Study Of Mahabharata

Author(s): J W De Jong
Publisher: J W De Jong
Catalog link: https://jainqq.org/explore/269714/1

Page #1

C'est un passionnant et instructif objet d'étude que l'effort des savants occidentaux aux prises avec le Mahabharata depuis cent cinquante ans.
Georges Dumézil*
THE STUDY OF THE MAHĀBHĀRATA
A brief survey (Part I)
by
J. W. de Jong
Since the Second World War the Indian epics, and especially the Mahābhārata, have become the focus of Indian studies. In the nineteenth century and the first half of this century the entire corpus of Vedic literature was studied by many prominent scholars to the relative neglect of the epic. In recent decades, however, interest has shifted more and more to the epic. Several scholars, dissatisfied with the results obtained in the past, have brought new methods of research to bear on the study of the epic. Then there is an increasing awareness of the fact that a better understanding of the nature of Hinduism and Indian society depends to a large measure on the interpretation of the Indian epic. From a more philological point of view, the recent completion of critical editions of the Mahābhārata in 1966 and of the Rāmāyaṇa in 1975 has drawn attention to textual problems in both epics.
The first edition of the Mahābhārata appeared in Calcutta, between 1834 and 1839, in four volumes together with a table of contents (sūcipatram). Before that time several parts of the epic had been translated separately, in the first place the Bhagavadgitā, the first English translation of which was published by Charles Wilkins (1750-1833) in 1785. Wilkins also translated other episodes of the Mahā. bhārata. Franz Bopp (1791-1867) was the first scholar to edit the Sanskrit texts of two episodes, "The Slaying of Hidimba" and "The Story of Nala”, in 1816 and 1819. Bopp is also probably the first Western scholar to have read the entire Mahābhārata. This was a remarkable achievement if one takes into account that he completed his reading of the Mahābhārata in 1825, less than ten years after having begun his Sanskrit studies in Paris, and that he had to work with manuscripts without the help of a dictionary. However, Bopp did not continue his study of the epic and his planned survey of the contents of the entire epic did not see

Page #2

It
R
(810)
the light of day.
The first scholar to undertake a systematic study of the Mahabharata was Christian Lassen (1800-1876), who wrote a series of articles entitled "Contributions to the knowledge of Indian Antiquity" (Beiträge zur Kunde des indischen Alterthums aus dem Mahābhārata), the first of which appeared in 1837 (Zeitschrift für die Kunde des Morgenlandes I). Lassen was interested in the Indian epic mainly as a source for the history of pre-Buddhist India. In the first two volumes of his Indische Alterthumskunde (1847 and 1852; second edition 1867 and 1873) he often refers to the two Indian epics. In the first parvan (anukramaniparvan) of the first book it is said that "there are brahmins who learn The Bhārata from Manu onward, others again from the tale of The Book of Astika onward, others again from The Tale of Uparicara onward” (I. 1.50; tr. van Buitenen, p. 22). According to Lassen, this passage shows that there were three beginnings or, in other words, three different recensions (II. p. 496). A verse in the same parvan mentions a Bhārata in 24,000 śloka-s: "First he i. e. Vyāsa composed the collection of the Bhārata in twenty-four thousand couplets, without the minor narratives; this much the learned call The Bhārata proper" (1. 1. 61; tr. van Buitenen, p. 22). Lassen thinks that this probably refers to the second recension, which was recited by Ugraśravas at the Twelve-year Session of The Family Chieftain Saunaka (cf. I. 1. 1 and I. 2. 29). He identifies the Saunaka mentioned in the Mahābhärata with the teacher of Āśvalāyana whose Grhyasūtra (III, 4. 4) mentions both a Bhärata and a Mahābhārata. Lassen dates the second recension about 400 or 460 B. C. (1, p. 598; 11, p. 499). The third and most detailed recension was compiled shortly after Asoka. According to Lassen, it is not possible to place this recension much later because in it Krşna is not yet represented as the dominating form of Vişnu as was the case at the time of Megasthenes (II, p. 495). However, Lassen does not exclude the possibility that in later times the third recension received several additions including the Bhagavadgitā.
In 1897 Auguste Barth published a long review of Dahlmann's Das Mahābhārata als Epos und Recht sbuch (Journal des Savants, 1897, pp. 221-236, 320-337, 427-449; reprinted in Oeuvres de Auguste Barth, Tome quatrième, 1918, pp. 347-403) in which he gives a brief survey of Mahābhārata studies (Oeuvres, IV, pp. 349-354). Barth remarked that, according to Lassen after the second recension was compiled, the only interpolations to find their way into the original poem, were those of a Krishnaite nature. Omitting such interpolations, the epic in its original shape may be looked upon as being pre-Buddhist. Barth added that Lassen's conclusions were very arbitrary. In discussing Lassen's contribution to the study of the Mahābhārata

Page #3

THE STUDY OF THE MAHABHARATA Held quotes from Barth Sukthankar's very disparaging remarks about Lassen's work seem to be based solely on the few lines written by Barth, although he does not even mention him by name. Sukthankar considered it necessary to add that Lassen's speculations appear crude and puerile in the extreme. He also speaks of Lassen's very perfunctory study of this prodigious poem, a gratuitous remark which only shows that he had not taken the trouble to study Lassen's Indische Alterthumskunde. It is interesting to note that in 1874, when reviewing the second edition of the second volume of Lassen's Indische Alterthumskunde, Barth expressed a much more favourable opinion: “Notwithstanding the objections which have been made to Lassen's opinion and which undoubtedly will be made, I do not hesitate to subscribe to it because I believe that of all the explications proposed it is still the one which corresponds the best to all the facts." It is quite possible that Barth's opinion of the value of Lassen's theories was more unfavourable in 1897 than it was in 1874. However, Barth's rapid summary of Lassen's work does not do him justice. It is necessary to point this out because both Held and Sukthankar repeat what Barth had written. Lassen's Indische Alterthumskunde has been a work of great importance and influence in the history of Indian studies. It is therefore necessary to rectify this kind of wrong judgment which tends to be reproduced uncritically. Lassen believed that the present text of the Mahābhārata represents essentially the third recension with some later additions. According to him very few fragments of the first and second recension have been preserved (II?, pp. 496-501). The third recension shows many traces of the veneration of Krsna which Burnouf attributed to a popular reaction against the worship of the Buddha. Lassen was of the opinion that such a reaction could only take place at a time when Buddhism had become very powerful (I', p. 591).
The Mahābhārata is for Lassen an important source for the history of preBuddhist India which is given in the second part of volume I of the Indische Alterthumskunde (I", pp. 568-1034). Lassen devotes many pages to the spread of the Aryans and the prehistory and history of the Pandavas. He pays particular attention to the genealogies of kings and to the historical background of legends (I,' p. 593). Regarding the saga of the Pāņdavas, Lassen writes that it is necessary to determine the meaning of their names, and to trace the older form of the saga and to eliminate later additions and distortions. Only when this has been done, he believes, will it be possible to bring together the remnants of the ancient and true history ("die Ueberreste alter, wirklicher Geschichte") (I', p. 775). In trying to unravel the historical background of the saga, Lassen pays particular attention to the etymology of names. Both Pandu and Arjuna mean "white", whereas Krşņa and Krşņā

Page #4

法華文化研究(第10号)
(Draupadi) mean "black". In the first edition of volume 1 of the Indische Alterthumskunde (1847), Lassen expressed the thought that these two colours indicated the opposition between the Aryans and the original inhabitants of India. In the second edition (1867) Lassen writes that this distinction is due to the fact that the Pañcālas and the Yadavas belonged to Aryan peoples who had immigrated before the Pandavas did (I, p. 791). Lassen explains the marriage of Draupadi, the daughter of the king of the Pañcālas, with the five Pāṇḍava brothers as reflecting an alliance between the Pandavas and the Pañcālas (I2, pp. 790-791). In other respects Lassen's attempts to make use of the epic are more fortunate. For instance, Lassen studied carefully the information found in the Mahabharata on the geography of ancient India, and the names of peoples and their location. He stresses in particular the importance of the digvijaya in Book II (I2, p. 655).
Lassen's work belongs to the period of the great pioneers of Indian studies. The first volume of his Indische Alterthumskunde appeared in 1847 and he did not change his point of view or his method during the twenty years which elapsed before the publication of the second edition. After Lassen, no other scholar has tried to make use of the epic as a historical source in the way he did. However, Lassen's work shows clearly the difficulties which one encounters in trying to discover a historical background for the epic saga. Later scholars have denied the presence of any historical facts in the epic. Others are convinced that the epic legends are not without a historical basis but that it is almost impossible to separate fact from fiction. Weber believed that the Mahabharata reflects a real battle between Aryan peoples which took place at a time when they had settled in India and the original inhabitants had been subjugated." Theodor Goldstücker (1821-1872) wrote: "The leading story is probably founded on real events in the oldest history of India, though in the epic narrative it will be difficult to disentangle the reality from the fiction. More positive is E. Washburn Hopkins, who wrote in 1915: "It [i. e. the Mahabharata] undoubtedly reflects some real contest, which may have taken place about a millennium before our era." Several years later Hopkins asserted: "Many of the characters of the Mahabharata appear to be real, historical figures." In a book published posthumously, Hermann Oldenberg wrote that the story of the Mahābhārata is based on events which took place at the end of the Rgvedic period about 1200 B. C.10 In 1947, Louis Renou wrote with regard to the Pāṇḍavas that there must be some historical connection and that there are here persons and facts of undeniable authenticity, however magnified and amplified they may be. Too many concrete details confirm this. According to him, the idea of a mythical Mahābharata cannot be maintained. The great feud must have taken place at the end
997

Page #5

THE STUDY OF THE MAHABHARATA
of the Ṛgvedic period, as the heroes of the epic are the direct ancestors of a King Janamejaya, the son of Parikṣit, who is mentioned as a contemporary figure in late Vedic texts and to whom the epic is held to have been recited for the first time."1 Recently this problem has been discussed by Grintser in a chapter of his book on the ancient Indian epic.12 He agrees with the opinion expressed by A. N. Veselovsky (1838-1906) who wrote that, taken as a whole, the history on which the epic is based is not identical with concrete events and that every epic blends together events separated by several centuries. Grintser remarks that the historical reality of the epic goes back to different epochs. It was not the task of the epic singers to write history and it was quite possible for them to combine reminiscences of events which belonged to different historical periods.14 In the introduction to his translation of the Udyogaparvan, van Buitenen raises the question: "Is there any reflection of real South Asian history in the Mahabharata ?" He continues: "The answer, now, from historians and structuralists alike, is a resounding no." However, van Buitenen tries to show that the composers of the Mahabharata were to a degree familiar with a King Paurava in northeastern Panjab, and with a town called Abhisări close to the seat of this Paurava realm, but clearly independent of him, since it is separately subjugated. He concludes: "I believe we can be confident that we have identified the King Poros who encountered Alexander the Great by the river Jhelum." These recent publications by Grintser and van Buitenen show that the problem of the relation of the epic to a historical reality is still being discussed. Lassen's approach is no longer acceptable, but one must not forget that whatever objections may justifiably be raised against his interpretations, he was fully aware of the fact that the relation between epic and historical reality is very complicated. Lassen wrote: "It goes without saying that the Indian saga has been subjected to all those transformations which are proper to an untrammelled tradition, the magnification of the characters of the heroes, the concentration of entire epochs, conditions and tendencies in the form of individual persons and in individual actions." Lassen was certainly not such a naive and credulous scholar as he is sometimes depicted.
In 1883 Lassen's treatment of the Mahabharata was severely criticised by Søren Sørensen (1848-1902) in his book: 'On the place of the Mahabharata in the Indian Literature. I. Attempt to separate the oldest elements' (Om Mahabharata's stilling i den indiske literatur. I. Forsøg på at udskille de ældste bestanddele). Due to the fact that Sørensen wrote in Danish, his work has not received due attention, for most scholars have based their opinion on the Latin summary in Sørensen's book (pp. 355-383). Sørensen did not believe that it was possible to discover a historical nucleus

Page #6

#tft* (#10) behind the many transformations which the legend had undergone, even if one supposed that such a nucleus had once existed: the point of departure could in inany instances just as well have been mythological (p. 68). Sørensen objected also to the use Lassen had made of genealogies and his allegorical explanation of proper names and events (pp. 65-66). He likewise rejected completely the way in which Lassen tried to reconstruct three different recensions. For instance, Lassen believed to have found an important fragment of the oldest recension in the Bhāratasūtra (Book I, parvan 61; 55 in the critical edition). According to Sørensen this text is nothing more than an extract summarising parts of the contents, and Lassen is wrong in trying to discover in it a part of the oldest form of the saga (p. 101).
Sørensen distinguished between the Mahābhārata saga, the Mahābhārata poem and the Mahābhārata compilation (p. 54). He believed that it would not be possible to uncover the legends which form the basis for the main story of the Mahābhārata because the Mahābhārata is the only source available. However, Sørensen considered it possible to separate, in great part, with absolute certainty the poem and the compilation (p. 55). However, in one place he concedes that in some instances the decision as to what constitutes an interpolation remains subjective and that there is a danger of moving in a circle by proceeding from assumptions which ought first to be proven or refuted independently (p. 133). The poem must have been created by a poet who made use of popular songs which were recited by different rhapsodes (p. 83). These songs belonged to the same main cycle of legends and were recited and revised so that they could be joined together as a whole dominated by one basic and all-embracing legend (p. 81). Here we see clearly the influence of similar theories which were held with regard to the formation of the Homeric epics on the basis of different lays (or Lieder). According to Sørensen, the poem was the result of the work of generations of poets, just as the Mahābhārata compilation was the work of several generations (pp. 81-82). These poets made use of popular legends about heroes. Under the influence of theological speculation, these heroes were made into sons of gods who fought with heavenly weapons against mythical beings. Sørensen believed that the poets who created the Mahābhārata poem already had written texts at their disposal (p. 86). Writing must have been used at an early period, because the works which were compiled by the brahmans were too large to be remembered by heart. Moreover, such studies as philosophy and grammar were impossible without writing. He also believed that the loose connection between the old parts and later additions of didactic and other materials indicated a written tradition. As far as I know no other scholar has considered it necessary to assume such an important role for

Page #7

THE STUDY OF THE MAHABHARATA writing in the history of the Mahābhārata, even though it is said in the first parvan of the first book that Ganesa is writing down the Mahābhārata as the saga Vyāsa dictates it (cf. Critical edition p. 884).
In the main part of his work, Sørensen examines the whole Mahābhārata book by book and eliminates those parts which he considers to be recent additions. He arrives at a total of about 27,000 verses spread over the first twelve books which contain the main story. He is guided in his choice by consideration of the following eight criteria : 1. the context; 2. rhetorical formulae such as questions and promises to the reader; 3. repetitions ; 4. long verses (triştubh); 5. style; 6. the divinity of Krşņa; 7. the presence of such seers as Vyāsa Nārada; 8. the presence and participation of divine beings (pp. 244-245). From these 27,000 verses Sørensen eliminates a further 20,000 verses on the basis of the occurrence of names of persons whom he considers not to have appeared in the original poem: the seers Nārada and Vyāsa; Keśava, Kșşņa, Krşņau (Krşņa and Arjuna), Madhusudhana, Hari, Govinda; King Janameya and foreign peoples such as the Saka, Yavana, Barbara, Mleccha,Cina, Hūņa, Hārahūņa, Pārasika, Pahlava and Pārada. In this way Sørensen arrives at what he calls 'the original poem' which, according to him, consisted of episodes without any other interconnection than the pure chronological succession of events belonging to the same cycle of legends. This poem was probably called the Bhārata and from it grew the present Mahābhārata which incorporated all the non-Buddhist poems composed in the course of several centuries. In the older parts, Brahman is the supreme god, but in the later ones Vişņu or Krsna and also, but to a lesser degree, Siva. Buddhism is not mentioned in the older parts, which must have been composed before 200 B. C. The parts in which the Greeks (Yavana) and the Scythians (Saka) are mentioned cannot be older than the first century B.C.
Sørensen's work is the most extreme example of the application of what has been called analytic theory.'' Barend van Nooten has recently stated that the most articulate spokesman for this theory. was the American scholar, E. Washburn Hopkins (1857-1932), whose first major work on the Mahābhārata, published five years after Sørensen's work, was 'The social and military position of the ruling caste in Ancient India as represented by the Sanskrit epic' (JAOS, 13, 1888, pp. 57-376; reprint Varanasi 1972). Van Buitenen, in his introduction to the first volume of his Mahābhārata translation, designates the methods of Hopkins and Holtzmann as historicist and that of Dahlmann as holistic. It is necessary to distinguish between a historicist and an analytical approach. The former tries to discover the historical nucleus of the poem or to uncover historical events on the basis of later

Page #8

***ILM# 105) legendary reworkings. The latter is primarily interested in the development of the epic, and not in the existence or absence of a historical background. Sørensen is an example of the second approach. Opposed to the analytical theory is the synthetic theory, an expression first used by Hopkins to designate Joseph Dahlminn's theory as developed by him in his book on the Mahābhārata published in 1895 (Das Mahābhārat a als Epos und Rechtsbuch. Ein Problem aus Altindiens Culturund Literaturgeschichte. Berlin, 1895). Opposed to the historicist theory is the mythological theory which was first applied to the Mahābhārata by Alfred Ludwig (1832-1911) in several publications which appeared in the years 1884 to 1900.
It is of course impossible to separate the different approaches (analytic, historicist, synthetic and mythological). The historicist is always at the same time an an ilyst because he believes it possible to distinguish older and newer elements in the epic in order to arrive at a historical background. The mythologist can. also be a historicist as one can see in the case of Alfred Ludwig.
It is not possible to study in detail all the different theories proposed since Lassen's, but it is necessary to examine briefly Hopkins's theories on account of the influence they have exercised up to the present. According to van Nooten, Hopkins's analytic theory is widely accepted.20 Van Buitenen states that "it is Hopkins's methods and views that have since been largely, if tacitly, accepted by scholarship."1 Hopkins developed his theory for the first time in his article on the ru ing caste. Later formulations can be found in The Great Epic of India (New York, 1901), an article on the Mahābhārata in the Encyclopaedia of Religion and EtPrics (Vol. VIII, 1915, pp. 325-327) and in his contribution to Volume One of the Ca nbridge History of India (Cambridge, 1922, pp. 251-276). In The Great Epic of India (pp. 397-398) Hopkins distinguishes the following stages : “Bhārata (Kuru) lay 3, perhaps combined into one, but with no evidence of an epic before 400 B.C. A Mahābhārata tale with Pandu heroes, lays and legends combined by the Puranic diaskeuasts, Krishna as a demigod (no evidence of didactic form or of Krishna's divine supremacy), 400-200 B.C. Remaking of the epic with Krishna as all-god, intrusion of masses of didactic matter, addition of Puranic material old and new; multiplication of exploits, 200 B. C. to 100-200 A. D. The last books added with the Introduction to the first book, the swollen Anuçāsana separated from Çānti and recognised as a separate book, 200 to 400 A.D.; and finally 400 A. D. +: occasional amplifications, the existence of which no one acquainted with Hindu literature would be disposed antecedently to doubt." Hopkins adds that the following facts are certain : "1. That the Pandu epic as we have it, or even without the masses of didactic material, was composed or compiled after the Greek invasion; 2. That

Page #9

THE STUDY OF THE MAHABHARATA this epic only secondarily developed its present masses of didactic material ; 3. That it did not become a specially religious propaganda of Krishnaism (in the accepted sense of that sect of Vaisnavas) till the first century B. C.; 4. That the epic was practically completed by 200 A. D.; 5. That there is no 'date of the epic' which will cover all its parts (though handbook makers may safely assign it in general to the second century B. C.)." In other publications Hopkins discusses in more detail the relative age of different parts of the epic.22 Hopkins is not only an analyst but also a historicist. He writes: "In so far, then, as we may discern a historical germ in the midst of poetic extravagance, it would seem that the poem represents an actual legend of a real tribe, and in so far as that legend persists in its adherence to polyandry as an essential part of the legend, a tribe which, like so many others in India, had been brahmanized and perhaps become allied by marriage to the old Bhārata tribe, whose legends were thus united with its own" (The Great Epic, p. 400).
We find the analytic theory also in Winternitz's famous History of Indian Literature. Moriz Winternitz (1863-1937) was one of the leading Mahābhārata scholars of his epoch and one of the advocates of a critical edition. In the second half of the first volume of his History, published in 1908, he discusses both Indian epics in great detail (pp. 259-440). Winternitz also distinguished different layers and stages: (1) A family feud in the royal house of the Kauravas leads to a bloody battle. The history of this bloody battle, which we shall probably have to regard as an historical event though we hear of it only in the Mahābhārata, was told in songs, and some great poet, whose name has been lost, combined these songs into an heroic poem of the great battle of the Kuru field. (2) In the course of centuries an enormous mass of the most diverse poetry has collected around this nucleus: numerous legends with a loose connection with the old heroic poem, and fragments of other heroic legends and cycles of legends. How much of this old bardic poetry already belonged to the original poem as secondary tales (episodes) and how much was added only later, will probably never be determined. In any case, our Mahābhārata is not only the heroic poem of the battle of the Bhāratas, but at the same time also a repertory of the whole of the old bardic poetry. (3) The brahmans incorporated into the epic legends of gods, mythological narratives of brahmanical origin and even didactic sections referring to brahmanical philosophy and ethics and brahmanical law. (4) However, the Mahābhārata was too much of a popular book for it ever to become the exclusive property of the brahmans. It was the purohita-s, the court-priests, who later on became temple-priests, who added numerous local myths in true Purāņa style : Vişnu and Siva myths, Purāņa-like cosmologies,

Page #10

10
法華文化研究 (第10号)
geographical lists and genealogies. (5) Attributable to ascetics, forest-hermits and mendicants were legends of saints, aphorisms, fables, parables, fairy-tales, and moral tales. As to the date of the Mahabharata, Winternitz arrived at the following conclusion: "1. Single myths, legends and poems which are included in the Mahābhārata, reach back to the time of the Veda. 2. An epic Bhārata' or 'Mahābhārata' did not exist in the Vedic period. 3. Many moral narratives and sayings which our Mahābhārata contains, belong to the ascetic-poetry, which was drawn upon, from the 6th century B. C. onwards, also by Buddhists and Jainas. 4. If an epic Mahābhārata already existed between the 6th and 4th centuries B. C., then it was but little known in the native land of Buddhism. 5. There is no certain testimony for an epic Mahābhārata before the 4th century B. C. 6. Between the 4th century B. C. and the 4th century A. D. the transformation of the epic Mahābhārata into our present compilation took place, probably gradually. 7. In the 4th century A. D. the work already had, on the whole, its present extent, contents and character. 8. Small alterations and additions still continued to be made, however, even in later centuries. 9. One date of the Mahabharata does not exist at all, but the date of every part must be determined on its own account." We have quoted the English translation of Winternitz which, though published in 1927, presents the same text with only some additional bibliographical references.
The most recent exposé of the analytical theory is to be found in van Buitenen's introduction to his translation of book I of the Mahabharata (Chicago, 1973, pp. xiii-xxv). Van Buitenen first outlines what he calls the central story of the Mahabharata, which possesses an extremely complex plot. Van Buitenen writes: "Whatever historical realities may also have been woven into the epic, it is not an accident of dynastic history; however fortuitous its career of expansion, the epic is not an accident of literary history. The grand framework was a design." This central story has proliferated to an extreme degree. We may, with greater or lesser certainty, identify the central parts of the narrative and "later" accretions to the central story. Still, the latter were understood within the framework of the former. The second perimeter contains the accretion of secondary materials to the central story. As an example van Buitenen mentions the divine birth of the Pandavas and other heroes. According to van Buitenen: "Such further elaborations are disappointing because they rob the human actors of much of their motivation.... The elaborations are disappointing also because they show little respect for the Gods themselves. The Gods pressed into service to explain human affairs are mostly of a venerable Vedic antiquity, but their mythology is decaying." The third perimeter consists of the brahminization of the epic by brahmans who

Page #11

THE STUDY OF THE MAHĀBHĀRATA expanded it with brahmanic stories and with interminable expositions on the dharma. Van Buitenen concludes: "Thus The Bhārata of 24,000 couplets grew to The Mahābhārata of 100,000. The original story was in the first phase of complication expanded from within, in the second phase mythologized, in the third phase brahminized. One might even discern a fourth phase, after the epic was first written down, when this collection of manuscripts became, as it were, a library to which new books could be added." The original poem which, according to van Buitenen was a conscious composition by one poet or a small group of poets must have been composed somewhere in the eighth or ninth century B. C. However, the oldest portions preserved are hardly older than 400 B.C. The didactic portions of what has been called the pseudo-epic were added to very late, perhaps as late as the fourth century A.D.
The analytic theory is probably the major contribution made in the nineteenth century to Mahābhārata studies. Even today this theory is far from being dead and buried. Therefore it seems justified to explain this theory in some detail. A variant of the analytic theory is the so-called inversion-theory which was first proclaimed by Adolf Holtzmann (1810-1870) in 1846 (Indische Sagen, II, Karlsruhe, 1846, p. 7). According to Holtzmann, the original poem was written for the glory of the Kurus. Later the poem was reworked in favour of Krşņa and the Pāņdayas, and Duryodhana and Karna were depicted in an unfavourable light. Holtzmann's theory was accepted by Lassen and Leopold von Schroeder,23 and further developed by his nephew Adolf Holtzmann (1838-1914).24 According to the latter the original poem was written by a Buddhist in the third century B.C. This poem was several times reworked and revised, the first time in favour of the Pāņdavas. In this recension the brahmans proclaimed the divinity of Krşņa. A second revision was marked by a reconciliation with Sivaism. The poem was transformed into a dharmaśāstra for those who were not allowed to study the Vedas or who, as for instance warriors, could not understand it. This transformation was executed during the tenth and eleventh centuries. The definitive fixation of the text took place a few centuries later. Holtzmann's theory contained many weak points and was totally rejected by Auguste Barth and Georg Bühler.25 Since then only very few scholars have been in favour of the inversion-theory in one form or another.26 However, even if the theory proposed by the younger Holtzmann is completely unacceptable, Held is right in pointing out that anomalies in the depiction of the Kurus and Pāņdavas still await a satisfactory solution. (cf. Held, op. cit., p. 11). In a recent article, John D. Smith remarked: "Much the most remarkable feature of the hero/villain antithesis in the two epics is, however, the frequent apparent reversal-in 'ethical

Page #12

**to010) terms-of the goodies and the baddies. In the Mahābhārata this is so extreme as to have led to the so-called inversion theory', according to which the Kauravas were actually the original heroes and the Pāņdavas the original villains. There can be few now who would adhere to this theory-Duryodhana and Duḥśāsana are ill equipped to serve as epic heroes--but the facts for which it sought to account remain to trouble us."27
Even less has to be said of the mythological theory of Alfred Ludwig (1832-1911), who thought that the original poem telling of the feud between two tribes was later transformed into a seasonal myth. Duryodhana represents winter while Bhima, the son of Vāyu, is spring, the period of the equinoctial storms. In the second half of the nineteenth century the nature-myth method was applied to Indian religion by Adalbert Kuhn and Max Müller. One finds it still accepted by Keith in his book on The Religion and Philosophy of the Veda and Upanishads (Cambridge, 1925, p. 58), and, although nowadays totally rejected by Indologists, it still crops up in general works on the history of religion.20
An entirely new vision on the epic is to be found in the abovementioned book by Joseph Dahlmann S. J. (1861-1930). Dahlmann's work caused a great stir in Indological circles. Winternitz wrote that by giving rise to a veritable "Dahlmannliterature" it revived studies on the epic.30 Dahlmann vigorously proclaimed the unity of the epic. The central theme of the Mahābhārata is the dharma, the religious and sacred law. The poet describes the battle between dharma and adharma. Sukthankar summarizes the main conclusions of Dahlmann as follows: “(1) The epic is a well defined unity. (2) All the different parts of the poem are joined together with some distinct
and definite purpose, and answering admirably that purpose. (3) The unity of plan and aim was conceived in the mind of one single individual,
who carried out the work in terms of this preconceived unity. (4) Therefore successive expansion, and one or more recasts of the poem are
out of the question. (5) The date of the poem as composed or compiled by the diaskeuast is cer
tainly not later than the fifth century B. C.1 According to Sukthankar, "of all foreign critics of the Mahābhārata (Dahlmann] inay be said to approach nearest to any real understanding of the Great Epic of India " (op. cit., p. 19).
On three points Sukthankar criticised Dahlmann. In the first place Dahlmann's views regarding the unity and homogeneity of the text were much exaggerated. Sukthankar states that the critical edition of the epic shows that large blocks of

Page #13

THE STUDY OF THE MAHĀBHĀRATA the text of the Vulgate must on incontrovertible evidence be excised as comparatively late interpolations. Secondly, Sukthankar does not believe that a popular work such as the Mahābhārata was composed with the exclusive object of expounding all the different aspects of Hindu Law. Thirdly, Sukthankar reproaches Dahlmann for having overlooked Sri Krşņa, the most characteristic creation of Indian genius, who was above Dharma and Adharma, beyond Good and Evil.
Contemporary critics of Dahlmann were at pains to point out that neither the early date proposed by Dahlmann for the compilation of the epic nor the unity of the epic could be accepted, because so many elements in the epic clearly belonged to different historical periods. Hopkins was willing to concede that Dahlmann had shown that epic legends and didactic motifs are closely united in the epic as it is today (The Great Epic of India, p. IX). At the end of his long and critical review, Auguste Barth also conceded that Dahlmann had shown up the fundamental unity, and that by eliminating the didactic parts one would mutilate the epic. Also according to Barth, Dahlmann had shown how futile attempts are to try and distinguish all kinds of layers, Brahmanical, Sivaite, Vişnuite (Barth, Oeuvres, t. IV. p. 402). However, Barth totally rejected Dahlmann's theory of the origin and age of the epic. Other critics were even less favourable than Hopkins and Barth. In a later work Dahlmann attempted to refute his critics, but by his many repetitions and exaggerations he only managed to undermine his own point of view (Genesis des Mahābhārata. Berlin, 1899). However, as in the case of the inversion-theory. Dahlmann had the merit of having posited a problem which even his severest critics did not succeed in eliminating.
As we see, the last twenty years of the nineteenth century witnessed great activity in Mahābhārata studies with the publication of works by Sørensen, Holtzmann, Ludwig, Dahlmann and Hopkins. The last-named continued his studies of the epic until his death in 1932, but without bringing any new point of view to its interpretation. Two important and useful tools for the study of the epic which were compiled in the last twenty years of the nineteenth century, appeared in 1903 and in the years 1904-1925 respectively. The first is Jacobi's careful and detailed survey of the contents of the epic: Mahābhārata, Inhaltsangabe, Index und Concordanz. Bonn, 1903. In this book he shows himself also a follower of the analytical method, as he indicates which chapters contain the main Mahābhārata legends, which narrative episodes, and which didactic teachings. The second work is Sørensen's Index to the Names in the Mahābhārata which was published posthumously in fascicles from 1904 to 1925 (reprint Delhi, 1963). This book is an indispensable reference work for the study of the epic.

Page #14

##xt* (410) In the first forty years of the twentieth century relatively little attention was paid to the Indian epic by Western scholars. At the beginning of the century three leading scholars, Heinrich Lüders, Hermann Jacobi and Moriz Winternitz, advocated the undertaking of a critical edition of the text of the Mahābhārata, but without success. Winternitz published his account of the epic in 1906, and a posthumous book on the Mahābhārata by the great Vedic scholar Hermann Oldenberg (1854-1920) was published in 1922, but neither publication broke new ground. It was only in 1935 that a completely new interpretation of the epic was advanced in a doctoral thesis entitled: The Mahābhārata, an ethnological study. The author was a Dutch cultural anthropologist or ethnologist, as one used to say at that time, G. J. Held (1906-1955). In his introduction, Held gave a short survey of epic research and remarked that Mahābhārata studies had arrived at an impasse because the epic had been studied too much from the philological side. Held declared himself fundamentally opposed to the analytical method which tried to discover the "original epic", a nucleus of an essentially narrative nature, and to eliminate later interpolations. He writes: "the discriminating between a number of different elements in the composition of the Epic must be devoid of all sense unless the student gets to know how it was that the various elements thus discriminated could have been brought together and united so as to constitute such a uniform whole" (p. 176). According to Held: “The word 'Epic' called up in the mind of scholars ideas analogous to those connected with the heroic literature of mediaeval Europe or of ancient Greece in the Homeric poems" (p. 28). However, " to know what an Epic is, we must know what place it occupied in that culture” (p. 30). A culture must be studied in its entirety, and this demands our considering the religious, social and economic manifestations alongside one another as three aspects of that culture (pp. 34 and 35). Held was much inspired by the writings of French ethnologists such as Émile Durkheim (1858-1917), the author of Les formes élémentaires de la vie religieuse (Paris, 1912), and Marcel Mauss (1872-1950), author of Essai sur le don (Paris, 1923). The idea of culture as a structure in which myth, ritual and society were closely interrelated and connected was developed in his lectures in Leiden by J.P. B. de Josselin de Jong (1886-1964).82 Of special importance for Held's work were the studies on the Javanese theatre published by W. H. Rassers (1877-1973).88
In Held's conception the Indian epic is a myth, connected with a ritual, which reflects a social organization similar to the division of a tribe into two moieties or phratries. The mutual relationship between the two phratries is always fluctuating between friendship and enmity. Both groups together constitute a complete unity, seemingly divided but forming in effect a totality (p. 296). “There is a certain

Page #15

THE STUDY OF THE MAHABHARATA
differentiation in status and in function to be discerned between the two phratries. One phratry is frequently more closely connected with the religious life, the other with the social life, the activities of the one more directly concern the sacral plane, the activities of the other the profane plane of the tribal existence. This social and functional discrimination is to be seen in the great tribal festive gatherings. The same differentiation is still further sustained. The more social side of the tribal life runs in the direction of the potlatch, while the more religious side is concentrated about the rites of initiation" (p. 310).
15
According to Held, the two parties in the epic, the Kauravas and the Pāṇḍavas, are opposed to each other as phratries, but are not phratries in the real sense of the word, that is to say, two exogamous moieties (cf. pp. 307-310). Held writes: "The tribal dualism is also met with in the structure of the society in another `shape, functionally taking the place of the older phratry-dualism. In every community where the society system obtains they are to be divided into two great societies, sometimes into two groups of societies. Such, in our opinion, is the relationship existing between the two parties of the Epic" (p. 308). The relation between the two phratries is at the same time one of complementarity and of rivalry. Held shows that this dual organization exists also in heaven in the contrast between deva-s and asura-s, because the social organization which prevails in heaven is conceived of as being the same as that upon earth. Deva-s and asura-s are rivals, but they cooperate with each other in the ritual of the churning of the drink of immortality (amṛtamanthana), Held considers this ritual to reflect a human society in which the circulatory connubial system prevailed. In the Indian mind, he writes, the actions of churning and fire-making by means of fire-sticks are connected with the idea of cosmic motion which is also found in the uninterrupted sequence of life and death, of activity (pravṛtti) and cessation (nivṛtti).
Held stresses that the gods must be studied by preference, side by side in a phratry-relation. Between Vişņu and Siva exists a religious identity: "Vişņu is Śiva, but that does not imply, reasoning upon a religious basis, that Vişņu is one and coincident with Siva; he is neither more nor less than Śiva; the fact is, he is a different being" (p. 168). Held compares this relation to that which exists between deva-s and asura-s: "The Devas are the same as the Asuras, but they are not identical with them" (p. 169). Held points out that in the classificatory system the two rival phratries are usually identified with all kinds of things constituting each other's natural complements, e. g. with male and female, light and dark, heaven and earth, etc. (p. 170). According to him, "In India, where, as we think we have made plausible, a patrilineal grouping is intersected by a latent mat

Page #16

#**it (105) rilineal grouping, each phratry by itself is composed of a male and a female half. In this respect, therefore, each phratry constitutes in itself a perfect whole, which, in the classificatory system is expressed, as we have already seen, by male-female. So it can be said, that the two phratries together constitute the tribe as a whole; it can also be maintained, that the whole is also constituted by each phratry in itself, by each half or moiety of the tribe. The two patrilineal phratries are to one another as man to woman. But when we turn to matrilineal descent we are equally justified in saying, that the two phratries are identical, the patrilineal dual division being intersected by the matrilineal. Here, then, with the literal facts of the case before us we are confronted with the to all appearances illogical logic of A being the same as but not identical with B." (pp. 170-171).
Many scholars have considered those portions of the epic in which divine homage is rendered to Krşņa as later interpolations. Held remarks that they are so closely related to, so essentially interwoven with the rest of the work that they do not admit of being separated from the context without further ado (p. 22). He points out that in the Purāņas Krsna and his brother Balarāma represent the two tribal moieties; they constitute the couple, of whom one is the benefactor, the other the arch-deceiver or trickster, the two culture heroes. However, in the epic Krşna himself is both benefactor and the trickster in one and the same person. He is a bene. factor because he assists Arjuna in battle and aids the Pāndavas with good advice. As Vişņu-Nārāyaṇa he is also the demiurge, the creative god himself, who, aroused from sleep, starts the Cosmos "turning again". He is a trickster and is continually represented as an imposter who manages to carry his point by employing all kinds of fraudulent means. He not only has a dual nature on the mental plane, but also on the physical plane because he is often represented as a child. According to Held, his doings as a child are not a later addition to the Krşņa legend and he points out that the Kamsavadha, the struggle between Krşna and his maternal uncle, is mentioned already by Patañjali. The Kamsavadha, in Held's view, is the dramatic representation of a ritual antagonism between the two phratries and calls to mind the ceremony of initiation in which the mother's brother is the initiatordesignate (p. 178). Arjuna is the initiate, being one with, and yet not the same as, the god-initiate, Kışņa (p. 186). According to Held, essential for the study of the epic is the identification of Krşņa and Arjuna with Nārāyaṇa and Nara, each couple being the essential complement of the other (p. 164). Kșşna is one with Arjuna as Nārāyaṇa is with Nara.
Held agrees with Sylvain Lévi, who considered the Bhagavadgitā to be the central part of the epic ("le coeur et le noyau de l'ouvrage "). In the Gitā, Krşņa

Page #17

THE STUDY OF THE MAHĀBHĀRATA instructs Arjuna, the initiate, as to the place man occupies in the universe about him and the true purport of all human action. The battle is a rite and the scene of the battle is not an earthly field of war but a seat of strife in the sacred world, where the initiate, as yet ignorant of the cosmic significance of the conflict, is to be released by his initiator from the bandages that blindfold him and given an insight into the cosmic order (p. 335).
Kșşņa and Arjuna are on the side of the Pāņdavas. In them the idea of initiation is concentrated. On the side of the Kauravas there is Rudra-Siva, the heavenly gambler. The two functions embodied in those two parts are not mutually exclusive, nor are they diametrically opposed to each other, but they are complements the one of the other just as are the rival moieties of a tribe one of which is more engaged in religious activity and the other more in social activities. The game of dice in the epic is not a game of luck or of ability, because it is won by the Kauravas by deceit just as the asura-s were deprived of the amyta by fraud on the part of the deva-s. Yudhișthira is the Dharmarāja not in spite of, but exactly because of, his passion for dice. The game of dice in the epic is a potlatch; and the prestige of the players is the prize played for. That is why the Dharmarāja accepted the challenge with becoming pride (p. 327). The deceit practised by Duryodhana's uncle, Sakuni, is not a psychological or historical problem. Held remarks that deceit occurs whenever the poet feels called upon to credit either of the two groups, related to one another as are two phratries, with enjoying a position of superiority in respect of the other (p. 306). The mutual spirit of emulation and the deceit practised by both parties alike are the concomitants of a phratry-relationship (p. 331),
In his final remarks Held points out that the didactic and epic elements belong to each other genetically because the task of instruction is part and parcel of the process of initiation. The two elements were not artificially brought together and fused by a diaskeuast as maintained by Dahlmann. The logical connection between the didactic and the epical elements is founded upon the internal unity of the culture represented by the epic. The unity still perceptible in the epic is the unity of the form of the society with which the epic is genetically connected. Held believes that the form of human society with which the epic is genetically connected, is probably to be found in the period of the Brāhmaṇas. He adds that if by the so-called "epic" period is meant the entire period in which the Mahā. bhārata assumed its present shape, the time round about the rise of Buddhism must be taken as the terminus a quo.

Page #18

#W
IL*
8109)
NOTES
The Mahābhārata, Part I
* Mythe et épopée, I. Paris, 1968, p. 44, n. 1. 1. Cf. J. A. B. van Buitenen, The Mahabharata. Volume 1. Chicago, 1973, p. xxxi. 2. Ernst Windisch, Geschichte der Sanskrit-Philologie und indischen Altertumskunde. Erster Teil.
Strassburg, 1917, pp. 68-69. 3. G. J. Held, The Mahabharata. An Ethnological Study. Amsterdam, 1935, pp. 3-4; V. S. Sukt.
hankar, On the Meaning of the Mahabharata, Bombay, 1957, p. 5. 4. “Malgré les objections qu'on a déjà faites à cette manière de voir et qu'on y fera sans doute encore, je n'hésite pas pour mon compte à m'y ranger, estimant que de toutes les explications proposées, c'est toujours encore celle-ci en somme qui répond le mieux à l'ensemble des faits" (Revue critique, 20 juin 1874, p. 389; Oeuvres, III, 1917, p. 124). 5. E. Burnouf, Introduction à l'histoire du Buddhisme indien. Tome Premier. Paris, 1844, p. 136, n. 1. Windisch errs in stating that Lassen omitted in the second edition Burnouf's suggestion according to which the veneration of Krsna was due to a popular reaction to the spread of Buddha's doctrine (op. cit., p. 168). 6. A. Weber, Akademische Vorlesungen über Indische Literaturgeschichte. 2. A., Belin, 1876 p.
204. 7. Literary Remains, vol. I. London, 1879, p. 103. 8. Encyclopaedia of Religion and Ethics, volume VIII. Edinburgh, 1915, p. 326. 9. The Cambridge History of India. Volume I. Cambridge, 1922, p. 257. 10. Das Mahabharata. Seine Entstehung, sein Inhalt, seine Form. Göttingen, 1922, p. 10. 11. Louis Renou et Jean Filliozat, L'Inde classique, tome I. Paris, 1947, p. 384. 12. P. A. Grincer, Drevneindijskij épos. Moskva, 1974, pp. 153-174. 13. Op.cit., p. 171. 14. Op.cit., p. 173. 15. The Mahabharata. Volume 3. Chicago, 1978, p. 154. 16. Op. cit., p. 158. 17. Indische Alterthumskunde, 12, p. 585: "Es versteht sich von selbst, dass die Indische Sage auch alle solche Umbildungen erlitten hat, wie sie der noch frei waltenden Ueberlieferung überhaupt eigenthümlich sind, die Vergrösserung der Heldengestalten, die Zusammen. drängung ganzer Zeiten, Zustände und Richtungen in die Gestalt einzelner Personen und
in einzelne Handlungen..." 18. Cf. Karl Lachmann, Betrachtungen über Homers Ilias. 1847. 19. Barend A, van Nooten, The Mahabharata. New York, 1971. p. 44. 20. op. cit., p. 45. 21. The Mahābhārata. Volume 1. Chicago, 1973, p. xxxiii. 22. Cf. Ruling Caste' pp. 67-68; The Bhārata and the great Bhärata', American Journal
of Philology, XIX (1898), pp. 1-24. 23. Cf. Adolf Holtzmann, Zur Geschichte und Kritik des Mahabharata. Kiel, 1892, pp. 11-12. 24. Ueber das alte indische Epos. Durlach, 1881; Das Mahabharata und seine Theile. 4 Bde. Kiel,
1892, 1893, 1894, 1895. 25. A. Barth, Oeuvres, t. III, pp. 434-438; t. II, pp. 150-151; t. IV, pp. 352-353; G. Bühler und
J. Kirste, Contributions to the History of the Mahabharata', Sitzungsberichte der Wiener Akademie der Wissenschaften, vol. 127, no. 2 (1892), pp. 1-58.

Page #19
19 THE STUDY OF THE MAHABHARATA 26. For the opinions of G. Grierson and J. J. Meyer, see G. J. Held, The Mahabharata. An Ethnological Study. Amsterdam, 1935, p. 11. 27. 'Old Indian, The Two Sanskrit Epics', Traditions of Heroic and Epic Poetry. Volume One. London, 1980, p. 66. . 28. Ueber das verhaltnis des mythischen elementes zu der historischen grundlage des Mahabharata. Prag, 1884; Ueber die mythische grundlage des Mahabharata. Prag, 1895. Das Mahabharata als Epos und als Rechtsbuch Prag, 1896. 29. Cf. a passage from a recent book by Ninian Smart quoted by Karl Werner, JRAS (1982), p. 15. 30. A History of Indian Literature. Vol. I. Calcutta, 1927, p. 316, n. 1. 31. V. S. Sukthankar, On the Meaning of the Mahabharata. Bombay, 1957, p. 21. 32. Held refers several times to his paper on 'The Origin of the Divine Trickster '(De oor sprong van den goddelijke bedrieger. Amsterdam, 1929). 33. De Pandji roman, Antwerpen, 1922; 'Over de zin van het Javaansche drama', BTLV 81 (1925), pp. 311-384; Over den oorsprong van het Javaansche toneel', BTLV 88 (1931), pp. 317-450. These two articles were published in English translation in Panji, the cultural hero. The Hague, 1959: 'On the meaning of Javanese drama', pp. 1-61; 'On the origin of the Javanese theatre', pp. 93-215.