Mahabharata Vol. 2 (Penguin Translated Texts) Read online

  The Mahabharata

  Volume 2

  (Sections 16 to 32)

  Translated By

  BIBEK DEBROY

  PENGUIN BOOKS

  Contents

  About the Translator

  Dedication

  Family Tree

  Map of Bharatavarsha

  Introduction

  SECTION SIXTEEN

  ARJUNA-VANAVASA PARVA

  Arjuna’s sojourn in the forest; begins with the reasons for his banishment

  SECTION SEVENTEEN

  SUBHADRA-HARANA PARVA

  Subhadra’s abduction

  SECTION EIGHTEEN

  HARANA HARIKA PARVA

  The giving of gifts that followed Subhadra’s abduction

  SECTION NINETEEN

  KHANDAVA-DAHA PARVA

  The burning of the Khandava forest by Arjuna and Krishna

  SABHA PARVA

  SECTION TWENTY

  SABHA PARVA

  The building of the assembly hall

  SECTION TWENTY-ONE

  MANTRA PARVA

  Consultation prior to the royal sacrifice

  SECTION TWENTY-TWO

  JARASANDHA-VADHA PARVA

  The killing of Jarasandha

  SECTION TWENTY-THREE

  DIGVIJAYA PARVA

  The conquest of the world by the Pandavas

  SECTION TWENTY-FOUR

  RAJASUYA PARVA

  The royal sacrifice

  SECTION TWENTY-FIVE

  ARGHABHIHARANA PARVA

  The offering of a gift and the consequent dispute

  SECTION TWENTY-SIX

  SHISHUPALA-VADHA PARVA

  The killing of Shishupala

  SECTION TWENTY-SEVEN

  DYUTA PARVA

  The gambling with dice

  SECTION TWENTY-EIGHT

  ANUDYUTA PARVA

  The aftermath of the gambling match

  ARANYAKA PARVA

  SECTION TWENTY-NINE

  ARANYAKA PARVA

  Pandavas’ sojourn in the forest

  SECTION THIRTY

  KIRMIRA-VADHA PARVA

  The killing of Kirmira

  SECTION THIRTY-ONE

  KAIRATA PARVA

  The story of Shiva assuming the role of a mountain-dweller to fight with Arjuna

  SECTION THIRTY-TWO

  INDRALOKABHIGAMANA PARVA

  Arjuna’s visit to Indra’s world and the Nala–Damayanti story

  Acknowledgements

  Copyright Page

  About the Translator

  Bibek Debroy is an economist and is Research Professor (Centre of Policy Research) and Contributing Editor (Indian Express group). He has worked in universities, research institutes, industry and for the government. He has published books, papers and popular articles in economics. But he has also published in Indology and translated (into English) the Vedas, the Puranas, the Upanishads and the Gita (Penguin India, 2005). His book Sarama and her Children: The Dog in Indian Myth (Penguin India, 2008) splices his interest in Hinduism with his love for dogs. He is currently translating the remaining volumes of the unabridged Mahabharata.

  For Suparna

  Ardha bhāryā manuṣyasya bhāryā śreṣṭhatamaḥ sakhā

  Bhāryā mulam trivargasya bhāryā mitram mariṣyataḥ

  Mahabharata (1/68/40)

  Family Tree

  Bharata/Puru Lineage

  Map of Bharatavarsha

  Bharatavarsha (sixth century BCE)

  Introduction

  The Hindu tradition has an amazingly large corpus of religious texts, spanning Vedas, Vedanta (brahmanas,1 aranyakas,2 Upanishads,), Vedangas,3 smritis, Puranas, dharmashastras and itihasa. For most of these texts, especially if one excludes classical Sanskrit literature, we don’t quite know when they were composed and by whom, not that one is looking for single authors. Some of the minor Puranas (Upa Purana) are of later vintage. For instance, the Bhavishya Purana (which is often listed as a major Purana or Maha Purana) mentions Queen Victoria.

  In the listing of the corpus above figures itihasa, translated into English as history. History doesn’t entirely capture the nuance of itihasa, which is better translated as ‘this is indeed what happened’. Itihasa isn’t myth or fiction. It is a chronicle of what happened; it is fact. Or so runs the belief. And itihasa consists of India’s two major epics, the Ramayana and the Mahabharata. The former is believed to have been composed as poetry and the latter as prose. This isn’t quite correct. The Ramayana has segments in prose and the Mahabharata has segments in poetry. Itihasa doesn’t quite belong to the category of religious texts in a way that the Vedas and Vedanta are religious. However, the dividing line between what is religious and what is not is fuzzy. After all, itihasa is also about attaining the objectives of dharma,4 artha,5 kama6 and moksha7 and the Mahabharata includes Hinduism’s most important spiritual text—the Bhagavad Gita.

  The epics are not part of the shruti tradition. That tradition is like revelation, without any composer. The epics are part of the smriti tradition. At the time they were composed, there was no question of texts being written down. They were recited, heard, memorized and passed down through the generations. But the smriti tradition had composers. The Ramayana was composed by Valmiki, regarded as the first poet or kavi. The word kavi has a secondary meaning as poet or rhymer. The primary meaning of kavi is someone who is wise. And in that sense, the composer of the Mahabharata was no less wise. This was Vedavyasa or Vyasadeva. He was so named because he classified (vyasa) the Vedas. Vedavyasa or Vyasadeva isn’t a proper name. It is a title. Once in a while, in accordance with the needs of the era, the Vedas need to be classified. Each such person obtains the title and there have been twenty-eight Vyasadevas so far.

  At one level, the question about who composed the Mahabharata is pointless. According to popular belief and according to what the Mahabharata itself states, it was composed by Krishna Dvaipayana Vedavyasa (Vyasadeva). But the text was not composed and cast in stone at a single point in time. Multiple authors kept adding layers and embellishing it. Sections just kept getting added and it is no one’s suggestion that Krishna Dvaipayana Vedavyasa composed the text of the Mahabharata as it stands today.

  Consequently, the Mahabharata is far more unstructured than the Ramayana. The major sections of the Ramayana are known as kandas and one meaning of the word kanda is the stem or trunk of a tree, suggesting solidity. The major sections of the Mahabharata are known as parvas and while one meaning of the word parva is limb or member or joint, in its nuance there is greater fluidity in the word parva than in kanda.

  The Vyasadeva we are concerned with had a proper name of Krishna Dvaipayana. He was born on an island (dvipa). That explains the Dvaipayana part of the name. He was dark. That explains the Krishna part of the name. (It wasn’t only the incarnation of Vishnu who had the name of Krishna.) Krishna Dvaipayana Vedavyasa was also related to the protagonists of the Mahabharata story. To go back to the origins, the Ramayana is about the solar dynasty, while the Mahabharata is about the lunar dynasty. As is to be expected, the lunar dynasty begins with Soma (the moon) and goes down through Pururava (who married the famous apsara Urvashi), Nahusha and Yayati. Yayati became old, but wasn’t ready to give up the pleasures of life. He asked his sons to temporarily loan him their youth. All but one refused. The ones who refused were cursed that they would never be kings, and this includes the Yadavas (descended from Yadu). The one who agreed was Puru and the lunar dynasty continued through him. Puru’s son Duhshanta was made famous by Kalidasa in the Duhshanta–Shakuntala story and their son was Bharata, contributing to the name of Bharatavarsha. B
harata’s grandson was Kuru. We often tend to think of the Kouravas as the evil protagonists in the Mahabharata story and the Pandavas as the good protagonists. Since Kuru was a common ancestor, the appellation Kourava applies equally to Yudhishthira and his brothers and Duryodhana and his brothers. Kuru’s grandson was Shantanu. Through Satyavati, Shantanu fathered Chitrangada and Vichitravirya. However, the sage Parashara had already fathered Krishna Dvaipayana through Satyavati. And Shantanu had already fathered Bhishma through Ganga. Dhritarasthra and Pandu were fathered on Vichitravirya’s wives by Krishna Dvaipayana.

  The story of the epic is also about these antecedents and consequents. The core Mahabharata story is known to every Indian and is normally understood as a dispute between the Kouravas (descended from Dhritarashtra) and the Pandavas (descended from Pandu). However, this is a distilled version, which really begins with Shantanu. The non-distilled version takes us to the roots of the genealogical tree and at several points along this tree we confront a problem with impotence/sterility/death, resulting in offspring through a surrogate father. Such sons were accepted in that day and age. Nor was this a lunar dynasty problem alone. In the Ramayana, Dasharatha of the solar dynasty also had an infertility problem, corrected through a sacrifice. To return to the genealogical tree, the Pandavas won the Kurukshetra war. However, their five sons through Droupadi were killed. So was Bhima’s son Ghatotkacha, fathered on Hidimba. As was Arjuna’s son Abhimanyu, fathered on Subhadra. Abhimanyu’s son Parikshit inherited the throne in Hastinapura, but was killed by a serpent. Parikshit’s son was Janamejaya.

  Krishna Dvaipayana Vedavyasa’s powers of composition were remarkable. Having classified the Vedas, he composed the Mahabharata in 100,000 shlokas or couplets. Today’s Mahabharata text doesn’t have that many shlokas, even if the Hari Vamsha (regarded as the epilogue to the Mahabharata) is included. One reaches around 90,000 shlokas. That too, is a gigantic number. (The Mahabharata is almost four times the size of the Ramayana and is longer than any other epic anywhere in the world.) For a count of 90,000 Sanskrit shlokas, we are talking about something in the neighbourhood of two million words. The text of the Mahabharata tells us that Krishna Dvaipayana finished this composition in three years. This doesn’t necessarily mean that he composed 90,000 shlokas. The text also tells us that there are three versions to the Mahabharata. The original version was called Jaya and had 8,800 shlokas. This was expanded to 24,000 shlokas and called Bharata. Finally, it was expanded to 90,000 (or 100,000) shlokas and called Mahabharata.

  Krishna Dvaipayana didn’t rest even after that. He composed the eighteen Maha Puranas, adding another 400,000 shlokas. Having composed the Mahabharata, he taught it to his disciple Vaishampayana. When Parikshit was killed by a serpent, Janamejaya organized a snake-sacrifice to destroy the serpents. With all the sages assembled there, Vaishampayana turned up and the assembled sages wanted to know the story of the Mahabharata, as composed by Krishna Dvaipayana. Janamejaya also wanted to know why Parikshit had been killed by the serpent. That’s the background against which the epic is recited. However, there is another round of recounting too. Much later, the sages assembled for a sacrifice in Naimisharanya and asked Lomaharshana (alternatively, Romaharshana) to recite what he had heard at Janamejaya’s snake-sacrifice. Lomaharshana was a suta, the sutas being charioteers and bards or raconteurs. As the son of a suta, Lomaharshana is also referred to as Souti. But Souti or Lomaharshana aren’t quite his proper names. His proper name is Ugrashrava. Souti refers to his birth. He owes the name Lomaharshana to the fact that the body-hair (loma or roma) stood up (harshana) on hearing his tales. Within the text therefore, two people are telling the tale. Sometimes it is Vaishampayana and sometimes it is Lomaharshana. Incidentally, the stories of the Puranas are recounted by Lomaharshana, without Vaishampayana intruding. Having composed the Puranas, Krishna Dvaipayana taught them to his disciple Lomaharshana. For what it is worth, there are scholars who have used statistical tests to try and identify the multiple authors of the Mahabharata.

  As we are certain there were multiple authors rather than a single one, the question of when the Mahabharata was composed is somewhat pointless. It wasn’t composed on a single date. It was composed over a span of more than 1000 years, perhaps between 800 BCE and 400 ACE. It is impossible to be more accurate than that. There is a difference between dating the composition and dating the incidents, such as the date of the Kurukshetra war. Dating the incidents is both subjective and controversial and irrelevant for the purposes of this translation. A timeline of 1000 years isn’t short. But even then, the size of the corpus is nothing short of amazing.

  Familiarity with Sanskrit is dying out. The first decades of the twenty-first century are quite unlike the first decades of the twentieth. Lamentation over what is inevitable serves no purpose. English is increasingly becoming the global language, courtesy colonies (North America, South Asia, East Asia, Australia, New Zealand, Africa) rather than the former colonizer. If familiarity with the corpus is not to die out, it needs to be accessible in English.

  There are many different versions or recensions of the Mahabharata. However, between 1919 and 1966, the Bhandarkar Oriental Research Institute (BORI) in Pune produced what has come to be known as the critical edition. This is an authenticated text produced by a board of scholars and seeks to eliminate later interpolations, unifying the text across the various regional versions. This is the text followed in this translation. One should also mention that the critical edition’s text is not invariably smooth. Sometimes, the transition from one shloka to another is abrupt, because the intervening shloka has been weeded out. With the intervening shloka included, a non-critical version of the text sometimes makes better sense. On a few occasions, I have had the temerity to point this out in the notes which I have included in my translation.

  It took a long time for this critical edition to be put together. The exercise began in 1919. Without the Hari Vamsha, the complete critical edition became available in 1966. And with the Hari Vamsha, the complete critical edition became available in 1970. Before this, there were regional variations in the text and the main versions were available from Bengal, Bombay and the south. However, now, one should stick to the critical edition, though there are occasional instances where there are reasons for dissatisfaction with what the scholars of the Bhandarkar Oriental Research Institute have accomplished. But in all fairness, there are two published versions of the critical edition. The first one has the bare bones of the critical edition’s text. The second has all the regional versions collated, with copious notes. The former is for the ordinary reader, assuming he/she knows Sanskrit. And the latter is for the scholar. Consequently, some popular beliefs no longer find a place in the critical edition’s text. For example, it is believed that Vedavyasa dictated the text to Ganesha, who wrote it down. But Ganesha had a condition before accepting. Vedavyasa would have to dictate continuously, without stopping. Vedavyasa threw in a counter-condition. Ganesha would have to understand each couplet before he wrote it down. To flummox Ganesha and give himself time to think, Vedavyasa threw in some cryptic verses. This attractive anecdote has been excised from the critical edition’s text. Barring material that is completely religious (specific hymns or the Bhagavad Gita), the Sanskrit text is reasonably easy to understand. Oddly, I have had the most difficulty with things that Vidura has sometimes said. Arya has today come to connote ethnicity. Originally, it meant language. That is, those who spoke Sanskrit were Aryas. Those who did not speak Sanskrit were mlecchas. Vidura is supposed to have been skilled in the mlechha language. Is that the reason why some of Vidura’s statements seem obscure? In similar vein, in popular renderings, when Droupadi is being disrobed, she prays to Krishna. Krishna provides the never-ending stream of garments that stump Duhshasana. The critical edition has excised the prayer to Krishna. The never-ending stream of garments is given as an extraordinary event. However, there is no intervention from Krishna.

  How is the Mahabharata classified? The core compone
nt is the couplet or shloka. Several such shlokas form a chapter or adhyaya. Several adhyayas form a parva. Most people probably think that the Mahabharata has eighteen parvas. This is true, but there is another 100-parva classification that is indicated in the text itself. That is, the adhyayas can be classified either according to eighteen parvas or according to 100 parvas. The table (given on pp. xxiii–xxvi), based on the critical edition, should make this clear. As the table shows, the present critical edition only has ninety-eight parvas of the 100-parva classification, though the 100 parvas are named in the text.

  Thus, interpreted in terms of BORI’s critical edition, the Mahabharata no longer possesses the 100,000 shlokas it is supposed to have. The figure is a little short of 75,000 (73,787 to be precise). Should the Hari Vamsha be included in a translation of the Mahabharata? It doesn’t quite belong. Yet, it is described as a khila or supplement to the Mahabharata and BORI includes it as part of the critical edition, though in a separate volume. Hence, I have included the Hari Vamsha in this translation as well. With the Hari Vamsha, the number of shlokas increases to a shade less than 80,000 (79,860 to be precise). However, in some of the regional versions the text of the Mahabharata proper is closer to 85,000 shlokas and with the Hari Vamsha included, one approaches 95,000, though one doesn’t quite touch 100,000.

  Why should there be another translation of the Mahabharata? Surely, it must have been translated innumerable times. Contrary to popular impression, unabridged translations of the Mahabharata in English are extremely rare. One should not confuse abridged translations with unabridged versions. There are only five unabridged translations—by Kisori Mohan Ganguly (1883–96), by Manmatha Nath Dutt (1895–1905), by the University of Chicago and J.A.B. van Buitenen (1973 onwards), by P. Lal and Writers Workshop (2005 onwards) and the Clay Sanskrit Library edition (2005 onwards). Of these, P. Lal is more a poetic trans-creation than a translation. The Clay Sanskrit Library edition is not based on the critical edition, deliberately so. In the days of Ganguly and Dutt, the critical edition didn’t exist. The language in these two versions is now archaic and there are some shlokas that these two translators decided not to include, believing them to be untranslatable in that day and age. Almost three decades later, the Chicago version is still not complete, and the Clay edition, not being translated in sequence, is still in progress. However, the primary reason for venturing into yet another translation is not just the vacuum that exists, but also reason for dissatisfaction with other attempts. Stated more explicitly, this translation, I believe, is better and more authentic—but I leave it to the reader to be the final judge. (While translating 80,000 shlokas is a hazardous venture, since Ganguly, Dutt and Lal are Bengalis, surely a fourth Bengali must also be preeminently qualified to embark on this venture!)