, with the streaming traffic in its main streets for its whirlpools of water, and courtyards for its still pools, like the horrible river of hell whose sharks are the nooses of Time and Death. Bereft of its heroes, devoid of joy, stripped of its glory, it was like a lotus pond in winter.

These images of a city made of water prepare us for the final scene, in which the city is actually flooded by the ocean, to remain submerged for ever, like Atlantis. The poetry is, in a very real sense, part of the plot.

And so is religion. Most scholars of the Mahabharata can be divided into two groups, those who care primarily about the political plot and those who care primarily about the religion. The first group regards the myths and the long sermons on dharma as digressions from the story of the human heroes, an intrusive padding awkwardly stuck around a zippy epic plot. But this is tantamount to saying that the arias in a Verdi opera are unwelcome interruptions of the libretto; Hindus regard the myths and the philosophical arguments, like the arias, as the centrepiece, for which the narration (the recitative) is merely the frame. The extreme version of the cut-to-the-chase (more precisely, cut-to-the-battle or cut-out-the-gods/myths) Mahabharata selection is the 1999 one-volume translation by Chakravarthi Narasimhan (who was under-secretary of the United Nations), which you can read from cover to cover without encountering any event that could not be reported, mutatis mutandis, in the Wall Street Journal. The historians of religion, on the other hand, generally zero in on the Bhagavad Gita (which many Hindus lift out of context and use as a self-contained sacred text); some also concentrate on the theology of Krishna and the other great god of the Mahabharata, Shiva, a darker and more violent god than Krishna.

Smith translates three of the 18 chapters of the Gita (summarising the rest), and many of the episodes in which Krishna appears (as a human more often than as a god), but he chooses not to translate many other passages full of interest for historians of religion, particularly passages dealing with Shiva, who is less central than Krishna to the plot of the Mahabharata but appears at several crucial moments and is absolutely central to its more tragic theology. The merely summarised passages include the story of the sacrifice of Daksha (a recurrent trope for the sacrifice of the human heroes in general), the battle between Arjuna and Shiva disguised as a low-caste hunter, and the hymn of the thousand names of Shiva (an important landmark in the development of sectarian Hinduism). Smith doesn’t translate a single syllable of books 12 and 13, though these raise issues that form the basis of many later discussions of Hindu law and philosophy. Nor does he translate the great stories that have continued to inspire Indian dramatists and poets through the ages, such as the tales of Shakuntala, Nala and Sibi (though he does translate the tale of Savitri). He merely summarises the story of Shikhandin’s magical cross-gendered birth and later gender transformation (at the hands of various gods and demigods), though he translates in full the passages in which Shikhandin is wounded in battle and dies (at the hands of his human opponents). By leaving out the poetry, the stories (often about and/or by women), the philosophy and the religion, Smith has cut away the soft, female parts of the book and left us primarily the hard, male, political and martial parts. The Hindus have a story about this: Bhringin, a devotee of Shiva, refused to honour Shiva’s wife, who therefore cursed him to be a skeleton, stripped of the flesh and blood that (in classical Indian embryology) is given by the mother and left only with the bones that are the father’s contribution.

But some such lacunae are inevitable in a massive undertaking such as this. If it is true that no Indian ever hears the Mahabharata for the first time, it is also true that none of us ever hears all of the Mahabharata, or hears the Mahabharata that anyone else hears; we all make our own selections, for warriors or for gods, for prose or for poetry. Just as the god Vishnu becomes incarnate in Krishna through what the Hindus call a ‘partial’ incarnation (while the full Vishnu remains intact in the world of the gods), so, too, any translation can only aspire to be a partial incarnation. There’s always a danger that a translator will abridge too far; Smith abridges just enough, perfectly capturing the Mahabharata that he most admires. Yet his version is sufficiently incomplete to allow him to live to a happy old age, and all of us to keep the volume safely in our homes.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences