Kr̥ṣṇa in the Harivaṁśa (Vol I): The Wonderful Play of a Cosmic Child

By André Couture

Written over a thirty-year period, the thirteen texts of this book — some of which have been updated, others translated from the original French — address various aspects of Krishna’s childhood in the Harivamsha. As a part of a continuous effort to better understand this oft-neglected complement to the Mahabharata, the present book demonstrates that these stories of Krishna’s childhood were carefully composed by brahmanas who knew fully well what they were doing.
During the ten or so years he spends as a herder in the forest surrounding Mathura, Krishna prepares himself to kill the evil king Kamsa: when packs of wolves spring from the hairs of his body, he manifests his destructive power; he appears as a true avatara when he dives into the Yamuna to subdue the snake Kaliya; he reveals himself as a new Brahma able to create a new world when he uplifts Mount Govardhana with which he has just identified himself, then sheltering cows and herders in his own body.
It is author’s contention throughout these chapters that these episodes cannot be dismissed as a hotchpotch of legends borrowed from the Abhiras or similar pastoral tribes. Neither does one do justice to the genre when one reinterprets the story symbolically, as if it were the product of an overactive imagination. Rejecting these positions, the author instead attempts to show here how these talented storytellers carefully crafted a narrative, often using material drawn from their own Vedic tradition, in order to address the new concerns of their audiences.

• Language: English
• Publisher: D.K. Printworld
• Release date: Mar 9, 2023
• ISBN: 9788124611920

Book preview

Front.jpg

Kr̥ṣṇa in the Harivaṁśa

Kr̥ṣṇa in the Harivaṁśa

Vol. 1

The Wonderful Play of a Cosmic Child

André Couture

Cataloging in Publication Data — DK

[Courtesy: D.K. Agencies (P) Ltd. ]

Couture, André, 1945- author.

Kr̥ṣṇa in the Harivaṁśa / André Couture.

volume 1 cm

Includes bibliographical references (pages ) and index.

Contents: vol. 1. The wonderful play of a cosmic child

ISBN 13: 9788124608241

. Mahābhārata. Harivaṁśa – Criticism, interpretation, etc.

. Krishna (Hindu deity) I. Title.

DDC 294.5923046 23

ISBN : 978-81-246-1192-0 (E-Book)

ISBN : 978-81-246-0824-1 (HB)

First published in India, 2015

© André Couture

All rights reserved. No part of this publication may be reproduced or transmitted, except brief quotations, in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior written permission of the copyright holders, indicated above, and the publishers.

Printed and published by:

D.K. Printworld (P) Ltd.

Regd. Office: Vedaśrī, F-395, Sudarshan Park

(Metro Station: Ramesh Nagar), New Delhi – 110 015

Phones: [phone], 2546 6019; Fax: [phone]

e-mail: [e-mail]

Website: www.dkprintworld.com

Contents

Volume 1

The Wonderful Play of a Cosmic Child

Abbreviations

Introduction

Part 1:The Harivaṁśa, Genre in Context

1.1 The Harivaṁśa: A Supplement to the Mahābhārata

1.2 Akrūra and the Bhāgavata Tradition according to the Harivaṁśa

1.3 The Harivaṁśa and the Notion of Purāṇa

1.4 Possibility of Assigning a Date for the Harivaṁśa

Part 2: Kr̥ṣṇa’s Birth in the Harivaṁśa

2.1 The Story of Saṁkarṣaṇa and Kr̥ṣṇa’s Births: A Drama Involving Embryos

2.2 Saṁkarṣaṇa and Kr̥ṣṇa’s Hair Births

2.3 The Problem of the Meaning of Yoganidrā’s Name

Part 3 Kr̥ṣṇa’s Marvellous Childhood in the Harivaṁśa

3.1 Cowherd Settlements and Forests in Three Ancient Versions of Kr̥ṣṇa’s Childhood

3.2 Kr̥ṣṇa’s Childhood according to the Harivaṁśa: Study of the Composition of the Narrative

3.3 Kubjā: The Hunchbacked Woman Straightened Up by Kr̥ṣṇa

3.4 Kr̥ṣṇa’s Strange Name of Dāmodara

3.5 The Winged Mountains: Variations on a Vedic Theme

Conclusion: Birds, Herders and Yogins in the Harivaṁśa

Bibliography

General Index

Index of Passages

Abbreviations

For an easy reading, only a few abbreviations have been adopted in this book.

Introduction

O

ver

a period of several decades, my research has been devoted primarily to the study of Kr̥ṣṇa and the Harivaṁśa (HV). The thirteen texts of this book, most of which were previously published either in French or English, form the first part of a collection in which the fruit of those years of study will be gathered together.

More than forty years ago, as I was looking for an avenue for future research, my Sanskrit teacher, Prof. Anne-Marie Esnoul, encouraged me to attend the seminar which Prof. Olivier Lacombe was scheduled to offer that following year (1972-73) at the University of Paris-Sorbonne on the Tenth Book of the Bhāgavata Purāṇa (BhP). In preparation for the seminar, I spent my summer vacation reading and re-reading Hauvette-Besnault’s French translation of Kr̥ṣṇa’s childhood. As a neophyte, I was unable to make much sense of its many philosophical digressions and descriptions of wonderful deeds. In my attempts to become better acquainted with Kr̥ṣṇa’s adventures, I eventually discovered Wilson’s old translation of the Fifth Book of the Viṣṇu Purāṇa (ViP), which provides a more straightforward version of Kr̥ṣṇa’s biography; one probably written a few centuries earlier than the BhP. Finally, I came across the first account of Kr̥ṣṇa’s life among the herders surrounding the city of Mathurā in the HV, a book considered to be a supplement to the Mahābhārata (Mbh). Given the secondary status accorded to the HV and the fact that A. Langlois’ early French translation (1834-35) of the text was more a paraphrase of the Sanskrit than a true rendering, it took me some time to realize how important this text was to an understanding of Kr̥ṣna’s childhood.

A few years after my Ph.D. dissertation (1977), the objective of which was to compare the three oldest Sanskrit texts dealing with Kr̥ṣna’s childhood, I decided to translate that part of the HV which deals with this period of Kr̥ṣṇa’s life, i.e. chapters 30 to 78 of the Critical Edition (Couture 1991). As the translation progressed, new hypotheses and insights gradually emerged which I shared at international conferences and in several publications.

The first four texts included in the present book deal with general issues that any serious reader of the HV will eventually have to address. The fourth of these texts, having reviewed the standard scholarship on the thorny question of the date of composition of the HV, proposes a down-to-earth approach (Possibility of Assigning a Date to the Harivaṁśa?). In its present form, this survey is a completely new essay, even if several pages overlap texts which I had already published on the subject or discuss hypotheses which I had previously formulated. The three other texts examine specific words which have become clichés in a large part of current scholarship, so much so that a fresh and completely new examination appeared to me to be the most promising and realistic way of arriving at an understanding of their underlying meaning. Khila is the first of these words. In spite of the common and apparently obvious translation, the term does not correspond to the English notion of appendix, but refers rather to a supplement received from elsewhere for the sake of completion, fulfilment or elucidation (The Harivaṁśa: A Supplement to the Mahābhārata, 1996). The next word is bhāgavata, a term encountered several times in the episode of Akrūra’s ablutions in the Yamunā waters. In a text entitled Akrūra and the Bhāgavata Tradition in the Harivaṁśa (1986, revised), I argue that, for the first time in history, the HV provides a real context for the use of this epithet. This socio-religious context, that should not be underestimated, clearly demonstrates that the term is used to designate those who accept Viṣṇu as the supreme deity, thus transposing their loyalty to their king (rājabhakti) into total subservience to the Bhagavat. The third word, purāṇa, is regularly used in the HV (The Harivaṁśa and the Notion of Purāṇa, 2003). Whereas important earlier research left the impression that the meaning of the word had already been clarified, a close reading of the HV makes it clear that the model of a Purāṇa as having five main components (pañcalakṣaṇa) does not make sense in that context. When the Purāṇas are examined with the sole specification that they should be considered to be a Vedic complement, it becomes clear that they constitute a set of legitimations developed over time in response to the challenges of an emerging complex urban civilization.

The second section of the book focuses on the births of Kr̥ṣṇa and his elder brother, Saṁkarṣaṇa. The first two texts concern the ways in which both of them are born, since, according to tradition, both brothers were born either from their natural parents, Vasudeva and Devakī (The Story of Saṁkarṣaṇa and Kr̥ṣṇa’s Births: A Drama Involving Embryos, 2009), or from a white and a black hair (Saṁkarṣaṇa and Kr̥ṣṇa’s ‘Hair’ Births, 2012). Here, these apparently disparate narratives are taken seriously, contextualized and appropriately analysed. Instead of being viewed as remnants of primitive tribes which defy interpretation, these stories reveal a rather precise structure and appear to be strangely connected to certain commentaries found in the Brāhmaṇas. The third text of this section (The Problem of the Meaning of Yoganidrā’s Name, 1999) deals with the compound yoganidrā used as a designation for Goddess Ekānaṁśā, the name given to Kr̥ṣṇa’s elder sister who was born in King Kaṁsa’s cowherd settlement to protect Kr̥ṣna and Saṁkarṣaṇa’s births. Instead of connecting yoganidrā to a particular state of meditation, I associate the word with that kind of fake sleep (yoganidrā) which is well attested to in classical texts. Even a king protected by his best fighters needs such a sleep, if he is to dominate the world.

The third section of the book deals with the period of childhood itself. This set of stories is very complex, and no one can pretend to encompass all of them. In choosing a select number of episodes and problems, I hoped to indirectly address some of the most important topics. The first breakthrough made in terms of an accurate reading of these ancient stories of childhood came when, in 1974, I realized that, in spite of the current translations and recent sectarian biases, it was no longer possible to defend the position that the Sanskrit word vraja refers to the region of Braj, or that gokula and vr̥ndāvana refer to the villages of Gokul and Brindaban. After a complete survey of the HV, ViP and BhP, I then decided to add a completely new chapter to my 1977 dissertation, a chapter that was published as Cowherd-Settlements and Forests in Three Ancient Versions of Kr̥ṣṇa’s Childhood in the Journal asiatique in 1982. The second breakthrough occurred when I began to examine the sequential arrangement of the various episodes of Kr̥ṣṇa’s childhood while translating the HV. The summary of these investigations gave rise to the second text of this section, Kr̥ṣṇa’s Childhood according to the Harivaṁśa: Study of the Composition of the Narrative (1989). I had realized that it was necessary to go beyond the distinction between royal and pastoral episodes emphasized by Charlotte Vaudeville, and came to the conclusion that the key episodes of Kr̥ṣṇa’s childhood — Kr̥ṣṇa’s transformation into packs of wolves, the submission of the Snake Kāliya and the uplifting of Mount Govardhana — should be linked to the figure of the supreme deity who destroys, maintains and recreates the whole world. No doubt, the young Kr̥ṣna who used to run in the forests surrounding Mathurā was the same unique god whom the sage Mārkaṇḍeya watched lying on the back on a banyan leaf in the middle of the cosmic ocean.

The remaining texts address a number of other problems. In the more recent article,Kubjā, the Hunchbacked Woman Uplifted by Kr̥ṣṇa (2011), I developed an observation made during the 1970s, that there is an underlying similarity between the way Kr̥ṣṇa heals the hunchback and the way he uplifts Mount Govardhana. As this long study evolved, other important feminine figures such as Pūtanā and Śūrpaṇakhā became sources of comparison. It may also be said in passing that, considered against the backdrop of this more recent research, my 1999 article, Kr̥ṣṇa’s Strange Name of Dāmodara (Brahmavidyā, Adyar), now appears far less speculative. The entire context of this episode, it seems to me, favours an interpretation of the mischievous child as a reference to the supreme Puruṣa, who grinds and devours the whole world (the mortar), and who stays in close relationship with the Snake which represents the Remainder of the world (the rope), even as he laughs in the middle of the dharmic and adharmic forces present in the universe (the arjuna trees).

I conclude this first set of texts with an article entitled Birds, Herders and Yogins in the Harivaṁśa. In this essay, I advance the novel idea that Kr̥ṣṇa’s childhood in the HV intermingles seamlessly winged mountains and wild geese, simple cowherds and detached yogins. Written in 2006, as part of a Festschrift in honour of Prof. Pierre-Sylvain Filliozat, neither this article nor the Festschrift was ever published. I include it here as a token of gratitude for this great scholar’s indefectible friendship and valuable help.

The articles in this collection which were published between 1995 and 2008 were funded by the Social Sciences and Humanities Research Council of Canada. I am indebted to those scholars who commented on drafts of the various chapters or debated this or that thesis defended in this book with me. At the risk of omitting other important contributions, I would particularly like to mention Christopher Austin, Madeleine Biardeau, Gérard Colas, Anne-Marie Esnoul, Pierre-Sylvain Filliozat, Phyllis Granoff, Marcelle Saindon, Charlotte Schmid and Ganesh Umakanth Thite. My sincere gratitude to all of these scholars and to all those who, knowingly or unknowingly, encouraged me to continue my research on the HV. Since English is not my mother tongue and since many of these articles were first published in French, I would also like to thank Robert Hurley, Christopher Austin and Élyssa Marcoux-Bissoondath for the linguistic revision of this book.

January 2015

André Couture

Part 1

The Harivaṁśa, Genre in Context

1.1

The Harivaṁśa

A Supplement to the Mahābhārata

¹

When

the Indian tradition speaks of the Mbh as a huge composition of about 100,000 stanzas (śatasāhasrī saṁhitā) in round figures, the HV occupies the last 16,000 and is divided into three parvans: the Harivaṁśa-Parvan (55 adhyāyas), the Viṣṇu-Parvan (128 adhyāyas) and the Bhaviṣya-Parvan (135 adhyāyas).² In the Critical Edition of P.L. Vaidya, this large book has been cut down to 118 adhyāyas and 6,073 ślokas. The new recension of the HV is supposed to present the oldest text available through a critical comparison of four printed editions and thirty-six manuscripts, written in eight different scripts.³ Vaidya insists that this new version is of course not the original text, which would be shorter still. Even if it is occasionally called the nineteenth book of the Mbh, the HV is more accurately referred to as a khila to the Mbh. But before moving on to the significance of the HV as a khila, it may be useful to start with an overview of the book.

The Harivaṁśa at a Glance

Even in the smaller size of the Critical Edition, the HV contains an array of themes which focus on Viṣṇu’s great manifestation (prādurbhāva) as Kr̥ṣṇa in the city of Mathurā. The story fits exactly into the same narrative pattern as the Mbh. Once more, at Śaunaka’s request, Ugrasravas repeats the stories told in former times by the brāhmaṇa Vaiśaṁpāyana at the request of King Janamejaya. This time, Vaiśaṁpāyana starts his narration with the creation of beings at the beginning of the kalpa, and proceeds with the story of King Pr̥thu (chaps. 4-6) and the description of the various Manus and manvantaras. After a few chapters relating the rise of the solar dynasty, he continues with a whole section dealing with the Fathers or Pitr̥s (chaps. 11-19) and their capacity to grant yoga as a new means of liberation.⁴ The next chapters deal with the origin of King Soma and describe the lineages of the lunar dynasty up to Yayāti, the Yādavas and the Vr̥ṣṇis, including Vasudeva and his son Kr̥ṣṇa Vāsudeva. In chaps. 28-29, we read the episode of the gem syamantaka. This is followed by a summary of Viṣṇu’s manifestations (chaps. 30-31), a description of the battle waged because of Tārā, Br̥haspati’s wife, and the account of the killing of the giant Kālanemi (chaps. 32-38). Vaiśaṁpāyana carries on with a description of the brahmaloka (chaps. 39-40), a report of the discussion between the Earth, Brahmā and Viṣṇu, and Viṣṇu’s promise to send many deities including himself and the Goddess Yoganidrā to defeat Kālanemi’s rebirth as King Kaṁsa and his host of asuras (chaps. 41-47). Next comes a detailed account of the famous childhood of Kr̥ṣṇa and his elder brother Saṁkarṣaṇa in the forests surrounding Mathurā which ends with Kaṁsa’s death and Ugrasena’s consecration as the king of Mathurā (chaps. 48-78).⁵

The next stage in Kr̥ṣṇa’s life begins with his initiation into the gamut of the sciences of his day and the tuition fee he agrees to pay his guru (chap. 79). A few chapters summarize Kr̥ṣṇa’s repeated fights with Jarāsandha to avenge Kaṁsa’s death (chaps. 80-82). The elder brother Saṁkarṣaṇa goes back to the Vr̥ndāvana forest, meets the cowherds and drags the Yamunā River with his ploughshare (chap. 83). The giant Kālayavana finally shows up: he will be killed by Kr̥ṣṇa, but the Yādavas decide to migrate to the western coast where they lay the foundation of the city of Dvāravatī. Then, the following episodes take place: the kidnapping of Rukmiṇī and the killing of her brother Rukmin (chaps. 87-89), the celebration of Saṁkarṣaṇa’s greatness (chap. 90), Kr̥ṣṇa’s victory over Naraka, the son of the Earth, and the liberation of 16,000 women, the snatch of celestial tree pārijāta (chap. 91), the construction of Dvāravatī (chap. 93), the entry of the Yādavas into the new city and Nārada’s discourse before the Yādavas gathered in a marvellous new hall. Continuing his description of the Vr̥ṣṇi lineage, Vaiśaṁpāyana introduces the main wives of Kr̥ṣṇa (chap. 98), recounts the killing of the asura Śambara by Pradyumna, the son of Kr̥ṣṇa and Rukmiṇī (chap. 99), praises Kr̥ṣṇa Vāsudeva for all his deeds (chaps. 100-04) and explains the victory of Aniruddha, the son of Pradyumna, over Bāṇa the father of his beloved Uṣā (chaps. 105-13).

The last part of the book is told by Ugraśravas who focuses on King Janamejaya, the son of Parīkṣit, performing a horse sacrifice during which the God Indra misbehaved with his wife Vapuṣṭamā (chap. 114). The incident came to a peaceful end (chap. 118), but serves as a pretext for a long description of the evils afflicting the society of the Kaliyuga. Before the final benedictions, the Vulgate provides a description of the cosmic dissolution and of the next creation by Brahmā as a lotus. This section, the long description of Viṣṇu’s avatāras as Varāha, Narasiṁha and Vāmana, along with several other episodes are unanimously considered as late additions, and are thus relegated in the Critical Edition to the second volume of appendices.

Even if incomplete, this summary should be sufficient to make it clear that the HV deals with cosmological myths, lineages, Viṣṇu’s avatāras, Kr̥ṣṇa’s deeds and so on, in the usual manner of the Purāṇas. In fact, this final poem (mahākāvyam, HV 118.43) is also called a Purāṇa (Mbh 1.2.69: harivaṁśas tatas parva purāṇaṁ khilasaṁjñitam); and as P.L. Vaidya remarks, it is the first and the oldest Purāṇa that has come down to us.⁶ But this same quotation from the Mbh also points to the fact that the HV is traditionally labelled a khila, a term in which conceals a certain ambiguity, and this is a major obstacle to overcome for a proper reading of the text. This is precisely the problem I would like to address in this short article.

The Meaning of Khila

Considered as an Appendix

According to Monier-Williams’ Sanskrit–English Dictionary,⁷ khila means (1) (m.) a piece of waste of uncultivated land situated between cultivated fields, desert, bare soil; (2) (n.) a space not filled up, gap, that which serves to fill up a gap, supplement (of a book, etc.), additional hymn appended to the regular collection. Khila is also translated by appendix, as if the Sanskrit word matches perfectly the definition one finds in Webster’s New Dictionary of Synonyms, according to which the word appendix is used of appended material which contributes (as by way of illustration, amplification, or citation of documents) to the effectiveness of a treatment that is still relatively complete in itself.⁸ This definition of appendix corresponds to some aspects of the word khila, but unfortunately it also suggests that the material gathered at the end of the Mbh is optional; in other words it is not really necessary for an understanding of the book, or it is a late addition which was more or less casually added for a fuller treatment of a question. Considered as an appendix, the HV is also often treated as if it were parts of the Mbh which exceed its regular length, i.e. pages excluded from the standard work, because they are redundant, superfluous, or have no obvious use, in the same way as those pieces of land which lie outside the parcels under cultivation in the first definition of khila. In the final analysis, the word became identified with leftovers or scraps. A khila is bound to be perceived in this way when treated as an appendix. Either the book is not considered at all, or it becomes an extra-section at the end of the Mbh which nobody cares to read, apart from a few specialists. Under such circumstances, it is not surprising that the HV has suffered what Daniel Ingalls called a curious neglect.⁹ But is this ambiguous meaning of appendix the real meaning conveyed by the term khila when used by the Indian tradition to qualify the HV? It is my contention that this is not the case and I suggest a return to a more traditional and positive meaning of the term.

The Harivaṁśa as a

Khila to the Mahābhārata

Whatever the meaning of khila, Occidental scholarship has been quasi-unanimous in underrating the importance of HV in the understanding of the Mbh. In his studies on the Mbh, E.W. Hopkins granted little importance to the HV. He contended that this addition was not only later than the earlier Mbh, but must be considered as a pseudo-epic, very probably coloured by Christian influences. The real epic should have dealt chiefly with heroic tales, noble deeds, hunting and fighting. The only explanation of its actual state comes from the fact it has been interpolated and altered with irrelevant legends and teachings.

Again, the presence of a huge volume of extraneous additions, containing both legends and didactic stuff, now tagged on to the epic as its nineteenth book and recognized in the last part of the epic itself, is an object-lesson in dynamic expansion which in itself shows how the pseudo-epic may with perfect regard to historic probability be supposed to have been added to the epic proper.¹⁰

M. Winternitz treats the HV as a sort of final compilation without any importance.

It is above all no ‘poetry’, in no sense the work of any poet but a heaping up or very loosely arranging side by side of texts — legends, myths, hymns — which serve to glorify God Viṣṇu.¹¹

Against the general trend and speaking from a literary point of view, Ingalls is alone in arguing that the very nature of the Harivaṁśa . . . is misconceived in many of our standard histories of Sanskrit literature.¹² He is certainly right when he insists that the HV should be characterized as a mahākāvya. But in my opinion this point does not mean that the Purāṇic elements of the book are still clearly inessential, as he pretends.¹³ The HV can be a mahākāvya, a purāṇa and a khila to the Mbh at the same time.

P.L. Vaidya’s Critical Edition should have helped balance these evaluations of the HV as a khila. But in spite of the great usefulness of his work, something in his introduction to the first volume does not capture the real purpose of the HV as it was brought to us. Of course, the learned editor describes the HV as the Khila-Parvan of the Mahābhārata,¹⁴ or as a Khila or supplement to the Mahābhārata.¹⁵ The Harivaṁśa being itself a Khila, he adds, has ample scope for additions and repetitious matter of episodes not connected with the main narrative of the text.¹⁶ But precisely, Vaidya seems more interested in discovering later additions than in appreciating the character of the HV as a khila. It must be emphasized that the Critical Edition of the Mbh itself seems to speak of this last book as a set of appendices:

khileṣu harivaṁśaś ca bhaviṣyac ca prakīrtitam.

— Mbh 1.1.233¹⁷

Among the Appendices are mentioned The Genealogy of Hari and The Book of the Future.

According to the commentator Nīlakaṇṭha (ad Mbh 1.2.379), these two examples have been selected among other possibilities by the narrator mainly to demonstrate the great esteem in which these books were held. As a matter of fact, the number of these appendices has clearly little to do with their proper function as khilas.

Consistently using the singular to designate the HV, P.L. Vaidya takes for granted that the result of the Critical Edition must point to a single text dealing with the genealogy of Hari.

[A]ll Harivaṁśa is a Khila capable of admitting extra matter at any place.¹⁸

While one may defend this point, a slight but significant shift in the meaning of the word khila must be noted. The tradition says there are many khilas to the Mbh, whereas Vaidya’s critical study leads him, even when he speaks of the whole HV as a bundle of khilas,¹⁹ to distinguish between a main khila which is the genuine HV, and all the other khilas (or pariśiṣṭas) which have little or no organic connection with the main topic of the work.²⁰ In fact, P.L. Vaidya is less interested in elucidating the relation between the HV and the Mbh, than in showing the lack of connection between on the one hand the passages he rejected as star-passages or on the other hand appendices, and the text he reconstituted. First of all, Vaidya’s HV is conceived of as an independent poem which was later appended to the Mbh. The following quotation confirms Vaidya’s particular bias:

I should like to note here, he says, that the tendency of redactors is normally in the direction of additions rather than in the direction of omissions. The late Dr. V.S. Agrawala termed this process as revision or bringing a work up to date. Of course, such a revised work cannot be termed as original. I should like to elaborate this point a bit further with special reference to additions in my Critical Text of Harivaṁśa. I hold the view that the original text of Harivaṁśa must have been what has been given in this edition or a bit smaller. Each redactor went on putting into the original Harivaṁśa what he thought, could, or even could not go with, with the result that its volume went on increasing till it reached the volume of the Vulgate.²¹

There is much truth in this evaluation of the successive improvements to this book. Nevertheless, Vaidya considers the HV to be an independent work with its own addenda, whereas the Indian tradition would rather consider the HV as a set of addenda which completes the Mbh.

At this point, we come face to face with the problem of the exact meaning of a khila in classical literature. Khilas, pariśiṣṭas, pariśeṣas, śeṣas are continuously attested to since the R̥gveda, and particularly in the śrauta- and gr̥hya-sūtras. A a khila, the HV is certainly not an isolated case. On the contrary, it is just one more Indian work using this category to qualify its relationship to tradition. The best description I have read of this sort of text occurs in the introduction to a work by C.G. Kashikar.²² It is worth reading the author’s precise wording:

The Pariśiṣṭa literature or supplementary texts constitute an important section of ancient Sanskrit literature. Sometimes these supplements are independent works, sometimes they are integrated with the original texts in such a way that it becomes difficult to find out their real character.

After presenting this literature in all its diversity, Kashikar establishes the function of these pariśiṣṭas:

The main object of a Pariśiṣṭa is to complete a work which might otherwise remain incomplete in some respect or the other. The incompleteness may either be original or may be newly felt on account of the changed circumstances. So far as the Śrauta Sūtras are concerned, the matters dealt within the Pariśiṣṭas cannot necessarily be regarded as late, because the possibility of a particular topic not having been discussed in sufficient details in the main work cannot be ruled out. . . . But in many cases the Pariśiṣṭas aim at making additions to the existing works. The additions can be of two kinds: (i) those dilating upon the subject-matter in the main work, and (ii) those introducing altogether new matter. . . . The Pariśiṣṭas contain certain matters which are introduced anew in course of time under different circumstances.²³

From Kashikar’s description, it is obvious that, first of all, a khila has an inherent relationship to some other main work. It may be a late addition intended to deliver the keys necessary to make an old text readable in a later context. It may also be nearly as old as the main text (or even earlier), provided the khila is used to complement another text, by adding, for example, the mythological support needed for its full understanding. In other words, khilas or pariśiṣṭas correspond exactly to what Jaimini called śeṣa in the ritual context of the Mīmāṁsā-Sūtras. A śeṣa is what it is, Jaimini says, because it exists for the sake of something else (śeṣaḥ parārthatvāt).²⁴ The complement (śeṣa) exists as a reality dependent on something else which may be dubbed a śeṣin or a pradhāna and which needs a complement to exist fully. If what is true with the ritual can also be applied to the texts describing or interpreting these rituals, then why would this not also be applicable to the case of the Mbh?

In his commentary on the Mbh and HV, Nīlakaṇṭha (c. 1650-90) suggests that the HV belongs to the Viṣṇu Purāṇa (ViP) properly, and that this ViP is called a khila to the Mbh when it is read after it to complete its meaning (ākāṅkṣāvaśāt, ad Mbh 1.2.82). Even if one considers it implausible to identify the HV with the ViP as Nīlakaṇṭha suggests in his commentary, the definition he gives of a khila goes straight to the point. According to the Vedic use of the word, he says, a khila is a text which belongs to another school and is read elsewhere because of some special requirement:

śākhāntarasthaṁ śākhāntare yad āpekṣavaśāt

paṭhyate tat khilam iti vaidikī prasiddhiḥ. — ad Mbh 1.2.82

The main purpose of a khila is to add a complement judged necessary for the full understanding of a text, whatever its provenance. Nīlakaṇṭha’s commentary to HV 1.1.1 is even more explicit:

A khila is something added for a particular reason to a given Vedic corpus, though properly speaking, it does not belong there. Such is the case with the Śrīsūkta and the Medhāsūkta, which have been included in the compilation known as the R̥gveda, as a khila. In the same way there is a khila to the eighteen books of the Mbh. This is the Viṣṇupurāṇa, a Bhaviṣyapurāṇa, which is added to the Mbh as a khila called the HV. It is added there because it completes [upabr̥ṁhaṇārtham] the glorification of the deeds of Kr̥ṣṇa recounted at various points throughout in the epic and because it enhances detachment, which is useful to the goal of mokṣa. We know this is the case because we see the word Purāṇa applied to the HV in the phrase the Purāṇa, called a khila [purāṇaṁ khilasaṁjñitam]. Thus, just as the somayāga enjoined in the Veda of the adhvaryu is not complete [na pūryate] without hymns and praise songs belonging to other Vedas, so the meaning of the Mbh is not complete without the HV.

Nīlakaṇṭha’s gloss agrees totally with Kashikar’s view. The relation between the HV and the Mbh matches similar cases in the Vedic tradition exactly. As such, the notion of khila does not imply later or earlier dating. It is a supplement received from elsewhere for the sake of completion, fulfilment, elucidation.

Madeleine Biardeau has a few remarks on the relationship between the HV and the Mbh which support the present analysis. From the outset, Biardeau affirms that the insurmountable contradictions which lead many Indologists to split Kr̥ṣṇa into different characters are the result of scholarship which originated in the nineteenth century and which is prevalent even today. Even if she does not choose to base her studies on the Poona Critical Edition, she notices that the text reconstituted in this edition does not support such a fragmentation of the texts either. Biardeau remarks that the epic considers the HV as a supplement even though it is presented as a Purāṇa and coins the phrase texte para-épique [paraepic text] for it.²⁵ She rightly argues that the connection of Kr̥ṣṇa with Arjuna is explicitly presented at the end of the Govardhana episode and shows the close interrelations between the HV and the Mbh. Even if the Critical Edition does its best to weed out its Mbh of all allusions to the pastoral biography, Kr̥ṣṇa’s titles like Govinda, the disguise of Kr̥ṣṇa’s sister Subhadrā as a gopī at the moment of her marriage with Arjuna and the intervention of odd-looking warriors called gopa Nārāyaṇa clearly point to the complementarity of both texts.

Here, within the epic myth a call for a complement is evident, unless a complement already exists.²⁶

In fact, the relatively late date usually accepted for the HV seems to beg the question. It is based on an informal de facto argument which considers everything that exceeds the epic frame as late, including of course all the didactic parts of the text. If the Mbh is assumed to have been composed between the fourth century

bce

and the fourth century

ce

, the HV when considered as a set of appendices must necessarily have been composed towards the end of this period. This is often the unstated reason given for the obvious assignment of the HV to a late date (fourth or fifth century

ce

). Ingalls finishes his article, generously alluding to a "general consensus that the Harivaṁśa dates from between the birth of Christ and the third century

ce

. But he adds, Stylistically it represents a survival from a much earlier age".²⁷ Actually, the main obstacle for an early dating of the HV may just as well be attributed to an underlying but unstated understanding of a khila as a late and secondary addition. Thoughtless reactions of this kind must be gotten rid of before