Pramāṇa · Search · About · API

Studies in the Epics and Puranas c13: Chapter VI: Epic and Purāṇic Studies

Studies in the Epics and Purāṇas c13

Language
English

Contents

Chapter VI: Epic and Purāṇic Studies

c13.2

Since the year 1917, the foundation of the Bhandarkar Oriental Research Institute, Poona, which has been taken as the starting point of this survey, there has been quite an amount of valuable addition of books and articles in the field of Epic and Purāṇic Studies, in common with other branches of Indology. These are necessarily of unequal merit and importance. In the following pages, it is proposed to review important publications concerning the *Mahābhārata*, *Bhagavadgītā*, *Rāmāyaṇa* and Purāṇas, which appeared during the last thirty-seven years.¹

c13.3

¹ I must at the outset record my deepest obligations to my revered *guru*, the late Dr. V. S. Sukthankar, who not only placed his unique collection of articles and papers at my disposal, but also gave me valuable suggestions and notes, outlining the plan, especially regarding the Mbh studies. But for his help, it would not have been possible for me to complete the original paper within the limited time. I also thank my friend Dr. R. N. Dandekar for giving me an opportunity of surveying the Epic and Purāṇic material. I have taken this opportunity of further revising and adding to my original paper and bringing it up-to-date. For the last ten years, I have referred only to the most important publications.

1. The Mahābhārata

c13.s1.1

One of the main tasks set before it by the Bhandarkar Oriental Research Institute was the preparation of a critical edition of the *Mahābhārata*, and the publication of the different parvans of the Mbh constitutes the most epoch-making, stupendous and colossal work in India in the domain of Indological studies during the present century. The Mbh has always held an important position in the Indian Culture, and the need of a critical edition of the national epic was always felt and pressed by eminent Indologists like Winternitz and Macdonell.² Earlier studies of the *Mbh* were vitiated, being based on a defective text. In matters of grammar, metrics, history, religion and sociology, it was quite uncertain whether conclusions were based on old and genuine material of the epic or on mere variant readings of no great antiquity and authority. In the light of the critical edition, however, we can now point out the lapses of earlier scholars. That the statistics and conclusions of Hopkins regarding the epic metre will require to be revised is patent on the face of it, as some of the stanzas which form the basis of these conclusions have been found to be interpolations, and readings of many others in the vulgate have been changed; Edgerton has referred to some such instances.³ Sukthankar has pointed out the blunder of Oldenberg in regarding one passage as a survival from the oldest form of the epic, which MS evidence proves to belong to the latest interpolations in the epic.⁴ Edgerton refers to the waste of valuable time in interpreting defective readings of the vulgate text.⁵ It has been found that the vulgate modernised many of the archaic lines of the original.⁶ The large number of interpolations of chapters printed at the end of the critical text and of individual lines or stanzas printed in the footnotes, amply shows the amount of uncertain matter that may have crept into earlier studies and conclusions.

c13.s1.2

² Winternitz, *HIL*, I, pp. 467 ff; Macdonell, *HSL*, pp. 282 ff.

c13.s1.3

³ *JAOS*, 59, p. 165.

c13.s1.4

⁴ *Kane Festschrift*, p. 474 (n 7).

c13.s1.5

⁵ *JAOS*, 59, p. 368.

c13.s1.6

⁶ Sukthankar, *Epic Studies* (I), *JBBRAS*, IV, p. 161.

c13.s1.7

There has been a steady and persistent demand for a critical edition of the *Mbh* since 1897. The International Association of the Academies of Europe and America undertook the preparation of an international edition of the epic in 1904, and some preliminary work was done, but the first world war prevented any further progress and the scheme had to be abandoned. The Bhandarkar Oriental Research Institute formed an Editorial Board of eminent scholars for preparing the critical edition of the *Mbh* with Utgikar as the Editor. The late Sir R. G. Bhandarkar, the then seniormost Ācārya of Indology in Western India, inaugurated the work in April 1919 by writing the first *maṅgala śloka*: *Nārāyaṇaṃ namaskṛtya*, etc. The work proceeded on systematic and scientific lines. Utgikar pointed out, on a study of select chapters of the *Mbh*, that, contrary to the views held thereto the Devanāgarī recension also had a longer and a shorter text, and that an older text than Nīlakaṇṭha can be arrived at.⁷ The preliminary work consisted of the preparation of tables of available MSS and collation of the text of the Parvasaṃgraha and consideration of the Javanese version of the *Virāṭaparvan* (which proved that the southern recension was studied in Java).⁸ Winternitz and Thomas approved of the scheme, and the tentative edition of the *Virāṭaparvan* based on sixteen MSS was published by the Institute in 1923. It contained 2050 stanzas, divided into 67 chapters. Orientalists the world over hailed the work with approval, and Winternitz in the course of his review offered some useful suggestions.⁹ Later researches have proved that Utgikar placed too much reliance on the Parvasaṃgraha data, believing it to be the same in both the Northern and Southern recensions, and relied on the Kumbhakonam Edition (which is really a curious blend of the N. and S. recensions) as the S. recension.

c13.s1.8

⁷ ABORI, I, pp. 145-155; II, pp. 73-77.

c13.s1.9

⁸ Utgikar, ABORI, II, pp. 155-188.

c13.s1.10

⁹ ABORI, V, pp. 19-30.

c13.s1.11

Utgikar, however, resigned in 1924, and next year Sukthankar took charge as the General Editor.

c13.s1.12

The first fascicule of the critical edition of the Ādiparvan appeared in 1927, and subsequent fascicules have followed at regular intervals. The complete Ādiparvan was published in 1933 along with the Prolegomena, which, though chiefly concerned with the Ādiparvan, covers practically all aspects of the Mbh text-criticism, and is a brilliant exposition of the entire text-problem. The publication of the critical edition of the Ādiparvan has rightly been acclaimed by Winternitz¹⁰ as "the most important event in the history of Sanskrit philology since the publication of Max Müller's edition of the Ṛgveda".

c13.s1.13

¹⁰ ABORI, XV, p. 159.

c13.s1.14

From the first the undertaking of the critical edition was an epoch-making event in a number of ways. It was a national work. Nothing of the kind had been attempted in India before; and when complete, this huge, herculean work will pale into insignificance any literary event accomplished hitherto in India. The publication has been uniformly greeted with applause and approbation by all Indologists, and the resolutions of the All-India Oriental Conference, the International Congress of Orientalists and the American Oriental Society, expressing their approval of "the eminently satisfactory manner in which the work is being done by the Institute", bear ample testimony to the immense value of the critical edition according to competent academic bodies. The Prolegomena, to which we shall presently refer, has vindicated Indian scholarship in the eyes of the European scholars. The colossal task demanded in the editor philological acumen, keen insight—almost a sixth sense for spotting the right reading,—and Sukthankar possessed these qualities to a very high degree, being qualified for this "monumental task by learning, training, skill and enthusiasm."¹¹ Indeed no better testimony may be found to Sukthankar's worth than the remarks of Winternitz who states: "Neither in India nor in Europe any one scholar will be found who could have done the work better than Dr. Sukthankar has done."¹²

c13.s1.15

¹¹ Dewherst in JRAS, 1931, p. 466.

c13.s1.16

¹² Indol. Prag., I, p. 67.

c13.s1.17

The *Prolegomena* is a model of good temper, moderation and objectivity. It will ever stand as the basis of textual criticism for all further works in Indian classics. The *Mbh* text has been definitely established, at all stages of its history, to have been "fluid and carelessly guarded", affording easy opportunities of addition, omission, alteration, conflation, athetisation, haphazard synthesis of divergent readings and versions and other diaskeuastic activities, not only of careless scribes but also of a host of scholars, poets and reciters. The *Mbh* had all along been a living and growing text. The text tradition was not simple and uniform, but multiple and polygenous; and in the successive revisional and amplificatory activities it is now impossible to discover the elusive lost archetype. Hence the *Mbh* problem, as has been often repeated, is a problem *sui generis*,¹³ and the general principles followed in critical editions of classical texts in Greek and Latin can be applied only with great limitations and reservations to those in Sanskrit. It is impossible to trace a genealogical tree of all MSS of the *Mahābhārata*. The peculiar conditions of transmission of the epic necessitate an eclectic but cautious utilization of all MSS classes.¹⁴ For text reconstruction a thorough critical study of the complicated text material of the epic is necessary.

c13.s1.18

¹³ JBRAS, IV, p. 157; ABORI, XI, p. 262; XV, p. 164; IHQ, XI, p. 598; &c.

c13.s1.19

¹⁴ *Prolegomena*, p. lxxxv f. As all my observations are based on the *Prolegomena*, I have not given references to it in order not to swell the footnotes.

c13.s1.20

The chief scripts represented by the MSS are Śāradā, Nepālī, Maithili, Bengali, Devanāgarī, Telugu, Grantha and Malayālam. The MSS divide themselves primarily into the Northern and Southern recensions, and each is further subdivided into a smaller number of groups. The N. recension resolves itself into the North-Western represented by the Śāradā version (with its Devanāgarī transcripts), and the Central represented by the versions in Nepālī, Maithilī, Bengalī and Devanāgarī; and the S. into the Telugu and Grantha, and Malayālam versions.

c13.s1.21

The secondary testimonia comprise the Javanese version (c. tenth century), the Andhra adaptation by the Telugu poets Nannaya Bhaṭṭa (eleventh century) and his successors, the Bhāratamañjarī of the Kashmīrian Kṣemendra (eleventh century), and the Persian translation made in the reign of Akbar. The versions of the different commentators such as Devabodha, Arjunamiśra, Ratnagarbha, Nīlakaṇṭha, Vādirāja and Caturbhuja also figure among those whose readings are taken into account in the critical notes of the different parvans.

c13.s1.22

There is a large number of "misch-codices", MSS giving text not only of one particular version, but a mixture of the readings belonging to different versions and even recensions. The MSS material has been classified according to scripts. One of the remarkable results of the investigations of Sukthankar is the establishment, for the first time, of a Kashmīrian version of the text. The "Śāradā Codex" was purchased by Bühler for the Bombay Government in 1875 and was lying unused on the shelves of the Deccan College Collection for 50 years until its resurrection by Sukthankar. The Kashmīrian version has been proved to be the best and most authentic of extant versions. It is the shortest of all versions, interpolations in it being relatively few. Edgerton also admits its importance by calling it the 'oldest and best' text, 15 and Winternitz justifies the preference given to the Kashmiri-an version. 16 It has, however, its own eccentricities. The Śāradā text thus is the *textus simplicior*, the southern text being the *textus ornatior*. The best representative of the southern recension is the Malabar version. The text has been constituted not on any subjective grounds, but on the clear evidence of the MSS themselves in order to purge it of numerous later additions, spurious *ślokas* and long passages.

c13.s1.23

The general principles enunciated in the *Prolegomena* have been proclaimed to be 'unquestionably sound' by all reviewers. 17 Sukthankar has referred to some important principles which have been admitted by all: 18 Agreement between the N. (Kashmirian) and the S. (Malabar) recensions has been taken as the greatest indication of originality. Where two classes of MSS agree in opposition to other two classes, preference has been given to that side on which the Kashmir MSS stand. Preference again has been given to the reading which best suggests how other readings might have arisen. Interpretation receives precedence over emendation, and the more difficult reading has been preferred to the simpler one. Hiatus is to be restored wherever we find variants in the MSS which may be explained by assuming them to be the different efforts by scribes unaccustomed to hiatus to remove the irregularity. Irregular *sandhis* also formed part of the original MSS tradition. 19 The grammatical and metrical irregularities of the original are responsible for quite a large fraction of the mass of variations which the editors came across in the MSS. Sastri's view that the northern text is 'mutilated' has been proved to be untenable.²⁰ The constituted text based on eclectic principles will, no doubt, be more faithful than any single codex. Greatest difference of opinion, however, is possible with reference to the readings "less than certain", which have been indicated by waved lines, as the balance of the MSS evidence is equally divided in their case. Many critics (though admitting the soundness of the general principles) have indicated passages where they would prefer other readings, most of which pertain to the waved line readings. Sukthankar in his brilliant "Epic Studies" has stated reasons which have guided him in preferring the particular readings. The Prolegomena has once for all proved the futility of the Parvasaṃgraha argument, since the Parvasaṃgraha has been tampered with in different versions, and hence we must clearly abandon any hope of deriving any really useful information from it regarding the extent of the text. The established text is indeed a distinct contribution to the philolgical history of the epic in India.²¹

c13.s1.24

¹⁵ JAOS, 49, p. 283.

c13.s1.25

¹⁶ ABORI, XV, p. 169.

c13.s1.26

¹⁷ Cf. Winternitz, ABORI, XV, p. 169; Keith, IC, III, p. 768; Edgerton, JAOS, 48, p. 188; Banerji-Sastri, JBORS, 1929, p. 283.

c13.s1.27

¹⁸ ABORI, XVI, pp. 90-91.

c13.s1.28

¹⁹ Keith objected to the restoration of hiatus and irregular Sandhi (IC, III, pp. 768-771); but the oldest Nepalese MS has confirmed Sukthankar's view (ABORI, XIX, pp. 211-15).

c13.s1.29

²⁰ Prolegomena, p. xxvi.

c13.s1.30

²¹ Ruben and Raghu Vira are following the same technique and principles for their critical editions of the Rāmāyaṇa, and no better testimony can be found for the excellence and soundness of the Prolegomena than this transformation of an erstwhile critic into a staunch supporter. This is the greatest triumph for Sukthankar.

c13.s1.31

In spite of the difficulties and limitations in the way of text reconstruction of such a fluid and constantly changing text, the case is not so discouraging as might appear at first sight. There is a considerable part of the text where the N. and S. recensions are in full agreement, where there are no variants or really no important variants at all. A considerable number of passages can be constituted with an amount of certainty by agreement of versions between which the chances of mutual borrowings or contamination are *prima facie* the least likely, such as e. g., those between Śāradā and Malayālam. Thus, a great deal of a fairly old text can be restored for further critical study, and this, in fact, has been done by the critical edition.

c13.s1.32

The constitution of the text of the critical edition has proceeded, as far as possible, on strictly objective evidence, all subjective considerations being eschewed as a matter of principle. As regards the comparative value, relative preference and importance of the different kinds of objective evidence, Sukthankar has formulated certain rules which may better be put in his own words:—

c13.s1.33

"The highest documental probability we can demand and expect is when all manuscripts of our critical apparatus—which is the same as saying, all our different versions—agree on a reading or a feature. We must accept this as original; at least we do not want to question it, at present. In the absence of such complete concord, the next best combination is the agreement between the Śāradā version and the Southern recension (against Bengali-Devanagari). Third in importance is, in my opinion, the concord between the Southern recension and the Bengali-cum-Devanagari version (against the Śāradā). Fourth in order stands the agreement between only Northern versions or only Southern versions *inter se*, which I consider, in general, of equal value. With a proviso that a passage or a stanza or even a line, which is not necessary to the context, may be rejected, if it is actually omitted entirely in even one of the important versions, since—as experience has shown—the chances of conflation are always very much greater than those of accidental or intentional omission."22

c13.s1.34

Seventy MSS in all were collated for the first two chapters of the Ādiparvan; but it was later found that five or six MSS of each class were sufficient to establish text of that version. Readings from fifty MSS have been given for the first two chapters, and from thirty-eight for the rest. The total number of ślokas in the Ādiparvan (according to the Parvasaṃgraha) is 7984. In the critical edition 121 long and 1634 short passages have been proved to be interpolations.

c13.s1.35

With regard to the readings adopted in the critical edition, it may be observed that subsequent researches have almost invariably confirmed the correctness of the readings of the constituted text. Belloni-Filippi, in his article on the Kadrū-Vinatā episode, justifies the excision of a passage (B. I. 22. 1-3 =No. 13 in App. I of the Ādiparvan), showing that the lacuna is only apparent and the textus simplicior is quite in order.²³ Lüders, on the ground of intrinsic probability, pleads for the preference of satyaṃ cāmṛtam eva ca, to which Edgerton had objected.²⁴ The oldest Nepālī MS of the Ādiparvan, 700 years old, confirms not only many of the disputed readings of the constituted text but justifies some of the emendations made by Sukthankar, and supports his views regarding hiatus and irregular sandhi.²⁵ This is unimpeachable evidence of the correctness of the method followed by Sukthankar. Katre has noticed that the use of the optative form iyāt is not due to an error or irregularity, but it was the regular optative form of the epic.²⁶ In the controversy between Johnston and De,²⁷ regarding kāñcana in Mbh V. 19. 15, the reading in the critical edition has been confirmed by kāñcana-druma-samnibha in Mbh. III. 40. 2. Edgerton in his "The Goat and the Knife" justifies the reading adopted by the critical text.²⁸ The division of the epic into 100 sub-parvans as listed by Sukthankar is corroborated by an old Gujarati translation of the Mbh.²⁹

c13.s1.36

²² Āraṇyakaparvan, Intr. p. xviii.

c13.s1.37

²³ Ascoli Mem. Vol., 1930, pp. 174-180.

c13.s1.38

²⁴ OLZ, 24, n. 1142 f.

c13.s1.39

²⁵ "Epic Studies (VII)", ABORI, XIX, pp. 201-262.

c13.s1.40

²⁶ JAOS, 57, pp. 316-17; NIA, I, p. 536; BDCRI, I, pp. 8-13.

c13.s1.41

²⁷ JRAS, 1939, p. 220 f; 1940, pp. 69 ff.

c13.s1.42

²⁸ JAOS, 59, pp. 366-368.

c13.s1.43

²⁹ Forbes Gujarati Grantha Mala, Nos. 15, 20. It is dated Samvat 1644 (= A.D. 1587); it supports the critical edition in omitting the Kanikaniti.

c13.s1.44

The recorded variants of the critical edition of the Mbh have opened up a new vista in the history of Indian Linguistics by paving a permanent way towards linguistic geography, and Indo-Aryan linguisticians were not slow to grasp the importance of the rich field awaiting investigation. A critical study of these epic variants is sure to yield fruitful results which will throw new light on the problem of epic language and literature. The critical edition has, in fact, re-awakened general interest in the Mbh and given impetus to a number of subsidiary studies by Edgerton, Katre, Kulkarni, Mehendale, Sen Gupta, Shende and others.³⁰ Sukthankar's "The Bhṛgus and the Bhārata" is a fine specimen of higher criticism.³¹

c13.s1.45

³⁰ Shende and Kulkarni obtained the Ph D. of the University of Bombay for their theses entitled "Bhṛgvāngirasa element in the Mbh" and "Epic Variants."

c13.s1.46

³¹ "Epic Studies (VI)", ABORI, XVIII, pp. 1-76.

c13.s1.47

Most of the reviews of the critical edition were critical and contained valuable material, instead of being merely formal; and in this connection, mention may particularly be made of those by Winternitz, Lüders, Keith, Edgerton, Weller, Lévi, De, Nag, Lesny, Belloni-Filippi and Pisani.

c13.s1.48

Following the same principles as laid down in the *Prolegomena* and under the direction of the first General Editor and with the help of his trained band of Pandits and scholars have appeared the subsequent volumes of the critical edition. Sukthankar's *Ādiparvan* was followed by the *Virāṭaparvan* in 1936 edited by Raghu Vira; and it speaks very highly for the thoroughness and triumph of the principles and methods outlined by the General Editor that the editor of the *Virāṭaparvan* should admit that "the technique of reconstruction has been perfected to such a degree that the personality of the editor has been almost eliminated". As rightly observed by De, however, the skill and judgment of the mind that guides a critical work like this can never be dispensed with.³² The *Virāṭaparvan* is the "maṅgala" of the *Mbh* reciters, and the problem of the *Virāṭaparvan* within the *Mbh* is *sui generis*. We are told that nowhere in the entire *Mbh* two recensions recede from each other so widely as in the *Virāṭa*, the shortest of the major Parvans, and this would be evident from the fact that out of 1834 stanzas in the text, the editor is sure about only 300; the rest bear wavy line. The editor has utilised thirty-two MSS; his testimonia comprise not less than fifteen commentaries, Kṣemendra's *Bhāratamañjarī*, Javanese version, etc. For the *Virāṭa*, the Śāradā and Kashmirian MSS do not form separate categories: they are equally good. There are 1178 interpolated passages in the footnotes and 62 long passages in Appendix I. Various concordances have been given; there is also a comparison between the Javanese version and the critical edition.

c13.s1.49

³² IHQ, XIII, p. 370.

c13.s1.50

The *Udyogaparvan* made its appearance next in 1940 under the editorship of that seasoned scholar and well-known Orientalist, S. K. De. Thirty-seven MSS were used for the *Udyoga*, and the testimonia include commentaries by Devabodha, Arjunamiśra, Sarvajñanārāyaṇa, Śaṃkara, and Nīlakaṇṭha. The number of short additional passages is 595, and of lengthy insertions 14. The wavy line is not much in evidence. The *Udyoga* differs from the *Ādi* and *Virāṭa* in that the N. and S. recensions do not recede very materially from each other with regard to omission, transposition and variation of continuous passages. Sukthankar's views about the flexibility of *sandhis* and metrically defective lines in the original text have been borne out by the *Udyoga* also. The critical edition gives the *Parvasaṃgraha* figure for the *Udyoga* as 6698, and the constituted text contains 6063 stanzas and 197 chapters. Readings adopted in the *Udyoga* on independent grounds are confirmed by the Javanese extracts supplied by Juynboll. De fully endorses the remarks of Sukthankar and Raghu Vira about Sastri's edition of the *Mbh*, which he calls a 'misch-edition', as the text is composite, containing much adventitious matter from the North. Sukthankar, Raghu Vira and De have definitely shown that Sastri places too much reliance on a single MS or a particular group of MSS which do not represent the best tradition of the recension, pins his faith to the *Parvasaṃgraha* figures of chapters and verses to which he makes his text conform by curiously strenuous manipulation of stanza and chapter division, against his MS evidence.³³ Thus Sastri's edition is far from being critical, but it gives a continuous southern version in a handy form.

c13.s1.51

³³ *Udyogaparvan*, Intr.

c13.s1.52

The first fascicule of the *Araṇyaka-parvan*, of which Sukthankar was the Editor, was published in 1941, and the final fascicule in January 1943, at the occasion of the Silver Jubilee of the Bhandarkar Oriental Research Institute, a few days before the sudden demise of Sukthankar. Twenty-eight MSS (plus two MSS not included in the Critical Apparatus) headed by the unique and valuable birch bark Śāradā codex, which presents the best and shortest version here also, were utilized for constituting the critical text. Two important testimonia, viz. Devabodha's commentary and the Javanese version, are lacking for the Āraṇyaka.

c13.s1.53

Considering the gigantic forest of ancient myths and legends in this parvan, its text is, relatively speaking, remarkably smooth, in relation to the Ādi and Virāṭa, as may be seen even from such exterior factors as comparative paucity of important variants and of the editorial wavy line indicative of disconcerting parallelisms and uncertainties. It is worth noting that no MS calls it Vana-parvan, which is the name by which it is known in the vulgate and other printed editions. The critical edition has 16 subparvans and 299 adhyāyas; it presents the shortest text compared with the Calcutta, Bombay and Madras editions.

c13.s1.54

The episodic material of this parvan, comprising ancient gāthās and anuvaṃśa stanzas, stories, descriptions, discourses on moral, ethical and philosophical topics, etc., is not a subsequent elaboration secondarily introduced, but formed part of the original epic, its purpose in the epic being to fill up the "temporal hiatuses" in the narrative (viz. twelve years' exile, etc.). Similar material in the Bhīṣma, Śānti and Anuśāsana was also in the original epic. It may be observed that there are close and striking resemblances between many of these stories as found in the Mbh and the Purāṇas. Though in a few cases the assumption of independent common source may be correct, in most cases compilers of the late versions of the Purāṇas have drawn their material from the Mbh narratives.

c13.s1.55

The *Araṇyaka* presents a curious phenomenon in that in this parvan, the N. version is considerably longer than the S.; the vulgate contains fourteen whole adhyāyas which are entirely absent in the S. edition as also in the critical edition. The important and interesting legends of Nala, Ṛṣyaśṛṅga and Sāvitri are critically edited here for the first time. The temptation of Arjuna by Urvaśi (App. No. 6), story of the killing of Naraka and rescue of the Earth (App. No. 16), and the visit of Durvāsas to Yudhiṣṭhira, etc. (App. No. 25) (and other lengthy passages totalling thirty-one) have been relegated to the Appendix on documentary and intrinsic evidence. It is interesting to note that the *Bhāratamañjarī* also omits these three incidents.³⁴

c13.s1.56

The *Sabhāparvan* was begun under Sukthankar as General Editor, and he had seen through the whole critical text, etc. Its final fascicule appeared under the General Editorship of S. K. Belvalkar.

c13.s1.57

Of the *Sabhāparvan*, edited by F. Edgerton, the first fascicule appeared in 1943 and the second in 1944. In all thirty MSS have been utilized for the *Sabhāparvan*. With regard to the classification of MSS, the Editor considers it more appropriate to assume three independent recensions instead of two, viz. Western (comprising Śāradā and K, i.e. Devanāgarī group allied to the Kashmir version), Eastern (comprising Nepāli, Maithili, Bengali and Devanāgarī other than K) and Southern (comprising Telugu, Grantha and Malayālam). Among the commentaries used, Devabodha's text shows close resemblance to the Kashmiriān recension, whereas Vādirāja's text belongs to the southern recension. Of the additional testimonia, the text of Bhāratamañjarī was of minimal value, and no Javanese version of the Sabhāparvan is yet reported.

c13.s1.58

The constituted text is divided into 72 adhyāyas, —precisely the number in the Parvasaṃgraha (I. 2. 103). The total number of ślokas attributed to the Sabhāparvan according to the Parvasaṃgraha is 2511, whereas the critical edition shows 2390 (or 2432, if extra-length stanzas are reckoned as one and a half). The number of stanzas in Bombay and Calcutta editions is almost identical, viz. 2751, nearly 300 more than the number in the critical edition. Sastri's text contains 4511 (or more than 2000 more than our text). The longest single insertion in Sastri's text is the glorification of Kṛṣṇa put into the mouth of Bhīṣma in the Śiśupālavadha episode (II. 34-40 in Sastri) which is not found in Northern MSS.

c13.s1.59

Among the notable omissions may be mentioned (1) the celebrated prayer of Draupadī to Kṛṣṇa (B. II. 68. 41-46) when Duḥśāsana was stripping off her garments,³⁵ and (2) the scene in which Kuntī takes leave of the exiled Pāṇḍavas (B. II. 79. 21-29), which have been excluded on the strength of conclusive MSS evidence. According to the Editor, from the artistic standpoint, the text seems to gain rather than lose by the exclusion of these.

c13.s1.60

³⁵ Sabhāparvan, Intr. It may be observed that Winternitz had already recognised this as an interpolation, partly on textual grounds, partly because of the plot of the Dūta-vākya (Festschrift Kuhn, pp. 299-304).

c13.s1.61

The reference to the city of Rome (II. 28. 49) indicates the period not earlier than the first century B.C. as the date of the original.

c13.s1.62

As already stated, Sukthankar suddenly passed away on 21st January 1943, and his mantle fell on the worthy shoulders of S. K. Belvalkar, and he has proved himself a capable successor by satisfactorily handling the problems relating to the Bhīṣma and other parvans. So far Belvalkar, as General Editor, has brought out the Bhīṣmaparvan and the Śāntiparvan (six fascicules) edited by himself, the Karṇaparvan (one fascicule) edited by P. L. Vaidya, and Droṇaparvan (one fascicule) edited by S. K. De. The Sauptikaparvan edited by H. D. Velankar and the Strīparvan edited by V. G. Paranjpe have been printed, but not yet released on account of some technical difficulties. R. N. Dandekar's Śalyaparvan is almost ready.

c13.s1.63

The Bhīṣmaparvan, edited by Belvalkar, is based on thirty-four (plus four more for supplementary notes) MSS and five (plus occasionally three more) commentaries. The commentary of Devabodha was available for the entire Bhīṣma excepting the Gītā. The Javanese version, two epitomes, and Telugu and Persian translations have been referred to for critical purposes. The critical text contains 117 adhyāyas, and of longer additional passages there are only six. So far as the Bhīṣma is concerned, it is found that the main recensions are only two and not three as assumed for the Sabhā.

c13.s1.64

The text of the Bhīṣma, except in some ten or twelve chapters, is fairly smooth and the wavy line has not been much used. The Durgāstotra at the beginning of the Bhg and the repetitious episode of the death of Śveta in the account of the first day's battle are among the important passages relegated to the Appendix. Some late additions such as the Viśvopākhyāna (B. VI. 65-68; Cr. Ed. VI. 61-64) and the reduplication of the Kṛṣṇapratijñābhaṅga episode in the third and the ninth day of the battle, are found in all MSS, and are retained in the Critical edition.³⁶

c13.s1.65

³⁶ Bhīṣmaparvan, Poona, 1945-47, intr.

c13.s1.66

Belvalkar has enhanced the value of the critical edition by introducing some new features in addition to those usual in the earlier volumes of the critical edition. The very useful "Pedigree of the Bhīṣmaparvan MSS" gives a clear idea of the interferences and contaminations between different branches of tradition and shows only a few MSS as the pure representatives of an unmixed transmission from the archetype. After the Appendices, there are "Critical Notes" dealing with textual criticism, interpretation of words and passages, comparison with other works, especially the Purāṇas, bibliography, etc.

c13.s1.67

Six fascicules of the Śāntiparvan and one each of the Karṇaparvan and the Droṇaparvan have been published so far. In the absence of exhaustive critical introduction and appendices, which will appear in the final fascicules of the respective parvans, it is not possible to deal with these volumes in detail. We shall content ourselves with a brief outline of the critical edition of these parvans.

c13.s1.68

The total number of individual MSS utilised for the entire Śāntiparvan is fifty-two for the text and eight for the commentaries. Three very rare and important MSS were used for the Śāntiparvan, viz. (1) a unique Śāradā MS belonging to the Bibliothèque Nationale of Paris; (2) a complete Bengali MS for the entire parvan bearing a date corresponding to A.D. 1687; and (3) the oldest MS for the Śānti,—the unique Nepali MS written in Maithili bearing a date corresponding to A.D. 1519. For the Rājadharma section, the testimonia comprise of four commentaries (Arjunamiśra's Bhāratārthadīpikā, Vimalabodha's Durghaṭārthaprakāśinī, Nīlakaṇṭha's Bhāratabhāvadīpa, and Vādirāja's Lakṣālaṃkāra), one epitome (Kṣemendra's Bhāratamañjarī) and one translation (Tikkana's Andhra Mahābhāratamu). The Mokṣadharma section, which contains several prose passages divided into paragraphs of varying lengths, has the same testimonia, with the only difference that the following five commentaries were used:—Arjunamiśra's *Bhāratārthadīpikā*, Nīlakaṇṭha's *Bhāratabhāvadīpa*, Paramānanda Bhaṭṭācārya's *Mokṣadharmatikā*, Vidyāsāgara's *Vyākhyānaratnāvali*, and Vādirāja's *Lakṣālaṃkāra*. The Appendixes to the parvan will consist of (1) Longer Passages, (2) Closer parallelisms between sections of Mokṣadharma and some Purāṇa texts, (3) Critical notes, (4) Concordance, etc. The General Introduction will deal with the critical estimate of the MS material used and a discussion of the several literary, historical and philosophical topics which will have to be considered in that connection. Belvalkar is the Editor of the *Śāntiparvan*.³⁷

c13.s1.69

³⁷ *Śāntiparvan*, Fasc. 18-10 (Rājadharma), Fasc. 21 (Āpaddharma), Fasc. 22, 23, 24 (Mokṣadharma). Poona, 1949-53.

c13.s1.70

The *Droṇaparvan* is being edited by the veteran Sanskritist and well-known Indologist, S. K. De, who has already edited the *Udyogaparvan*. The constituted text, which approaches much nearer the Parvasaṃgraha figure of 170 adhyāyas as against 202 of the vulgate, is based on thirty-five MSS of different recensions and versions, the standard and method of text-reconstitution being almost the same as in the earlier parvans. Five commentaries, viz. by Arjunamiśra (incomplete), Caturbhuja, Devabodha, Nīlakaṇṭha and Vādirāja, have been used as testimonia. The largest insertion consists of twenty adhyāyas in the vulgate containing the legends of Mṛtyu (chs. 52-4), of Sṛñjaya and his son Suvarṇasthivin (ch. 55) and the lengthy episode of Ṣoḍaśarājakiya (chs. 56-71), which are all really the secondary elaborations of the same legends and episodes found in the *Śāntiparvan* (Cr. Ed. 29-31; 248-50), and hence omitted from the Droṇaparvan. The printed fascicule runs up to Abhimanyuvadha.³⁸

c13.s1.71

³⁸ Droṇaparvan, Fasc. 25, 1953.

c13.s1.72

Thirty-two out of a total of fifty-six MSS were used for the Karṇaparvan edited by P. L. Vaidya, of which one fascicule has been published. The testimonia consist of four commentaries, viz. those by Arjunamiśra, Caturbhujamiśra, Nīlakaṇṭha, and Vādirāja, three epitomes, viz. Kṣemendra's Bhāratamañjarī, Maladhārī Devaprabha's Pāṇḍavacarita, and Amaracandrasūri's Bālabhārata, and one translation, viz. Tikkana's Āndhra Bhāratamu. Several MSS, both in the N. and S. recensions, are not consistent in their values. An interesting feature of this parvan, as observed by Vaidya, is that some MSS, which in earlier stages, represented the northern recension, developed a tendency to go over to the southern recension.³⁹

c13.s1.73

³⁹ Karṇaparvan, Fasc. 20, 1950.

c13.s1.74

We may now refer to some special features of the critical edition. Interpolations form an important aspect in this connection. We are apt to miss in the critical edition some of the interesting episodes which we were used to associate with the Mbh; but really speaking, we miss nothing of any importance in the critical edition. As regards interpolations, I may draw attention to the following that are the most interesting: Gaṇeśa episode; dramatic scene at the svayaṃvara of Draupadī; and Kaṇikanīti from the Ādiparvan; Durgā-hymns in the Virāṭa and the Bhīṣma; Śrī Kṛṣṇa clothing Draupadī in the Sabhā; story of Durvāsas, when he comes to the Pāṇḍavas for dinner; and Urvaśī's advances to Arjuna in the Āraṇyaka; one adhyāya from the Sanatsujātīya in the Udyoga, which has not been commented on by Śaṅkarācārya. Spuriousness of every one of these passages has been thoroughly established on indubitable MS evidence. As rightly observed by Winternitz, not even one of the passages relegated to the footnotes or the Appendix "will be found to be a real loss to the epic as poetry."[40]

c13.s1.75

Attention may also be drawn to the important discoveries of Sukthankar. He brought to light the old "Śāradā Codex" and established the "K" version. Then he pointed out the importance of Devabodha's commentary, which was collated for the first time. One of the fruitful results of the search for old MSS of the Mbh carried on by the B. O. R. I. was the discovery of the oldest Nepali MS of the Adiparvan by Rajguru Hemraja Pāṇḍit.

c13.s1.76

Before proceeding to the articles, let me emphasise the importance of the critical edition for scholars. Reference has already been made to the serious drawbacks of earlier studies owing to want of a critical edition of the text. Keith rightly stresses the value of the critical edition by observing that in grammatical, metrical, historical, religious or sociological studies it is important to see whether conclusions are based on old material or on mere various readings of no great authority. The critical edition supplies material for solving all these questions in an effective shape.[41] Citations from the Mbh would nowadays be useless unless they are authenticated, so to say certified, by the critical edition. Therefore for all scholarly purposes the critical edition should always be consulted and cited for the parvans so far published by the B. O. R. Institute; otherwise the citations are as good as useless. Besides giving the critical text, the critical edition gives a digest of the manuscript tradition of the Mbh extending over nearly a thousand years. In the critical edition, as observed by Pisani, the editors have tried "to reach the form which the poem had before its spreading through India.... the archetype which stands at the basis of the different recensions and branches of tradition."42

c13.s1.77

⁴⁰ ABORI, XV, p. 172.

c13.s1.78

⁴¹ Keith, IC, III, p. 767.

c13.s1.79

Inseparably connected with the question of the critical edition is the series of "Epic Studies" by Sukthankar, which attempts to answer with great precision and in detail, some of the fundamental criticisms levelled against the principles enunciated in the *Prolegomena* or their application in particular contexts. In the first, entitled "Some Aspects of the Mahābhārata Canon,"43 Sukthankar sets forth his reasons for adopting particular readings in the text objected to by Weller and Edgerton, and maintains the enumeration of exactly 100 sub-parvans in the Parvasaṃgraha list. Epic Studies II: "Further Text-Critical Notes"44 states the reasons that guided him in his choice of readings in question, which have been criticized by Winternitz. It has rightly been observed that we are apt to prefer the readings of the vulgate which appear at first sight to be better than the constituted text. Ruben's criticism⁴⁵ of the principles underlying the preparation of the edition and the constitution of the text has been effectively met in Epic Studies III.⁴⁶ Ruben states the aims and methods of classical philology, and writes about their application to the problems of the *Mbh* textual criticism; but these cannot be applied to the *Mbh* in *toto*. Ruben further contends that the Śāradā MS as a matter of fact does not differ from other MSS, and that every contextual criterion is problematic. The first evidently is untenable; and after detailed discussion, Sukthankar has shown that the criterion of intrinsic probability is not problematic at all, but quite definite and conclusive. "More Text-Critical Notes"47 form the next part of the Epic Studies (IV). After referring to the points from the Prolegomena on which there was consensus of opinion among scholars, Sukthankar sets forth in detail his reasons for the choice of the readings in the critical text in view of Winternitz's preference for certain other readings in his review of the Ādiparvan. We shall later refer to "Mahābhārata Commentators" (Epic Studies V).⁴⁸ In "The Bhṛgus and the Bhārata" (Epic Studies VI),⁴⁹ Sukthankar has given myths and legends relating to the Bhṛgus occurring in the Mbh to show the vital part played by the Bhṛgus in the shaping and development of the epic. It is also shown that the original Bhārata became the Mahābhārata at the hands of the Bhṛgus. This highly illuminating paper has served as the basis of other articles and dissertations which have pursued the subject further.⁵⁰ In the next paper on Epic Studies (VII), "The Oldest Extant MS of the Ādiparvan,"51 which is 700 years old and hails from Nepal, is described with a facsimile plate, collated with the published text, and it is shown that the MS almost invariably supports the critical text.

c13.s1.80

⁴² ABORI, XXIX, p. 303.

c13.s1.81

⁴³ JBBRAS, N.S., IV, pp. 157-178.

c13.s1.82

⁴⁴ ABORI, XI, pp. 165-191.

c13.s1.83

⁴⁵ "Schwierigkeiten der Textkritik des Mahābhārata", AO, 8, pp. 240-256.

c13.s1.84

⁴⁶ "Dr. Ruben on the Critical Edition of the Mahābhārata", ABORI, XI, pp. 259-283.

c13.s1.85

⁴⁸ ABORI, XVII, pp. 185-202.

c13.s1.86

⁴⁹ ABORI, XVIII, pp. 1-76.

c13.s1.87

⁵⁰ Weller, "Who were the Bhriguids?", ABORI, XVIII, pp. 296-320; A. P. Karmarkar, "Dr. Sukthankar's Theory..... and the light it throws on the Dravidian Problem", ABORI, XX, pp. 21-24; Agrawala, "Bhṛguvaṁśa aur Bhārata", NPP, Śrāvaṇa 1997, pp. 105-162.

c13.s1.88

⁵¹ ABORI, XIX, pp. 201-262; "Epic Studies (VIII)" dealing with "Rāmopākhyāna and Rāmāyaṇa" is mentioned later on.

c13.s1.89

In connection with the reading *hāsyarūpeṇa Śaṃkaraḥ* (I. 57. 21) as against the vulgate *haṃsarūpeṇa ceśvaraḥ* supported by Winternitz and Meyer, Sukthankar has written a note under "Epic Questions (I): Does Indra assume the form of a Swan?"⁵² Additional evidence in the form of the newly discovered Nepalese MS has been adduced. It is doubted whether Indra in the context could have had any reason for assuming the form of Haṃsa as suggested by Meyer. The Haṃsa incarnation of Indra is shown to be "nothing more than a canard." The second article in the series, unfortunately the last published article from Sukthankar, appeared in the Silver Jubilee Number of the *Annals*.⁵³ It deals with the Parvasaṃgraha figures and shows how an exaggerated reliance on them would lead to misleading results and curious conclusions. A conspectus of the figures for the Adhyāyas and Ślokas of the eighteen parvans of the Mbh from the Parvasaṃgraha chapter according to the critical edition, the Javanese version, and the Southern recension by P. P. S. Sastri have been given in a tabular form, and observations on the figures of the table follow. There is another table giving the figures for the Adhyāyas and ślokas for the eighteen parvans arrived at by a collation of various extant versions of the Parvasaṃgraha chapter of the Ādiparvan.

c13.s1.90

⁵² BDCRI, I, pp. 1-7.

c13.s1.91

⁵³ ABORI, XXIII, pp. 549-558.

c13.s1.92

Edgerton has started "Epic Studies: Second Series" of which Nos. 1, 2 and 3 deal respectively with (1) Grammar and Syntax, (2) Metre, and (3) Vocabulary in the Sabhāparvan.⁵⁴ S. K. De has given a list of words and phrases peculiar to epic usage in "Some Lexicographical Notes on the Udvogaparvan of the Mbh", the main purpose of the notes being to register and tabulate usages, rather than to attempt authoritative interpretation. The critical text has been used throughout.54* In his "Mbh Notes," of which three have so far appeared in the *Annals*, and in another article, V. S. Agrawala discusses the words occurring in the Mbh. His "Śiṁśumāraśiraḥ" disputes the reading *śiṁśumāra-puram* in the critical edition (I. 176. 15), and maintains the reading *Śiṁśumāraśiraḥ* as preserved in the vulgate (I. 185. 16) on the ground that *śiṁśumāraśiraḥ* means *makaratoraṇa*, "the architrave of the gateway adorned with fish-tailed crocodile."56 "Mbh Notes I" deals with (i) *vāraṇau ṣaṣṭihāyanau*, (ii) *dvaipa* and *vaiyāghra*, (iii) *upasṛtāḥ*, (iv) *haraṇa*, and (v) story of Yavakrita. It is shown that (i) at sixty an elephant is full grown; (ii) *dvaipa* and *vaiyāghra* mean chariots covered with the skins of *dvipin* and *vyāghra* respectively; (iii) suggests *upasṛtāḥ* for *upasṛtān* in the vulgate (III. 240. 4-5), observing that Sukthankar has preferred *upasṛtāḥ* (Cr. Ed., III. 229. 5) as "undoubtedly the correct reading" "though rather feebly supported by the MS evidence"; (iv) *haraṇa* means a dowry or nuptial presents.⁵⁶ In "Mbh Notes II" Agrawala justifies the reading *prākāravapra* in the critical edition (IV. 10. 1) in preference to *prākāravakṣa* of the Kashmiri MS. It is shown that the term *prākāravapra* as the qualifying epithet of *kuṇḍala* was used in a technical sense as the name of a particular kind of ear-ornament (*kuṇḍala*). Illustrations of *prākāravapra kuṇḍala*, which is not found in post-Śuṅga art, have been given from Bharhut.⁵⁷ "Mbh Notes III" deals with (i) *Śiṁśumāraśiraḥ*, (ii) *vaiyāghra*, (iii) *Pūrva-Yāvāta* and *Uttara-Yāvāta* legends, (iv) *praveritā*, (v) *mahānārāhna*, and (vi) *prati*. Agrawala shows that (i) *Śiṁśumāraśiraḥ* is to be preferred to *Śiṃśumārapuraḥ* stating that it is a very fit example of *lectio difficilior* and that the oldest extant MS of the Ādiparvan from Nepal reads *śiṃśumārasiraḥ* and sets the seal of final approval on this debated reading. It is pointed out that feeble textual support is not the final argument in rejecting an otherwise superior reading; (ii) there are other references to *vaiyāghra* in the epic; (iii) Pūrva-Yāyāta and Uttara-Yāyāta legends are authenticated by the critical edition; (iv) *praveritā* in I. 68. 73, meaning "thrown about, cast hither and thither, cast away", has connection with the dialectical Hindi root *pabeḍanā* also written as *paveranā*; (v) *mahāparāhṇa* in I. 181. 40 is the Āṣāḍhī day; and (vi) *prati* in II. 5. 68 is the name of a coin.⁵⁸

c13.s1.93

⁵⁴ BDCRI, V, pp. 1-12.

c13.s1.94

54a Dikshit Mem. Vol. (BDCRI, VIII), pp. 1-33.

c13.s1.95

⁵⁵ JISOA, 1939, pp. 168-9.

c13.s1.96

⁵⁶ ABORI, XXI, pp. 280-4.

c13.s1.97

⁵⁷ ABORI, XXIII, pp. 19-22.

c13.s1.98

⁵⁸ ABORI, XXVI, pp. 283-297.

c13.s1.99

Belvalkar has started a new series dealing with "Some Interesting Problems in *Mbh* Text-Transmission" that appealed to him while preparing the critical edition of the *Bhīṣmaparvan*. His studies of four problems have been published so far. Problem No. 1 deals with the repetition of two stanzas (B. 14. 25-26, absent in the Critical Edition, and B. 14. 57 cd-59 ab= Cr. Ed. 15. 53-54) which is found in ten out of fifty-nine MSS examined for the critical edition of the *Bhīṣmaparvan*. The problem considered is how the stanzas were shifted from the later to the earlier place in the *Mbh*. The explanation given is that the twenty MSS (10+9+1) which exhibit the three accidents of repetition, transposition and omission presuppose a parent MS with a loosened folio with missing folio number, which got misplaced. From the circumstance that no MS representing the Southern Recession is involved in these three more or less related accidents, but representatives of the Northern Recension (Śāradā, Kashmiri, Devanāgarī and even Bengali) alone are so involved, support is found for the theory of only two main Recensions of the Mahābhārata text transmission adopted in the Critical Edition.⁵⁹ Problem No. 2 deals with the reading अन्तपीडाविकपिणः adopted in Cr. Ed. VI. 55. 16 b. The stanza is B. VI. 59. 17=Sastri, VI. 54. 17. The vulgate reads:

c13.s1.100

⁵⁹ ABORI, XXV, pp. 82-7.

c13.s1.101

विनिभिन्नाः शरैः केचिद् अन्त्रापीडप्रकम्पिनः ।
अभीताः समरे शत्रून् अभ्यधावन्त दर्पिताः ॥

c13.s1.102

There are no less than 22 variants (besides six found in MSS not included in the Critical Apparatus), of which classification according to the versions has been given. Only 9 of these 28 readings seem possible as they make some sense after slight correction. Rules about objective evidence have been formulated. But in cases such as the present one, where variants offered are so diverse and where nearly a dozen different interpretations of the pāda are possible, the problem is: Can we always avoid bringing in subjective considerations and choosing a reading which gives "the best" sense? The critical edition accepts the reading of the best MS, viz. Ś₁ (अन्तपीडाविकर्षिणः) needing only a slight correction in the last two letters which was made on the basis of the readings of the allied MSS K 1-3 (पिणः for षिणः) This is a legitimate procedure. Belvalkar rightly remarks that "a passage like the above is the despair of the text-critic".⁶⁰ Problem No. 3 deals with the repetition of the incident of Kṛṣṇa's rushing against Bhīṣma, which first appears in the third day's battle (in 55. 89-92, 98), in the tenth day's battle (in 102. 55-58, 64). It is shown that the reduplication of the incident on the third day is secondary, being an afterthought with the express purpose of engrafting the sectarian Viśvopākhyāna on the main story. Incidentally it is hinted that the third was the penultimate day of Bhīṣma's generalship in an earlier stage of the epic. Belvalkar observes "that the present 18-day battle is frightfully padded up by reduplications, exaggerations and inventions of the miraculous is generally conceded."⁶¹ Problem No. 4 considers the question raised in the review of the Bhīṣmaparvan (JAOS, 70, pp. 117-119) whether, in admitting into the constituted text about a dozen passages omitted by practically the entire Southern recension, the principle enunciated by Sukthankar (Prolegomena, p. xcv)⁶² has not been seriously violated. The article deals exhaustively with all twelve cases and advances arguments in justification of their inclusion.⁶³

c13.s1.103

⁶⁰ ABORI, XXV, pp. 239-43.

c13.s1.104

⁶¹ ABORI, XXVI, pp. 106-119.

c13.s1.105

⁶² "All lines belonging to one recension only, and a fortiori such as pertain to a combination of MSS amounting to less than a recension and which are not necessary for the context" have to be considered as additions and "placed in the footnotes or the Appendix, pending further inquiry regarding their credentials."

First 107 of 908 passages. Read on

Linked texts (37)