Recent Russian Publications On Indian Epic

Author(s): J W De Jong
Publisher: J W De Jong
Catalog link: https://jainqq.org/explore/269242/1

Page #1
7. W. DE JONG ECENT RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 1. Ja. V. Vasilkov, "The Mahabharata and oral epic poetry', Narody Azii i Afriki, 1971, no. 4, pp. 95-106. 2. Ja. V. Vasilkov, 'Elements of oral-poetic technique in the Mahabharata', Literatury Indii. Stat'i i soobscenija (Moscow, 1973), pp. 3-23. 3. P. A. Grintser, The old Indian epic. Genesis and typology. Moscow, 1974, 419 pp. The critical study of the Mahabharata (henceforth abridged M) and the Ramayana (abridged R) began in Europe in the beginning of the nineteenth century. The history of the study of the Indian epic has still to be written. Most of the older publications have been mentioned and discussed by Ernst Windischand by Moritz Winternitz. For many years Auguste Barth critically analysed the literature on the Indian epic.3 1 Geschichte der Sanskrit-Philologie und indischen Altertumskunde. I. Strassburg, 1917; II. Berlin-Leipzig, 1920. 2 A History of Indian Literature. I. Calcutta, 1927. * Cf. Oeuvres de Auguste Barth, vol. 5, Index general.

Page #2
2. THE ADYAR LIBRARY BULLETIN Of particular importance is his long article on J. Dahlmann's Das Mahabharata als Epos und Rechtsbuch which contains a brief history of M studies. It is more difficult to find information on the publications which have appeared since 1920. Very useful in this respect is an article by Oscar Botto on the Origins and development of the Indian epic' but he does not mention the work done by G. J. Held, Stig Wikander and Georges Dumezil.2 In the nineteenth century most European scholars were well versed in Greek and Roman literature. For them the classical epic par excellence was Homer's Iliad and Odyssey. It is therefore not surprising that in the study of the Indian epic the same methods were applied as those which were current among classical scholars. In his little book on the Indian Epic Poetry (London, 1863, pp. 9-10) Monier Williams expressed the opinion that the Ramayana had grown out of songs which chanted the praise of the chief of a body of Aryan invaders who forced their way into the peninsula of India as far as Sri Lanka. A similar theory had been 1 Op. cit., vol. 3 (Paris, 1917), pp. 347-403 (first published in 1897). For older literature on M see also Adolf Holtzmann, Das Mahabharata im Osten und Westen (Kiel, 1895), pp. 165-204. 2 Oscar Botto, Origini e sviluppo dell'epica indiana ', Atti del Convegno Internazionale sur tema: 'La Poesia epica e la sua formazione (Roma. 1970), pp. 655-77; G. J. Held, The Mahabharata. An Ethnological Study. Amsterdam, 1935; Wikander's articles are mentioned by Georges Dumezil, cf. Indo-Iranian Journal, vol. 3 (1959), pp. . 1-16; Georges Dumezil, Mythe et epopee, I, II, III, Paris, 1968, 1971, 1973.

Page #3
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 3 put forward by Karl Lachmann (1793-1851) with regard to the Iliad which he considered to be an amalgam of popular lays. Later Homeric scholars rejected Lachmann's theory and tried to discover the original nucleus of the Homeric poems. Similar attempts have been made by S. Soerensen in his Om Mahabharata's stilling i den indiske literatur (Copenhagen, 1883) in which he arrived at a nucleus of about 7000 sloka-s. Well known is also Jacobi's attempt to exclude many interpolated passages from R.1 Soerensen and other scholars such as A. Holtzmann jun. and E. W. Hopkins who, although in a much less extreme way, tried to separate older and younger layers in the epic, applied the same methods as the ones used by Homeric specialists with regard to the Iliad and the Odyssey. In the domain of Homeric studies the 'Analysts' held the field until the end of the First World War, although unitarianism had never been entirely dead. The unity of M was vigorously defended at the end of the nineteenth century by Joseph Dahlmann.3 Dahlmann's synthetic theory, as it was called by E. W. Hopkins in the preface to his book, The Great Epic of India (New York, 1901), failed to convince the great majority of scholars and no attempts were made to revive the unitarian theory until recently. Held, an ethnologist, assumed that . Das Ramayana, Bonn, 1893; The Ramayana, Baroda, 1960.. 2 Cf. E. R. Dodds, 'Homer ', Fifty Years of Classical Scholarship (London, 1954), pp. 1-17 and 31-35. 3 Das Mahabharata als Epos und Rechtsbuch, Berlin, 1895; Genesis des Mahabharata, Berlin, 1899.

Page #4
THE ADYAR LIBRARY BULLETIN M reflected a dualistic organization of the society in which it originated. In recent years Wikander and Dumezil have tried to point out the structural unity of M by applying the theory of the tripartite ideology which Dumezil has developed in a long series of publications which culminate in the three volumes mentioned above (note 2, p. 2). In the field of Homeric scholarship the publication of two books by a young American scholar Milman Parry (1902-35) marked a new epoch. Both books were published in 1971 in an English translation in The Making of Homeric Verse. The Collected Papers of Milman Parry (Oxford University Press, 1971). In The Traditional Epithet in Homer, Parry examined the numerous epithets or adjectival phrases which together with the names of the chief characters of the epic form fixed formulas. According to Parry's definition a formula is a group of words which is regularly employed under the same metrical condition to express a given essential idea'.2 Parry stressed the fact that the traditional character of the formulas lies in the fact that they constitute a system distinguished at once by a great extension and by great simplicity (op. cit., p. 16). Replacing the terms 'extension' and simplicity' by length and thrift Parry describes this system as follows: 'The length of a system consists very obviously in the number of formulas which make it 1 L'Epithete traditionnelle dans Homere, Paris, 1928; Les formules et la metrique d'Homere, Paris, 1928. 2 The Making of Homeric Verse, p. 272.

Page #5
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 5 up. The thrift of a system lies in the degree in which it is free of phrases which, having the same metrical value and expressing the same idea, could replace one another.' (op. cit., p. 276). Parry pointed out that this complicated technique could only be the product of a time in which the bardic tradition was in its full vigour (op. cit., p. 175). Parry's ideas, when first published in 1928, did not meet with much approval from the part of classical scholars with very few exceptions. However, since the Second World War, the situation has changed completely and in 1954 E.R. Dodds declared without hesitation that Parry had furnished the decisive proof that the Homeric poems are oral compositions (op. cit., p. 13). In his very instructive introduction to Milman Parry's Collected Papers, Adam Parry points out that it is more correct to say that Parry proved that the style of the Iliad and the Odyssey is typical of oral poetry (p. lxi, n. 1). Parry used the word 'Homer' as meaning sometimes the poet' or 'the poets of the Iliad and Odyssey' and sometimes the traditional text of the Iliad and Odyssey'. (op. cit., p. 191, n. 2). Parry's studies did not prove that Homer himself was an oral poet. According to Adam Parry it is quite conceivable that Homer made use of writing to compose a poem in a style which had been developed by an oral tradition (op. cit., p. 1xi, n. 1). Pierre Chantraine, Revue de philologie, 55 (1929), pp. 294-300; Martin P. Nilsson, Homer and Mycenae (London, 1933), pp. 179 ff.

Page #6
THE ADYAR LIBRARY BULLETIN Although the main conclusions of Parry's earlier work on The Traditional Epithet in Homer and on Homeric Formulae and Homeric Metre have been accepted by most of the Homeric specialists, a lively discussion is still continuing over theories developed by Parry in his later work and further extended by A. B. Lord, J. A. Notopoulos and others. In 1930 Parry published a long article, entitled "Studies in the Epic Technique of Oral Verse-Making. I. Homer and Homeric Style'. Part II: "The Homeric Language as the Language of an Oral Poetry' appeared two years later.1 In this article Parry tried to show that the Iliad and the Odyssey are composed in a traditional style and are composed orally. Secondly he examined the differences between oral and written poetry in style and form. Although Parry maintained the same definition of the formula as given by him in his study on the traditional epithet, in practice he extended the concept of the formula as is obvious from his analysis of the first twenty-five verses of the Iliad and the Odyssey. (op. cit., pp. 301-4). Parry underlines expressions which are found elsewhere in Homer and puts a broken line under phrases which are of the same type as others. In this way Parry 'brought into the category of formulas not only the repeated expressions, but those which are of the same type as others' (op. cit., p. 313). Parry adds that the formulaic element is not limited to those formulas 1 Reprinted in The Making of Homeric Verse, pp. 266-324 and 325-64.

Page #7
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 7 but that there are more general types of formulas. In this way Parry showed that almost the whole of Homer is traditional and oral. Scholars have objected against the use of the term 'formulaic' by Parry.1 Other scholars have been even more bold than Parry. In his analysis of the first fifteen lines of the Iliad Lord even considers very current combinations of pronouns and particles as formulaic.2 J. A. Notopoulos and G. S. Kirk have extended the use of the term 'formula' to single words.3 Parry regarded as formulas, or possible formulas, only expressions made up of at least four words or five syllables, with the exception of noun-epithet phrases (op. cit., p. 275, n. 1). It is certainly necessary to stress the fact that there is still a great measure of disagreement between Homeric scholars as to the definitions of the terms ' formula' and 'formulaic' because it has become usual to refer to the Parry-Lord theory as having been generally accepted. As we have seen, Parry himself modified his theories between 1928 and 1930-32 and 1 Cf. A. Hoekstra, Homeric Modifications of Formulaic Prototypes. Amsterdam, 1965, p. 15; J. B. Hainsworth, The Flexibility of the Homeric Formula, Oxford, 1968, pp. 16-17. 2 A. B. Lord, The Singer of Tales (Cambridge, Mass., 1960), p. 143. Cf. Hoekstra, op. cit., p. 15. 3G. S. Kirk, The Songs of Homer (Cambridge, 1962), p. 67; J. A. Notopoulos, Am. J. Phil., 83 (1962), pp. 355 ff. and Harvard Studies of Classical Philology, 68 (1964), pp. 28 ff. Cf. the critical remarks made by Hainsworth, op. cit., p. 35, n. 4 and Hoekstra, op. cit., pp. 14, 15 and 20. 4 Cf. N. Sen, JAOS,, 86 (1966), p. 399.

Page #8
& THE ADYAR LIBRARY BULLETIN his pupil A. B. Lord has further developed Parry's theories. A very lucid discussion of Parry's theories, their reception and the reactions which they have provoked is to be found in Adam Parry's introduction to Parry's collected papers. In the concluding paragraph of his studies on the epic technique of oral verse-making Parry pointed out the necessity of studying other oral poetry where the processes of composition can be observed in actual practice. In 1933-5 Parry collected more than 12,500 texts in Jugoslavia. In the autumn of 1935 he did begin a book to which he gave the title "The Singer of Tales' but his death on 3rd December 1935 interrupted his work, when he had completed only seven pages which have been published by Lord.1 Lord has continued Parry's work and in 1960 he published The Singer of Tales in which he studies the processes of composition of oral narrative poetry, the manner in which the singers compose, learn and transmit their epics (cf. Foreword, p. vii). Lord's book is well known and is often quoted by scholars. He devotes one chapter of the book to the theme, a topic which Parry had already studied. Following Parry, Lord calls a group of ideas regularly used in telling a tale in the formulaic style of traditional songs, a theme. To quote Adam Parry's introduction: The theme is a sort of basic unit of narration in an oral poem. It may be a unit of action: a single combat, the calling 1 The Making of Homeric Verse, pp. 469-73.

Page #9
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 9 of an assembly, the arrival at a palace; or it may be a description of arms, or a chariot, or a feast. It is clear that such themes recur often in the Iliad and Odyssey, indeed that the poems are to some extent made up of them. Parry, as we can see clearly in his review of Walter Arend in this volume, and others after him, saw this as a distinguishing characteristic of oral poetry.' (p. xli). The importance of the theme in oral poetry had already been pointed out very clearly by the Russian scholar V. V. Radlov in the introduction to the fifth volume of his Proben der Volksliteratur der turkischen Stamme Sudsibiriens (St. Petersburg, 1885, pp. xvi ff.). Radlov's description of the performance of a Kirghiz singer is quoted by Parry (op. cit., p. 334, n. 1). Nora Chadwick gives an English translation from which I reproduce the following passage: Every minstrel who has any skill at all always improvises his songs according to the inspiration of the moment, so that he is not in a position to recite a song twice in exactly the same form; but one must not suppose that this process of improvisation involves composing a new poem every time. The procedure of the improvising minstrel is exactly like that of the pianist. As the latter puts together into a harmonious form different runs which are known to him, transitions and motifs according to the inspiration of the moment, and thus makes up the new from the old which is familiar to him, so also does the minstrel of epic poems. Through an extensive practice in production, he has whole series of elements of production', if I may so express

Page #10
10 THE ADYAR LIBRARY BULLETIN it, in readiness, which he puts together in suitable manner according to the course of the narrative. Such ' elements of production consist of pictures of certain occurrences and situations, such as the birth of a hero, the growing up of the hero, the glories of weapons, preparations for battle, the storm of battle, the conversations between heroes before battle, the depicting of characteristics of persons and of horses, the characterization of the well-known heroes, the praise of the beauty of the bride. . . The art of the singer consists only in arranging all these static component parts of pictures with one another as circumstances require, and in connecting them with lines invented for the occasion. "Now the minstrel can utilize in his singing all the formative elements specified above in very different ways. He knows how to represent one and the same picture in a few short strokes. He can depict it more fully, or he can go into a very detailed description with epic fullness. The greater the number of different formative elements at the disposal of the minstrel, the more diversified will be his performance, and the longer will he be able to sing without tiring his listeners by the monotony of his descriptions. The amount of the formative elements and the skill in putting them together is the measure of the skill of the minstrel. A skilled minstrel can recite any theme he wants, any story that is desired, extempore, provided that the course of events is clear to him. When I asked one of the most accomplished minstrels whom I had got to know if he could sing this or that song, he answered

Page #11
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 11 me: 'I can sing any song whatever; for God has implanted this gift of song in my heart. He gives me the word on my tongue, without my having to seek it. I have learnt none of my songs. All spring from my inner self.' And the man was right. The improvising minstrel sings without reflection, simply from his inner being, that which is known to him, as soon as the incentive to singing comes to him from without, just as the words flow from the tongue of a speaker without his producing intentionally and consciously the articulations necessary to produce them, as soon as the course of his thoughts requires this or that word. The accomplished minstrel can sing a day, or a week, or a month, just as he can speak, and narrate all the time.'1 Radlov's 'elements of production' (Vortragsteile in German) is exactly the theme as defined by Lord. Following Lord we can therefore distinguish three elements in oral poetry: formulas, patterns and themes. Instead of patterns Lord uses the term "formulaic expression' to denote a line or a half line constructed on the pattern of the formula (op. cit., p. 4). As Hainsworth remarks, the term is not very happy, as being virtually tautologous with 'formula' (op. cit., p. 18, n. 3). He uses 'formula' for the repetition of a word-group and pattern ' and 'theme' for repetitions respectively in the arrangement of words and the subject-matter of the poems (op. cit., p. 35). 1 Nora K. Chadwick and Victor Zhirmunsky, Oral Epics of Central Asia (Cambridge, 1969), pp. 222-4.

Page #12
12 THE ADYAR LIBRARY BULLETIN Parry's theories on the oral nature of the Homeric poems have been applied by other scholars to the Beowulf, the Chanson de Roland, and the Bible. C. M. Bowra in his Heroic Poetry (London, 1952) has also been greatly influenced by Parry's ideas. However, Bowra excluded the Indian epics from his study. In 1958 M. B. Emeneau drew attention to the marks of oral composition and of a very early synthesis of numerous oral recitations into one unified text in both M and R,1 and in 1964 R. K. Sharma devoted a brief chapter to the technique of oral poetry in M, but the first systematic application of Parry's formulaic analysis is to be found in an article published by N. Sen in 1966.3 Sen selected five sections of twenty-six lines each. He underlined the groups of words that are repeated elsewhere in the same Kanda and put a broken line under the 'formulaic'expressions (i.e. those constructed in the same syntactic, rhythmic, and word-boundary patterns as another line, half-line, or part of a line, with at least one common word in a similar position metrically). Finally, he indicated with dots the words which show a distinct formulaic tendency. In his article on M and oral epic poetry Vasil'kov refers to the studies by Emeneau, Sharma and Sen mentioned 1. Oral Poets of South India: The Todas', Journal of American Folklore, 71 (1958), pp. 313-4. 2 Elements of Poetry in the Mahabh rata (Berkeley and Los Angeles, 1964), pp. 167-75. 3 Comparative Studies in Oral Epic and the Valmiki Rama. yana: A Report on the Balakanda', JAOS, 86 (1966), pp. 397-409.

Page #13
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 13 . above. He draws attention to the fact that the literary historian A. N. Veselovsky (1838-1906) pointed out that it was wrong to transfer problems of purely literary criticism to problems of folk-poetry in the ancient period and that it was necessary to take as a startingpoint the epic which was still being sung and to study thoroughly its structure and the stages of its development. However, Vasilkov's main source of inspiration for the study of the characteristics of oral poetry in Mis Lord's The Singer of Tales. Lord, and before him, Parry distinguished unperiodic enjambement from periodic or necessary enjambement. The first is characteristic of oral composition. The sentence, at the verse end, gives already a complete thought but it continues in the next verse, freely adding ideas by new word groups. When there is no complete thought expressed at the verse end, enjambement is necessary. This is characteristic of literary style. As examples of unperiodic enjambement in M, Vasil'kov adduces 3.104.10-11 and 3.39.12-13 (crit. ed.). In more detail Vasil'kov studies the importance of thematic analysis and enumerates such important themes as duels, 1 Victor Zhirmunsky quoted the same passage in N. K. Chadwick and V. Zhirmunsky, Oral Epics of Central Asia (Cambridge 1969), p. 319. 2 Parry, 'The Distinctive Character of Enjambement in Homeric Verse', The Making of Homeric Verse, pp. 251-65; A. B. Lord, 'Homer and Huso III: Enjambement in Greek and Southslavic Heroic Song', Trans. Am. Phil. Ass., 79 (1948), pp. 113-24; Lord, The Singer of Tales, pp. 54 and 131.

Page #14
14 THE ADYAR LIBRARY BULLETIN ascetic exploits, requests made to story-tellers, receptions of guests, descriptions of nature, etc. He points out that a theme can be briefly stated in one or two sloka-s or developed at great length. The themes are often introduced by a formulaic expression. For instance, the theme asceticism' is introduced by a formulaic expression containing the word tapas, the theme ' battle' by an expression containing the word yuddham. A theme such as that of asceticism' is further developed with the help of formulas and formulaic expressions. By way of example Vasil'kov analyses adhyaya 163 of book 3 into the following themes: request to the story-teller, asceticism, battle, the god-donor (varada). The same topic is repeated in adhyaya-s 39 and 40 of book 3 but with the addition of other themes: description of nature and the request of the rsi-s to Siva for protection. Vasil'kov draws attention to the fact that themes are traditionally associated in the mind of the singers: 'In a traditional poem, therefore, there is a pull in two directions: one is towards the song being sung and the other is toward the previous uses of the same theme. The result is that characteristic of oral poetry which literary scholars have found hardest to understand and to accept, namely an occasional inconsistency, the famous nod of Homer.'1 Vasil'kov remarks that from the point of view of literary criticism one of two conflicting versions of the same story must be the original. However, oral performance admits Lord, The Singer of Tales, p. 94.

Page #15
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 15 the possibility of inconsistencies, but oral composition makes them inevitable. An oral poet, while relating again the same story, chooses a different way of elaborating on it. As examples of inconsistencies in the treatment of the same topic, Vasilkov refers to the descriptions of the cremation of Pannu in adhyaya-s 116 and 118 of book 1 and to the descriptions of the flight of the Pandava-s from the jatugrha in 1.136 and 3.13. The bard, the singer (Greek: aoidos) improvises with the help of formulas, formulaic expressions and themes. The rhapsode, to the contrary, recites songs which he has learnt by heart.1 Vasilkov remarks that in the past scholars tended to limit the concept of oral poetry to poetry recited by rhapsodes. However, in his opinion, there is no doubt that the M itself refers to bards and not to rhapsodes. Stories which are repeated in M can be divided into two groups according to Vasil'kov. Firstly, those which are told in chapters which are separated from each other. Secondly, stories which are first told briefly and, subsequently, in a more developed form. As an example of the first category Vasil'kov refers to 3.163 and 3.39-40 which both tell the story of the battle of Arjuna with the Kirata. Just as a theme is often introduced by a formulaic expression, the kernel of the topic to be related is often indicated in the first verses (cf. 3.163.3 and 3.39.1,3). As an example of the second category 1 The function of the rhapsode in ancient Greece is not completely clear (cf. J. B. Hainsworth, op. cit., p. 2, n. 2).

Page #16
16 : THE ADYAR LIBRARY BULLETIN Vasil'kov refers to the story of Astika, first told briefly in 1.13, and, thereupon in great detail (1.14-53) just as the topic of M itself is first told 'briefly in 1.55 and, thereupon, in the 18 books of the epic. Vasil'kov remarks that it would be impossible for a rhapsode to memorize the complete text of M, but that it is quite possible for a bard to sing a poem of this length. For instance, the Kirghiz bard Orozbakov (1876-1930) sang a version of the epic "Manas? which contained about 200,000 lines and also the sequels which are of about the same length.1 In M it is often said that the singer satisfies two conditions. He tells his story truthfully (tattvena, yathatatham, yathavittam, yathatattvam, etc.) and in detail (vistarena). In the description of battle scenes the singer frequently uses the word tatah and repeats a word or words from the preceding sloka. According to Vasil'kov these devices facilitate the choice of formulaic expressions when the tempo of the narration is accelerated. 1 Vasil'kov gives a reference to Zhirmunsky's paper on the epic folk-singers in Central Asia (VII Congres international des sciences anthropologiques et ethnologiques. Moscou 3 aout-10 aout 1964, volume VI, Moskva, 1969, pp. 234-41). However, according to Zhirmunsky Saghymbai Orozbakov dictated by memory about 200,000 verses of his version of the epic 'Manas' whereas Sayakbai Karalayev (born in 1904) can recite the same number of verses of this epic and also about 200,000 verses of the sequels, which deal with the son and the grandson of Manas, the 'Semetei ' and the Seyetek (p. 236). Elsewhere Zhirmunsky attributes about 250 000 verses to the versions of the epic 'Manas' dictated by these two singers (Nora K. Chadwick and V'ctor Zhirmunsky, Oral Epics of Central Asia, pp. 279-80).

Page #17
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 17 Vasilkov's article is of special importance for his remarks on the treatment of the theme in M and on the stylistic devices used by the bards. In his second article Vasil'kov examines the formulaic technique of the epic. He refers to an article by P. Grintser on the epic formula in M, published in 1971.1 In this article Grintser points out the importance of the formulaic endings of pada-s and of the function of supporting' words in the formation of these endings. According to Vasil'kov a supporting word can belong to two categories: semantically positive words, for instance, a subject in the nominative (maharaja, etc.) or semantically neutral, additional words of which the only function consists in completing the pada. As already remarked by Grintser the formulas of the epic fall into two groups, the first occurring in uneven pada-s (a and c) and the second in even pada-s (b and d). As examples of words of the first category in a and c Vasil'kov quotes nouns: raja, virah, samgramah, yuddham and proper names: parthah, kaunteyah, and as examples of the second category vocatives: rajan, maharaja and locatives: yuddhe, samgrame. To both categories belong Tatpurusa-s: ksatriyasresthah, yadusresthah, Bahuvrihi-s: dharmatma and compounds with maha as first element: mahabahuh, mahaprajnah, etc. Furthermore participles and adjectives can be used as supporting words. For instance, kruddha tends to be used with proper names: arjuno bhrsasamkruddhah, arjunas capi samkruddhah, bhimasenas Published in a volume entitled Struktura drevneindijskogo teksta. Tezisy which I have not been able to consult.

Page #18
18 THE ADYAR LIBRARY BULLETIN tu samkruddhah, or with the verb abhi-dru- and synonyms: abhyadravanta samkruddhah, abhyadhavanta samkruddhah, etc. Also often used at the end of pada-s are ornamental epithets: ghora, dhimat, srimat and Srestha (in combination with balinam, vadatam, dhanvinam, etc.). Adverbs are also used as supporting words: vegena in abhidudrava vegena, abhyadravanta vegena, etc.; turnam in aruroha ratham turnam, athasya ratham turnam, etc. Finally, the last word of an uneven pada is often a gerund: adaya, asthaya, asadya, krtva, drstva, gatva, etc.: divyam tad dhanur adaya, punah svaratham asthaya, etc. In even pada-s the following categories of words are to be found: compounds with maha- as first element (with a short penultimate syllable): mahabala, maharatha, expressions with a short penultimate syllable: tapodhana, visampate, marisa; adjectives: uttama, viryavan, pratapavan; ' additional ' locatives: maharane, mahahave, mtdhe, samyuge; adverbs: bhrsam, dudham and adverbs ending in -sah: sarvasah, bhagasah, anekasah, etc. As example of a sloka in which all pada-s end with a 'supporting word ' Vasil'kov quotes M 8.11.37: tatah kruddhau MAHARAJA banau glhya MAHAHAVE I ubhau ciksipatus TURNAM anyonyasya VADHAISINAU 11 Supporting words tend to be used with certain categories of words. This explains the formation of pure formulas ', i.e. formulas which occupy an entire pada, for instance, athanyad dhanur adaya, sataso 'tha sahasrasah, tasthau giririvacalah, ksitikampe yathacalah, chinnamula iva drumah, vatoddhuta iva drumah. Vasilkov

Page #19
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 19 remarks that the number of pure formulas' in M is not very great and is extremely small when compared to the number of pada-s which contain only a supporting word as formulaic element. In the sloka especially the fourth pada contains formulas, for instance: tad adbhutam ivabhavat, param kautuhalam hi me. Often this pada contains comparisons: simhah ksudram/gam yatha (-mtgan iva), garutman iva pannagam, etc. Developed formulas can occupy two pada-s, three pada-s or even an entire sloka. Vasil'kov divides them into two groups of which the first consists of often used phrases (relating to battles, asceticism, descriptions of nature, etc.). The component parts are formulaic pada-s which can be used also independently. Sometimes a developed formula is a more 'rigid' formula-cliche, for instance: atrapy udaharantimam itihasam puratanam prahladasya ca samvadam muner angirasasya ca (2.61.58) baler vairocanasya ca (3. 29. 1) muner ajagarasya ca (12.172.2) Other proper names can be combined with the supporting word samvada in pada c: indrakasyapasamvadam (12.173.4), bihaspates ca samvadam (14.6.1). The supporting word can even be absent as in the following two sloka-s: atrapy udaharantiMAM ITIHASAN PURATANAM I agastyasya mahayajne puravsTTAM arimdama 11 (14.95.4)

Page #20
20 THE ADYAR LIBRARY BULLETIN srnu me vistareNEMAM ITIHASAM PURATANAM | bhratrbhih sahitah partha yathavRTTAM yudhisthira || (1.201.1) An even greater variety in wording is to be found in formulaic expressions relating to battles. In many of these expressions the only common elements are: the word puddha or a synonym; the predicate there was, there happened' (asit, abhut, pravavrte, etc.) and finally, words relating to the word 'battle'. These words are of three types: (I) names or epithets of participants in the battle in the genitive: kurunam pandavanam ca, kurupandavasainyanam, etc.; (II) adjectives qualifying the word 'battle': tumulam lomaharsanam, ghorarupam bhayanakam, etc.; (III) traditional comparisons of the following type: indra (sakra) prahladayor iva, bali (vrtra)vasavayor iva. Instead of the word yuddha, samgrama can be used: tatah pravavrte bhuyah samgramo lomaharsanah tavakanam paresam ca samare vijigisatam || (6.53.34) On the model of such verses, other verses can be formed in order to express entirely different ideas. For instance, yuddha can be replaced by sabda: tatah sabdo mahan asit tumulo lomaharsanah (3.270.9) and by varsam: tato varsam pradur abhut sumahal lomaharsanam (3.168.12). The second class of developed formulas is represented by some very rare examples of formulas which are almost cliches and which are used with reference to certain individuals. The first example refers to Bhima:

Page #21
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 21 vinadantam mahanadam bhinnabherisamasvanam i bhramayam asa suciram visphurantam acetasam 11 (3.12.62) This sloka is found with the variant nadantam ca for vinadantam in the Bombay edition 4.21.75. The second example is found in a speech addressed by Vidura to Dhitarastra (2.57.15): na sreyase niyate mandabuddhih stri srotriyasyeva glhe pradusta i dhruvam na roced bharatarsabhasya patih kumarya iva sastivarsah 11. The same formula occurs in 3.6.15 with the variant jatasatro for mandabuddhih and bruvan na rucyai for dhruvam na roced. Improvisation is characteristic for the oral poetic tradition. However, formulas of the second class seem to have been repeated almost literally. Are these formulas the product of a literary tradition or did they arise during the rhapsodic period of existence of the epic when the text was learnt by heart by rhapsodes ? Vasilkov believes that both possibilities have to be ruled out. The first because the variations (dhruvam na roced--bruvan na rucyai) are phonetic and phonetic variations are more characteristic for the oral tradition than for the written tradition. The second is to be excluded because cliches do not extend over more lengthy passages as could be expected in the case of a text learnt by rhapsodes. Vasilkoy concludes by pointing out that the basic elements of the oral poetic

Page #22
22 THE ADYAR LIBRARY BULLETIN technique in M are the supporting words whereas pure formulas play only a subordinate role. We have analysed in some detail Vasil'kov's two articles. It will not be possible to do the same for Grintser's book. His book is divided into two parts. The first is entitled 'oral and written tradition in the old Indian epic', the second 'typology of the old Indian epic'. The second part is a wide ranging study of the Indian epic in comparison with other epics in which fundamental problems concerning the relations between folktale, myth and epic are studied. This part of Grintser's book is much more speculative than the first part. Even a bare summary would be beyond the scope of the present article. For this reason we will limit ourselves mainly to the first part in which Grintser examines the characteristics of oral epic technique in the two Indian epics. Grintser has published in 1960 a translation of the Vikramacarita according to the Southern recension, in 1963 a study on the Indian frame-story and in 1970 a little book for the general reader on the Mahabharata and the Ramayana.1 From 1969 onwards he has published several articles on the epic which foreshadow the main conclusions at which he has arrived in his book. The first chapter of part 1 deals with the oral performance of the epic and the performers. Most of the 1 Zhizn' Vikramy ili 32 istorii tsars' ogo trona. Moskva, 1960; Drevne-indijskaja proza (obramlennaja povest'). Moskva, 1963; 'Mahabharata' i 'Ramajana'. Moskva, 1970.

Page #23
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 23 materials collected by G. (=Grintser) on the suta-s, magadha-s, etc. and the recitation of the epic during sacrifices, etc. have already been amply discussed by previous scholars. More important is the second chapter on the epic formulas, which contains the most detailed formulaic analysis of the Indian epic yet undertaken. G. divides the formulas into six groups: (I) attributive formulas: kuntiputro yudhisthirah, bhimo bhimaparakramah, purusavyaghrah, nararsabhah, mahabalah, mahesvasah, dharmajnah, etc.; (II) narrative formulas: svabahubalam asritah, kytva pradaksinam, vavarsa saravarsani, abhidudrava vegena, tatah samabhavad yuddham tumulam lomaharsanam, etc.; (III) auxiliary formulas: etasminn eva kale tu, atha ratryam vyatitayam, sataso 'tha sahasrasah; (IV) formulas of direct speech: nesyami/nayami yamasadanam, param kautuhalam hi me, tistha tistheti cabravit, srotum icchami, srnu rajan, etc.; (V) maxims: esa dharmah sanatanah, yato dharmas tato jayah, sarve ksayanta nicayah patanantah samucchrayah|samyoga viprayoganta maranantam ca jivitam, punnamno narakad yasmat pitaram trayate sutah | tasmat putra iti proktah pitin yah pati sarvatah, etc.; (VI) similes: purnacandranibhanana, parnendusadesanana, devagarbhabhah, dandapanir ivantakah, etc. I have quoted only a few examples from the many given by G. In an appendix G. lists 594 formulas, indicating whether they occur in even or uneven pada-s. In a brief section on the metre in the epic (pp. 47-52) G., referring to calculations made by Hopkins, points out that 87.5 per cent of the sloka-s are pathya-s. Consequently most uneven pada-s end on v-. whereas all even

Page #24
24 THE ADYAR LIBRARY BULLETIN pada-s end on vov. Formulas are to be found mainly in the second halves of the pada-s because of the greater rigidity of their metrical scheme. Moreover, the metrical difference between the even and uneven pada-s results in the presence of two different sets of formulas. G. lists a number of examples on pp. 53-7 and adds a more complete list of 97 examples in a second appendix. Let us quote a few examples (the first is to be found in even pada-s, the second in uneven pada-s): x x yudhisthiro raja-dharmaputro (or kuntiputro) yudhisthirah; laksmano laksmisampannah--laksmano laksmivardhanah; x x x nara-(or manuja- or kuru-sardulah; X X X nara-(or manuja- or kururam) rsabhah; rajan-mahipate; rajendra-naradhipa; turnam--bhrsam; drstva--pasyatah. As is obvious from the examples, G. considers even single words as formulas as do Kirk and Sen. G. remarks that usually the basic element of the formula is a supporting word of two or four syllables in the second half of the pada, for instance: isabha with which the following formulas are formed: purusarsabha, nararsabha, etc.; sardula--nrpasardula, rajasardula, etc.; vara--(sarva-) sastrabhitam varah, jayatam varah, etc.; madhyama --sumadhyama; bhusitah--sarvabharanabhusitah; locana-kamalalocanah, rajivalocanah, abravit-idam vacanam abravit, etc. G. remarks that many formulas are not only synonymous but also metrically equivalent. The formulaic system of the Indian epic is not characterized by simplicity as is the Homeric system of name-epithet formulas. Moreover, the Indian epic allows the construction of formulas of different meaning by means of the same supporting

Page #25
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 25. word, for example: qsabha-bharatarsabhah, vanararsabhah; sattama--dvijasattamah, rajasattamah; sampanna--sarvalaksanasampannah, rupayauvanasampannah; tejas-amitatejasah, agnisamatejasah, etc. Finally, G. examines the formation of formulaic expressions on the model of the usual formulas: abhyadhavata samkruddhah (formula)-pidayamasa samkruddho (formulaic expression); x x x nihatam dustva--Saktim vinihatam drstva, vinitan vrsabhan dustva; srotum icchami-gantum icchasi, sthatum icchati, yoddhum icchami. Furthermore, the following supporting words serve for the formulation of formulas: turnam-avatirya rathat turnam; sahasa--vinardamanah sahasa; kathamcana--na gacchami kathamcana, etc. Parry had noted the fact that the use of formulas for metrical convenience resulted sometimes in their being applied in cases in which their meaning does not fit the context. Examples can also be found in the Indian epic, for instance: paradarasmi bhadram te (M 4.13.14a), the comparison of Uttara who follows Arjuna to a gajavadhu: sisum gajavadhur iva (M 4.35.9). However, G. does not agree with several examples adduced by Sharma (op. cit., p. 11) and S. N. Gajendragadkar.1 According to him the surprising fact is not that formulas are used inappropriately but that this happens only very rarely. In the following chapter G. analyses a chapter of M in the same way as has been done by Parry and 1'A Study in Mahabharata Similes', Journal of the University of Bombay, vol. 19, part 2 (1950).

Page #26
26 THE ADYAR LIBRARY BULLETIN Lord. G. has selected for his purpose chapter 49 of book 6 which describes a duel between Drona and Dhrstadyumna. G. indicates by bold type fixed formulas, by spacing formulaic expressions and by italics words which on the basis of the materials, collected by him, cannot with certainty be said to be true formulas but of which the metrical position is sufficiently fixed. G. arrives at the conclusion that only 23 out of the 164 pada-s are entirely free from formulas and that 77 per cent of the entire passage consists of formulas and formulaic expressions. About the same percentage is found by him in other battle chapters (book 5 chapter 185: 84 per cent) whereas the percentage is slightly smaller in battle chapters of R (book 6, chapters 44 and 98: 80 per cent). According to G. the number of formulas is the greatest in battle scenes. In narrative chapters the percentages found by him are 40 per cent (M book 4, chapter 22; book 3, chapter 22) and 50 per cent (R 2.59 and 6.61). The didactic portions are poorer in formulas and formulaic expressions but they contain many maxims which occur in the same or in slightly different form in other texts such as Purana-s, Sastra-s, frame-stories, etc. For instance, according to Buhler about one-tenth of the Manusmrti is found in the third, twelfth and thirteenth books of M.1 G. draws attention to the fact that Sen had arrived 1 The Laws of Manu (Oxford, 1886), p. lxxx. See also E. W. Hopkins, The Great Epic of India (New York, 1902), pp. 19-23; A. Holtzmann, Das Mahabharata im Osten und Westen (Kiel, 1895), pp. 114-50.

Page #27
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 27 at a lower percentage of formulas, i.e. 39.6 per cent but he remarks that Sen has used only one book (R book 1). Moreover, G. taxes him with inconsistency in the application of his principles, for instance in saying that the use of dhitiman (R 1.1.8d: dyutiman dhrtiman vasi) is formulaic in 1.21.6c (dhitiman suvratah sriman) and in 1.28.3d (dhitiman suvratah sucih) because in these cases dhitiman occurs either in different pada-s or in different positions within the pada. G. is even less convinced by the calculations made by J. L. Brockington with regard to stereotyped expressions in books 2, 3, and 4 of R.2 According to Brockington more than 1 in 30 pada-s in book 2 (3.4 per cent) are stereotyped, about 1 in 21 pada-s in book 3 (4.9 per cent) and over 1 in 16 in book 4 (6.3 per cent). According to G. the great number of formulaic phrases in the Indian epic proves its oral character. Formulas, remembered by the singer, and formulaic expressions, created according to established models, were an indispensable instrument for the composition of epic verses in a given metre. The formulaic style is not characteristic of classical poetry. G. points out that comparisons of the eyes with lotuses, etc. are not formulas but cliches, standard images. Also the limitations which the choice of a metre imposes on a poet must not be confounded with the use of formulaic 1 G. mentions also 1.71.7c (mahaviryasya dhitiman) but Sen does not say that here its use is formulaic (op. cit., p. 402). 2. Stereotyped Expressions in the Ramayana', JAOS, 90 (1970), pp. 210-27.

Page #28
28 THE ADYAR LIBRARY BULLETIN expressions. As example G. mentions Kalidasa's Meghaduta, in which words such as bhavanti, etc. can be used only in the second part of the verse. G. adds that many epic formulas do not necessarily have to be used only in the metrical position for which they are destined, for instance mahabala can also be used in the first half of a pada: mahabalam mahabahum, mahabalam mahaprajnam, etc. However, although the number of such deviations is considerable, nevertheless, on the whole, it is insignificantly small in comparison with the number of formulas which are found in their own' usual positions. In the last two pages of this chapter G. raises a very important problem on which the opinions of scholars differ greatly, i.e. the problem of transitional texts. According to G. the text of the Indian epics, in the form in which it has been transmitted in manuscripts, is not a simple transcription of some oral performance. G. remarks that several Homeric scholars believe that between the period in which the oral poetry flourished and the period in which the epic text was written, there was a more or less lengthy period of time in the course of which the poem received a fixed 'transitional' form. During that period the poem was transmitted by rhapsodes who knew it by heart. This theory has been rejected by other scholars. In the case of the Indian epic this theory is even more improbable because it would be impossible to transmit such vast epics in a rigorously fixed form. A second group of scholars believes that transitional texts are texts

Page #29
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 29 written in imitation of the oral tradition. G. mentions that according to Lord oral and written techniques are mutually exclusive (op. cit., p. 129). However, G. remarks that, for instance, in the Beowulf there are definite characteristics of the written tradition and that the same applies to the Indian epic and, especially, to R. According to G. this is not the result of an imitation of oral versions of the epic but, once the epic was written down, the manuscripts were subject to changes, contaminations and other editorial activities.1 However, G. is convinced that, although there are traces of the influence of a written tradition, both epics existed already as fully completed poems in the oral stage of their composition. In the fourth chapter G. studies the technique of the narrative: themes, repetitions, catalogues and inserted episodes. We mentioned already Lord's definition of the theme, Radlov's elements of production' and Vasil'kov's remarks on the treatment of the theme in M. G. remarks that the extensive treatment of the theme is a characteristic of the Indian epic. A 1 Cf. the remarks made by Hopkins: ' In what shape has epic poetry come down to us? A text that is no text, enlarged and altered in every recension, chapter after chapter recognized even by native commentaries as praksipta, in a land without historical sense or care for the preservation of popular monuments, where no check was put on any reciter or copyist who might add what beauties or polish what parts he would, where it was a merit to add a glory to the pet god, where every popular poem was ha dled freely and is so to this day.' (op. cit., p. 400).

Page #30
30 THE ADYAR LIBRARY BULLETIN theme is constructed on the basis of a determined sequence of traditional elements and contains a standard choice of formulas but, although there is a great degree of uniformity, variations are far from being excluded.1 Lord, for instance, analyses four descriptions of the arming of the hero in the Iliad (op. cit., pp: 89-91). G. quotes a typical duel scene in the epic as an example. The use of traditional themes leads to repetitions which are often literal. G. remarks that these repetitions are conditioned by the requirements of the composition of the epic and by the formulatory nature of its language. It is not correct to maintain that a repeated passage is unauthentic as has been done by Jacobi. According to G. great care must be taken in framing hypotheses on the dependence of one epic on the other on account of the occurrence of the same theme. Similarities are not caused by borrowings but traditional epic themes are elaborated in similar ways in both epics. Just as themes are repeated, formulas are also repeated because they remain in the memory of the singer (for instance M 4. 8.4 and 8. 8; 5. 29. 47 and 37.51; 1. 1. 96ab and 184ab; 4.36.4 and 38: 4). Sometimes longer passages are repeated almost literally (M 6.17.26-30 and 18.17cd-21ab). The epic singer is especially fond of having a remark repeated by somebody to whom the same remark was addressed before (cf. M 4. 41. 11 and 13). Verses are also repeated without 1 Cf. W. Arend, Die typischen Scenen bei Homer. Berlin, 1933, and M. Parry's review, op. cit., pp. 404-7.

Page #31
31 similarity of theme in different parts of the epic and without being connected with the concrete epic situation. This occurs in the case of groups of maxims, of stereotyped glorifications of narrators and of references to widely known legends or myths. Not only single verses but also whole passages are repeated when the same episode is being told twice or more. Often a story is first told by the narrator and later repeated in the same or almost the same words by one of the heroes. In other instances, the same events are described both in a detailed and a short version. In such cases the short version is, as it were assembled' out of the verses of the detailed version (cf. the birth of Drona and his enmity with king Drupada in M 1. 121-128 and 1. 154). The cause of the duplication of an episode can be simply the desire of the poet to single out the most important and pathetic scenes but, more probably, repetitions result from the special nature of the oral performance of the epic. Two possibilities must be taken into account. 1. The epic singer continues his performance during many days. Therefore it is very probable that at the beginning of a new session and before a partly new audience he repeats an episode which he had already told before in order to connect the events of the story with each other. 2. The same traditional epic topics are related by many different singers and often in substantially different versions. The singer who knows these versions sometimes wishes to incorporate in his performance two or three versions of the same epic event. RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 6

Page #32
32 THE ADYAR LIBRARY BULLETIN Previous scholars have often drawn attention to contradictions and inconsistencies in the epic.1 ACcording to G. many contradictions can be explained by the fact that, depending on the requirements of the story, the epic singers each time propose an explanation which corresponds to a certain moment in the story or by the fact that they put on the first plan a detail which is most appropriate in the context. It happens also that the same event is described twice. However, in agreement with the well-known epic law of chronological incompatibility parallel events are described as being consecutive. Another characteristic of oral poetry is the occurrence of enumerations of people, phenomena and objects (cf. M 3. 80-83; 4. 40-43). The presence of enumerations is to be explained by the fact that the knowledge of them constitutes an indispensable part of the training of the epic poet. The tables of contents which are found in both Indian epics fulfill a similar function. One of the most important characteristics of the Indian epic is the abundance of all kinds of inserted episodes of different contents and size. Those episodes which are unconnected with the basic narrative constitute more than half of M. Normally, they are introduced in M in the following way. During a dialogue or the telling of a story one of the heroes mentions a legend and the second participant or the listener requests For instance Hopkins, op. cit., pp. 369-75; N. K. Sidhanta, The Heroic Age of India (London, 1929), pp. 17-19.

Page #33
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 33 him to tell the legend in detail. In R, as a rule, the second participant in the dialogue does not make such requests and the mention of the legend serves no other purpose than a simple reference or a comparison. This becomes later one of the favourite devices of classical Sanskrit poetry. G. mentions with approval Pisani's theory according to which inserted episodes do not appear in a chaotic way but are meant to fill the great temporal hiatuses which occur when the tempo of the main narrative is considerably slowed down.1 Often inserted stories do not simply duplicate or illustrate an epic event or supplement the characterization of some person, but they reflect one or other essential facet of the conception of the epic. Finally, G. remarks that inserted episodes are also to be found in other epics such as the Beowulf, Gilgamesh and the Homeric poems. In the next three chapters Grintser examines topics already studied in the past by many scholars. It will therefore be sufficient to indicate briefly the contents of these chapters and to call attention to some points of particular interest which are discussed by G. Chapter 5 deals with the recensions and versions of the epic. G. draws attention to the fact that also in the case of other old epics several versions exist, for instance, the Akkadian poem on Gilgamesh is known in three versions. 1 'The Rise of the Mahabharata', A Volume of Eastern and Indian Studies presented to Professor F. W. Thomas (Bombay, 1939), pp. 166 ff. The original Italian version has been published in his book, Mahabharata, Episodi scelti (Torino, 1954), pp. 41-59.

Page #34
34 THE ADYAR LIBRARY BULLETIN Both M and R contain indications as to the existence of versions of different size and to their transmission from singer to singer. G. admits that the first and seventh book of R just as the Anusasanaparvan and some other sections of M are later in character but he does not consider them as interpolations which are unsupported by tradition and even less as belonging only to the written redaction of the epos. According to him we see here the consequence of a contamination of several versions of the epic narrative which existed parallel to each other although, probably, some of these versions are older as to their origin. G. sees also traces of the combination of several versions in the fact that the fourth chapter of book 1 of M begins with the same sentence as chapter 1. It seems justified to him to consider the text of M which has been transmitted to us as an extension of a narrative which was much shorter and which concentrated on the battle on the Kuruksetra. This shorter version was later enlarged by means of stories and teachings told by Bhargava-s. G. remarks that the Odyssey also seems to reveal traces of a much shorter version. When the Indian epic was finally written down, the consequences were different for M and R. In M the didactic principle was even more reinforced, but in Rwe observe a transition from the objective epical narrative to a subjective and emotional interpretation, from the poetry of action to the poetry of feeling. The creators of R made a much greater use than earlier poets of all kinds of tropes and rhetorical figures, of lyrical

Page #35
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 35 descriptions of nature and of indications of the psychological motives of the heroes. In general this happened without destroying the formulaic basis of the style of R but sometimes formulas lost their functional role, i.e. to serve as means for the construction of metrically correct verses and were used for purely ornamental purposes especially as metaphors or comparisons. G. gives several examples of which I quote only one. The usual comparison garudah pannagan iva is elaborated to the length of an entire sloka: raksasendramahasarpan sa ramagarudo mahani uddharisyati vegena vainateya ivoragan 11 (Bombay ed. 5. 21. 27) In the last chapter of part 2: The Ramayana: from heroic epic to literary epic', G. examines in great detail the stylistic characteristics of R.' Chapter 6: "Absolute and relative chronology' studies the date of both epics, the different theories on the relation of the Ramopakhyana to R, other episodes common to both epics, historical data in R, the influence of the epic on later Sanskrit literature, geographic differences between the two epics and the terminus post quem of both epics. Chapter 7 examines the historical foundation of the epics and the different layers of which they consist. G. discusses briefly the mythical and allegorical explanations given by several scholars. However, he agrees with the majority of the scholars who believe that both epics have a real historical basis. G. remarks that the historical basis of R is

Page #36
36 THE ADYAR LIBRARY BULLETIN much more elusive than that of M. Nevertheless, G. maintains that there is hardly a justification for doubts about its historical basis or at least about the fact that indirectly it reflects certain real events. In this connection G. points out that in many other epics historical facts and figures can be traced. In the history of many peoples there existed a heroic age. In India this heroic age must have lasted roughly from the 14th to the 10th century B.C. Between the heroic age and the epics themselves there is normally a period of several centuries (sometimes up to five or six centuries). During that period of time in India the exploits of Rama, of the Kaurava-s and of Arjuna were sung and transmitted. G. remarks that it is not surprising that in the course of time historical events were distorted. He adduces several examples from other epics such as the Chanson de Roland, the Nibelungenlied and the Russian byliny (epic songs). The Sanskrit epics contain chronologically incompatible elements as is obvious from the different social, political and cultural conditions which are described in it. G. quotes A. N. Veselovsky who declared that the history on which the epic is based, cannot on the whole be identified with certain specific events and that each epic contains a stratification of facts and a blending of elements separated by several centuries. G. points out that neither the metre nor the language of the epic is a product of one period. Older and freer metres are combined with classical metres in which the number of syllables is fixed. The language of the epic is not spoken Sanskrit but contains

Page #37
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 37 both archaic and relatively new elements. The epic poetry of most peoples is transmitted in a language which is more or less artificial. G. points out that, for instance, the Anglo-Saxon poems and the Homeric poems mix forms from different dialects. G. concludes by remarking that, although the realia which are reflected in the epics belong to different historical epochs, nevertheless the epics remain an integral whole. It was not the task of the epic singers to relate history and it was easy for them to amalgamate for instance heroic stories and mythological topics with ethical doctrines and lyrical descriptions which belong to an entirely different epoch and a different worldview. The Indian epics absorbed different levels of experience of the Indian people and they did not simply leave the tradition of many centuries unchanged in its diversity but, each time, they synthesized the traditional elements in the spirit of the ethical and ideological demands of that moment. Vasilkov's two articles and Grintser's book show clearly the importance of the study of the Indian epics as oral epics. With a few exceptions scholars in the past have been quite willing to acknowledge the fact that the Indian epics were transmitted orally during a long period. Even the expression epic formula' is already to be found in Hopkins's The Great Epic of India and an appendix to his book lists parallel phrases in the two epics. Many of them are epic formulas in the technical meaning given to the word 'formula' by Parry and other scholars. Also Hopkins remarked

Page #38
38 THE ADYAR LIBRARY BULLETIN already that many verses end with the same diambic form of one compound: mahabala, paramtapa, arimdama, and the less frequent (because less needed) common form of the prior pada's pathya ending: mahabaho, mahaprajna, etc. (op. cit., p. 67). However, Hopkins was not primarily interested in the technique of the oral epic and did not elaborate a formulaic analysis of the epic diction. As mentioned above Grintser operates with a concept of the formula which includes single words. Apropos of Homer, Hoekstra has rightly pointed out that, since obviously most single elements of the vocabulary of a language are traditional and since all Greek hexameter poetry from Homer to Theocritus shows preferred 'positions for definite metrical wordtypes, this kind of 'formulaic analysis ' is virtually a reductio ad absurdum of the theory (op. cit., p. 14). Although Parry in his first publication defined a formula as an expression ', later he replaced expression by a group of words (four words or five syllables, see above). Grintser does not hesitate to quote disyllabic words as formulas. In Sanskrit many words of four syllables fill half a pada and they are of course to be found in metrically preferred' positions. It would be preferable to extend the length of a formula to a group of words or a compound of at least five syllables, in order to make a clear distinction between traditional vocabulary and formulaic diction. It is certainly not necessary to establish a minimum percentage of formulaic content to prove the oral and traditional nature of a text. The formulaic contents of oral compositions vary

Page #39
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 39 greatly. According to Emeneau the Toda songs seem to be almost entirely formulaic. In his analysis of the first 15 verses of Homer's Iliad, Lord arrived at the conclusion that 90 per cent of the text was formulaic. However, many scholars have rightly rejected Lord's method in arriving at this figure. Lord stated dogmatically that 'an oral text will yield a predominance of clearly demonstrable formulas, with the bulk of the remainder " formulaic", and a small number of nonformulaic expressions. A literary text will show a predominance of nonformulaic expressions, with some formulaic expressions, and very few clear forumlas.' (op. cit., p. 130). Hainsworth has rightly pointed out that the difference between Homer and Homeric imitators such as Apollonius and Virgil is not a difference of degree but one of kind. Neither Apollonius nor Virgil shows any economy or extension in their limited collections of formulae. Their choice of words is made on entirely different principles (op. cit., p. 9). However, what may be true of Homer and his imitators, may not apply to the oral and written traditions in other literatures. It remains to be seen whether in India there is a difference in degree or a difference in kind between the epic and texts belonging to the written tradition. The Indian epic does not show an organization of the diction into economical and extensive systems and types 1 See M. B. Emeneau, Oral Poets of South India--the Todas', Journal of American Folklore, 71 (1958), pp. 312-24; 'Style and Meaning in an Oral Literature ', Language, 42 (1966), pp. 323. 45; Toda Songs, Oxford University Press.

Page #40
40 THE ADYAR LIBRARY BULLETIN as is found in Homer. Grintser compares briefly the epic formulaic system with the phraseology of Kalidasa's Meghaduta which is composed in a highly artificial metre, the Mandakranta. It would be more instructive to compare the epic with works written in the epic sloka, such as, for instance, the Kathasaritsagara. To what extent does the metre in this text determine the use of certain groups of words and compounds at the end of the pada? Grintser's concept of the supporting' word is undoubtedly extremely useful because with the help of these supporting words the improvisor is able to create patterns (to use Hainsworth's terminology). The epic singer anticipates the supporting word at the end of the pada, and, in this way, composes pada-s which are similar in structure to other pada-s of which the scheme is present in his memory. Formulaic analysis will show how far the same themes are developed in the same traditional way. In Homeric studies very important research has already been undertaken in this direction. It will be more difficult to apply the same methods to the Indian epics because of their size and the great variety between the different recensions which have to be taken into account. Homeric specialists have at their disposal several dictionaries, indices and concordances but there is nothing similar for the Indian epic apart from a pada-index. However, it will certainly be possible to make a beginning with a systematic thematic analysis of both Indian epics. This will be of great help in tracing the construction and

Page #41
RUSSIAN PUBLICATIONS ON THE INDIAN EPIC 41 the history of the epics. As Grintser remarks several times, the existence of several recensions of one epic is typical for the oral epic. In the past analysts have tried to discover the original nucleus of the epic and the accretions which were added to it in later times. Dodds has drawn attention to the fact that in oral transmission successive poems not only incorporate but as a rule remodel early work. He adds that older analysts were misled by the analogy of excavation, in which it is often possible to expose untouched layers of earlier material (op. cit., p. 16). Analysis will still be possible and even indispensable. It belongs to the very nature of an orally transmitted poem to incorporate both older and newer elements. However, a critical analysis will have to take into account the fact that often the oldest elements of an oral poem have been retouched more than later elements. Johnston rightly remarked that the late date of a passage is often obvious from the fact that there is a high degree of agreement between different recensions. It is usually thought that agreement between recensions is an indication as to the original, common element but this rule is certainly much more valid for a work belonging to the written tradition than for an oral composition. Grintser's bibliography lists 463 items and he seems to have missed very little of importance for the study of the Indian epics. When discussing the relation between R and Asvaghosa's Buddhacarita, Grintser remarks that 1 The Buddhacarita, part II (Calcutta, 1936), p. xlix.

Page #42
42 THE ADYAR LIBRARY BULLETIN the description of the scene in the harem in Canto V of the Buddhacarita resembles a similar .description in R (5. 9-11). Referring to Winternitz's History of Indian Literature Grintser declares that in this case it is rather the epic which has been the imitator. Winternitz based himself upon Cowell's statement in the introduction to his edition of the Buddhacarita (Oxford, 1893). Since Cowell many scholars have studied the relations between R and the Buddhacarita. Recently V. Raghavan expressed his conviction that there is no doubt that Asvaghosa knew all three cantos of R (5. 9-11). However, the great value of Grintser's book is not impaired by the fact that, in this point as in some others, it is possible to disagree with him. Let us conclude by expressing the hope that his book will soon be translated into English. Every student of the Indian epic will be able to learn much from Grintser's work which undoubtedly will become a landmark in the history of the study of the Indian epic.3 1 See Johnston op. cit., p. xlviii, n. 1. ? " Buddhological Texts and the Epics', The Adyar Library Bulletin, 20 (1956), pp. 349-59. 3 I am much obliged to my colleague Igor Serebryakov for having presented me with a copy of Grintser's book.