1. Varuna and Vidusaka on The origin of Sanskrit Drama Kuiper F.B.J..djvu

Verhandelingen der Koninklijke Nederlandse Akademie van Wetenschappen, Afd. Letterkunde, Nieuwe Reeks, deel 100

VARUNA AND VIDŪSAKA

On the origin of the Sanskrit drama

by

F. B. J. KUIPER

North-Holland Publishing Company, Amsterdam, Oxford, New York, 1979

ISBN 0 7204 8452 9

1979. 3716

STAATS- U. UNIVERSITATS BIBLIOTHEK HAMBURG

AANGEBODEN IN DE VERGADERING VAN 12 SEPTEMBER 1977

CONTENTS

CHAPTER I: VARUṆA: ASURA AND ĀDITYA

1. Introductory Remarks . . . . . . . . . . . . . . . . . . . 5

2. Varuṇa as an Asura . . . . . . . . . . . . . . . . . . . . . 6

3. Cosmogony and Asuras . . . . . . . . . . . . . . . . . . . . 10

4. Varuṇa and the Father Asura . . . . . . . . . . . . . . . . . 13

4a. antár asmin . . . . . . . . . . . . . . . . . . . . . . . . 16

4b. pitáram் jahāmi . . . . . . . . . . . . . . . . . . . . . . 19

4c. Varuṇa's ádhipatyam . . . . . . . . . . . . . . . . . . . . 22

5. The Ādityas I . . . . . . . . . . . . . . . . . . . . . . . . 32

6. The Ādityas II: Āditya and Asura . . . . . . . . . . . . . . . 41

7. The Ādityas III: the Rigveda . . . . . . . . . . . . . . . . . 46

8. The Ādityas IV: other Vedic texts. . . . . . . . . . . . . . . 52

9. The Ādityas V: later texts . . . . . . . . . . . . . . . . . . 57

10. Ādityas and Aṅgirasas . . . . . . . . . . . . . . . . . . . . 60

11. Varuṇa as a demoniacal figure and as the god of Death . . 67

12. Varuṇa in the epic I . . . . . . . . . . . . . . . . . . . . 74

13. Varuṇa in the epic II: the Lord of the water . . . . . . . . 77

14. Varuṇa in the epic III: the nether world . . . . . . . . . . 81

15. Varuṇa in the epic IV: the Asuras in Varuṇa's world. . . . 88

16. Ambiguous figures: Uśanā Kāvya and Viśvarūpa . . . . . . . . 93

17. The evocatio (upamantrana) . . . . . . . . . . . . . . . . . 101

18. Varuṇa and the Churning of the Ocean . . . . . . . . . . . . 104

CHAPTER II: THE VIDŪṢAKA

1. Introductory Remarks . . . . . . . . . . . . . . . . . . . . 110

2. The Bhāratīya Nāṭyaśāstra: the religious character of the dramatic performance . . . . . . . . . . . . . . . . . . . . . 118

3. The Bhāratīya Nāṭyaśāstra ch. I: date and particulars of the legendary first performance . . . . . . . . . . . . . . . . . 128

4. The "first" dramatic performance (NŚ. 1.51-69) . . . . . . . 142

5. The presents of the gods . . . . . . . . . . . . . . . . . . . 144

6. Description of the playhouse . . . . . . . . . . . . . . . . . 147

7. The Consecration of the Theatre I . . . . . . . . . . . . . . 153

8. The Consecration of the Theatre II . . . . . . . . . . . . . . 154

9. The role of the jarjara in the Consecration . . . . . . . . . . 157

10. The conclusion of the Consecration: Cosmogonical traits . . . 162

11. The jarjaraprayoga in the pūrvaraṅga . . . . . . . . . . . . . 166

12. Nāyaka, Vidūṣaka and Trigata . . . . . . . . . . . . . . . . . 171

13. The Trigata I . . . . . . . . . . . . . . . . . . . . . . . . 180

14. The Trigata II . . . . . . . . . . . . . . . . . . . . . . . . 189

15. The Churning of the Ocean . . . . . . . . . . . . . . . . . . 193

16. The Character of the Vidūṣaka in the Sanskrit Drama . . . . 199

3

17. Dramas without a Vidūṣaka . . . . . . . . . . . . . . 210

18. Vidūṣaka and Jumbakā . . . . . . . . . . . . . . . . 213

19. Vidūṣaka and Viṭa . . . . . . . . . . . . . . . . . . 223

20. The Nāyikā as the Leading Lady . . . . . . . . . . . . 236

Excursus: The Sādhyas . . . . . . . . . . . . . . . . . 242

General Index . . . . . . . . . . . . . . . . . . . . . 245

Index of Text-places . . . . . . . . . . . . . . . . . . . 250

CHAPTER I

VARUṆA: ASURA AND ĀDITYA1

1. INTRODUCTORY REMARKS

The place of Varuṇa in the Vedic pantheon is one of the most difficult problems of Rigvedic mythology, "le point névralgique des études védiques" as Renou has rightly characterized it2. The fundamental difficulty would seem to be that the modern approach, which tries to give an uncontradictory definition of his character, such as "god of the night-sky"; "god of the oath", fails to do justice to the intrinsical double-sidedness of the god. Defining Varuṇa as a numinous god wrongly suggests that a solution has been given while the very search for it has still to begin. The only thing which this formula rightly expresses is the difficulty which the notion of a divine-demoniacal god presents to rational thinking, whereas myth can express the double-sidedness of such a god in its own direct way, which does not dissimulate the logical contradictions involved.

One of the mythological traits in which Varuṇa's ambiguous nature manifests itself is the fact that in the Rigveda he is the Asura par excellence (and as such is rightly equated to Ahura Mazdā of the Old Iranian religion) while, at the same time, as the most prominent of the Ādityas, he is one of the Devas. Since it cannot be doubted that the Asuras and Devas represent the two fundamentally contrasting moieties of a dualistic cosmos3, it is clear that this double-sided character of Varuṇa, his being an Asura as well as a Deva, constitutes one of the basic problems of Vedic mythology. It is curious, therefore, that only a few scholars have recognized this difficulty and taken it seriously. In the first place two men of genius must be mentioned: Abel Bergaigne, who clearly formulated one aspect of the problem but could only explain it as the expression of two different conceptions of the divine which co-existed in the Vedic religion (see p. 45) and, secondly, Peter von Bradke, who pointed out that there are two different conceptions of the Asuras in the Rigvedic material, while rightly rejecting the possibility that the one has developed

1 When in references to some Saṃhitās of the Black Yajurveda the numbers for page and line are added, the number of the line indicates the beginning of the passage quoted. In view of the character of this study it has been considered unnecessary to increase the typographical difficulties of this text by maintaining in the transscription of the Yajurvedic texts the distinction between ardhacandra and "anusvāra". which differs from one text to the other. See, e.g., Macdonell, Vedic Grammar, p. 53, Renou, Grammaire de la langue védique, p. 19.

2 Renou, Festschrift Lommel, p. 122.

3 Out of the many publications special mention may be made of G. J. Held, The Mahābhārata, An ethnological Study (Amsterdam, 1935) and F. D. K. Bosch, The Golden Germ. An Introduction to Indian Symbolism (Indo-Iranian Monographs, vol. II, 's-Gravenhage 1960.

from the other. His own conclusions, however, are unacceptable 4. Further,

Eggeling, SBE. 44, p. XX ff., made some interesting observations on

the problem. Of the present generation of scholars W. Norman Brown,

in one of his earliest papers, was the first to draw attention to the problem 5.

In the German-speaking countries, on the other hand, the discussion of

the Asura problem and the relation between Ahura and Daēva in Ancient

Iran, has not, from Haug down to Paul Horsch 6 led to notable results.

This is mainly due to the fact that the contrast between Devas and Asuras

was interpreted in terms of "divine" versus "demonical" 7 and that the

difficulties inherent in the conception of the Asuras were rather concealed

than explained by an evolutionistic approach 8.

2. VARUṆA AS AN ASURA

The Vedic passages where Varuṇa, either alone or conjointly with Mitra,

is called ásura are not numerous 9. Sometimes this term occurs in a context

where Varuṇa's ominous and inauspicious character is prominent. This is

particularly the case in II.28.7 "Strike us not, O Varuṇa, with thy weapons

which wound the sinful man, O Asura, while thou art searching for him,"

I.24.14 "We deprecate thy wrath, O Varuṇa, by our worship, our sacrifices

and oblations. Thou, who hast the power, O wise Asura, O king, untie

the sins that we have committed" and Ath. S. V.11.1 "How unto the

great Asura didst thou speak here? how with shining manliness unto the

yellow father? having given, O Varuṇa, a spotted (cow) as sacrificial fee,

thou hast with the mind intended re-bestowal" (Whitney) 10. In the other

passages it is rather Varuṇa's all-powerfulness that is stressed; cf. VIII.42.1

"He, the all-knowing Asura, has propped up the sky, he has measured

out the extent of the earth . . . All those things are (the effect of) Varuṇa's

vows," II.27.10 "Thou, O Varuṇa, art king over all, over the gods,

4 See P. v. Bradke, Dyâus Asura, Ahura Mazdâ und die Asuras. Studien und Ver-

suche auf dem Gebiete alt-indogermanischer Religionsgeschichte. Halle, 1885.

5 W. Norman Brown, "Prosoltyting the Asuras (A Note on Rig Veda 10,124)",

JAOS. 39 (1919), pp. 100–103 (cf. also JIES. 2 [1973]).

6 Paul Horsch, Die vedische Gāthā- und Śloka-Literatur. Bern, 1966.

7 Cf., e.g., Hillebrandt, Vedische Mythologie II2 (1929), p. 402ff.: "Manen, Dämonen,

Asuras".

8 This is also true of Hillebrandt, Ved. Myth. II2, p. 420 n. 1 (see below, p. 32)

and of Renou, Études védiques et pāṇinéennes [EVP] 15, p. 171 "Juxtaposition

qui tournera vite en une opposition tranchée".

9 P. von Bradke, op.c., pp. 72, 120. Dandekar, "Asura Varuṇa", Annals Bhandarkar

Oriental Research Institute XXI (1940), pp. 178–79 (whose characterization of

Varuṇa as "Creator and constructor of the Universe", p. 181, needs some speci-

fication).

10 Cf. II.28.7 má no vadhair Varuṇa yé ta iṣṭáv énah kṛtvántam asura bhrīṇánti,

I.24.14 áva te hélo Varuṇa námobhir áva yajñébhir imahe havírbhih, kṣáyann

asmábhyam asura pracetā rájan énāmsi śiśrathah kṛtáni, AS. V.11.1 katháṁ mahé

ásurāyā 'bravīr ihá katháṁ pitré hárāye tveṣánrmṇah, pŕśniṁ Varuṇa dákṣinām

dadāván punarmaghatvāṁ mánasā 'cikitsih. For the last hymn see Hillebrandt.

Varuṇa und Mitra, pp. 88, 110, von Bradke, op. c., p. 100, Bloomfield, The Atharva-

veda, p. 77, Renou, Festgabe f. H. Lommel, p. 126.

O Asura (!), and over the mortals", X.132.4 (cf. II.27.10) "The other one, yonder Heaven, was given birth to, O Asura. Thou, O Varuṇa, art king over all", further V.63.3, where Geldner gives an alternative explanation of ásurasya māyáyā as referring to Varuṇa, and Ath. S. IV.15.12 "(. . .) Pouring down waters, our father Asura . . . Let the gurgles of the waters puff, O Varuṇa—let go the waters that hang down (from the sky); let the frogs whose arms are speckled croak among the water courses" 11, Mitra and Varuṇa are addressed by the dual vocative asurā in I.151.4 "That people is foremost that is very dear to you, O Asuras, you announce the high order, you (gods) who are ṛtāvan", and in VII.36.2 "This newest hymn of praise I make for you, O Asuras, Mitra and Varuṇa" 12.

Particularly interesting are the passages where Mitra and Varuṇa are called deváv ásurā or devánām ásurāḥ, cf. VIII.25.4 "The two great sovereigns (samrājā), Mitra and Varuṇa, the gods who are Asuras" 13 and VII.65.2 "They are the two Asuras of the Devas, they are the Aryas" 14. The exact meaning of this term, which only occurs in the Rigveda, has to my knowledge never been correctly explained. It apparently implies that there are some Devas who are Asuras, in contrast with others, such as Indra, who are not. In a few passages, it is true, Indra is also called an Asura, but it is generally conceded that this is due to a secondary extension 15. In the family collections Indra is an Asura-killer (VI.22.4 asurahán) but never an Asura 16. In VIII.96.9 he is exhorted to slay the

11 VIII.42.1 ástabhnād dyám ásuro viśvávedāḥ, áminīita varimāṇam prthivyáḥ, . . . viśvét táni Várunasyā vratáni, II.27.10 tvám viśvagām Varuṇā 'si rājā yé ca devā asura yé ca mártāḥ (cf. Renou, EVP. 7, p. 90), X.132.4 asā́v anyó asura sū́yata dyá́us tvā́m viśvagām Varuṇā 'si rājā, V.63.3 samrājā ugrā́ vṛṣabhá divā́ś pā́t prthivyā́ Mitrā-várunā vicarṣaṇī, citrébhir abhraiḥ ṭ́ṣatho rá́vadyam vā́rṣayatho ásurasya māyáyā, AS. IV.15.12 apó niṣiñcínn ásuraḥ pítā nah śvásantu gā́rgara apā́m Varuṇa áva ní́cīr apā́h srja vádantu práñībhāvo maṇḍúkā iriná 'nu (for ní́cīr see India Maior, Congratulatory Volume J. Gonda, p. 152). Cf. also, e.g., MS. I.6.11 (104,2) Váruno vai devā́nā́m rājā (see n. 74).

12 I.151.4 prá só kṣitír asura yá máhi priyá ṭtāvānā́m rtā́m á ghoṣátho bhŕát, VII.36.2 imā́m vā́m Mitrāvaruṇā suvrktím iṣáṃ ná kṣṇve asura̐ nā́viyah. Cf. also Bernhard Geiger, Die Ameṣ̌a Spentas, p. 222 n. 1.

13 mahā́ntā Mitrā́varuṇā samrājā devā́v ásurā "(qui sont à la fois) dieux (et) Asura's," Renou, EVP. 5, p. 91 (cf. 7, p. 68), like Geldner: "die Götter und Asura's sind." What is actually meant is "Asuras who have become Devas", "Devas of Asuric origin". Different is X.82.5 paró devébhir ásurair yád ásti "what is beyond Devas and Asuras", where two different groups are meant.

14 tá hí devā́nām ásurā tá aryāḥ "die sind die Asura unter den Göttern, die freund-lichen" (Ludwig, similarly Bergaigne III, p. 87, and Renou). Incorrect von Bradke, op. c., p. 87 and Geldner: "sie sind die beiden Gebieter (Asuras), die Herren unter den Göttern", since all Ādityas must have been devā́nām ásurāḥ. Cf. VIII.27.20, Bergaigne 1.c. and, for Varuṇa as king of the gods, n. 296. Different is, of course, ayā́m devā́nām ásuro ví rājati "This Asura rules over the gods". See n. 74.

15 Cf. I.174.1 (an imitation of II.27.10, where Varuṇa is addressed), VIII.90.6 (in a fixed formula, see von Bradke, p. 61) and two passages in the tenth maṇḍala (X.96.11 and 99.12, see von Bradke, p. 73). See also Bergaigne III, pp. 71ff., 80, B. Geiger, Die Ameṣ̌a Spentas, p. 205 n. 2, Kuiper, IIJ. 3, p. 187.

16 Hillebrandt, Ved. Myth. II2, p. 427f.

7

asurā adevdh 17. If, then, the Rigvedic poets distinguished between two classes of Asuras, the question naturally arises why none of the later texts (except some Atharvavedic passages, cf. p. 41) make this distinction and why they no longer refer to Varuna as an Asura.

The contrast between Devas and Asuras, as we find it in the brāhmaṇas, is only one of the forms in which the dualistic world order of upper world and nether world manifests itself. (See, e.g., Hist. of Rel. 15, p. 107 ff.). The same dualism was recognized in the contrast of the first half of the year (devayāna) versus the second (pitryāna), of the first half of the month with the waxing moon (śuklapakṣa) versus the second with the waning moon (kṛṣṇapakṣa), of day versus night (e.g., TS. I.5.9.7), of light versus darkness (e.g., ŚB. II.4.2.2; 5) and, on the social level, of Ārya versus Dāsa (IIJ.12, p. 282). Cf., e.g., “The Śūdra and the Ārya were created, Day and Night were their adhipatis” 18.

The fact that the dualism is in this manner only expressed in the brāhmaṇas is generally regarded as a proof of its post-Rigvedic character. It has, however, been objected that the same dualism underlies the whole Rigvedic mythology and clearly stems from the proto-Indo-Iranian religion 19.

The difference which seemingly exists in this respect between the brāhmaṇas and the Rigveda is rather due to the circumstance that the authors of the brāhmaṇas were theologians 20, who handled the mythic tradition in a systematic manner, whereas the poets of the Rigvedic hymns rather presupposed it and, besides, expressed themselves in the traditional poetical terminology.

The traditional theory of a “Verteufelung” of the Asuras in a later stage of development of Vedic mythology 21 is no doubt incorrect. It is due to a failure of former generations to realize the mythological meaning of the strife between the Devas and their elder brothers 22, the Asuras, “both of them sons of Prajāpati” as the Śatapatha Brāhmaṇa significantly adds 23.

In the ancient conception of the world the strife between the two parties of gods was necessary for the generation and continuation of life, and for the same reason it was necessary that it ended with the gods subduing their opponents.

This idea, however, is seemingly at variance with another one, quite

17 Von Bradke, op. c., p. 86.

18 śūdrāryā asrīyetiām, ahorātre adhipatī astām MS II.8.6. (p. 110, 14), KS. XVII.5 (248,23), KKS. XVI.4 (107,9/124,19), TS. IV.3.10.2. See also G. J. Held, The Mahābhārata, p. 169 and cf. IIJ. 12, p. 282.

19 See IIJ. 3, p. 211.

20 Similarly von Bradke, p. 90.

21 Thus, e.g., Macdonell, Vedic Mythology p. 156. See above, n. 8 and Renou, EVP. 13, p. 145: “acceptation péjorative d'Asura.”

22 See von Bradke, p. 89, my note in IIJ. 8, p. 106 and cf. Rām. VII.11.14.

23 ubhāye prājāpatyāḥ ŚB. II.4.2.8, I.2.5.1, I.5.3.2, I.7.2.22, II.2.2.8, II.4.3.2, III.4.4.3, III.6.1.8, IV.2.2.11, V.1.1.1, IX.2.3.8, IX.5.1.12, XI.1.8.1, XI.5.6.3. See Lévi, La doctrine du sacrifice, p. 44, Minard, Trois Enigmes II, p. 325. For the same reason it can be said (VS. 30.22) asūdrā abrahmaṇāś té prājāpatyāḥ, for Prajāpati is undefined (ŚB. VI.4.1.6), or he is both defined and undefined (VII.2.4.30), limited and unlimited (VI.8.1.4).

common in the brāhmaṇas, according to which the Devas drove the

Asuras away from these worlds 24. At first sight this might look as though

the dualism of the cosmic contest had been replaced by a monism in which

the dark, demonic aspects of the world had been pushed away. Such a

conclusion, however, would be incorrect. There are certainly, in late

24 The Ṛigveda only refers to the victory of the Devas over the Asuras: X.53.4

tád adyá vācáh prathamám masíya, yénā 'surān abhí devā́ ásāma, 157.4 hatvóya devā́n

ásurān yád áyan, cf. asuráhām, epithet of Agni (VII.13.1) and Indra (VI.22.4).

Its reticence (except VIII.96.9 cakréṇa tám ápa vapa, see von Bradke, p. 86) is

noteworthy in view of AS. IV.19.4 yád adó devā́ ásurāms tváyā 'gre niráakurvata

"when of yore, in the beginning, the gods drove out the Asuras with thee", IX.2.17

yéna devā́ ásurān pránudanta, yénė 'ndro dáisyūn adhamám tám no nind́aya, 18 (babādhé

for nind́aya), X.3.1 sá me sátrūn ví bādhātām Índro dáisyūn ivā́ 'surān, perhaps

also VI.7.3 (Paipp.XIX.3.12) yéna devā́ ásurānām ójijñayā ávriddham (“repelled” ??).

Cf. furtherMS. I.10.5 (145,1) devā́s ca vā́ ásurā́s cā 'smṛ́l lokā́ āsan. sá Prajā́patir

akāmayata: prá 'surān nudeya . . . sò 'surān pránudanta, II.5.3 (50,3) táto devā́ ásurān

ebhyó lokébhyah pránudanta, II.5.9 (60,8), III.6.6 (66,14) devā́ ásurān hatvá 'bhyó

lokébhyah pránudanta, III.2.1 (14,8) tair ásurān ebhyó lokébhyo nirabādhan-ta (15,8)

tair ásurān ebhyó lokébhyah pránudanta, III.8.1 (92,1) ásurāṇā́m vā eṣù lokéṣu púra

āsan . . . tán ebhyó lokébhyah pránudanta, IV.1.10 (13,14) ásurāṇā́m vā iyám prthivī

āsīt, té devā́ abruvan: dattā́ no 'syā́h prthivyā́ íti . . . (after the four quarters have

been assigned to the Vasus, Rudras, Ādityas and Maruts, the text continues, p. 14,1)

táto devā́ ásurān ebhyó lokébhyo nirabhajan; KS. VIII.5 (88,11) stomapurogā́ vai devā́

ebhyo lokébhyo 'surān pránudanta (otherwise MS. I.6.4: 92,3 ásurān abhyàjayan),

IX.14 (117,2) pañcahotrā́ vai devā́ asurān pránudanta, X.7 (132,15) táto devā́ ásurān

ajayan, té 'surā́n nítvā́ rákṣāmsy ápanudanta, (132,19) yántvā purástād rákṣāmsy

ásāms, tánī téna pránudanta . . . yány evóbhītā ásāms, tánī téna vyábādhanta, XIII.3

(181,14), XXV.2 (104,13) Viṣṇur (vā́) imā́mī lokān udajayat, sá ebhyo lokébhyo 'surān

pránudata, XXI.4 (41,20) prajāpatir vai devā́ ásurān pránudanta, XXVI.1 (121,10)

prā́cībhir vā́ āhutíbhir devā́ ánusurān pránudanta, XXXII.5 (23.9) Viṣṇumukhā́ vai

devā́s candobhír ebhyo lokébhyo 'surān pránudanta (cf. TS. I.7.5.4), XXXV.20 (66,12)

sá [Prajāpatíḥ] cāturmáṣyair evā́ 'surān pránudata (67,2) Índro Vṛtrám ahan,

vaiśvadevéna vai só 'surān pránudata, XXXVII.12 (93,6) tenai 'nān pránudanta;

TS. II.6.1.3 té devā́ḥ prayājair ebhyó lokébhyo 'surān pránudanta, VI.2.3.2 tám Rudró

'vāsrjat, sá tisráḥ pūro bhittvā 'bhyo lokébhyo 'surān pránudata, VI.3.1.4, 4.10.3-4

yé purástād ásurā́ āsan, táḿs tābhir ápanudanta, yáḥ pratīcír yé paścād ásurā́ ásan,

tā́ḿs tābhir ápanudanta; AB. I.23.2 té [devā́ḥ] yám eva porathāmám upasádam

upā́yāms, tayaí 'vainām asmāl lokād anudanta, yám dvitīyām . . . II.11.1 té [devā́ḥ]

'gninaitva purastād asurarakṣāmsy apāghnataí 'gninā paścāt, V.11.1 té vai devā́ḥ

saṣṭhenaivā 'hnai 'bhyo lokébhyo 'surān pránudanta, VI.14.9-10 (nirjaghnuh, prajā́-

gāya); KB. VIII.8 [= 9.10] etena vai devā́ḥ pañcadaśena vajrenai 'bhyo lokébhyo

'surān anudanta, XII.3.9 etena vai devā́s triṣamrddhena vajrenai 'bhyo lokébhyo

'surān anudanta; PB. VIII.9.1 asurā́ vā eṣu lokéṣv āsāms, tā́n devā́ hariśriyam ī́ty

asmāl l. pránudanta (etc.), IX.2.11 asurā́ vā eṣu lokéṣv āsāms, tā́n devā́ ūrdhvāsaad-

manenai 'bhyo lokébhyah pránudanta; GB. II.2.7 té ebhyo lokébhyo niraghanan, ekayā́

'muṣmāl lokād, ekayā́ 'ntarikṣād, ekayā́ prthivyā́ḥ; ŚB. IX.2.3.8 devā́s cā́ 'surās co

'bháye prajāpatyā́ díkṣu despar-dhanta, té devā́ ásurāṇā́m díso 'vṛṇjata, XIII.8.1.5

té devā́ sapátnān bhrātrvyān díg-bhyo 'nudanta, té 'dikkṣu pārábhavan, XIII.8.2.1

devā́s cā́surās co 'bháye prajāpatyád asmī́l loké 'spardhanta, té devā́ ásurān sapátnān

bhrātrvyān asmāl lokā́d anudanta, etc. - These passages are important in more

than one respect. KS. XXIII.8 (83,10), XXVIII.2 (154,1), 3 (156,9) devā́ vā asurā́n

hatvā vairadeyād iṣamā́nā́ḥ . . . and similar passages point to the conclusion that

Vedic texts, traces of a tendency to deny the reality of the cosmic contest (see p. 40) but here we are obviously dealing with a completely different mythological notion. A brief sketch of the Vedic cosmogonical myth may be helpful to explain this 25.

3. COSMOGONY AND ASURAS

According to the cosmogonical myth, as it can be reconstructed from the scattered indications in the Vedic texts, this world arose in two successive phases. In the first, the primordial hill rose from the bottom of the primeval waters and drifted about on their surface. It contained, as a potentiality, life and light and all the goods of human existence, but a strong force of resistance (vṛtrá), impersonated by a serpent, prevented them from spreading into the world of darkness outside the hill. The Asuras were the gods of this rudis indigestaque moles, this world of undifferentiated unity. The second phase is initiated by the sudden emergence of Indra as the protagonist of the Devas. Indra, who apparently comes from the darkness outside the hill but whose origin is unknown (apart from some references to his mother), throws his vájra at the floating hill, which thereby becomes fixed and, at the same time, is split open. He pushes heaven and earth asunder (which symbolizes the creation of the dual cosmos) and the cosmic tree rises from the hill, as the central pillar of the universe, to support the sky. Only at this moment (later re-enacted in Indra's banner festival) the god and the world pillar are identical. As for the opening of the hill, however, this meant conquering

mythologically there was little or no difference between driving away and killing the Asuras. The difficulty which the driving away presented to rational thought is openly stated in ŚB. IX.2.3.2-4, which first relates how Indra and Bṛhaspati, before the sacrifice, drove the Asuras away in the south but then goes on to say: “Now what the gods did then, that is done on this occasion. Those fiends, it is true, have now been chased away by the gods themselves, but when he does this, he does so thinking, 'I will do what the gods did;' and having had the Asuras, the mischievous fiends, chased away in the south by Indra and Bṛhaspati, he performs this sacrifice in a place free from danger and devilry” (transl. Eggeling). The sacrificial act thus repeats the primordial events. The question then arises whether the place to which the Asuras had been relegated, although different from imé lokáh, still formed part of the universe. According to the brāhmaṇas it was outside the trailokya, that is, heaven, atmosphere and earth (AB. I.23.2, GB. II.2.7), but according to the epic the Asuras were in Varuṇa's underworld (see p. 81, RS. X.82.5 and History of Religions 15, p. 115). The construction of the theatre, however, which represents the sacred world as protected against the evil influences from outside, shows that the Asuras (Dait yas, Dānavas) were considered to be outside the cosmos, in contrast with the nāgas and other inhabitants of the pātāla or rasātala (Bhāratīya Nātyaśāstra 1.95, 3.62), which protect the dramatic performance “from below” (adhastād raṅgapīthasya) and reside in the lowest joint of the jarjara (IIJ. 16, p. 252 and see below). In view of the relationship between Nāgas and Asuras (see p. 88), this functional contrast should be noted. No weight should be laid on the commentary which paraphrases Kā d. 227 line 2 asura by pātālavāsi-devānām.

25 The Indian creation myth is treated in greater detail in History of Religions 10 (1971), pp. 91-110 (only this part !) and 15 (1975), pp. 107-120.

the power of resistance of the realm of the Asuras, that is, subduing the Asuras themselves.

It has been argued elsewhere 26 that this cosmogonical fight had to be repeated annually, at the beginning of the new year, in order to renew life and society, and that the Ṛigveda in its essence is a collection of hymns of the New Year festival. The obvious implication is that Indra was a seasonal god 27. Much more important, however, is the question which has never explicitly been posed so far, as to what became of the Asuras. In the next section it will be shown that some of them went over to the party of their opponents and thus became devā́ asurāḥ. The reason why this term is no longer used in the post-Ṛigvedic literature will become clear when the problem of the Ādityas is discussed. Whether intentionally or not, their origin is no longer mentioned, once they have been incorporated in the Devas and have become part of the dual cosmos.

It will be shown that although their original name of Asura is no longer used, they actually retain an ambiguous character up to the period of the epics (and later). As for the rest of the Asuras, who are sometimes called ásurā adevā́ḥ in the hymns 28, they are chased away from this world as a result of Indra’s victory. Since, however, Indra and the Devas have to fight them at the beginning of every new year they must apparently return into this world during the period of crisis which precedes the new strife. For a short period, therefore, the banished demons again become seasonal anti-gods. In later sections we shall have to return to these Asuras. In the present context it is sufficient to remark that their so-called “Verteufelung” is consequently not the result of a historical development of the Vedic religion but is part of the cosmogonical myth.

Here again, Bergaigne has made some interesting remarks on the problem (vol. III, p. 115), although he has not recognized the seasonal character of “les démons . . . vaincus, frappés, foudroyés, déchiquetés par leurs adversaires”. Therefore, the question remains to be answered what exactly is the position of the devā́ asurāḥ during the same period. It has been suggested that in the period for which most of the Ṛigvedic hymns were apparently composed, Varuṇa was particularly dreaded because the inauspicious side of his character was prominent 29. At this point it will suffice to point to this problem (see below, pp. 24, 42).

26 See “The Ancient Aryan Verbal Contest”, IIJ. 4, pp. 217–281.

27 Cf. Kālikāpurāṇa 90.52ff. (J. J. Meyer, Trilogie altindischer Mächte und Feste der Vegetation III, p. 113). The much discussed line (ṚS. II.12.5) “of whom some say ‘he is not here (ásáṅ)’ ” may refer to his supposed absence. It would seem entirely out of the question that the poet here refers to “atheists” (nāstika), as is still the current opinion. See further p. 25 and n. 135.

28 See von Bradke, p. 105.

29 See, e.g., S. Lévi, Doctrine du sacrifice, p. 167f., Bergaigne, Rel. Véd. III, pp. 142, 157, Hillebrandt, Ved. Myth. II2, p. 33, Lommel, Symbolon 4, p. 158ff., Renou, Festschrift Lommel p. 125, Kuiper, IIJ. 5, pp. 53, 55; 8, p. 109 n. 68. According to Renou there is a “durcissement du personnage” in the Atharvaveda, but in view of the reticence of the Ṛigveda regarding Varuṇa a different explanation is possible. In this connection attention may be drawn to Hillebrandt’s words (II2,

11

As for the fact that the fight between gods and Asuras is constantly referred to in the brāhmaṇas but only seldom in the Ṛigveda, this is primarily a matter of difference of style. The authors of the brāhmaṇas, in order to explain the aim and effectiveness of a certain rite, had to recur to the beginning of the world, that is, the cosmogony, which was a guarantee for the efficacy of the ritual. For this purpose they used the fixed formula (the exact wording of which differs from school to school): “The Devas and the Asuras were at strife”. When, on the other hand, the Vedic poets relate, over and over again, how Indra slew the serpent, they refer to the same cosmogonial fight in a stylistically different form. Mythologically, this is equivalent to the formula of the theologians.

To conclude this section a few words may be devoted to the relation between the Asuras and Death. In the brāhmaṇas Varuṇa is sometimes identified with pāpmán 30 and mṛtyú “death” (see p. 72). This shows that, as a result of Varuṇa's incorporation in the pantheon of the Devas, also Death and the ominous powers have been integrated with the cosmic order. In the same way the Ṛigvedic poet addresses the dead in the words “Thou shalt see the two kings, revelling each in his own wise, Yama and the deva Varuṇa” 31. While in later times a distinction was made between the world of Yama and that of Varuṇa (see p. 74), the Ṛigvedic poet here regards Varuṇa as the god of the world of the dead. The latter continue to belong to the organized cosmos: “Les Pères sont des morts, mais des morts demeurés mortels; ils n'ont pas rejeté le mal” 32. They are in the charge of the devā ásurāh. This explains the difference with TS. II.4.1.1, a passage more than once quoted in the handbooks 33. Here three contrasting pairs are enumerated, viz.

devā́h : ásurāḥ manusyà̀h : ráksāṁsi pitáraḥ : piśā́caḥ

Undue prominence is usually given to the horizontal pairs. The real importance of this passage is, however, in the vertical groups, and particularly in the fact that the dead (pitáraḥ) are here classed together with the Devas and men. This is apparent from KS. X. 7 (132,10) and JB. I.154 (line 10), where the pitáraḥ and the manuṣyà̀h have changed places and the former, accordingly, are opposed to the ráksāṁsi. Although this order also occurs in TS. VI.1.1.1 in the tetrad devā́ḥ – pitáraḥ – manuṣyà̀h – rudràḥ 34, the common order (corresponding to heaven –

p. 34): “Keine Erklärung Varuṇa's darf als ausreichend angesehen werden, welche nicht dem ṚV., dem Ritual oder dem klassischen Sanskrit gerecht wird und nicht die Vermittlung zwischen den drei Auffassungen gewährt. . .” Only his character of a god of the subterranean world can account for the different aspects, cf. IIJ. 8, p. 115, Hist. of Rel. 15, pp. 114f., 116.

30 Cf. Pāli Māro pāpimā. 31 ubhā́ rājānā́ svadháyā́ mádantā́ Yamáṃ pabyā́si Várụṇaṃ ca deváṃ. Cf. IIJ. 8, p. 109. 32 Lévi, Doctrine du sacrifice, p. 98. 33 E.g., Macdonell, Vedic Mythology, p. 164, Hillebrandt, Ved. Myth. II2, pp. 405, 421 (cf. Delbrück, Altindische Syntax, p. 215). 34 Arbman, Rudra, p. 154 n. 2.

earth - nether world) must have been devāh - manusyāḥ - pitáraḥ, which

also occurs, e.g., in ŚB. III.6.2.26 and which is sometimes abridged to

devamanusyāḥ, a term which then also includes the pitáraḥ 35. The Asuras

mentioned in this passage are real demons, who stand outside the cosmic

order. Logic would demand that Varuṇa, the former Asura who had

become a Deva, is their adversary. It is the object of this study to show

that this is only part of the truth and that in fact Varuṇa is, even in

his relation to his banished brothers, basically ambiguous.

4. VARUṆA AND THE FATHER ASURA

Varuṇa's fundamental ambiguity is particularly apparent in his relation

to the 'Father Asura' in the cosmogony. The critical point in this relation

is the moment when Indra, by breaking the 'obstruction' of the

undifferentiated primeval world, makes the dual cosmos arise. As we have

seen, this act was not so much a real creation as rather a transformation

of the primordial Chaos into an organized world. It did not put an end

to the existence of the Asuras. A re-grouping takes place. While the

majority of the Asuras are chased away from this world and, according

to later texts such as the Mahābhārata, continue to live 'under the earth'

or in the nether world (see p. 88), some become Devas and are yajñíya

"worthy of sacrifice" and yajatá "worthy of worship" (although it is

not certain whether all Devas were necessarily yajñíya, see n. 36). The

most prominent among them (apart from Agni and Soma) are Mitra and

Varuṇa, the devā́ asurā́ (VIII.25.4). It is of fundamental importance,

therefore, to make a clear distinction between, on the hand, the undivided

group of Asuras as gods of the undifferentiated primordial world - seldom

referred to, it is true, in our texts - and, on the other hand, the later

two groups of Asuras, some of whom belonged to the ordered cosmos,

while the others subsisted outside it, on the fringes of the 'formed' world.

The main points of Vedic cosmogony can be summarized in the three

stages I. waters II ásat III the duality of sát and ásat. Apart from the

famous nāsadasíya-hymn (RS.X.129) the Rigveda has very little to say

about stage I but the more so about the primordial hill, which arose from

the waters and symbolized the undifferentiated world of ásat against which

Indra directed his attacks (Vṛtra-slaying). The theological speculations of

the Veda centre round this undifferentiated character of ásat, in the

Brāhmaṇas often impersonated by Prajāpati who is prior to the dual

world of sát (III) but continues to exist in it as the representative of its

totality (see below). The world of ásat, which may be roughly named

chaos after its Greek homologue, is said to have been 'one', 'without

35 Delbrück, Altindische Syntax, p. 57. It is true that at ŚB. II.1.3.1 the Devas

are opposed to the Pitaras, but cf. XIII.8.1.6 'He thus gives him [viz. the deceased]

a share of the world of men''. Is the different order in the Black Yajurveda due to a

tendency to bring the dead in closer contact with the gods? See about this tendency

Caland, 'Een indogermaansch lustratiegebruik', Meded. Kon. Akad. Wet., Afd.

Letterk. 1898, pp. 279–80.

13

a second”, “unformed”, “undifferentiated (undiscriminated)”, “un-measured or immeasurable (limitless)” and “not defined (by a name)”, which according to L. Renou and L. Silburn, Sarūpa-Bhāratī 1954, p. 68ff. is equivalent to “unshaped, unbounded, unorganized”. A few illustrative passages may be quoted: “That breathed according to its nature, although there was no wind. (It was) one, there was not further any other thing whatsoever”, “Asat, verily, was this in the beginning”, “In the beginning there was nothing here, neither heaven nor earth or atmosphere. This, being āsat, thought: May I be”, “Prajāpati, verily was alone this (world): there was neither day nor night”, “Prajāpati was alone this (world). Vāc was his only (possession). Vāc was his second”, “Asat was this in the beginning, one without a second”, “Unformed, verily, was this”, “It was neither day nor night, (it was) undifferentiated”. The rise of the cosmos from the primordial world is sometimes described in the abstract terms of the theologians as a birth, as in “Sát was born from āsat in the first period of the Devas”, or as an emanation, e.g. “Prajāpati, verily, was alone this (world). Only Vāc was his own (possession). Vāc was his second. He thought: This Vāc I will emit (visrjā iti)”. Cf. “The Devas are on this side (that is, also a product) of the emanation of this (world)”; or, finally, as a multiplication, e.g. “This, which was one, has developed to all (that exists)”, “Prajāpati is manifoldly born”, etc. 36.

After this digression we can return to the problem of the Asuras who partly belong to the dual world of the cosmos, partly seem to have been relegated to a non-defined world beyond the pale (Hist. of Rel. 15, p. 112). Two restrictions must, however, be made. First, the terminological distinction is only relevant for the Rigveda, since in later texts the term devá ásurāḥ does no longer occur. The first group is then only denoted

36 See, e.g., von Bradke, pp. 59, 104ff. (where the relation between the Devá ásurāḥ and the Ásurā adevāḥ is discussed), RS. X.124.3, which stresses the transition from ayajñiyá to yajñíya, and, on the other hand, X.19.7 yé devāḥ ké ca yajñiyāś te rayyá sám srjantu nah, which seems to imply the existence of Devas who are not yajñíya. Note in this connection that devá ásura is also used with reference to Rudra, cf. V.42.11 námobhir deváṃ ásuram duvasya. The passage might be of importance for determining the place of Rudra in the Vedic pantheon. - The passages quoted are the following: RS. X.129.2 ánīd avātám svadháyā tád ékāṃ tásmād dhā 'nyán nā párāḥ kim̉ canā 'sa, ŚB. VI.1.1.1 āsad vā idāṃ ágra āsīt, TB. II.2.9.1 idāṃ vā ágre naivá kim̉ canā 'sīt. nā dyáur āsīt, nā prthiví, nā 'ntárikṣam. tád āsad evá sán māno 'kuruta : syáṃ iti, PB. XVI.1.1 Prajāpatir vā idam eka āsīt, nā 'har āsin na rātrir āsīt, XX.14.2 Prajāpatir vā idam eka āsīt, tasyā Vāg eva svam āsīt, Vāg dvitīyā (otherwise RS. XII.2: 167,15), ChU. VI.2.1 asad eva 'sīd āgra āsīt ekam eva 'dvitīyam. tasmād asataḥ sad ajāyata, MS. III.1.1 (75,1) áklptam vā idāṃ ásīt (cf. TS. VI.1.5.3 kl'p̣tyāi "for bringing into harmony" and the fixed phrase (ékam sántam) bahudhā kalpáyati, RS. VIII.58.1, X.114.5, cf. I.164.46 vadanti, AB. III.4.9 viharantī), TS. V.3.4.7, VI.4.3.3 má vā idāṃ dívā nā náktaṃ āsīt, ávyāvṛttam (cf. vyāvartayati "to distinguish, separate"), ŚB. II.3.1.7, TB. I.1.8.1, etc.), RS. X.72.2 devā́nāṃ pūrvé yugé 'sataḥ sád ajāyata, cf. 6 iyáṃ visṛṣṭib, VIII.58.2 ekāṃ vā idāṃ ví babhūva sárvam, AS. X.8.13 Prajāpatir bahudhā ví jāyate and the well-known formulaic phrase Prajāpatir akāmayata : prajāyeya, bhūyān syām iti (Deussen, Allg. Gesch. d. Phil. I.1, p. 181ff.), e.g., PB. VI.1.1 Prajāpatir akāmayata : bahu syāṃ, prajāyeye 'ti. For āsat see also n. 121.

14

by the specific name Ādityas, whereas the name Asuras is exclusively used with reference to the Asuras who have become demons. Secondly, it seems that the distinction is no longer made during the period of crisis which preceded the new year. As will be shown below, there are reasons to suppose that in those days there was a regression to the situation before the “second creation”, in which case all Asuras must again have formed one group against the Devas. With these restrictions the use of the term ásurā adevāh to denote all Asuras who are opposed to the Devas (von Bradke, p. 105) can be accepted, although it occurs only once in RS. VIII.96.9 ánāyudhāso ásurā adevās cakréna tám ápa vapa ṛṣiṣin “the

Asuras that have not become Devas are without weapons, disperse them, O ṛṣiṣin, with thy disk”

In this connection a passage in RS. X.124 is of particular importance. It has been called a difficult hymn37, but its difficulty is mainly due to the fact that earlier scholars failed to recognize its place in Vedic cosmogony. This is particularly true of Geldner’s interpretation (Übersetzung III, p. 353). He explained it as a finding again of Agni, which is a late reflex of some misconceptions of the nineteenth century. Even in the most recent discussion by Horsch the hymn is, in spite of some correct observations, finally placed in a wrong context.

The passage referred to above is found in the stanzas 4 and 5:

4. bahvīḥ sáṃā akaram antár asminn Índram vṛṇānāḥ pitáraṃ jahāmi, Agníḥ Somo Váruṇas té cyavante rāṣṭráṃ tád avāmy āyán.

37 Hillebrandt, Ved. Myth. II2, p. 61 “In diesem schwierigen Liede”, Paul Horsch, Die vedische Gāthā- und Śloka-Literatur (1966), p. 239 with note 3: “Dem schwierigen Liede 10,124 liegt gerade der Übergang der Weltherrschaft von den Asuras unter Varuṇa zu den Devas unter Indra zugrunde, wobei jedoch charakteristischerweise eine Versöhnung zwischen Varuṇa und Indra stattfindet: Varuṇa tritt damit zu den Devas über.” In spite of this correct remark Horsch’s total conception of this hymn is, I am afraid, beside the mark. Of the secondary literature on this hymn the following may be mentioned in chronological order; 1877: A. Hillebrandt, Varuṇa and Mitra, p. 107f.; 1882: A. Bergaigne, Religion védique III, pp. 129, 145ff. (on Varuṇa: Vṛtra !); 1885: P. von Bradke, Dyāus Asura, p. 97ff.; 1885: H. Oldenberg, ZDMG. 39, p. 68ff. [= Kleine Schriften, p. 490ff.]; 1890–1897: H. S. Vodskov, Sjaeledyrkelse og Naturdyrkelse (I, Rig-Veda og Edda], p. 210ff. (“en forholdsvis markedlig god oversaettelse af den så mishandlede RV. X.124,”

S. Sørensen, Norsk Tidsskrift for Filologi 3R VI, 1897–98, p. 137 n. 2); 1893: E. Hardy, Die Vedisch-brahmanische Periode der Religion des alten Indiens, p. 59; 1897: Geldner, Vedische Studien II, p. 292ff. (translation of the hymn on p. 295f.); 1908: Leopold von Schroeder, Mysterium und Mimus im Rigveda, p. 196ff.; 1912: Oldenberg, Rigveda, textkritische und exegetische Noten II, pp. 342–44; 1913: A. Hillebrandt, Lieder des Ṛgveda, p. 21ff. (devayāna taking the place of pitryāṇa ); 1920: J. Charpentier, Die Suparnasage, p. 118; 1923: Oldenberg, Die Religion des Veda3–4, pp. 94 n. 2, 162ff.; 1929: A. Hillebrandt, Vedische Mythologie II2, p. 61ff.; 1951: Geldner, Übersetzung III, p. 352ff. (like Geldner 1897: Agni’s flight; see also Hillebrandt, Ved. Myth. II2, pp. 62, 211); 1965: L. Renou, EVP. 14, p. 29 (translation), p. 97 (notes); 1966: P. Horsch, Die vedische Gāthā- und Śloka-Literatur, p. 239, and the references given in IIJ. 5 (1961), p. 53, Etudes Asiatiques 25 (1971), pp. 87, 97. The following publications were inaccessible to me: Perry, JAOS. XI, p. 159f., T. Segerstedt, RHR. 57, p. 174f. (quoted by Oldenberg) and S. Sørensen, “Til spergsmaalet om Aditya’erne,” Festskrift til Vilhelm Thomsen (1894), p. 335.

15

5. nirmāyā u tyé asurā abhūvan tvám ca mā Varuṇa kāmá́yāse,

rténa rājanm anṛtam viviñcán máma rāṣṭrásya 'dhipatyam éhi.

(4) [Agni:] "Many years have I passed within him; I now leave the

father and choose Indra. Agni, Soma and Varuṇa (now) secede. The

sovereignty has now turned: I accede to that one and support it".

(5) [Indra:] "The Asuras have just lost their magic power (māyā). If

thou, O Varuṇa, lovest me, assume, O king, the ādhipatya of my empire,

dividing ṛtá from what is against the ṛtá".

The first stanza is entirely clear except akaram, whose exact meaning

is obscure 38. There are, however, three expressions which call for some

comment, viz. antár asmin, indram vrṇānā́h and pitā́ram jahā́mi.

4a. antár asmin.

As has been noted long ago 39, in the Rigveda antár is still an adverb.

When followed or preceded by a locative it means, accordingly, "within"

(Dutch binnen in). Its function of stressing the meaning of the locative

is apparent in, e.g., MS. II.4.3 (40,19) ásti vắ idā́m tyásminn antár vīryà̀m

versus TS. II.4.12. 4–5 ásti vắ idā́m, má́yi vīryà̀m, KS. XII.3 (164,18)

vīryà̀m vā idā́n mayy asti "there is, indeed, this strength in me".

Grassmann's assumption of a special meaning "in seiner Gemeinschaft"

for this passage and for VII.86.2 40 has rightly been rejected by Geldner,

Übersetzung III, p. 353. It is due to an insufficient insight into the

mythological implications of the context.

Since the primordial world was undifferentiated, its most prominent

feature was the absence of all contrasts which characterize the phenomenal

dual world. Thus it lacked, for instance, one of the basic contrasts, that

of male versus female. As a consequence the poets, when referring to the

"first creation", were compelled to have recourse to paradoxes as a means

of expressing what cannot be expressed. Hence it is that they speak

alternately of a female being, a primeval Mother, of the womb of a male

being, of a bisexual bull, a cow which is at the same time a bull. Cf., e.g.,

III.29.14 "Day after day, the joyous one (Agni) does not close his eyes,

38 akaram : for the function of the aorist see K. Hoffmann, Injunktiv, p. 157. The

meaning "durchmachen, zubringen", conjectured by Grassmann, is uncertain. It

has only been recorded from Mhbh. XV.1.6 crit. ed. (PW. II, col. 83) but from

RS. I.33.15c jyók cid ātrá tasthivā́mso akran "Schon lange hatten sie hier verweilend

gesäumt" (Geldner) it may be inferred that in Vedic idiomatic usage kr̥-, when

used with a word for a certain span of time, had actually that meaning. For

this passage only earlier translators followed Grassmann: Hillebrandt, Lieder, p. 21 and

Hardy, p. 59 ("viele Jahre verbrachte ich bei ihm"), W. Norman Brown, JAOS.

39 (1919), p. 102 ("Many years have I passed within him"). Geldner's translation

"Viele Jahre war ich in ihm tätig" clearly presents difficulties. Hence Renou:

"Durant de nombreuses années j'ai travaillé chez lui", K. Hoffmann "viele Jahre

habe ich (bis jetzt) bei ihm gewirkt," which does, not, however, solve the problem.

39 See Roth, PW. I, col. 239, Delbrück, Altindische Syntax, p. 446.

40 Accepted by Delbrück, Altind. Syntax, p. 446 (cf. also his earlier work Ablativ,

Localis, Instrumentalis, 1867, p. 43), and more recently by Hoffmann and Renou

in their translations (see n. 38).

16

ever since he was born from the womb of the Asura" 41.

This primeval world can also be represented by the specific power of obstruction (vrtrá) which is peculiar to it. Cf., e.g., I.54.10 42 "There was the darkness of the waters, which made their foundation (?) vacillate; the mountain was in the womb of the vrtrá" 43, that is, the primordial hill was present as a potentiality in the womb of Chaos. Whether or not the poet here refers to an abstract obstruction or, more likely, to the serpent as its personification, is not relevant. The Chaos consisted of the primeval waters (ápah, dánu), whose child was the serpent (dánavá). Since in the "first creation" the primordial hill (párvala) rose from the bottom of these waters, it must have been present there. After the hill had risen above the surface of these waters, the power of "resistance" was concentrated in it, rather than in the waters. This explains why Indra's attacks are as much directed against the hill as against the serpent 44.

Parallel to these Rigvedic passages, where the poets refer to the womb of the Asura or Vṛtra, there are several passages in the brāhmaṇas where, in the more direct diction of their authors, the phrase "within Vṛtra" is used. Geldner (l.c.) rightly points to TS. II.5.2.3, which relates that Agni and Soma were still in Vṛtra at the moment Indra was going to strike and that they said to him "Strike not, we are within" (má prá hār, āvám antáh svah). Parallel passages 45 are KB.2 III.7.17ff. 46 "As for his offering to Agni and Soma at the Full Moon sacrifice: Agni and Soma were within Vṛtra; against them Indra could not throw his vajra. For them he instituted this share [a cake] at the Full Moon sacrifice" 47, ibid. XV.3.5 (= KB.1 XV.2) "Agni and Soma were within Vṛtra: against them Indra could not throw his vajra. They came outside to this share

41 ná ní miṣati suráṇo divé-dive yád ásurasya jathárād ajāyata. On the "Father Asura" (whose place was to be taken by Prajāpati in the world of order) see, e.g., Hillebrandt II2, pp. 63f., 423 n. 4 and cf. Etudes Asiatiques 25, pp. 87, 97f.

42 apám atiṣṭhad dharaṇáhvaram támo 'ntár vṛtrásya jathárareṣu párvataḥ. Cf. Bergaigne II, p. 201, III, p. 146.

43 For the accent see Wackernagel, Altindische Grammatik II/1, p. 220 ("im Behälter schwankend"), but the meaning "receptacle", which Grassmann gives for dharúṇa, is probably incorrect, see Lüders, Varuṇa, p. 172 n. 2) and note the contrast with agnihvára YV. (op. c., p. 18). Renou, EVP. 17, p. 20: "Les ténèbres se tenaient (là), bouleversant le fondement des eaux."

44 See Lüders, Varuna, p. 170f and, for a more detailed discussion of the hill in Vedic cosmogony, cf. History of Religions 10 (1970), p. 106f., India Maior, p. 144ff. Cf. also the short formula girir vai vṛtró "Vṛtra is a mountain" in MS. IV.5.1 (62,15). This is not the place to discuss Ivanov and Toporov's etymological connection of párvata with Slavic Perúnŭ.

45 See Leo Buschardt, Vṛtra, Det vediske Dæmondrab, pp. 55, 72f.

46 KB2 refers to the new edition of the Kauṣītaki Brāhmaṇa by E. R. Sreekrishna Sarma (Verzeichnis der orientalischen HSS. in Deutschland, Suppl. Band 9,1), Wiesbaden 1968.

47 (17) atha yat paurnamāsyām Agniṣomau yajati (18) Agniṣomau vā antar Vṛtra āstām (19) tā́v Indro nāsaknod abhi vajram prahartum (20) ābhyām etám bhāgam akalpayat paurnamāseam.

17

of theirs at the full moon sacrifice" 48, ŚB. I.6.3.13 "While Indra thus

went forward (pursuing Vṛtra) he said to Agni and Soma: 'You belong

to me, and I to you. He is nothing to you. Why then do you aid this

Dasyu against me? Come over to me! . . .'”. This passage ends with the

words "They went over to him, and after them came all gods, all sciences,

all glory, all food, all prosperity" 49. The TS. passage quoted above also

ends with Indra's words "You belong to me, come over to me", whereupon

all kinds of gods of Life followed Agni and Soma. In paragraph 5 of that

passage there is a variant: "Heaven and earth said 'Strike not, for he

is lying upon us' " 50. In this connection attention may be drawn in

passing to the wording of 4: sá devátā vṛtrán nirhūya vártraghnai̇ havíḥ

pūrṇámāse nír avāpad . . . "After he had called off the deities from Vṛtra

he offered the vártraghna libation at full moon". This practice of calling off

someone will be discussed in a later section (see p. 103).

Passages where antár asmin, antár várune refer to the dual cosmos are

not relevant in this connection 51.

48 (5) Agniṣomau vā antar Vṛtra āstām (6) tāv Indra nāśaknod abhí vajram prahartum

(7) tāv etám bhāgam upanicakrāmatām [v.l. upanircakrāmatām] (8) yaś cainayor

asau paurnamāse. Read upanirakrāmatām?

49 ŚB. I.6.3.13 sá vā Índras táthāivor nuttás cāran, Agniṣomā upamantrayāḿ cakré:

'gniṣomau yuvāḿ vai māma stho yuváyor aháḿ nā yuváyor eṣá kim caná kāḿ

ma imáḿ dásyum vardhayatha, úpa mā 'vartethām iti.

50 TS. II.5.2.3-4 ténā 'bhy àyata tāv abrūtām Agniṣómau: má prá hār āvám antáh

sva íti, máma vā yuvám sthó 'bravin, máṁ dhóy itam íti.

51 It is not always possible to draw a sharp dividing line between the undifferentiated

primordial world and the nether world of the organized cosmos, as the latter preserves

some characteristics of the first (cf. Hist. of Rel. 15, p. 118 n. 22). In VII.87.5 tisró

dyávo nihitā antár asmin tisró bhūmir úparāḥ śávidhāináḥ, gŕtso rájā 'Varuṇas

cakra etám . . . "Three heavens are placed within him, three earths below them,

arranged as six, King Varuṇa, able (?), has made this [golden swing]" there seems

to be, in the words antár asmin, a reminiscence of Varuṇa as a potentiality.

older world of the Asuras, in which the world was contained as potentiality.

In V.47.3 ukṣá samudró arusáḥ suparnáh párvasya yónim pitúr á viveśa

"The bull, the reddish eagle, entered the womb of the primordial father"

the poet means the nether world, whatever the exact meaning of samudró may be (Geldner: "Meer des

Lichts", Hillebrandt, Ved. Myth. I2, p. 323: the moon). In VII.86.2, a Varuṇa

hymn which immediately precedes that quoted above, the text reads: utá sváyā

tanvò sám vade tát kadā nv antár Várune bhuváni, kim me havyám áhrṇāno juṣeta

kadā mṛḷikáḿ sumánā abhí khyam. The poet no doubt speaks of his entering. In a

visionary state of mind, the palace of Varuṇa, which is described in VII.88.5

brhántam mānam . . . sahásradvāram . . . grhám. Renou, EVP. V, p. 70 rightly renders

the stanza "Alors je me consulte en moi-même: quand donc serai-je dedans Varuṇa?

Est-ce qu'il daignera agréer mon offrande sans être courroucé? Quand apercevrai-je

(sa) compassion, (en sorte que j'aie) l'âme apaisée?" [Otherwise Renou, La pensée

religieuse de l'Inde ancienne, 1942, p. 60f. "Quand donc serai-je à nouveau proche

de Varuṇa?",] Geldner: "Wann werde ich wohl dem Varuṇa nahe kommen?" (cf.

also Siebenzig Lieder des Rigveda, 1875, p. 6), K. Hoffmann, Injunktiv, p. 246

"Wann werde ich drinnen bei Varuṇa sein?". Thieme, Mitra and Aryaman, p. 65

translated "When shall I get inside of True-Speech?" with the following comment:

"It literally says that the poet wants to reach that point where he is never again to

We can now return to X.124.4a bahvìḥ sámā akaràm antár asmin. Since the poet refers to the undifferentiated sexless primordial world, Vodskov 52 was right in comparing V.2.2 pūrvîr hî gárbhah śarádo vavárdha

'paśyam jātám yád ásūta mā́tā 'Many years the foetus had been growing, I saw him when his mother brought him forth so that he was born'. Mythologically, antár asmin "within him" (that is, in the Father Asura) is equivalent to ásurasya jatharād (III.29.14), vṛtrásya jatháreṣu (I.54.10), antar vṛtre (in the brāhmaṇas) and to the mother. They are all expressions for the mythological notion of an unorganized world which was darkness (támas). Therefore, Bergaigne, Rel. védique III, p. 146, was perfectly right in translating "dans son sein". This explains why in stanza 1 of the same hymn X.124 Agni is addressed with the words jyók támā āśayiṣṭhāḥ "for a long time thou hast been laying in the darkness" 53. It is the darkness out of which light was to be born.

4b pitáram jahāmi

As we have seen in the preceding section, at the moment when the Devas appear as an entirely new element outside the world of the Asuras, some entities go over from the Asuras to the new gods. This process of reorganization of the original world is described in various ways, which may here be briefly summarized before some details are discussed more thoroughly : A. Indra releases the entities from the bonds of the primordial world by breaking its power of obstruction (vṛtrá). B. The entities go over to the gods, even before Indra accomplishes his demiurgic act, either of their own accord (see, e.g., p. 16) or because they have been called forth by Indra (cf. nirhā́ya, p. 18). C. They are stolen from the primordial world before Indra's exploit, by a mythological figure who comes "from outside" (his place of origin is never specified) and overcomes the fiendish powers of that world: thus Mātariśvan steals Agni, and the śyená or suparná, Soma. For a more ample discussion see below, p. 102. The idea which underlies the different processes of B and C was apparently that Indra was unable to accomplish his act unless he was first strengthened by the powers of Soma. Hence the logical contradiction of a god who must drink Soma to acquire the power needed to deliver Soma. In the

leave the sphere of true speech, never again to be involved in untruth". In IIJ. 8, p. 109 n. 68, the question was raised whether in antár vā́rune the word may still have been used as an appellative for the nether world, its application as a proper name to the god being due to taboo. It may be objected that as early as circa 1380 B.C., in the treaty of the Mitanni king Šattiwaza, Varuṇa was already used as a proper name. However that may be, the inference drawn from the Vedic material that Varuṇa's palace was in the nether world (IIJ. 8, p. 107ff.) is confirmed by the epic, which locates it in the Pātāla (cf. Mbh. V.96.5. avagāhya bhūmim and see n. 305).

52 See Vodskov, Sjæledyrkelse og Naturdyrkelse p. 211.

53 Cf. X.51.5 támasi kséṣy agne. For the function and the form of the aorist see K. Hoffmann, Injunktiv, p. 157 (with n. 100), and Johanna Narten, Sigmatsche Aoriste, p. 60, 255 respectively.

Yajurveda there is a fixed formula which says that Indra slew Vṛtra with the help of Agni and Soma 54.

In X.124 we are concerned with the evocatio, the calling off of Soma, along with the change of sides of Varuṇa, Agni and Soma. The poet apparently did not regard the two notions as basically different. The evocatio occurs in 6c, where Indra addresses Soma with the words "Let us slay the vṛtrá (Vṛtrá). Come out, O Soma !" (hánāva vṛtrám niréhi Soma).

On the other hand, the gods go over of their own accord in 4c "Agni, Soma and Varuṇa, they go over" (Agníh Sómā Várụṇas té cyavante) and in 4b "preferring Indra I leave the father" (Índraṃ vṛṇānā́h pitaráṃ jahāmi). The last pāda shows a well-known pattern, which Geldner 55 has rightly recognized in Atharvaveda XII.1.37 "The pure earth starts in fright away from the serpent (sarpá́)" and "choosing Indra, not Vṛtra, she, the earth, adheres to Śakra, the lusty bull" 56. Pāda c Índraṃ vṛṇānā́ prthiví na Vṛtráṃ should especially be noted 57. Geldner also points to RS. IX.97.41 "That great deed the bull Soma has performed that he, the embryo of the waters, chose the gods" (apáṃ yád gárbhó 'vṛṇīta devā́n).

A similar situation is found as late as the Mahābhārata : while in the myth of the Churning of the Ocean Śrī is among the first gods of Creation that emerge from the primeval waters through the joint effort of Asuras and Devas 58, in her conversation with Indra (XII.221.26-27) she herself gives an entirely different account of her transition to the gods: 59 "Formerly I lived among the Asuras, although I am the origin of Truth and Law. When, however, I saw that they were wicked, I preferred to live with thee". (Śakra said:) "What was the behaviour of the Daityas among whom thou livedst, O fine-faced woman? And what didst thou see which made thee come here, leaving the Daiteyas and Dānavas?" 60

A more or less parallel passage (Mbhb. XII.91.21-22) describes how she left Bali: "Then Śrī went away from him, as she lacked majesty (living) with him. (22) Thereupon, having left him, she went to Pākaśāsana [Indra]. Afterwards he [Bali] regretted it when he saw Śrī (living) with Puramdara

54 Cf. MS. II.1.3 (5,1), KS. XXIV.7 (97,18), KKS. XXXVII.8 (202,20) agniṣómā-bhyāṃ vai vīryèṇa 'ndro vṛtrám ahan, TS. I.6.11.6, VI.1.11.6, AB. II.3.12 agniṣómā-bhyāṃ vá Indro vṛtrám ahan. In ŚB. XI.1.6.19 Agni and Soma are called Indra's brothers. See also RS. IX.61.22 sá pavasva yá āvithe 'ndraṃ vṛtrā́ya hántave, and cf. Etudes asiatiques 25 (1971), p. 87.

55 In the introduction to his translation of X.124 (Übersetzung III, p. 352f.)

56 yápa sarpáṃ vijámānā vimírgvari yáśyāṃ déann agnáyo ápe antáh, párā dásyūn dádāti devapīyúṃ Índraṃ vṛṇānā́ prthiví na vṛtráṃ, śakrá́ya dādhre vŕṣaṇāya vŕṣṇe.

57 Cf. MS. III.7.3 (77, 16), KS. XXIV.1 (90,7), KKS. XXXVII.2 (195,17/228,1), TS. VI.1.6.5.

58 Mbbh. I.16.34 crit.ed. (= Matsyapurāṇa 250.3) Śrīr anantarām utpannā ghṛtāt paṃduravāsinī. Cf. Mbhh. XII.34.13, Bhāgavata Purāṇa VIII. 8.3 lataś cāvir abhūt sākṣāc Chri ramā.

59 Hopkins, Epic Mythology, p. 47, wrongly characterized this passage as a late moralization.

60 asureṣv avasāṃ pūrvāṃ satyadharmanibandhanā, viparītām̌s tu tān buddhvā tvayi vāsaṃ arocayam (Śakra uvāca) kathamvṛtteṣu daiteyṣu tvam avāsiṛ varānane, dṛṣṭvā ca kim ihāgās tvaiṃ hitvā daiteyadānavān?

[Indra]"61. Another illustration of this mythical pattern, viz. Uśanā Kāvya's going over from the Asuras to the Devas, will be discussed on p. 93ff.

After these introductory remarks stanza 2 of X.124 will be clear: "Now that I, being a Deva, go stealthily away from him who is no Deva, I go, looking forward, to Life: while leaving, being unfriendly myself, him who was friendly (towards me), I go over from my own party to a foreign lineage" 62. The implication is that, before the Devas arrived on the scene as a foreign element, Agni, as a child of the primordial world and born from the womb of the Asura 63, was himself an Asura. In several passages he is, indeed, denoted as such 64. Cf., e.g., IV.2.5 "This (sacrifice), O Asura, comprises oblations, it secures us offspring, it is long-extended wealth with a broad foundation, which secures (the possession of, or predominance in) a sabhā" 65. Like Varuṇa and Soma, Agni has an ambiguous character.

Just as in the ritual Soma is tied up (úpanaddha) as long as he is in Varuṇa's power, there is also an Agnír úpanaddhah. See Hillebrandt, Ved. Myth. II2, p. 60f. and for the interpretation IIJ.15, p. 231 66. Such statements as JB. II.155 line 7 tasmād āhur agníṣomāv asuryāv iti refer to this particular mythological situation, which has its counterpart in the ritual, e.g., when the Soma-plant has just been bought from the Soma vendor. Cf., e.g., KS. XXXIV.3 (37,19) somo vā eṣo asuryà iti tu, tasmān nā 'bhiṣutyah 67. These mythological notions of a nether world aspect of Agni and Soma have been preserved with a remarkable tenacity. In the

61 (21) athā 'smác (viz. Bali) chrīr apākrāmad yā 'min nā 'sīt pratāpinī (22) tatas tasmād apakramya sā 'gachat pākaśāsanam, atha sa 'nvatapat pāścāc Chriyam drstvā puramdare. I take pratāpin in the sense "majestic, powerful" since yā 'sminn āsīt pratāpinī ("burning, scorching, paining") would not seem to make sense. In V.106.12 atra pātālam āśritya Varuṇah śriyam āpa ca the word śrī is apparently used as an appellative.

62 (2) ádevād deváh pracatā gúhā yán prapásyāmano amrtatvám emi, śivám yát sántam ásivo jāhāmi svát sakhyád āráṇiṁ nábhim emi. It should be noted that amrtatvám can be construed with emi (Vodskov) or with prapásyāmano (von Bradke, Hillebrandt, Geldner). For sakhyá in the sense of "party" see IIJ. 4, pp. 237, 250.

63 III.29.14 yád ásurasya jathárād ájāyata (see n. 41) and 11a gárbha āsuró with Geldner's note (Übersetzung I, p. 363) "Asurisch (asurá) heisst Agni, solange er verborgen ist." Cf. also Renou, EVP. 12, pp. 70, 126.

64 See Bergaigne, Rel. véd. III, p. 145, Geldner, Ved. Studien II, p. 299 n. 1.

65 īlāvám eṣó asura prajā́vān dīrghó rayíb prthubudhnáh sabhā́vān. Cf. Renou, EVP. 13, p. 4 and his comment (p. 88): "Agni comme Asura, càd. 'maître', emploi reflétant un état antérieur à l'accession de devā." See also ibid., p. 145 (on Agni as an Asura).

66 In view of ŚB. I.4.1.34 the three Agnis mentioned in TS. II.5.8.6 (viz. of the gods, the pitṛ́s and the Asuras) are probably due to a later systematization : tráyo vá agnáyo, havyaváhanó devā́nām, kavyaváhanáh pitṛ́nā́m, saháraksā ásurāṇām. The name saharakṣás "accompanied by Rākṣasas", although correctly given as such by Eggeling, ŚBE. XII, p. 110f. (I.4.1.36), has caused some slips: Ludwig, Der Rig Veda IV, p. XVII (sahasrákṣah), Keith ("guardian").

67 In this connection the identification of Soma with Vṛtra should be considered. Cf. MS. III.7.8 (87,17) Sómo vái Vṛtráh, KS. XII.3 (165,1), ŚB. III.4.3.13; 9.4.2, IV.1.4.8; 2.5.15 Vṛtró vái Sómāśit. Geldner refers to RS. IX.74.7; 99.1 (Vedische Studien II, p. 299 n. 1). See Hillebrandt, Ved. Myth. III1, pp. 234–236 (= II2, p. 186) and cf. History of Religions 10 (1970), p. 107.

21

epic description of the town Pātāla in Varuṇa's subterraneous nāgaloka (Mhbh. V.97.3f.) both an āsuro 'gniḥ and the amṛta are said to be present 68.

4c. Varuṇa's ādhipatyam

Before the creation of the organized dualistic cosmos as well as after it, Varuṇa's position appears to have been, like that of Agni and Soma (and perhaps even more so), basically ambiguous. This ambiguity, however, is of a characteristically subtle nature. Very seldom does he overtly act as an opponent of Indra, and never does he fight the Devas as a warrior. In the cosmogonical fight against the powers of the primordial world Indra's attacks are exclusively directed against Vṛtra and the primordial hill, in which the force of obstruction is concentrated 69. As far back as the eighties of the nineteenth century, however, Bergaigne and von Bradke had rightly stressed the antagonism between the two gods in the Rigvedic hymns IV.42 and X.124. The latter hymn, however, deals with the moment when the fight was ended and the Devas had acquired the supremacy. Its importance lies in the fact that it shows how the poets conceived the organization of the cosmos after the defeat of the Asuras, and Varuṇa's incorporation in the world of the "second creation". As for IV.42, it does show the antagonism between Indra and Varuṇa during the cosmogonical fight but in the form of a bragging contest (vivāc). Whether or not the hymn has been composed for a royal consecration, this does not alter the fact that the general structure of vv. 1–6 points unmistakingly to a verbal contest 70. Cf. the first words of these verses: máma (1), ahám (2), ahám (3), ahám (4), máṃ (5), ahám (6). In the first three verses Varuṇa is speaking. He refers to the primordial character of his sovereignty: 'I am king Varuṇa, to me belonged that primordial sovereignty (āsuryāṇi prathamā)' (v. 2).

68 atrā 'suro 'gniḥ satatam dī̄pyate vāribhojanāḥ, vyāpṛṇa dhṛtātmānaṃ nibaddhāṃ samābudhyata (4) atrā 'ṃṛtaṃ suraiḥ pīvā nihitam nihatāribhīḥ. Cf. also I.19.6 pātāla-jvalanāvāsam asurāṇāṃ ca bandhanam. In Mhbh. III.221.15 Bomb. Night is said to have given birth to Agni and Soma (which would be reminiscent of Hesiod, Theogonia 107, 123f., 211f.): niśā tv ajanayat kanyāṃ Agniṣomāv ubhau tathā, Manar evā 'bhavad bhāryā suguve pañca pāvakāṃ, but the correct reading is without any doubt niśāṃ (III.211.15 crit. ed.).

69 For VI.68.2 vṛtratúrā, said of Indra and Varuṇa, see p. 44 and Bergaigne III, pp. 140, 142. This is a case of incidental assimilation, just as when Indra and Viṣṇu are said to stride out intoxicated by Soma (RS. VI.69.5, one of the most instructive passages in this respect), or when Indra and Varuṇa are implored to give their mṛḍiká "mercy" (VII.82.8, cf. Yt. 10.5 āca.nō jāmyaṭ mṛḍikā). Varuṇa's natural position, however, is outside the battle-field, see pp. 38, 106f., 108f. At best, he helps the Devas, e.g., JB. I.180. As a rule, however, the Vedic Varuṇa does not fight and cannot, therefore, be defeated. In the epic he fights the Daiteyas with his nooses (Mhbh. III.42.27–29). Different are Harivaṃśa App. I. 42.2209 (but cf. 42.2146 !) and the Purāṇas (nn. 445, 448).

70 See Bergaigne III, p. 142, von Bradke, p. 31. The interpretation of IV.42 as a cosmic vivāc was first suggested in IIJ. 4 (1960), p. 270 n. 80b and Numen 8 (1961), p. 38, that as the hymn of a "Königsweihe" by Lomme1, Festschrift Schubring, p. 32ff. and accepted by Schlerath, Königtum, p. 160. But see p. 42 and Gonda, Dual Deities, pp. 234–248.

22

Varuṇa tries to disguise the difference between himself and Indra: “I, Varuṇa, am (also) Indra . . . I have united these two wide, deep and well-founded worlds and have supported them” (v. 3). Indra, who comes second as the new-comer and pretender in this world and who cannot base his claims on old rights, is speaking in the next three verses (4–6). He starts with disputing Varuṇa's last statement: “I have supported (dhāráyam) the sky in the seat of the Ṛtá” (v. 4b), which clearly refers to 3d sám airayaṃ ródasī dhāráyam ca. He then goes on by stressing the difference between Varuṇa and him: “Aditi's son is righteous (ṛtávan) according to the Ṛtá; he has also extended the earth in a threefold way” (v. 4), “(but) me men with good horses invoke in their contests . . .” (v. 5). The verses 5–6 have an unconcealed bragging character: “I have performed all these (acts), no divine force can restrain me, as I am irresistible . . . both endless worlds fear (me) when (etc.)” (v. 6). Such passages have often been misinterpreted as though Indra were the parvenu among the gods. In fact, his position is comparable to that of, e.g., Tiṣtrya in the Avesta (see IIJ. 4, p. 255). He is trying to win a new status, which implicitly amounts to disputing Varuṇa's. It is significant that, while Varuṇa's words are taken for granted, a human speaker, immediately after Indra's ostentation, accepts his claims: “Thine (acts) all beings know; thou art proclaiming (prá braviṣi) them to Varuṇa, O vedhas; thou art famous for having struck down the obstructions (vṛtrā́ṇi), thou hast made the rivers stream, which were obstructed (vṛtā́n)” (v. 7). The repetition of “thou” (tvā́m) is important, since it corresponds to ahā́m (4a 5d 6a) and máṃ (5ab) in Indra's words. What is at stake is the justness, not of Varuṇa's reference to his old rights but of Indra's claims, which he proclaims to Varuṇa and which the poet confirms. From now on the situation in the cosmos has fundamentally changed, and the two gods are further invoked together (Indrāvaruṇā 9b, 10c). The last fact would seem to exclude Hillebrandt's interpretation of Varuṇa as the god of the old year (Ved. Myth. I2, p. 26), since in that case Varuṇa would have been replaced by Indra. Instead of it, we find the same situation as that presupposed in X.124. On the other hand, the question as to a possible ritual context in which this cosmic vívāc may have been dramatized is not relevant for our purpose; see p. 43. The only thing that matters here is, how this contest was visualized.

To some extent the rareness of the references to this antagonism in the Rigveda may be due to an intentional reticence on the part of the poets 71. The assumption of a taboo, which withheld the poets from clearly expressing their ideas is not entirely unfounded. If the theory that at the beginning of every new year Indra had to re-enact his cosmogonial fight is correct, it follows that not only the powers of Chaos (the ásurā adevdh) temporarily returned into this world from the place to which

71 See Bergaigne III. p. 76, Lévi, Doctrine du sacrifice p. 167f., Hillebrandt, Ved. Myth. II2, p. 127, Gösta Johnsen, IIJ. 9, p. 260 n. 74 and cf. IIJ. 3, p. 211, 5, p. 52f., and 15, p. 226.

23

they had fled after their defeat, but that also Varuṇa in this critical period must have reassumed his Asura-character of the primeval world. This means that at that time of the year he was not a Devá ásura but a dangerous Asura tout court. He must have been dreaded all the more as also his introduction into the world of the cosmos apparently did not result in an unambiguous position. Even in sources as late as the Mahābhārata there are indications to show that Varuṇa, although a god of this organized world, continued to maintain secret relations with the suppressed demons and the “night-side” of this world (see p. 90). His majestic character, too, was based upon his partly transcending the boundaries of the organized cosmos (see p. 38).

In this connection it may be useful to consider more closely X.124.5, quoted above:

nírmāyā u tyé ásurā abhūvan tvám ca mā Varuṇa kāmáyāse, ṛténa rājan ánṛtam viviñcán māma rāṣṭrásya ‘dhipatyam éhi.

From the translations 72 it is apparent that the chief difficulty of this verse lies in the last words máma rāṣṭrásya ‘dhipatyam éhi. Most translators seem to assume that with these words Indra offers Varuṇa, more or less as a compensation, a subordinate position in his realm. Oldenberg, Noten a.l., rejected this interpretation and translated, in essentially the same manner as Geldner in 1897, “komm her zu dem (d.h. erkenne an das)

72 1876: Ludwig, “mein ist des reiches oberherrlichkeit; komme [lasz es gut sein]”; 1877: Hillebrandt, Varuṇa und Mitra, p. 108 “komme heran zur Oberherrschaft über mein Reich”; 1877: Grassmann, “komm her in meines Reiches Oberherrschaft”; 1882: Bergaigne, Rel. Véd. III, p. 146 “deviens le suzerain de mon royaume”; 1885: P. von Bradke, o. c., p. 99 “und wenn du, o Varuṇa, nach mir verlangen trägst, so komme, o König, der du Recht vom Unrecht unterscheidest, und sei Oberherr über mein Reich (oder über mich und das Reich)”; 1894: S. Sørensen. Festskrift til Vilhelm Thomsen, p. 335f. (not accessible to me); 1897: Vodskov, “kom du hid til mit Riges Overherredømme”; 1897: Geldner, Vedische Studien II, p. 298 “und wenn du, o Varuṇa, der Wahrheit von der Lüge scheidest, o König, mich als den Oberherrn meines Reiches magst, so komm!”; 1913: Hillebrandt, Lieder des Ṛgveda, p. 21 “und du, willst du mich lieben, o Varuṇa, tritt, Recht und Unrecht scheidend, an die Oberherrschaft über mein Reich” (otherwise 1927: Ved. Myth. II2, p. 63 n. 2, where he disregards Delbrück’s observations on the use of ca with the subjunctive, see Altindische Syntax, p. 329f.); 1951: Geldner, Übersetzung III, p. 354 “Wenn du, Varuṇa, mich lieben willst, so tritt, o König, der das Rechte scheidet, die Oberherrschaft meines Reiches an !”; 1965: Renou. EVP. 14, p. 29 “Si toi, tu m’aimes, ô Varuṇa, toi qui discrimines, ô Roi, d‘avec l’Ordre ce qui est non conforme à l’Ordre, adhère à la souveraineté sur mon empire !” Recent translations (e.g., P. Horsch, Die vedische Gāthā- und Śloka-Literatur [1966], p. 239 n. 3) are virtually identical with Geldner’s. J. J. Meyer’s interpretation of this passage in Trilogie altindischer Mächte und Feste der Vegetation (1937), III, p. 213 (cf. p. 210) is based on a misunderstanding of the mythological context. Renou’s comment (p. 98) “éhi, nuance conforme à emi 2b” is not quite clear to me: in 2b Agni leaves the father Asura, whereas here Soma is invited to come to Indra (a + i-). In any case, it explains nothing about the ádhipatya.

ddhipatyam, das ich in meiner Königsherrschaft ausübe" 73. Several objections must be raised to this interpretation.

First, the words máma rāṣṭrásya are more problematic than is commonly realized. They seem to suggest that Indra is the king of the new order which he has inaugurated. This, however, is clearly contradicted by the Vedic texts. Our reconstruction of the Vedic cosmogonical myth implies that Indra is a seasonal god. He mainly functions in the short period during which in later times the Indradhvaja was erected, that is, about seven days. Only then is he the principal god who restores life and renews the cosmic order (as the symbolism of the tree demonstrates). No more than Varuṇa, however, is Indra a god of the totality. During his festival, it is true, the tree (with which he is identified and which in classical texts is even called Indra) is the cosmic centre. In the system of classification, however, he is never localized in the centre.

It need hardly be pointed out that rājan is the characteristic function of Mitra and Varuṇa 74, whose mother Aditi therefore bears the title rājaputrā "mother of kings". Theirs is the kṣatra 75, just as Ahura Mazdā's power is symbolized by his xšaθra 76. Indra's relation to the Devas, on the other hand, is characterized by his being jyeṣṭhá. Cf., e.g., VIII.63.12 indrajyeṣṭhā asmān avantu devāḥ 77, ŚB. III.4.2.2 indrajyeṣṭhā devā́ iti. Only once, in one of the late Vedic texts (TB. I.5.6.4) does the term indrarājan occur, preluding the classical use of such terms as devādhipa, amararāja, etc. 78. Indra, it is true, is often called a king 79, but he is a

73 There is no reason to doubt (with Oldenberg, Noten) the correctness of ādhipatyam, the analysis of the Padapāṭha. See Wackernagel-Debrunner, Altindische Grammatik II/2, p. 817. Incorrect Edgerton, Language 19 (1948), p. 112. In such matters we may trust the Padakāra's knowledge of the Vedic language, since he would scarcely have invented a non-existent ādhipatyam instead of ādhipatyam (AS. VS. MS., etc.).

74 Whatever the exact meaning of the term samrāj, which is the characteristic epithet of Varuṇa (VII.82.2), the latter's kingship is sometimes said to be universal, as in II.27.10 tvám viśveṣām Varuṇā 'si rājā yé ca devā́ asura yé ca mártāḥ (see p. 39), AS. I.10.1 ayám devānām ásuro ví rājati, vā́stā hi satyá Varuṇasya rājñāḥ, MS. I.6.11 (104,2), II.2.1 (15,18), ŚB. XII.8.3.10 Varuṇo vai devānām rājā. Of a special nature are Varuṇa's words in his bragging contest with Indra IV.42.1 viśvāyor viśve amŕtā yáthā nah, krátum sacante Varuṇasya devā́ rājāmi krṣṭír upamásya vavréḥ (see p. 22). Although Varuṇa begins with stressing his position as rājā (a: máma dvítā rāṣṭrám kṣatriyasya), this rāṣṭrám is specified with a reference to his krátū (cf. Ahura Mazdā's xratu). It is also Varuṇa who lays the krátū in the hearts of men, cf. V.85.2 hŕtsú krátum Varuṇo apsv Agním dví sūryam adadhāt Sómam ā́drav. The exact nature of this kingship, which Varuṇa possesses "from of old" (?divi) and which he seems to trace back to his position as supreme Asura in the primordial world, must still be defined in detail. It is not easy to see how the adhipati of the nether world could rule over the Devas, unless this was meant in a special sense.

75 RS. V.68.3 māhí vāṃ kṣatrā́m deveṣu, VI.67.5 víśve yád vāṃ . . . kṣatrā́m devā́so 'dadhuḥ sajóṣāḥ.

76 See B. Geiger, Die Aməša Spəntas, pp. 204-232.

77 For indrajyeṣṭhá(sa)h in the Rigveda see I.23.8, II.41.15, VI.51.15, VII.11.5. In PB. VII.8.3 Indra is bṛṣṭha of the gods. Cf. JB. II.128, 141, etc.

78 Mbh. V.10.7, I.83.6, 12 and Hopkins, Epic Mythology p. 122, Sørensen, Index to the Names in the Mahābhārata, p. 333ff.

79 rāṭ I.121.3, rājā I.174.1, 178.2, V.40.4, VI.24.1, VII.18.11, 27.3, rājānam VII.31.12, rājan I.63.7, VI.19.10, 39.5. In VII.84.1 rājānau is used for addressing Indra and Varuṇa.

25

svardj, a king in his own right, not a samrdj, which is the specific title of Mitra and Varuṇa. There can be no doubt that the poets and the priests chose their terms very carefully. In TS. V.5.9.4 the Vasus, pitaras, Ādityas and Maruts are invoked for protection in their respective quarters, after which the prayer ends with the words (5) devds tré 'ndrajyeṣṭhā várunarājano 'dhisthāc copáriṣṭāc pāntu (see nn. 183, 328). It is likely, therefore, that an ancient distinction has been preserved in the rajus agnihvarébhyas tvā rtāyúbhyā indrajyeṣṭhebhyo várunarājabhyo vdtāpibhyah parjányātmabhyah 80. In the light of such passages the terminology of RS. X.66 is particularly interesting. The poet, after invoking in verse 1 the gods "whose senior (or superior) is Indra" (deván huvé... yé... índrajyeṣṭhāso 'mŕtā stāvidhāḥ) goes on to say in Verse 2 índrapraśiṣṭā yé sáryasya jyótiṣo bhāgám ānaśúḥ '(the Maruts) who, impelled by Indra and by the command of Varuṇa, have got their share of the sunlight" 81. This does not necessarily point to a difference of rank, one being superior to the other, but it certainly proves that Indra was not, in Vedic mythology, the rājan of the Devas, even though he more than once receives that title.

When, therefore, Indra speaks of his rāṣṭrá, this cannot simply mean his "Königsherrschaft", unless one should limit this term to the very beginning of the organized cosmos. Nor can Geldner's translation "die Oberherrschaft meines Reiches" be correct.

When Indra has established the New Order, his role dwindles down to the comparatively modest one of a dikpāla, a protector of one of the quarters. The Indra festival, as we know it from later sources, illustrates his character of a seasonal god. It is inconceivable that Indra should offer Varuṇa a high rank for the brief period that he himself is the central god. Indra is speaking on behalf of the Devas to an Asura and the only sense that the poet can have attached to the words máma rāṣṭrásya is "the (new) reign [of the Devas] that has been inaugurated by me".

Secondly, it has often been demonstrated that as early as the Rigveda Varuṇa was closely associated with the waters 82. There is no reason to

80 MS. I.3.35 (41,16) in the Dadhiigraha. Cf. also KS. XXIV.5 (173,9), KKS. XLV (XLIV).6 (272,14/319,20) and TS. III.5.3a, which reads agnijitrébhyas tvā (with a curious accent) and has a slightly different word order. Note RS. X.132.4 (= II.27.10) tvám viśveṣām Varuṇā 'si rājā (with Oldenberg's note), VIII.93.11 and Bergaigne III, p. 76. See n. 74.

81 Otherwise Geldner: "Die von Indra befehligt, von Varuṇa unterwiesen, Anteil an dem Sonnenlicht erlangt haben." The two epithets are hapaxes in the Rigveda.

A late variation occurs in Kauś. 3.3 indraprasisthā varuṇaprasūtā āpah samudrād divam ud vahantu (see Lüders, Varuṇa, p. 104f.).

82 See, e.g., Lüders, Varuṇa, pp. 9-13, and Dandekar, Annals Bhandarkar Oriental Research Institute XXI (1940), p. 190, who refers to II.28.4, 38.8, V.85.3, 6, VII.87.1, VIII.41.2, X.75.2, AS. VII.88 (83), 1. The last passage runs as follows: apśú te rājan Varuṇa grhó hiranyáyo mitáh, táto dhrtávrato rājā sārvá dámani (MSS.: dhámani) muñcátu. Cf. MS. IV.7.8 (104,9) samudró vái Varuṇo, daksinardhá ālabhyáte yajamānasya nirvaruṇatvayai, 'tád vái eṣá 'bhyánūktā (RS. VIII.69.12): sádevo asi Varuṇa yásya te saptá eṣṭhávaḥ, ánukṣáranti kákūdam súrmyām suṣirám iva, iti, MS. II.5.6 (55,4) etá vái pratyańkṣam várunir yád āpah,

consider this the result of a later development 83. On the contrary, it must have been the original function of Varuṇa to represent or impersonate these waters. What earlier scholars have failed to recognize is that before Indra's demiurgic act Varuṇa and the Asuras were the gods of the primordial world which consisted of the waters. After the emergence of the earth floating on the waters and the subsequent creation of the organized world, the waters (āpah, salilá, samudrá) were thought of as being under the earth, as its foundation (pratiṣṭhá) 84, as well as surrounding it. Varuṇa's association with the waters, therefore, is due to the fact that the primeval waters, along with their lord Varuṇa, have been incorporated in the cosmos as part of the nether world. Hence it is that Varuṇa is said to dwell amidst his seven sisters 85 and that these seven rivers, when they flow from the central mountain over the earth, are said to emerge (from the samudrá in the nether world) through Varuṇa's throat as through a hollow reed: VIII.69.12 sudevó asi varuṇa yásya te saptá síndhavah, anuṣáranti kākúdạṃ súrmyạṃ suṣirdạṃ iva 86. A late confirmation of this interpretation (see n. 316) can be seen in the name Varuṇadeva, which in the thirteenth century was given to the stone slabs of water wells in Chamba State 87. See p. 84.

In none of the discussions of X.124.5 has the technical meaning of the term ādhipati been given due consideration. Oldenberg, Noten II, p. 124, even regarded it as a mere synonym of rájā. In Vedic cosmogony, however, ādhipati is the specific term for a god or group of gods who, in the system of classification, have the function of protector(s) of one of the points of the compass. As such they exist as soon as the Cosmos arises from Chaos, for the world order is inconceivable without a system of classification. Cf. VS. XIV.30, ŚB. VIII.4.3.16 dyávāprthiví vy àitām, Vásavo Rudrá Ādityá anuvy àyáṃs, tá evá 'dhipatayā āsan, which Eggeling renders as follows: "Heaven and Earth went asunder, the Vasus, Rudras and Ādityas separated along with them: they indeed were the lords!" 88 A similar description occurs in JB. II.141: Prajāpati first created the Vasus,

KS. XIII.2 (180,20) ápo vai Varuṇah, TS. VI.6.3.4 samudré hy àntár Várunah. In the Rigvedic passage quoted above the word sudeva should be noted: in his auspicious aspect as sudeva Varuṇa, although normally the god who withholds the water (see e.g. IIJ. 5, pp. 47, 51), is expected to let the water flow from the subterranean samudrá, see below, p. 27. See also Renou's remark on Varuṇa as "dieu bloqueur" in EVP. 7, p. 82f. In his auspicious aspect Varuṇa is identical with Mitra: Lüders, Varuṇa, p. 715, rightly remarks that whenever Varuṇa gives rain, Mitra is always mentioned together with him.

83 Cf. IIJ. 8, p. 115.

84 For references see, e.g., India Maior, p. 145 n. 1.

85 See, e.g., Lüders, Varuṇa, pp. 51, 154, 274, and cf. India Maior, p. 151.

86 For Geldner's entirely different translation see n. 316.

87 See J. Ph. Vogel, Antiquities of Chamba State I, p. 31ff., and cf. IIJ. 4, p. 249, 5, p. 52, 8, p. 112 n. 91.

88 Cf. KKS. XXVII.4 (107,13/124,23) and TS. IV.3.10.2 tésāṃ ādhipatyam āsīt. MS. II.8.6 (110,19) and KS. XVII.5 (249,4) only mention Vásavo Rudráḥ. Cf. ChU. III.6.4 (7,4, 8,4), see p. 59.

27

Rudras and Ādityas. They called themselves adhipatayah, since they were superior and senior (śreṣṭhāḥ and jyeṣṭhāḥ). Then, however, Prajāpati created Indra to be a chief and adhipati over them (eteṣām . . . śraiṣṭhyāyā'dhipatyāya) but Indra objects "How could I be a chief, an adhipati over them?" (katham eṣām śreṣṭho 'dhipatiḥ syām). The word usually denotes a function, which must be specified: Agni is the adhipati of Heaven, and Yama of the next world 89.

Since adhipati is especially used with reference to a function in the system of classification, it should be stressed that, in spite of the fact that only a few traces of the Vedic system can be found in the Rigveda, its antiquity cannot reasonably be questioned. Similar systems are well known from various archaic cultures 90. It is inconceivable, therefore, that later theologians could have created the Vedic system ex nihilo : at best they may have elaborated it.

The Atharvaveda-Saṃhitā, which is much closer to the brāhmaṇas in this respect, mentions the adhipatis in several passages. In AS. III.27.1–6 they have the function of the dikpālas of classical literature: Agni in the East, Indra in the South, Varuṇa in the West, Soma in the North, Viṣṇu in the centre (Sāyaṇa: nadir) and Brhaspati in the zenith. The Yajurveda has a series of formulas in which each of the quarters is addressed with its own specific name and its specific adhipati is mentioned. For the western quarter it has samrād asi pratīci dig Ādityās te devādhipatayaḥ 91.

In the system of classification the gods are also adhipatis of all kinds of entities and it is in this connection that Varuṇa is mentioned as the adhipati of the waters. Cf. AS. V.24.4 Váruno 'pám adhipatiḥ sá mā 'vatu (etc.), 5 Mitrávarunau vrṣṭyádhipatī tau mā 'vatām (etc.), TS. III.4.5.1 Agnír bhūtānām adhipatiḥ sá mā 'vatv, Índro jyeṣṭhānām, Yamáh prthivyád . . . Mitráḥ satyānām, Váruno 'pám. This implies that Varuṇa's sphere is different from Indra's and that conflicts might arise between the two domains. Aquatic animals are not subject to Indra and this is obviously the background of a cult legend told in JB. III.193. It relates that once upon a time all creatures praised Indra except the dolphin, who declared: "I will not praise thee. I move within the waters diving near (thee?) [or: "I continue diving near (thee?) within the waters"] and only so much I praise thee" (apsu antaś carāmy upanimajjann etāvato ahaṃ tvāṃ

89 Cf., e.g., MS. III.2.3 (18,2) Yamó 'múṣyasa lokasyā'dhipatyam ānase (last word as corrected by Caland, ZDMG. 72 [1918], p. 7), AB. III.43.1 Agnir vai svargasya lokasyā 'dhipatis, tám Vasavah prathamād āgachams . . .

90 For references see Rodney Needham's "Introduction" to Primitive Classification by Emile Durkheim and Marcel Mauss (The University of Chicago Press, 1963), which is a translation of "De quelques formes primitives de classification: contribution à l'étude des représentations collectives", Année Sociologique VI (1901-1902), Paris 1903, p. 1-72.

91 MS. II.8.9 (113,14), KS. XVII.8 (251,5), KKS. XXVI.7 (109.12/127,11), TS. IV.4.2.1–2, VS. XV.12, etc. For a different arrangement in the Atharvaveda and the Yajurveda see below, p. 56f.

stuyām). Cf. upanyāmajjat TB. I.1.3.6 (samīpe Sāyaṇa), KS. VIII.2 (84,14),

KKS. VI.7 (66,4/77,10), but here the exact reference of upa- is not clear.

Varuṇa moves through the sea (RS. I.164.14 adbhír yā́ti Várunah samudré).

In the same way the dolphin belongs to the sea (MS. VS. samudrā́ya

śiśumáro) and as such refuses to praise Indra in the common way. Although

its diving is vaguely reminiscent of “Varuṇa’s trick” referred to on p. 77,

it does not justify its refusal by a reference to Varuṇa.

Although in the epic Varuṇa is mostly denoted as apā́m patī́h, it is

actually his ādhipatyam which forms part of the religious tradition. As

far back as 1885 Oldenberg pointed to Mbhb. V.16.33-34 92. In vv. 27-28

Kubera, Yama, Soma and Varuṇa come to Indra to congratulate him

with slaying Vṛtra. Indra then asks them to help him in killing the

rājā devānām Nahuṣah and they ask him what their reward will be. He

replies

(31) bhavatu bhāvān apā́m patir, Yamah Kuberaś ca mahābhiṣekam,

samprāpnuvantu adya sahaiva tena 93. The episode then ends with the

two ślokas

(33) evam saṃcintya bhagavān Mahendraḥ pākaśāsanaḥ

Kuberam sarvayakṣāṇām dhanānām ca prabhuṃ tathā

(34) Vaivasvatam pitṝṇām ca Varuṇam cāpy apā́m tathā

ādhipatyaṃ dadau Śakraḥ satkṛtya varadas tada

Cf. also V.109.8, where Kubera is said to have been installed as an

adhipati over Rākṣasas, Yakṣas and Gandharvas (ādhipatyena . . .

abhiṣecitah, whereas Varuṇa was consecrated as king of the Yādasas and

as protector of the waters (V.108.3), and Indra as king of the Gods

(V.106.7). See p. 79f.

The relevance of these facts for the interpretation of X.124.5d máma

rāṣṭrāsyáḍ ‘dhipatyam éhi is clear. According to the Yajurveda, then, the

three groups of gods were already present as ādhipatis at the moment

when Heaven and Earth went asunder (see p. 27), that is, at the moment

of Indra’s demiurgic act. Since the Atharvaveda already refers to Varuṇa

as apā́m ādhipatiḥ, it can safely be assumed that the ādhipatya which

Indra offers Varuṇa immediately after the defeat of the Asuras (when

Varuṇa has gone over to the “foreign lineage” of the Devas), is the

ādhipatya over the waters and the western quarter. The fact that this

is related in the tenth book of the Rigveda should not be interpreted

chronologically, as a proof that this is a notion which has developed in

a period posterior to the family collections, but rather in the light of

the general character of the tenth book, which is closer to that of the

Atharvaveda.

On the basis of the preceding interpretation it cannot be said that

92 In “Ākhyāna-Hymnen im Ṛigveda,” ZDMG. 39, p. 71 n. 1 (= Kleine Schriften,

p. 493 n. 1). In his Noten II, p. 344, he wrongly withdrew this reference on account

of Geldner’s remark that “es sich dort nur um das bekannte ādhipatyam über die

Wasser handelt” (Vedische Studien II, p. 302 n. 1). This, however, is the very

thing that RS. X.124.5 refers to.

93 For Varuṇa's abhiṣeka see below, p. 79f.

Varuna is subordinated to Indra. They are on the same level but their

fields of actions are entirely different. Bergaigne's definition of Varuṇa's

character as "d'une suprématie nominale et d'une autonomie morale"

(Rel. védique III, p. 146), although thought-provoking like all this

sharp-sighted scholar has written on Varuṇa, does not go to the root of

the matter. Still less so does Hillebrandt's characterization of Varuṇa as

"Herrscher" and of Indra as "Feldherr" 94. As far as their rank is concerned

they appear to be equals. In the second part of this study an attempt

will be made to show that a similar essential equality is characteristic

of the relation of vidūṣaka and nāyaka in the Sanskrit drama.

In this light, I think, such a passage as RS. I.101.3 has to be interpreted 95.

The stanza reads as follows:

yásya dyávāprthiví paúṃsyam mahád

yásya vraté Váruṇo yásya Súryah,

yásyé'ndrasya sīndhavah sáścati vratám

marútvantam sakhyáya havāmahe

"Indra, to whose great manly power Heaven and Earth (conform),

under whose command Varuṇa and Sūrya (stand), whose command the

rivers obey, him who is accompanied by the Maruts we invoke for alliance".

It is possible that this passage belongs in the context of the New Year

festival. According to the Taittirīya Saṃhitā Indra was born on New Year's

eve (ekāṣṭakā) 96 but his festival lasted only five to seven days, at least

in the account of Purāṇic texts, and the Indra tree, which symbolized

the cosmic centre and the world pillar, was at the end of the festival

removed and thrown into a river. Although nothing is known about a

Vedic Indra festival, it is quite likely that the Vedic contests of a

potlatch-like character took place about the beginning of a new year

(IIJ. 5, p. 169ff.) and stood under the patronage of Indra. However that

may be, only a special occasion can have induced a poet to describe

the relation between Indra and Varuṇa in such terms. In any case, the

passage quoted, isolated as it is, is insufficient to prove Varuṇa's

subordination to Indra 97.

A third and last objection that must be raised to Oldenberg's inter-

94 In one of his earliest publications, Varuna und Mitra (1877), p. 103.

95 See Bergaigne, Rel. véd. II, p. 188, III, pp. 142, 219.

96 ekāṣṭakā : Cf. AS. III.10.12, Paipp. I.106.4, KS. XXXIX.10 (127,7), TS. IV.3.11.3

ekāṣṭaká tāpasā tápyamānā jajānā gárbham mahimánam Indram, téna dasyūn ŕy

ásahanta devā́ hantá 'surāṇām abhavac chácibhiḥ, KS. XIII.3 (182,22) etasyā́m vā

Indro 'jāyata antā́ mahimánam Indram. téna dasyūn ... (rest of the text is similar)

Hillebrandt, Ved. Myth. I2, p. 30, Caland ad PB.V.9.1 and cf. IIJ. 4, p. 223.

97 See n. 74. It is not necessary in this connection to discuss such passages as VI.20.2

where Indra's asuryàm is mentioned, or VIII.15.9 (AS. XX.106.3) tvám Viṣṇur brhán

kṣáyo Mitró grṇā́ti Várunah. For the first passage (read asuryám, von Bradke,

Oldenberg, ZDMG. 55, p. 325f.= Kleine Schriften, p. 784f.) see von Bradke, op.c.,

pp. 32, 44, E. Hardy, Die Vedisch-brahmanische Periode der Religion des alten

Indiens (1893), p. 23 and Geldner's translation, for the latter, e.g., Hillebrandt,

Varuna und Mitra, p. 105.

pretation (p. 24) is that the mythological conception of figures going

over from the Asuras to the Devas has a strictly fixed pattern. Above

(p. 20) the epic tale of how Śrī left the Asuras and came to Indra was

quoted. There is, however, a closer parallel in the Veda, as von Bradke

(p. 101) has pointed out. In the Atharvaveda (ASS. XIX.56, ASP.III.8)

it is related that Svápna, Sleep, originally belonged to the world of the

Asuras 98 but that, when he went over to the Devas, he was offered an

ádhipatyam. The hymn begins with the words “Out of Yama's world

(Yamásya lokád) hast thou come hither”. Verse 3 may here be quoted

in full:

br̥hádgāvā́ 'surebhyó 'dhī devā́n úpāvartata mahimnám īcán

tásmai svā́pnāya dadhur ádhipatyam trayastrimśád asā́h svàr ānā́ndh

“He, . . . , turned from the Asuras to the gods, seeking greatness; for

this Sleep the thirty-three [gods] established an ádhipatya after they

had reached heaven”. In several archaic religions Sleep is thought of as

residing, together with Death, in the primeval waters or “outside the

finite world”, as de Buck put it. Every night man is sleeping there and

his awakening is a new birth 99. In India a well-known illustration of

this belief is Viṣṇu's sleep, during which he is resting, every year, in the

waters under the earth upon the cosmic serpent Śeṣa 100. In the Rigveda

the words “like sleeping in the womb of Destruction, resting in the

darkness like the sun” 101, which describe something (or some one?) that

is buried (nikhāta) may not be significant in this context. When, however,

in the cosmogonical myth the dragon is said to be fast asleep at the moment

Indra slays him 102, this is, isolated though this description is, certainly

more than a fancy of an individual poet. Sleep is, indeed, closely related

to the destructive forces of the nether world, as is clearly stated in

AS. XI.8(10).19

svápno vái tamríir nirr̥tīḥ pāpmáno nā́ma devátāḥ

jará khā́latyā́m pdlityā́m śárīram ánu prdviśan

“Sleep, weariness, destruction, the deities named pāpmán (evils) 103, old

age, baldness, hoariness entered the body”. The dragon belongs by nature

to this world of sleep and mṛ́radevas. Sleep is associated with the west

98 Cf. Sāyaṇa ad v.3: svayam Asurapakṣīyaḥ san.

99 See W. B. Kristensen, Symbool en Werkelijkheid, p. 19, A. de Buck, De Gods-

dienstige Opvatting van den Slaap, inzonderheid in het oude Egypte. Inaugurele

Oratie [= Mededelingen en Verhandelingen No. 4 van het Vooraziatisch-Egyptisch

Gezelschap “Ex Oriente Lux”], Leiden (Brill), 1939.

100 Cf. History of Religions 10, p. 125.

101 RS. I.117.5 suṣupā́mā́m ná nirr̥ter upásthe sāryā́m ná dasrā́ támasi kṣiyántam.

The mythological allusion in sūryā́m ná . . . támasi should in any case be noted

(cf. IIJ. 8, p. 111ff.). Cf. Manu 1.5 prasuptam iva.

102 RS. IV.19.3 . . . abudhyā́m ábudhyamānā́m suṣupāṇā́m Indra, . . . áhim vájreṇa

ví rṇā́ apartván.

103 For the connection between pāpmán and mr̥tyú see n. 30 and below, p. 72.

in TS. V.5.10.4. Cf. also Mhbh. V.108.4,7,8, where Night and Sleep (nidrā)

are said to be in Varuṇa's western quarter and the Daityas are described

as sleeping there.

5. THE ĀDITYAS II

The fundamental problem, accordingly, is that Varuṇa although

becoming a god, does not entirely lose his character of an Asura. He

and Mitra are deváú ásurā (see p. 7), gods who are at the same time

Asuras and are still addressed as such (VIII.27.20). As Devas, however,

they form a special group, the Ādityas, the principal of whom is Varuṇa.

Their name, "sons of Aditi", characterizes them as sprung from the

undifferentiated primeval world of which the World mother Aditi is one

of the most prominent personifications. She cannot, however, be discussed

here in detail. This raises the question about the status of the Ādityas

in the Vedic pantheon. Hillebrandt quite incidentally refers to this

problem when, in a foot-note, he observes "denn gerade die Bezeichnung

Asura's gibt dem Namen im RV. nichts Dämonisches und als die Asura's

zu Dämonen geworden waren, ist Varuṇa-kein Asura mehr" 104. Here he

seems to be fairly aware of the complexity of the problem but at the

same time he has recourse to an evolutionistic scheme to explain it away.

It may be asked, indeed, if anything is explained by stating that at a

later stage of the evolution the Asuras became demons but that Varuṇa

was not affected by this change. Is this anything more than formulating

the problem again in other words, leaving the basic difficulty as it was

before? 105 The hypothetical historical development, handled as a last

irreducible fact, remains unexplained, its main function being that of a

deus ex machina. As for RS. X.124, Hillebrandt's ritualistic interpretation

104 Ved. Myth. II2, p. 420 n. 1.

105 It should be noted that it is often hard to ascertain what modern scholars, when

discussing this matter, actually mean to say. Thus A. K. Coomaraswamy, The

Yaksas II, p. 27, refers to ŚB. XIII.8.2.1 (one of the many passages where the

Asuras are said to have been driven away from this world by the Devas, see above

n. 24) and then remarks: "Varuṇa, indeed, escapes this fate for he is accepted as a

Deva, and his asuric character is forgotten . . . ". The last words cannot be accepted

without some modification but the main question is: by whom was Varuṇa, in

Coomaraswamy's opinion, accepted as a Deva? Apparently he thinks that this was

a voluntary act of the believers, which, then, must have reflected a fundamental

change in their mythological conceptions and in the religious attitude of Vedic man.

If this interpretation of his words is correct, they contain an arbitrary statement

which lacks a solid foundation in facts. William Norman Brown, in his juvenile paper

"Proselyting the Asuras (A Note on Rig Veda 10.124)," JAOS. 39 (1919), pp. 100–123,

had probably the deepest insight into the problem, but even in his article there is

still a lack of clearness in that he introduces the notion "proselyting" so as to

account for the fundamental change. But see his last article in JIES. 2 (1974).

In this connection it should be noted that Sāyaṇa's gloss pūrvadevāḥ for Asurāḥ

(ad AS. VI.109.13) is mere guess-work: in the preceding hymn he paraphrased the

same word by Dānavāḥ (VI.108.3) and elsewhere he uses other glosses again (II.3.3,

XIX.66.1). As for the púrve devāḥ of RS. VII.21.7 and X.109.4 (cf. X.90.7), they

are no doubt the Sādhyas; see Geldner ad I.164.50 and the Excursus below.

32

of it as referring to the beginning of the new sacrificial year 106 curiously disregards the words bahvīḥ sā́mā́ḥ "many years" in verse 4. A strictly mythological explanation was only given by Bergaigne, who quite rightly recognized in this hymn a description of the defeat of the Asuras by the Devas 107.

It follows from the preceding discussion that the basic problem is neither one of a historical change in religious belief (the older Asuras having been replaced by the younger Devas) 108, nor one of a contrast between two different conceptions of the Divine (a unitarian versus a dualistic one) existing side by side 109. As was pointed out above (p. 11), Bergaigne is the only scholar who has recognized the basic difference between the Asuras and Devas, although he was not aware of the fact that we are not so much concerned with two contemporaneous conceptions of the Godhead as with two different stages in the cosmogonical process (see n. 107).

106 See Lieder des Ṛgveda (1913), p. 21: "Auch dieses Lied knüpft wie X, 51–53 an den Wechsel von Devayāna and Pitryāna und die Neubelebung des Opfers am Jahresanfang an . . . Mit dem Übergang von der opferlosen Zeit des Pitryāna zu der Devayāna assoziiert sich der neu auftauchende Gegensatz von Devas und Asuras."; Ved. Myth. II 2 (1929), p. 62: "Wiederbeginn der Opfer nach der langen Zeit des Tamas." Some other passages Hillebrandt interprets (as Geldner interprets this one) as referring to Agni's flight and the effort of the gods to induce him to return (Ved. Myth. III, pp. 62, 211). See above, n. 37. (Geldner's interpretation of arkā́ at X.157.5 is very doubtful!).

107 Bergaigne, Rel. véd. III, p. 145ff. (on X.124): “. . . indication, et cette fois tout à fait claire de la rivalité de Varuṇa et d'Indra, et de l'usurpation consommée par le second [cf. Vodskov, p. 210 n. 1!] . . . la révolution qui lui a fait perdre la puis- sance en même temps qu'à Soma and Varuṇa . . . Agni, Soma, Varuṇa sont renversés; la révolution est faite” [= paryávar ráṣṭrā́m ]; pp. 74, 76, 147 (defeat of the Asuras, victory of the Devas), p. 84 (the name deva characteristic of “un dieu de la conception dualiste comme Indra”) and p. 3 where Bergaigne rightly opposed the gods of the dualistic cosmos to those of the undifferentiated world, although he did not yet situate this contrast in a cosmogonical context. He rightly stressed the ambiguous character of the Asuras (“caractère équivoque”, pp. 69, 70, 72, 76, 84, 115, 148 etc.). Only the disintegration within the world of the Asuras brought about by the creation of the cosmos he has failed to recognize. It should be added, however, that even he had not entirely abandoned the theory of the “Verteufelung” of the Asuras as a historical process of the Vedic religion; cf. vol. III, p. 68f., p. 69 n. 2: “les rapports ordinaires d'Indra avec les Asuras, même avec les Asuras conservant encore le caractère de dieux.”

108 Thus, e.g., Kasten Rönnow, Trita Āptya I (Uppsala Universitets Årsskrift 1927), p. 10ff., Paul Horsch. Die vedische Gāthā- und Śloka-Literatur (1966), pp. 238, 240ff., especially p. 243 and passim.

109 See Bergaigne, Rel. véd. III, p. 3ff., p. 142 and cf. especially p. 4: “Mais ce mythe . . . s'il consacre l'antériorité mythique de ces mêmes dieux sur Indra, ne saurait passer pour une preuve de leur antériorité historique, je veux dire d'une antiquité plus haute de la conception religieuse à laquelle ils appartiennent” and p. 149: “. . . ces deux personnages [viz. Indra and Varuṇa] qui, d'après le vers VI,68,4, dépassent en grandeur tous les dieux, résument les deux conceptions de la divinité entre lesquelles semble osciller la conscience religieuse des Āryas védiques.” Instructive is the reaction of an outstanding contemporary scholar to this structural approach, which can be found in P. von Bradke's Preface to his Dyāus Asura (1885), p. XIIIf. It was written three years after the publication of vol. III of Bergaigne's work.

This interpretation, however, while solving an old problem of Vedic religion, at the same time reveals the real complexity of the Vedic conception of the world. In the mythological way in which the Vedic poets express themselves, Indra breaks the power of resistance (vṛtrahátya). As we have seen above, this had constantly to be repeated, probably at the beginning of every year. Not only Indra, accordingly, but also the powers of resistance (vṛtrá), impersonated by the serpent (Vṛtrá), must have been seasonal deities 110. If the strife between Gods and Asuras had been decided once for all by Indra's victory, the brāhmaṇas would not have referred to this primordial myth for every detail of the ritual. In fact, however, not only the ritual as a re-enactment of the mythical prototype, but any action in this world could be seen in the light of that myth because the fight between the two powers is never ended. The locus classicus for this idea is the commentary on TS. I.2.12.2. In the Vedic age, the conception of time as a cyclical process within the limits of a year must have led to the belief that after the crisis which marked the transition from the old year to the new one the Asuras were for the time being eliminated. It is significant that in the brāhmaṇas it is said, as a stereotyped expression of their defeat, that the Asuras were driven out from this world (see n. 24). There is a marked contrast, accordingly, with the Rigvedic dragon (the impersonation of the power of resistance of the realm of the Asuras), which is slain 111. The implication is that the organized cosmos does not represent the total reality. Along its borders this cosmos is threatened by the demoniacal powers which, although driven away from this world into the darkness of the nether world, constantly imperil its existence. In psychological terms this myth might express a repression of the dark aspects of existence into the unconscious mind (n. 24) with, as its natural consequence, the fear engendered by that repression.

This fear is clearly expressed in such a passage as Śat. Br. I.2.4.8–11 (in Eggeling's translation): “The gods and the Asuras, both of them sprung from Prajāpati, were contending for superiority. The gods vanquished the Asuras; and yet these afterwards harassed them again. The gods then said: ‘We do, no doubt, vanquish the Asuras, but nevertheless they afterwards again harass us. How then can we vanquish them so that afterwards again—’

110 Bergaigne, who rightly stated that there was a mythological relationship between Varuṇa and Vṛtra, was wrong in identifying them directly (Rel. véd. III, p. 147f.). Actually the dragon is killed, whereas Varuṇa is incorporated in the group of Devas. 111 Bergaigne, III, p. 115: “Supposera-t-on qu'il enferme les démons des ténèbres ou de la sécheresse? Mais ce serait là une conception peu conforme à l'esprit général de la mythologie védique. Les démons sont vaincus, frappés, foudroyés, déchiquetés par leurs adversaires: il n'est pas ordinairement question de les enfermer.” This is true but a distinction should be made between the Asuras and Vṛtra (whom Bergaigne reckons among the Asuras). In the Mahābhārata the Asuras, driven away into the nether world of Varuṇa, are actually fettered (see below). As for the commentary on TS. I.2.12.2 (p. 321), it runs as follows: tasmād ihaive 'dānim eva tadupāvartanāt prāg eva devān vijayāmaha iti vicintya vajram udyatya devān abhilakṣya prahartum āgatāḥ.

we need not fight them again?' Agni then said: 'By fleeing northwards they escape from us'. By fleeing northwards they had indeed escaped from them. Agni said: 'I will go round to the northern side, and you will then shut them in from here; and whilst shutting them in, we will put them down by these (three) worlds; and from what fourth world there is beyond these (three) they will not be able to rise again' 112. In these words we find an awareness of the existence of a world beyond the world of order, a world of disorder and formless Chaos, over which gods and men had no control. With this fourth world, however, the reflections of the Vedic theologians had reached a limit. What was beyond the three worlds was unimaginable and the author declares himself to be uncertain about the existence of this fourth world (I.2.1.12, ŚBK. II.1.4.10): "What fourth (world) there is or is not beyond these (three) worlds, by that indeed he thereby chases away the spiteful enemy. Uncertain, no doubt, is what fourth (world) there is or is not beyond these (three) worlds . . ." (Eggeling). When this fourth world is incidentally identified with Prajāpati (ŚB. XI.1.2.8), this is entirely on the lines of the numerical speculations in which the sum total plus one symbolizes the all-embracing totality, personified by Prajāpati (cf. also Śrautakośa II, p. 781).

The same idea is implied in the conception of gods of the totality of all that is. In them a belief in the ultimate oneness of Order and Chaos, of Good and Evil, is expressed. If, however, they are a guarantee for the final victory of Order over Chaos remains questionable. Although they sometimes help the Devas to vanquish their adversaries 113, their

112 (8) Devá ca vā ásurās ca, ubhā́ve prajā́patyau pasprdhi­re. té ha sma yád devā́ ásurān jayanti tā́to ha smaí’nā́n púnar upóttisṭhanti (9) té ha devā́ ūcuḥ, jayāmo vá ásurā̄ms, tā́tas te èrá nah púnar upóttisṭhanti. kathā́m nv ènā́n anapajayyayā́m jayemé 'ti (10) sá hā 'gnir uvā́ca, údai’nco vai nah palā́yya mucyante (11) sá hā 'gnir uvā́ca, ahā́m uttaratáḥ pāryeṣyā́my, átha yūyám itā́ upasā́miṅrotsyatha. tā́nt saṁrúdhyai 'bhís ca lokā́r abhī́nīdhāsyā́mo yád u cemā́ṅl lokā́n áti caturthā́m tatáḥ na sā́ṁhāsyanta iti.

113 E.g., PB. VIII.3.1 'The Gods and the Asuras contended for (the possession of) these worlds. The Gods resorted to Prajāpati; he gave them this sāman, (saying): 'By means of this sāman ye will be able to drive them away from these worlds" (translation by Caland): devā́s ca vā ásurās caiṣu lokeṣv aspardhanta. te devā́ḥ Prajā́patim upā́dhācaiḥ. tebhya etat sā́ma prā́yacchad : etenā́’nā́n ebhyao lokebhyo kā́layiṣyadhva[m] iti. tenai 'nā́n kālayiṣyadhva iti; Caland, ZDMG. 72 [1918], p. 20 and pāddham, Kashi Skt Ser.: kālayiṣyadhram; Caland, ZDMG. 72 [1918], p. 20 and

translation. p. 167 emended this to kālayīyadhrā iti with the support of the Leiden manuscript. H. C. Patyal: "On the Modal Forms of the Simple Future in the Veda", Poona 1969 (Public. Centre Advanced Study in Sanskrit, No. 32), p. 83f maintains the form in -am. The manuscripts will have to decide which reading is correct). Other instances are JB. III.104. where the Devas ask Prajāpati for help against the Asuras: Devāsurā aspardhanta. te devā́ḥ Prajā́patim upā́dhāvaraiḥ: tan naś chandah prayacchā 'yātayā́ma yenā́ 'surān abhíharā́me 'ti. sa tapo 'tapyata. sa tapas tap-tvā 'tman simṛīmayad ardpayat, and TS. II.5.2.2, where Prajāpati dips the vajra so as to enable Indra to slay Vṛtra (after Tvaṣṭṛ had created the Indraśatru to avenge Indra's killing his son Viśvarūpa): ká Prajā́patim upā́dhāravac : chā́trur me 'jani 'ti. táṣmai vájraḿ sikté́ práyachad : eténa jahí 'ti. Otherwise MS. III.4.3. See further n. 408. A later instance is Maitri Up VII.10, where Brahmā gives the correct ex-planation of the ātman only to the gods, but not to the Asuras.

own position is one above the contending parties. Hence it is that Western scholars have often misinterpreted their role as due to weakness of character. Instructive, both for the position of the High God and for modern misunderstanding of his character are the words in which Hopkins describes Brahmā's role in the epics: “He is a god of asceticism, he is father of gods and demons. Therefore, to win his favor, gods and demons practice asceticism, and because he is an impartial father he grants invulnerability, etc., to either god or demon indifferently. As the demons always take advantage of this weakness, Brahman is ever engaged in preserving the world from the result of his own folly. One cannot call it ignorance, for he is prescient. He is ‘equable to all’, that is his boast and glory . . . but he is also well-disposed, suhrd, toward both demons and gods . . ., as being equally his children” 114.

A striking illustration of this position of Brahmā is found in the Bhāratīya Nāṭyaśāstra. After Indra's victory, his killing of the Asuras and Dānavas was celebrated in the assembly of the gods by Indra's banner festival (1.55 tasmin dhvajamahe nihatāsuradānave). During this festival, before an audience which consisted of gods, Gandharvas, Yakṣas, Rākṣasas and Snakes (1.62), Bharata and his sons gave a dramatic performance which reproduced “how the Daityas had been vanquished by the gods” (1.57). Since, however, the theme of this performance was the extinction of Daityas and Dānavas (1.64 prayoge . . . daityadānava-nāśane), “all Daityas who were there assembled” were enraged at it and, at their instigation, the personified Obstacles (Vighna), together with the Asuras (1.66), obstructed the performance. Although Indra drives them away 115 with his vajra, they come back and Bharata, unable to bring the performance to an end, asks god Brahmā for a means of protection (rakṣāvidhi). Brahmā, who was also the inventor of the dramatic art

114 Epic Mythology, p. 195, where he refers to Mhbh. V.76.7 ed. Pāṇḍavānām Kurūnām ca bhavān paramakah suhrt, surānām asurānām ca yathā vīra Prajāpatih. Cf. p. 49 “Not unlike the relations of the Rākṣasas to the Great-Father (god) is that of the Asuras, who also are continually receiving boons from Brahman” (with examples), p. 50 “Brahman gives the Asuras all boons except immortality, for immortality is withheld from them . . . The Asuras have no father-god except Brahman, for Brahman remains also their “father” and gives them not only boons but also good advice . . .”.

115 1.71 gateṣu teṣu vighneṣu sarveṣu saha Dānavaih ed. Ghosh, in accordance with the MS. na of the second edition of the Baroda text. The other editions read nihatēṣu ca sarveṣu vighneṣu (niḥteṣu in the ed. Raghuvaṃśa). M. M. Ghosh, The Nāṭyaśāstra, The Original Sanskrit Text, vol. I2, Introduction, p. xxxiv, gives the very arguments in favour of this reading which may at one time have induced a copyist to substitute gateṣu as an emendation for nihatēṣu. In view, however, of verse 74 drṣṭvaiva jarjaram te 'pi gamiṣyanti evam eva tu, Viṣṇudharmottara II.154.112 drṣṭamātreṇa yene 'ha vidravisyanti, Hemādri, Caturvargacintāmaṇi II,2, p. 401 na tathā 'pi kṣayam gatāḥ (cf. p. 402 tam drṣṭvā sarbalam sainyam bhagnam sa ca nipātitah) there is a possibility that gateṣu is the authentic reading. For the rest, it is possible, judging by the internal evidence, that the passages vv. 58-63 and 67-76a are interpolations but these questions of higher criticism are here left out of consideration. In any case the lines form part of the text as it has traditionally been handed down.

(nāṭyaveda) and who had instructed Bharata, now conceives the idea of creating a theatre, which Viśvakarman builds at his command. The cosmic symbolism of this building will be discussed below. Here it may be sufficient to state that it represents the world of order, protected against the demoniacal forces which threaten it from without (the “fourth world” of the Śatapatha Brāhmaṇa). As the demons are offended at this device which is directed against them, Brahmā tries to conciliate them by kind words (1.100 ff.): “Having heard these words of the gods Brahmā said to the Vighnas: ‘Why have ye risen to destroy the performance?’ Having heard this word of Brahmā Virūpākṣa, along with the Daityas (and) the hosts of Vighnas, spoke the following conciliatory words: ‘By creating this Nāṭyaveda, in accordance with the wish of the gods, Thy Holiness has rejected us. It is for the benefit of the gods. Thou shouldst not do this, O Grandfather of the World; no less than the Devas, the Daityas have sprung from Thee in the beginning (v.1.: all)’. When Brahmā heard these words of Virūpākṣa, he said: ‘Don’t ye be angry or dejected, O faultless ones! I have created the Nāṭyaveda, which shows the alternatives of good and evil 116 of you and the gods and which takes into account the connection 117 between deeds (karman) and mental states. It does not onesidedly indicate the feelings of you or the gods but it is a representation of the state of the whole of the three worlds’ ” 118. It should be noted that there is some similarity between the theatre as a protected area and the vāstu, which is the sacred ground plan for temples and palaces. V. S. Agrawala defines it as “the emergence of a divine Kshetra in the midst of the Āsuric universal. The forces of all the Devas are concentrated on the limited sanctified area of the Vāstu”, the latter being “a complete specimen or replica of the cosmos in which light and darkness exist together” 119.

As was stated above, the epic evidence, which will be discussed below in detail, shows that Varuṇa, although incorporated in the cosmos, still maintained secret relations with that which is beyond the bounds of it. In this connection it may be asked whether the many translators of

116 Abhinavagupta (Baroda ed. I 2, p. 34 line 16f.) has the following comment: subham aśubham ca dharmādharmarūpam sukhaduhkhalphalāvena vibhedana kalpāyati, adhyavasāyayati nāṭyavedah.

117 Abhinavagupta, p. 34 line 19 takes anvaya in the sense of descendance: anvayo 'bhijanaḥ āryāvartādi-brāhmaṇyā . . . śreti.

118 devānāṁ vacanaṁ śrutvā Brahmā Vighnān uvāca ha, kasmād bhavanto nāṭyasya vināśāya samutthitāḥ (102) Brahmaṇo vacanaṁ śrutvā Virūpākṣo 'bravid vacah, Daityair Vighnaganaiḥ sārddhaṁ sāmapūrvam idaṁ tataḥ (v.l. vacah) (103) yo 'yaṁ bhagavatā srṣṭo nāṭyavedah śurecchayā, pratyādeśo 'yam asmākaṁ surārthaṁ bhavatā kṛtaḥ (104) tan naitad evaṁ kartavyam் tvayā lokapitāmaha, yathā devās tathā Daityās tvattaḥ pūrvavinirgataḥ (105) Virūpākṣavacaḥ śrutvā Brahmā vacanam abravīt, alam vo manyunā Daityā viṣādām் tyajatāṁ 'naghāḥ (106) bhavatām devatānāṁ ca śubhāśubhavikalpakaḥ, karmabhāvānvayāpekṣī nāṭyedo mayā kṛtaḥ, (107) naikāntato 'tra bhavatām devānāṁ cā 'nubhāvanam (-tra bhāvanam), trailokyasyā 'sya sarvasya nāṭyavān bhāvānukīrtanam.

119 Vasudeva S. Agrawala, Matsya Purāṇa – A Study (Ramnagar, Varanasi, 1963), p. 345.

Atharvaveda Saṁhitā IV.16 120 have grasped the meaning of verse 5a sárvam̀ tád rájā Váruno ví caṣṭe yád antará róda­sī yát párāṣṭāt “King Varuṇa perceives all that is between heaven and earth and what is beyond”.

It seems that no one has ever wondered what the poet may have meant with this “bey­ond”. In the cosmogonical myth of the Rigveda (X.124) we are told that only some of the Asuras went over to the party of the gods but that the other Asuras, having lost their magic power (māyā́), renounced their sovereignty (paryá­vard rāṣṭrám).

The brāhmaṇas relate that the Asuras (that is, the ásurā adevā́h of the Rigveda) were driven away from “these worlds”. We are probably not wide of the mark when we say that by Indra’s demiurgic act ásat was converted into sát (RS. VII.24.5, X 72.2) but that part of ásat continued to exist, on the fringe of the organized cosmos, as an unformed (rather than non-)being 121.

Only the all-embracing transcendental world of the highest gods, “in the highest heaven, in the lap of Aditi” (X.5.7), encompassed both ásat and sát. It has mostly been insufficiently realized, owing to the exalted wording of such hymns as AS. IV.16, that Varuṇa is not a high god in the strict sense of the word.

The difference between Varuṇa and Prajāpati, who is “the limited and the unlimited” (ŚB. VII.2.4.30) and who encompasses all, like the year, is obvious. Similarly that between Varuṇa and Viṣṇu, who with his three strides covers (and creates) nether world, upper world and the transcendental third world of the cosmic totality122.

As an Āditya, Varuṇa belongs to the Devas, who are “on this side of Creation” (X.129.6). What is “beyond these (three) worlds” 123, is the fourth world to which the Śatapatha Brāhmaṇa refers.

Still, the Atharvavedic stanza quoted above cannot mean anything else but that Varuṇa had some connections with the world of ásat. If this conclusion is correct, the poet here openly declares what is mostly only indicated in an indirect way (see below, p. 68).

It will also be clear that the way in which out of the proto-Indo-Iranian *Asura the pre-Zoroastrian high god Ahura Mazdā has developed in Iran thus poses a special problem.

Some passages might easily lead to faulty conclusions with regard to Varuṇa. Thus RS. X.132.4ab “The other one, yonder Heaven, O Asura,

120 See the references in Bloomfield, SBE. 42, p. 390, Whitney-Lanman, Atharva-Veda, Translation and Notes (HOS vol. 7), p. 176 and H. Lüders, Varuṇa, p. 30f.

121 Only with regard to this “unorganized rest” of the universe can ásat be said to be “destructive” (Gonda, Religionen Indiens I, p. 181). For ásat = “undifferentiated, formless” see particularly H. Oertel, New Ind. Ant. I (1938), pp. 317–321 and the latest discussion by R. Ambrosini, Studia classica et Orientalia Antonino Pagliaro oblata I, p. 96ff. Cf. also R. N. Dandekar, India Maior, p. 93. For MS. III.7.1 (75,1) áklptam̀ vá idám ásit, etc. see n. 36.

122 This notion, it is true, does not seem to have been formulated expressis verbis before the classical period but it is so intimately interwoven with the essence of Viṣṇu’s character and with his three steps that it cannot well be due to a later elaboration. See Hist. of Rel. 15, p. 117f.

123 imá̀n lokā́n áti. Cf. RS. X.125.8 pará divá pará enā́ prthivyā́ pará devébhir ásurair yád ásti and see n. 112.

was given birth to: thou, O Varuṇa, art king over all (beings)" 124 should not be interpreted as meaning that Varuṇa is a god of totality. The poet who in a late period composed this stanza has here made use of II.27.10ab "Thou art king over all, (namely) over the Devas, O Asura, and the mortal beings"125, which restricts the range of Varuṇa's power to "this side of Creation" (X.129.6). Even more restricted it is in VII.83.5, where Varuṇa and Indra are addressed with the words "For ye two have power over the twofold wealth" (i.e., that of the upper world, which is Indra's domain, and that of the nether world, which is Varuṇa's) 126. In the former of these two stanzas Varuṇa is contrasted with (the Asura) Heaven 127, who "was given birth to" (sūyata). In view of Peter von Bradke's theory of Dyaus pitṛ́n it should be observed that, while Varuṇa was present in the primordial world as the eldest son of the World mother Aditi, heaven existed only as a potentiality. The latter's "birth", therefore, can only be due to Indra's vrtrahátya. This situation is clearly expressed in TS. II.5.2.5 "Heaven and earth said 'Strike not, for he [Vṛtra] is lying upon us' ". The cosmic duality of heaven and earth was only a potentiality in the undifferentiated unity of the primordial world, the unfolding of their dual character being prevented by the repressive power of Vṛtra.

The description found in other brāhmaṇas 128 (and also well-known from Egypt) to the effect that heaven originally lay upon the earth is only a different manner of expressing the same idea of undifferentiated unity, to which Indra put an end by "propping up" the sky 129. What the poet may have meant is that, in contrast with Dyaus, who owed his existence to Indra's intervention, Varuṇa was a god in his own right.

Bergaigne has seen the character of Varuṇa much more sharply than those scholars of the following generation who exalted the lofty character of the high ethical god. Right at the beginning of his chapter on "les dieux souverains et les dieux pères" he states "qu'à l'idée de la paternalité est souvent étroitement unie, dans les personnages en question, celle de la malveillance" (III, p. 2) and on p. 84 he gives the following striking characteristic of the Asuras: "Les Asuras devaient être des dieux auxquels les croyances des ṛishis n'opposaient pas de démons; dieux puissants pour le mal comme pour le bien; tantôt manifestés sous la forme brillante que semble impliquer le mot deva . . . tantôt au contraire cachés et cachant eux-mêmes la lumière et les eaux, tous les trésors que l'homme attend du ciel; enfin, en tant qu'opposés à Indra, assimilables dans une plus ou moins forte mesure aux démons combattus et vaincus par ce dieu.

124 asâv anyó asura sūyata dyaús tvám viśveṣām Varuṇā 'si rājā. Cf. Renou, EVP. 5, p. 89 "Cet autre là-haut, le Ciel, ô Asura, fut mis au monde; (mais) toi, ô Varuṇa, tu es le roi de tous (les êtres)," and see above, p. 25.

125 tvám viśveṣām Varuṇā 'si rājā yé ca devā asura yé ca mārtāḥ.

126 See above, p. 27. Similarly VII.82.4 īśānā vásva ubháyasya kārāva Indrāvaruṇā suhávā havāmahe.

127 Although P. von Bradke's translation (p. 71) is incorrect, there can be no doubt that Dyaus was also an Asura.

128 See n. 50.

129 See History of Religions 10, p. 104f.

Les Asuras en général, comme Tvashṭri en particulier, paraissent avoir eu dès l'origine un caractère équivoque, susceptible de s'accuser dans le sens démoniaque, comme de devenir l'expression la plus complète et la plus auguste de l'idée de la divinité". From a different starting-point and working along different lines we have arrived at results which are in some respects remarkably similar to those of Bergaigne. It is in this context that the passage of the early-classical Nāṭyaśāstra has been quoted to illustrate the complex nature of the relation between the Asuras and the Devas of the pantheon. The archaic character of the first introductory chapter of the Nāṭyaśāstra is also apparent from the fact that Indra is here still represented as the central figure among the Devas.

The later development in religious thought may either be described in terms of a shift from the worship of a seasonal god to that of the gods of totality (Prajāpati-Brahmā, Viṣṇu, probably also Śiva), which may then be explained from the seasonal rites getting into disuse, or, in a more theological way, as a shift from the exponent of the formed world that is threatened by the powers of the unformed (ásat) to the all-embracing gods who transcend this dualism and the cosmic contest (as is illustrated by the characteristic rôle of Brahmā and Viṣṇu in the epic tale of the Churning of the Ocean (see p. 104ff., Hist. of Religions 15, p. 119).

The earliest trace of such a theological re-evaluation of the idea of Indra's contest occurs in ŚB. XI.1.6.9, where it is said in a gāthā: "Never didst thou fight, nor hast thou, O Maghavan, any enemy. What is told of thy fights is only an illusion which thou hast created. Neither today nor formerly hast thou fought against an enemy". I do not think that it is right to describe this development as a tendency towards monism 130. Nor would it seem correct to characterize Indian religious thought in general as tending towards a one-sided emphasis on the non-formed. Although there has certainly been a shift of emphasis from the formed to the non-formed, still, in the older stage of Indian religion,

130 (9) tásmād āhur naistād asti yád daivāsurám yád idám anvākhyāne tvad udyādata itihāse vat tāto hy èvá tán Prajāpatih pāpmánā 'vidhyat té tā evá parābhavann iti (10) tásmād etád ṛṣiná 'bhyāuktam, ná tvám yuyutse katamó caná 'har ná te 'mitró maghavan kás caná 'sti, máyēt sā te yámi yuddhány āhúr nédyá sātrum ná nú purā yugutsá iti. For the interpretation see Minard, Trois Enigmes II, §§885ff., 872f., (relation to RS. X.54.2, for which see Bergaigne III, p. 83, Gonda, Four Studies in the Language of the Veda, p. 135f.). From the remarks made above it will be clear why I cannot accept Horsch's explanation in Die vedische Gāthā- und Śloka-Literatur (1966), p. 242: "Dieses merkwürdige Relikt einer verlorenen, mytholo-gischen Tradition besiegelt zugleich die Alleinherrschaft des deva Indra, der bald vom Lichtglanz der neuen Götter des Hinduismus, die ihre Macht grossenteils den niedern, z.T. vorarischen Volksschichten verdanken, verdunkelt wird." From a different point of view Heesterman has drawn attention to the elimination of the element of contest from the ritual ; see "Brahmin, Ritual and Renouncer," WZKSO. 8 (1964), p. 1ff.

man looked upon the phenomenal world as the result of a periodical strife between the powers of Chaos and Cosmos. In spite of characteristic differences there are, I think, significant similarities between Vedic and Old Germanic religious thought 131.

6. THE ĀDITYAS II: ĀDITYA AND ASURA

On the one hand, accordingly, Varuṇa is in the Ṛigveda still called an Asura, while, on the other hand, he is, as the principal of the Ādityas, a Deva. The single passage where the Ādityas are called Asuras 132 shows, it is true, that their character as devā asurāḥ was still known to the poets. Everywhere else, however, the Ādityas are simply called devāḥ, e.g., II.27.4. See p. 11.

The problem as to why the name Asura has fallen into disuse in the post-Ṛigvedic period 133 as an epithet of Varuṇa is certainly important, but this cannot be dissociated from another one, equally important, viz., Why is it that the Ṛigvedic poets still continue to call Varuṇa an Asura, whereas he has become a Deva, that is, "worthy of sacrifice" ? 134 In other words, What is the relation between the two terms Asura and Āditya and what were the respective functions of these two categories?

As for the first problem, this seems purely terminological. No indication can be found in our texts for a material change in Varuṇa's status and function during the whole Vedic period. The disappearance of Asura as a title of Varuṇa must, therefore, be due to causes which lay outside the god's personality. As far as I can see, the last trace of this use of Asura occurs in the Atharvaveda. Cf. I.10.1 "He rules, as an Asura, over the Devas" (n. 74) and, for the Asuras in general, VI.108.3 "The wisdom (medhā) that the Ṛbhus know, the wisdom that the Asuras know, the salutary wisdom that the Ṛṣis know".

Any attempt to answer the second question can at present only be tentative. It has been pointed out above (pp. 11, 23) that if the theory of an annual reiteration of Indra's Vṛtra-slaying and his fight against the Asuras as gods of the primeval world is correct, it then follows, first,

131 See G. H. van Senden, Het Werelddrama 2 (Amsterdam 1965), p. 74ff.: "Het Vormbesef in verschillende Cultuurgebieden", whose characterization of various cultures in terms of their position towards "formed" versus "non-formed" is important for a deeper insight into the background of Vedic religious thought (which he does not discuss).

132 VIII.27.20 (to Aryamā Mitró Váruṇaḥ, see v. 17) yád vā 'bhipitvó asurā ṛtám yaté chardír yemá ví dāśúṣe. See von Bradke, p. 75, Bergaigne III, p. 87. On the asuryà of the Ādityas (only II.27.4 dīrghádhiyo rákṣamāṇā asuryàm) see von Bradke, p. 33f., Bergaigne I.c.

133 Cf., e.g., Paul Horsch, Die vedische Gāthā- und Śloka-Literatur, p. 236. On p. 238 n. 2 there is a reference to Hillebrandt, Vedische Mythologie III 1, p. 439, who concluded from the rare occurrence of the name Asura in the family books of the Ṛigveda that the "Asura-cult" was already waning.

134 Cf. Arbman, Rudra, p. 152, who cites instances of the use of the word deva for demons in later literature.

41

that Indra was a seasonal god 135 and, secondly, that there must have been

a belief which held that in the critical period before the beginning of

the New Year the world had returned to the state of the primordial

Chaos, a belief well-known from other archaic religions. In other words,

in this period Devas and Asuras were again opposed, as they had been

in the beginning, and Indra had to repeat his contest for supremacy

with Varuṇa, who at this moment was again the protagonist of the world

of undifferentiated unity (see p. 24). To the Rigvedic poets this contest

was not a mere legend from a dim past but a reality of their actual world.

In RS. IV.42, which contains a bragging contest between the two gods

(see p. 22), Indra says (among other things) that he institutes the race

(ājí), which was a form of a socio-religious contest 136. If it should be

true that the hymn had been composed for a special ritual purpose, this

would only prove the more forcibly that the cosmogonical event, as the

paradigm for actual social life, was conceived as being reiterated continually

and (probably) periodically. According to J. C. Heesterman, The Ancient

Indian Royal Consecration (1957), p. 222, the rājasūya must originally

have been one of the yearly festivals which “cluster round the turning

135 For the uncertainty about Indra's origin (who apparently comes “from nowhere”)

see Minard, Trois Énigmes II § 786a. Cf. AS. III.10.12 (etc.), where Indra is said

to have been generated by the New Year's eve and to have slain the Dasyus or

Asuras (see n. 96), and further RS. I.132.2ab svarjeṣé bháre . . . áhann Indro yáthā

vidé śiṛṣṇá-śīrṣṇo 'pavácyah “Im Kampf, wenn es gilt das Sonnenlicht zu gewinnen . . .,

an dem Tage ist Indra, wie bekannt, Haupt für Haupt anzurufen” (Geldner, “en

(ce) jour EYP. 17, p. 46, instead of “erschlug”, “tötete”, Grassmann,

Ludwig, Pischel, Ved. Studien 1, p. 68). It remains to be seen to what extent the

seasonal character of Indra affords an explanation for the well-known passage II.12.5

yám smā prchánti kúha séti ghorám utém āhur naíṣó asti ‘ty enam “He about whom

(people) ask ‘Where is he?’ and about whom they say ‘This one is not (present)’.”

The Dutch equivalent would be “Hij is er niet” (German: “Er ist nicht da”). For

the continuation of the relative clause see Caland, Zur Syntax der Pronomina im

Avesta (Amsterdam, 1891), p. 46, Chantraine, Grammaire homérique II 3, p. 243.

Modern commentators who refer to the later nāstikas disregard the fact that in

parallel passages the poets do not question Indra's existence, their main concern

being who has seen him, or his act (V.30.1 kvà syá vīráḥ kó apaśyad Índram, VI.27.3

Índra nákir dadṛśa indriyám utém dadarśa), and where he is (II.12.5 kúha

sáti, VIII.64.7 kvà syá vṛṣabhó yūvā and V.30.1 quoted above). Hence the answer

ihá in VIII.100.4 ayám asmi jaritáh páśya mehá viśvā jātány abhy àśmi mahná “Here

I am, O singer, see me here, I surpass by my greatness all races”, which is a reaction

to 3 “There is no Indra”, some one says, ‘Who has seen him, whom shall we

worship?’ ” People do not know (that is, have not yet experienced) the maghavattvá

of the maghávan, because no one has yet seen his Indric force (VI.27.3), but his

existence is obviously not questioned. What matters is the god's epiphany, which

becomes manifest by his divine acts. In II.12 the poet refutes the assertion of

“some ones” by merely pointing to the god's accomplishing again his exploits.

This method of rebuttal would not seem favourable to the theory that “There is

no Indra” is the counterthesis in a verbal contest (see J. C. Heesterman, WZKSO.

XII-XIII [1968], p. 180). For the different problem posed by X.54.2 māyét sā te

yádni yuddhány āhúḥ “c'est (un tour de) ta magie, les combats qu'on dit (que tu as)

livrés” (Minard, Trois Énigmes II, p. 286), see p. 40 and n. 130.

136 See v. 5 krṇómy ājím maghávā 'hám Indrah (kṛ, not yā !).

of the year and are all in some way or other concerned with the regeneration

of the year, which is conceived of both as an actual time unit and as a

cosmological entity”. A striking confirmation of this conclusion is furnished

by the fact that an annually repeated abhiṣeka (samvatsarābhiṣeka),

usually performed about the time of the first (pūrvābhiṣeka, prathamā-

bhiṣeka), still exists. See M. R. Pant, J. Nep. Res. Centre I (1977), p. 96.

Therefore, the idea of an annual reiteration of the cosmogonial strife

seems well-founded, and there is no contradiction between Lommel’s

interpretation of the ritual use made of RS. IV.42 and the cosmogonial

explanation as a vivāc. No poet would have thought of composing a

hymn in which the primordial strife for supremacy is so dramatically

depicted if that contest had once for all been decided in the mythical

beginning. If, however, the contest was repeated every year, nothing is

more natural than that it took place during a renewal of the king’s status.

On the other hand, the hymns show beyond doubt that, in terms of

mythology, Indra had regularly to repeat his exploit of the vṛtrahátya.

It may, therefore, be considered plausible that also his contest with

Varuṇa, which resulted in the latter’s adoption into the ordered cosmos,

was repeated. Dandekar, India Maior, p. 93, argues that Vedic texts

do not speak of repeated creations and refers to RV. VI.48.22 “Only

once was the heaven born, once the earth”. It is certainly true, that

the texts do not say that heaven and earth were “born anew” like Agni

(VIII.43.9 gárbhe sāñ jāyase púnah) but they do not contradict the

conclusion of an annual renewal. What the poet stresses at the end of

VI. 48 is that Pṛśni has only once given birth (to the Maruts) and that

no one else was born after (tad anyó nánu jāyate) and in confirmation

of his statement he refers to heaven and earth. It can, indeed, be stated

that after creation no other heaven or earth has arisen, but this does not

contradict the obvious fact that the cosmogony was periodically reiterated.

The question then arises if Varuṇa bears the title Asura in the Rigveda

because this collection of hymns was in its essence (that is, in the family

books) designed for the celebration of the New Year period of contests

(see IIJ.4, p. 217ff.). This cannot definitely be proved, but since there

are in the Rigveda traces of a competitiveness between Indra and

Varuṇa, this must have recurred in the period of crisis at the end of

every year, just before Indra’s victory was celebrated by the erection

of his banner (see p. 129ff.). If at that time Varuṇa was again a true

Asura of the primeval world, his “losing” that epithet in later times

may possibly have been connected with the circumstance that the ancient

seasonal ritual gradually lost its importance. Further research is needed

to decide whether this suggestion is correct.

In close connection with the preceding remarks the often discussed

contrast between Varuṇa and Indra, as elaborated in the hymns VII.82–85,

may be briefly summarized 137. The most significant detail is that Varuṇa

137 See above, p. 25 and cf. Muir, Original Sanskrit Texts V (1874), p. 122ff.

Hillebrandt, Varuṇa und Mitra, p. 102f., Renou, EVP. 7, p. 6, Bergaigne III, p. 118

(on II.27.14, where he says about Indra “son nom . . . ne figure que dans la formule

only very rarely is said to join Indra in the latter's fight. Only once are they jointly addressed as vṛtratūrā (dual), "striking down obstacles" 138. Varuna's epithet samráj, as against Indra's title svaráj (VII.82.2), has often been commented upon 139. In 82.4 and 83.5 the two gods are said to have power over goods of "both sorts" (vásva ubháyasya). The word ubháya can refer to the two moieties, as in the stereotyped phrase, used with reference to Asuras and Devas, ubháye Prajāpatyàḥ "both of them sons of Prajāpati" (n. 23). The phrase has, therefore, been interpreted above as saying that since Indra has power over the goods of the upper world, and Varuna over those of the nether world, they conjointly have at their command all vásu of the world 140. In 82.5 Mitra honours Varuna with peace (kséma) 141, while Indra drives about as a mighty one (details not clear). While 82.6 is insufficiently clear 142, 83.9 says that the one slays the vṛtráni and the other watches over the vratáni (vows). A characteristic antithesis also occurs in 84.2: "May Varuna's wrath avoid us, may Indra give us space". The greatest blessing that is expected from Varuna is negative, his not striking men with his wrath. On the other hand there are indications in the Rigveda which justify the assumption that, apart from the cosmic and social aspects of Indra's vṛtrahátya, its religious importance was individually experienced as a break-through of áṃhas, a widening of consciousness 143. In 84.4 we are told that the Āditya gets round (?) the áṃrtā (plural), whereas the hero distributes goods (vásūni) 144.

de demande, et non dans la formule de déprécation"), and more in general III, p. 140ff., where also VI.68 (cf. Geldner, Vedische Studien 1, p. 142) is discussed. See further Dandekar, Cama Oriental Institute Golden Jubilee Volume (1969), pp. 237–248. For the hymns VII.82–85 see Renou, EVP. 5, pp. 99–102 (translation) and 7, pp. 81–87 (commentary). In the Atharvaveda Indra and Varuna are not (no longer?) mentioned conjointly, see Renou, Festgabe Lommei, p. 122f.

138 See note 69.

139 Dandekar, Cama Oriental Institute Golden Jubilee Volume (1969), p. 242f., holds that Varuna as a samráj is superior to both Indra and Mitra. From the analysis of X.124.5 (above, p. 24f.) it follows that the relation between Indra and Varuna cannot simply be defined in terms of superiority.

140 Sāyaṇa (ad VII.82.4): divyasya pärthivasya ca vasuno, dhanasya (cf. VII.83.5, VI.19.10). Against him Geldner: "Eher feindliches und eigenes". Renou: "Vous qui disposez d'un double bien-matériel", resp. "Puisque vous régissez l'un et l'autre biens (céleste et terrestre)". Cf. n. 126 and n. 23.

141 See Renou, EVP. 7, p. 82 Cf., however, dhruvákṣema as an epithet of Mitra and Varuna (IIJ. 8, p. 110 n. 68).

142 yáṃim anváh śnáthāyantam átirad dabhrébhir anyáh prá vrnoti bhúyasah. Cf. Bergaigne III, p. 113, Renou, EVP. 5, p. 99 "recouvre", 7, p. 82f. "dieu bloqueur".

143 For the nature of Varuna's blessing see Bergaigne III, p. 142 and cf. IIJ. 5, p. 52. Indra's vṛtrahátya, in so far as it has a bearing on the individual poet as the breaking of an inner obstruction and the opening of the well of inspiration, was discussed in IIJ. 4, p. 248. For a psychological evaluation of the experience of space see M. Lietaert Peerbolte, Psychic Energy (Wassenaar, 1975), p. 145f. The notion of áṃhas, discussed by J. Gonda, IIJ. 1, p. 33ff., might be considered in the light of Peerbolte's views.

144 Cf. VII.65.3 (Mitra and Varuṇa) tá bhû́ripā́sāv áṃrtasya sétū. Cf. Bergaigne, Rel. véd. III, p. 142, Renou, EVP. 7, p. 86.

In 85.3 the translation is not certain: "The one controls (? dhārayati) the chosen (?) people, the other breaks through the barriers (vrtráni) which do not offer resistance"145. Somewhat different is VI.68.3 "The one forcefully slays Vṛtra, the other accompanies (Indra?) as an inspired speaker"146, which leaves the question open as to who is accompanied by Varuṇa. The god who is ordinarily expected to inspire the speaker is Agni147, not the remote Varuṇa. This group of hymns is interesting because the Vasiṣṭhas, traditional Varuṇa-worshippers, have here tried to do full justice to the specific character of each of the two gods.

It is signficant that the Rigvedic religious thinkers who thus defined the two gods in terms of a contrast were aware of the systematic character of their theology. The first Western scholar to follow them in this respect was Bergaigne. Almost a century ago he described Vedic mythology "comme un systéme achevé de toutes piéces" (Religion védique I [1878], p. III). He recognized the necessity of a "systématisation" (p. II) of the data and thus became the first exponent of a structural approach avant la lettre. In contrast with Geldner's evolutionistic theories, which were in conformity with the spirit of the age, Bergaigne interpreted the difference between Devas and Asuras as "une opposition née de la rencontre des deux conceptions unitairé et dualiste, sur l'ancienneté relative desquelles je n'ai d'ailleurs pas cru pouvoir me prononcer" (III, p. 142).

It is natural that some of Bergaigne's conclusions are no longer acceptable. This is especially true of the relation between "unitaire" (which he considered characteristic of the Asuras and Ādityas) and "dualiste" (which he rightly recognized in Indra). Such remarks as that about "les deux conceptions de la divinité entre lesquelles semble osciller la conscience religieuse des Āryas védiques" (III, p. 159) show that he regarded them as two coexisting conceptions. In fact, however, they characterized two successive stages of the cosmic process, viz. the undifferentiated primordial world and the dualistic cosmos (if it is permitted, for a moment, to reduce the complex state of things to an over-simplified formula). The antithesis as he put it is, therefore, only correct for brief (presumably annual) period during which the original Asuras (who, in a way, were "jenseits des Guten und Bösen") and the Devas were again fighting for the supremacy over this world. As for the "dieux pères", however, the problem must be formulated in a different way. Bergaigne is right in describing Varuṇa and the "sovereign gods"

145 krṣṭír anyó dhāráyati práviktā vrtráni anyó apratíni hanti. For právikta different interpretations have been proposed: "auserwählte" (Geldner, Renou, EVP. 7, p. 87), "(aux résidences) séparées" (Renou 5, p. 102), "erschreckt" (Hillebrandt, Lieder; similarly Oldenberg), "fortgesprungene (Grenze)" (Thieme). Cf. Eggeling, SBE. 44, p. xxii, Dandekar, Cama Or. Inst. Golden Jubilee Vol. (1969), p. 248, n. 26, P. Thieme, Kl. Schr. p. 254 n. 2.

146 vájrenā 'nyáh sávasā hánti vrtrám siṣakty anyó vrjáneṣu vípral. See Renou, EVP. 7, p. 80 (also on vrjáṇeṣu).

147 See IIJ. 4, p. 248ff.

as comprising within themselves the opposition of divine and demoniacal

forces but, owing to his excluding the theological speculations of the

brāhmaṇas from his study (vol. I, p. IV), he has overlooked the fact

that, even in the Rigveda, the Ādityas did not stand apart in isolation

but were incorporated in the system of the dualistic cosmos. In this world

of order the original antithesis of Devas versus Asuras recurs in that

of upper world versus underworld, the latter being impersonated by the

Ādityas, the “Asuras among the gods” (devānām ásurāḥ). In a way,

therefore, the old antithesis of the formed versus the non-formed seems

to reappear in the world of the formed. As for the Ādityas, this adds

a new complication to their complex character, already recognized a

century ago by Bergaigne.

7. THE ĀDITYAS III: THE RIGVEDA.

For many details concerning the Ādityas as a group it is sufficient

to refer to the handbooks 148. For the members of their group the texts

give different numbers 149 but, in any case, they have always been

considered a closed group, in contrast with the limitless Asuras 150. When

one Āditya is referred to, this is always Varuṇa. Two Ādityas refer to

Varuṇa and Mitra, three to Varuṇa, Mitra and Aryaman.

In none of the handbooks, however, the place of the Ādityas in the

pantheon is given due consideration 151. From the oldest texts downwards

the pantheon is considered to consist of three groups of gods, the Vasus,

Rudras and Ādityas. There are a few other groups, which, however,

are not included in this fundamental tripartite system. Many gods, such

as Indra, Agni, Varuṇa, belong to one of the three groups. There are,

besides, a few individual gods who, in some way or other, stand for the

totality of the world and, accordingly, transcend this system. To them

belongs Prajāpati and originally, it seems, Viṣṇu. It is significant that,

when they are incorporated in the tripartite system, they are assigned

to the group of Ādityas. Cf., e.g., Viṣṇu. The same, however, is true of

Indra who in the brāhmaṇas is often counted among the Ādityas, although

148 See, e.g., Bergaigne III, p. 98ff., Macdonell, Vedic Mythology, p. 130, who

disposes of the Ādityas in half a page (under the heading “Groups of Gods”), in

contrast with, e.g. the Maruts (pp. 77–81).

149 Seven according to Bergaigne III, p. 98, but see Hillebrandt, Ved. Myth. II 2,

p. 87. For priestly speculations cf., e.g., MS. II.2.1 (15,2) tráyas vá saptá saptā 'dityā́ḥ,

KS. XI.6 (142,3) trír vā Ādityā́s saptá saptā, yāvanta evá 'dityā́s tā́n badhnāti; for

metrical speculations (eight Vasus: gāyatrī; eleven Rudras: tristubh; twelve Ādityas:

jagatī), see Oldenberg, Kleine Schriften, p. 696 n. 2 and, e.g., the Vedic Concordance

sub vasavas tvā gāyatrena. Different, again, is ŚB. XI.6.3.6; cf. also AB. VIII.12.4,

JB. I.283,1, II.101,2,4,7.

150 For the Asuras as an open group, see Oldenberg, “Varuṇa und die Ādityas”,

ZDMG. 50 (1896), p. 46 n. 3 (= Kleine Schriften, p. 691 n. 3); their number is

legion: F. D. K. Bosch, The Golden Germ, p. 87 (on the Kauravas).

151 See, e.g., Muir, Orig. Skt Texts V, p. 54ff., Oldenberg, l.c., pp. 43–68 (= Kleine

Schriften, pp. 688–713), Hillebrandt, Ved. Myth. II 2, p. 87. The latter two were

mainly concerned with the dispute on the Zoroastrian Ameša Spentas. Arbman’s

discussion (Rudra, p. 308) is unsatisfactory.

46

originally he was the leader of the Vasus. Other mythological figures,

such as Tvaṣṭr, seem in fact to have belonged to the undifferentiated

world. As will be shown below, the order in which the three groups were

enumerated was fixed ever since the Ṛigveda.

For the present the true meaning of this remarkable tripartition must

remain an open question. Here it can only be pointed out in passing

that during the cosmogonial fight Indra, as the leader of the Vasus

(note 153), with the help of the Rudras (=Maruts), gains the victory

over the Asuras of the primordial world, as a result of which some of

the last group join the Devas. The nature of the tripartition might thus

be formalized as “Vasus-Rudras+Ādityas”. It might be tempting to

suggest that, owing to the ambiguous character of the Ādityas (which

will be demonstrated in this chapter), they add a new dimension to the

group of Vasus-Rudras. However, not before a much better insight has

been gained into the nature of the Rudras, the most elusive of these

groups, will it be possible to determine the relation between Vasus

and Rudras.

Our task is more modest, viz. to answer the question how this tripartite

system was fitted into the normal quadripartite system of classification,

which turns clockwise from the East to the North. The Vasus are always

associated with the East, the Rudras with the South and, with rare

exceptions, the Ādityas with the West. The problem of finding a fourth

group for the North has been solved in different ways, which will be

discussed below.

The first question that arises is whether these fixed associations were

due to the fixed order of the three groups (as is, no doubt, the case in

some priestly speculations)152 or if, inversely, the association was from

the beginning a given fact and thus led to enumerating the groups in

the order of the classification system. As far as I can see, the second

alternative is correct because for Indra153 and Agni154, who were not

only Vasus but even their leaders, only the association with the East

is natural155. On the other hand, the connection of Varuna, the principal

152 Cf. the ascending order of metres (Gāyatrī = 8, Triṣṭubh = 11, Jagatī = 12),

for which see above, n. 149; or the stereotyped ascending order of earth, air and sky

(e.g., JB. II.141 see below, note 159, and ŚB. VI.1.2.10, see note 163).

153 VII.10.4, 35.6, X.6.3.

154 E.g., I.44.3 Agnim purupriyám adyá dūtám vrṇīmahe Vásum, I. 127.1 Agním

hótāram manye davantam Vásum sūnúm sāhaso jātávedasam, V.51.13 vaiśvānará

Vásur Agníh. Cf. MS. III.4.2 (47,1) Agnir vai Vásuh and even in the epic, Mhbh.

XII.122.31 Vasiṣṭham íśam viprāṇām Vasūnāṃ jātavedasam (viz. vidadhe prabhum).

155 For the classification system the fact that both Indra and Agni belonged to the

East (Indra: TS. II.4.14, guardian of the East; Bhar. NŚ. 3.23f., 5.95f., Purāṇas;

Agni: cf., e.g., TS. I.8.7, ŚB. IV.4.9, ŚB., Gobh.GS., Mhbh., Rām.) created diffi-

culties. Therefore, Agni is mostly associated with the Vasus and the East, while

Indra is relegated to the South and functions as leader of the Rudras. The original

identity of Rudras and Maruts (cf., e.g., III.32.3, X.32.5) and Indra's close connec-

tion with the Maruts may have contributed to this makeshift but do not explain it,

as Agni was also connected with the Rudras (VII.5.9, VIII.103.14). Anyway, a

47

Āditya, with the setting sun and Death, shows that the relation of the Ādityas to the West was not arbitrary. In the light of the following material it is unexplainable how one of the greatest Vedic scholars could write that in the Rigveda the Vasus are not yet a separate class 156.

A second question is, whether the few exceptions in the Rigveda to this order (III.8.8, VII.35.14, X.66.12) prove that the order was not yet entirely fixed at that time. I think, such a conclusion would not be warranted. The poets were much freer in handling the traditional material of their religion than the systematic theologians of the brāhmaṇas, and that they made use of this freedom for prosodical purposes is apparent from a comparison of III.20.5 with III.8.8: whenever the nominative or vocative plural of Vasu was required, the poets of the family books preferred to put this after the caesura. Cf. III.8.8 Ādityā Rudrā Vāsavah sunītháh (VII.35.14 Ā. R. Vāsavo jusanta, X.66.12 Ā. Rú. Vāsavah sūdānavah) as against III.20.5 Vásūn Rudrám Ādityám ihá huve. Similarly in the genitive plural in X.48.11 Āditydnām Vásūnām rudriyāṇām. Only in the latest book we find in an irregular pāda (X.128.9) Vásavo Rudrā Ādityā uparisphśam. A similar case is CII. III, p. 8, where Samudragupta is said to be Dhanada-Varune-ndrāntaka-sama (sequence N.W.E.S.).

A third question concerns the exact relation of Indra to the three groups. As for X.48.11, where Indra says "I, as a Deva, do not infringe the law (?) of the Devas, viz. the Ādityas; Vasus and Rudriyas" 157, this would seem to stress the solidarity among the Devas (whatever dhāman may here mean) rather than to express a subordination of Indra to the totality of the three groups. Such an inferior rank of Indra cannot be inferred either from an isolated brāhmaṇa-passage, where Indra is said to have been later born than the Devas 158. If any importance can be attached to this passage, “Devas” must here be a general term for the Devas Ādityas only (cf., e.g., X.65.9, 66.4). Such a cult legend as JB. II.141–142, where the Vasus, Rudras and Ādityas do not accept Indra's supremacy because they claim, as jyeṣṭha, to rank above him,

makeshift it was. Elsewhere the East is assigned to Indra and the South-East to Agni (e.g., Hopkins, Epic Mythology p. 151. Gonda, JAOS. 87 [1967], p. 423b), e.g. in the Gobhila Grhyasūtra and the Purāṇas, while in the epics Agni is often located in the East (Hopkins, pp. 56, 104, 151). Indra's connection with the Vasus remains unaffected by the artificial solutions of the cosmic classification, cf. Mhbh. XIII.150.79 Bomb. ed. sāvitrīṃ adhigamya Śakravasubhīḥ kṛtsnā jitā Dānavāḥ. The derivative vāsava, an epithet of the name Indra in AS. VI.82.1 (see Bloomfield, SBE. 42, p. 95, Ludwig, Rigveda III, p. 470, etc.) becomes a name of the god in the epics.

156 Hillebrandt, Ved. Myth. II 3, p. 387.

157 Ādityā́nām Vásūnā́m Rudrā́ṇā́m devó devā́nā́ṃ ná minā́mi dhā́ma. The translation given above is Geldner's, but dhāman is a notorious crux. Cf. Gonda, Dhāman, p. 38f.: “Projections of the divinity of his colleagues.”

158 Cf. TS. II.3.4.2 devā́nāṃ ānujā́varāḥ. But in the parallel version of KS. XI.4 (147,13) it is Brhaspati who is the ānujāvaráḥ. Cf. TS. V.5.9.5, where devā́ḥ in the classification system can only mean the devā́ Ādityā́ḥ (RS. X.65.9, 66.4, etc.).

may have arisen because of the necessity of justifying the Vighana-

sacrifice 159.

In the Ṛigveda we find either the three groups of gods alone 160, or

with their leaders, respectively Indra, Rudra and Varuṇa (or Aditi herself).

If the poet gives a quadripartite classification, the fourth is either

Br̥haspati with the "singers" (= Aṅgirasas) or Tvaṣṭṛ. Although the

Ṛigveda seldom refers to the system of classification expressis verbis

(as the Atharvaveda does), there can be no doubt that this complementation

of the triad was due to the need of adapting it to the cosmic classification

(expressed, e.g., at VIII.28.3). The poets, however, were not systematic:

sometimes only a single deity is mentioned in connection with three

groups, such as Mitrā-Vāruṇā or Agni (Indra in VS. 38.8).

The following is the material of the Ṛigveda: 161

III.20.5 Vásūn Rudrám Ádityám ihá huve

(cf. TS. III.5.1.2 V. R. Á. ihá jínvatam).

III.8.8 Ádityā Rudrā Vásavah suníthā

dyāvākṣámā prthiví antárikṣam

VII.35.14 Ádityā Rudrā Vásavo juṣante

'dám bráhma kriyamānáṃ návīyah

X.66.12 Ádityā Rúdā Vásavah sūdánavā

imá bráhma sasyamānā́ni jinvata

X.66.4 Áditir dyāvāprthiví rtám mahád

Índrāviṣṇū Marútah svàr br̥hát,

devā́n Ádityā́n ávase havāmahe

Vásān Rudrā́n Savitā́ram suáṃsasam

159 Prajāpati, after creating the Vasus on earth, the Rudras in the air and the

Ādityas in yonder world, creates Indra, for whom no room is left. Prajāpati then

says "I have created thee for the highest respect and for the overlordship over them"

(eteṣā́m tvā . . . śraiṣṭhyāyādhipatyā́yā 'rksi). The gods do not accept this: "We,

verily, are most esteemed, we are overlords because we are superior" (vayaḿ vā́va

śreṣṭhā[s] smo, vayam adhipatayo ye jyeṣṭhā[s] sma iti nā 'jānan). Prajāpati then

"sees" the Vighana-sacrifice and performs it for Indra, whereupon the latter slays

all his enemies (mṛdhah). The normal situation would rather seem to be that Indra,

as an ádhipati ("overlord"), meets the Vasus, Rudras and Ādityas (who are also

overlords, see below VS. XIV.30, ŚB. VIII.4.3.16) on equal terms. He sometimes

claims, it is true, to be śreṣṭha (PB. VII.8.2 ahaṁ vah śreṣṭho 'smi) but in the light

of the older mythology it is hard to determine what the mythological meaning

of Indra's becoming a śreṣṭho 'dhipatih may have been. With the jyaiṣṭhyam śraiṣṭhyam

of JB. II.67 cf., e.g., Śāṅkh. Ār. IX.8 jyeṣṭhāya śreṣṭhāya svāhā, Tā. X.44, ŚB.

XIV.9.3,4, AB. IV.25.7-9, etc.

160 Variations are rare. In I.58.3 Rudrébhir Vásubhir only two groups are mentioned.

See also n. 166. The variation usually concerns the Rudras, who are replaced by

another group, cf. VII.44.4 Ādityébhir Vásubhir Aṅgirobhir (although, properly

speaking, the Aṅgirasas are not gods, see n. 208), II.3.4 ghrténā 'ktā́m Vasavah

sídate 'dáṃ Viśve Devā Ādityā yajñíyāsah. Similarly AS. I.9.1: Vasus, Ādityas,

Viśve Devāḥ.

161 Cf. Arbman, Rudra, p. 157 n. 1.

The triad has been extended to four in:

VII.10.4 Indram no Agne Vásubhih sajóṣā

Rudrám Rudrébhir á vahā brhántam,

Ādityébhir Áditim viśvajanyām

Brhaspátim t́kvabhir viśvávāram

śám no Índro Vásubhir devó astu

śám Ādityébhir Várunah suśastíṣah,

śám no Rudró Rudrébhir jálāṣaḥ

śám nas Tváṣṭā gnábhir ihá śṛṇotu

X.66.3 Índro Vásubhih pá̄ri pātu no gáyam

Ādityair vá Áditih śárma yachatu,

Rudró Rudrébhir devó mṛlayāti nas

Tváṣṭā no gnábhir suvitā́ya jinvatu

X.125.1 ahám Rudrébhir Vásubhiś carāmy

ahám Ādityair utá Viśvádevaiḥ,

ahám Mitrāváruṇo 'bhā bibharmi

ahám Indrāgní ahám Aśvínā 'bhā

This is the only place in the Rigveda where the Viśve Devas are introduced as a fourth group, which is sufficient to show the secondary character of this quadripartite system. In the family books they once take the place of the Rudras in II.3.4cd

ghrténā 'ktám Vasavah sídáte 'dám

Viśve Devā́ Ādityā yajñíyāsaḥ

The nature and the function of the Viśve Devāh are still obscure 162, although they were clearly connected with the idea of totality, cf. RS. I.89.10cd (= AS VS)

Viśve Devā́ Āditih páñca jánā

Āditir jātám Āditir janitvam

In the brāhmaṇas, where they often occur together with Brhaspati in the cosmic classification this character is still more perspicuous.

162 Caland's suggestion that they were the deified forefathers (deváh pitárah, see Altindischer Ahnenkult, p. 182f.) does not seem to be confirmed by the data, except Mhbh. V.107.3, XII.165.440* (see Serensen). Nor are they "a factitious sacrificial group meant to represent all the gods" (Macdonell, Vedic Mythology, p. 130). For a recent discussion see Renou, EVP. 4, pp. 1-11. On the basis of X.65.11 he characterizes them as "une pluralisation anticipée de Prajāpati" (p. 9 n. 1). They are "impelled by Indra, commanded by Varuṇa-prasiṣṭāḥ), which may indicate that they were connected with both upper and world and thus represented aspects of both moieties (cf. n. 183). As this antithesis shows, the alternative translation of prasūtā as "hervorgebracht" (H. P. Schmidt, Brhaspati und Indra, p. 129) cannot be correct, cf. also vājaprasūta, hāryaśvaprasūta. For II.3.4, where the Viśve Devāḥ take the place of the Rudras, see note 160. I doubt if Renou is right (pp. 3,8) in taking viśve Vásavah, viśve Marútah and viśve Ādityā́ḥ as standing for the Viśve Devāḥ.

50

Only Agni or the dual group Mitrāvaruṇā is mentioned in connection with the three groups in the following stanzas:

X.150.1 sámiddhaś cit sám idhyase

devébhyo havyavāhana,

Ādityai Rudraír Vásubhir na à gahi

mrḷikáya na à gahi

II.31.1 asmákam Mitrāvaruṇā 'vatam rátham

Ādityai Rudraír Vásubhiḥ sacābhúvā,

the second pāda of which recurs in the hymn to the Aśvins:

VIII.35.1 Agnímé 'ndrena Várunena Viṣṇunā

'dityai Rudraír Vásubhiḥ sacābhúvā,

sajóṣasā Uṣása Sūryeṇa ca

sómam pibatam Aśvinā.

Finally the three groups occur once in the Rigveda in a very special context in VIII.101.15, where Aditi is said to be the daughter of the Vasus, the sister [!] of the Ādityas and the mother of the Rudras. The poet can hardly have meant (in spite of X.48.11, discussed above, p. 48) that all Devas descended from the primordial world, just like the Ādityas, sons of Aditi 163. At best he may have referred to the relationship that existed between the Devas, as “younger brothers” of the Asuras, and the World Mother, who was the female manifestation of the Father Asura in the primordial world (where male and female were still interchangeable as a consequence of its undifferentiated state), before she, like the Ādityas, became part of the dual world and thus gave rise to Diti as her counterpart among the banished Asuras 164.

It must be admitted, however, that what can be inferred from the texts about the origin of the Devas is very vague and it may be questioned if the Rigvedic poets themselves had clear-cut ideas on this point. This is particularly true of Indra. In the Vedic myth his appearance in the undifferentiated world had an irrational character and it is not surprising

163 A different idea occurs in ŚB. VI.1.2.10. Here Prajāpati first creates Agni and the Vasus, who are located on earth, then Vāyu and the Rudras in the air, thereupon the Sun and the Ādityas in the sky, and finally the Moon and the Viśve Devāḥ in the quarters (diśas), which again points to an association with cosmic totality. Cf. also AS. I.9.1 (Vasus, Ādityas, Viśve Devāḥ).

164 For references see Indol. Studies – W. Norman Brown, p. 146 n. 59 and further Max Müller, Hymns to the Maruts (1869), pp. 230–251 (cf. SBE. 32 [1891], pp. 241–263), J. Muir, Original Sanskrit Texts (1874), pp. 35–53, Oppert, ZDMG. 57 (1903), pp. 508–519, H. Güntert, Der arische Weltkönig und Heiland (1923), p. 201ff. Heinrich Zimmer, Eranos Jahrbuch 6 (1938), pp. 175–220, L. Buschardt, Vṛtra, Det rituelle Dæmondrab (1945), p. 97f. Aditi must have “gone over” like Varuṇa, but she represents the dual cosmos in its totality. She knows “both worlds” (KS. XXIII.8: 83,19 aham evemā ubhau lokau veda, KKS. XXXVI.5: 191,23/223,14. aham evemau). Hence her being ubhayādatáḥśṛṇī? (different explanations at ŚB. III.2.4.16, TS. VI.1.7.5, MS. III.7.5: 82.1). As devī she is opposed to the Asuras (MS. IV.2.3: 24,17 Ādítir devéṣv ásīt, Kūṣṭā 'suresu). The banished Asuras got Diti for their mother (A. B. Keith, Religion and Philosophy, p. 217f.).

51

that the Rigvedic references to his mother (VIII.77.1–2, etc.) are very

vague. In the light of such reflections it would seem very doubtful that

originally Indra was regarded as the last-born son Aditi, as he is sometimes

represented in the Yajurveda (but see n. 96). Since the scarce evidence

leaves too much room for subjective interpretations, no decision is possible.

The stanza reads as follows:

mātá Rudrāṇāṃ duhitá Vásūnām

svásā 'dityānām amŕtasya nā́bhih,

prá nu vocam cikituṣe jánāya

má gám ánāgām Áditim vadhiṣṭa.

8. THE ĀDITYAS IV: OTHER VEDIC TEXTS.

In the brāhmaṇas, too, the three groups Vasus, Rudras and Ādityas

occur. In the Black Yajurveda the earth, identified with Aditi, is said to

be the seat (sádas) of the three classes of gods 165, and the somakráyaṇī-cow

is addressed with the words “You are a Vasvī, a Rudrā, Aditi, an

Ādityā . . .” 166.

The problem of how to extend this system by a fourth group is solved

in different ways, depending on whether a fifth group is added for the

centre or not. Thus the Aitareya Brāhmaṇa follows the tradition of

RS. X.125.1 by adding the Viśve Devas with Brhaspati 167. As for the

Yajurveda, in one passage of the Taittirīya Saṃhitā and the Taittirīya

Brāhmaṇa the same extension is found but here it can be shown to be

165 MS. IV.1.13 (18, 6ff.) iyám vấ Aditir, asyắm evainam asisadad, Vásūnām Rudrā́ṇām

Ādityā́nām sádo 'sity, etấ vấ etám agre devátām abhyàjayayains, tábhya evainat sádah

karoti, KS. XXXI.10 (13,4), KKS. XLVII.10 (294,11/345,20) Vásūnām Rudrā́ṇām

Ādityā́nām sádanam así 'ti, yấ evấ devatấ abhyàjayayāms tábhya enat sádanam karoti.

Cf. MS. I.1.12 (7,16), KS. I.11 (6,10), KKS. I.11 (8,9), Vedic Concordance, p. 850a,

B. Geiger, Die Ameša Spentas, p. 205.

166 TS. I.2,5, VI.1.8.1, ĀpŚS. X.22.11 Vásvy asi, Rudrá 'sy, Āditir asy, Ādityā́ 'si

sukrá 'si, candrá 'sī. Caland’s translation “Die Gute bist du” for vásvy asi (Āpastamba

Śrautasūtra, l.c.) is a slip of the mind. In this connection the question may be asked

whether (in spite of the fact that the Aśvins are a dual deity, which RS. X.125.1

puts on a par with Mitrā́várunā and Indrā́gṇī) the words vásū Rudrá purumántū

directed to them (I.158.1) do mean “Gut sind die beiden Rudras” (Geldner).

Since I.58.3 Rudrébhir Vásubhir (n. 160) refers to the two classes and VII.59.8

Maruto . . . Vasavo means “O Maruts . . . O Vasus” (Geldner, Renou) it is more

likely that the Aśvins are referred to as “the Vasu and the Rudra with many

thoughts” (cf. Renou, EVP. 16, p. 24). That there is a structural contrast between

the twins has been demonstrated (convincingly in my opinion) by Stig Wikander,

“Nakula et Sahadeva”, Orientalia Suecana 6 (1957), pp. 66–96.

167 AB. I.24.4 te devā abībhayur : asmakam vipramānām anu idam asurā ābhaviṣyanti

'ti. te vyutkramyā́ 'mantrayantā́, 'gnir Vasubhir udakrā́mad, Indro Rudrair, Varuṇa

Ādityair, Brhaspatir Viśvair Devaiḥ, JB. I.283 Vasavo gāyatrím samabharhan . . .

Rudrās triṣṭubhám samabharhan . . . Ādityā jagatím samabharhan . . . Sādhyā́s cā́

'ptyā́s cā́ 'ticchandasám samabharhan. For Brhaspati cf. RS. IV.50.6 pitré viśvádevā́ya

and see H.-P. Schmidt, Brhaspati und Indra, p. 218. In general cf. Bergaigne II,

pp. 114–156 (“arithmétique mythologique”).

52

secondary 168. The passage of the AB. is interesting because parallel

versions occur in the Black Yajurveda. The tale contains an aetiological

myth which explains the system of classification from a quarrel that once

arose among the Devas, while they were fighting against the Asuras.

Because no group was willing to accept another group's superiority

(śrī, śraisṭhya, jyaiṣṭhya), they split into four (or five) separate groups,

which dispersed into the four directions 169. In one version of the

Yajurveda the Maruts, who in the Rigveda are identical with the

Rudras 170, have been split off as a separate group, which here takes

the place of the Viśve Devas. Although in other passages the normal

order is preserved 171, in this version Ādityas and Maruts have changed

places, for which I cannot offer an explanation 172.

More normal, however, is that the Maruts are located in the North

when five groups are distinguished. In that case the Viśve Devāh are

located in the centre, symbolizing the totality. This is found in the Taittirīya

Samhitā, the Śatapatha Brāhmaṇa and the Gopatha Brāhmaṇa. 173. In

this centre, however, the dualism recurs in the vertical opposition of

168 TS. III.5.6.2 Vásubhyo Rudrébhya Ādityébhyo Viśvebhyo vo Devébhyaḥ pannéjanīr

grhṇāmi, TB. II.2.10.5–6 tám devāḥ samantám páryaviṣan, Vásavaḥ purástāt, Rudrá

dakṣiṇatáh, Ādityáh pascát, Viśve Devá uttarátah, Angirasah pratyañcam (6) Sādhyáh

párāñcam (S. Lévi, Doctrine du sacrifice, p. 73). That the Viśve Devāh have later

been added, appears from the parallel versions MS. I.3.1, (29,10), KS. III.9 (28,4),

KKS. II.16 (24,8/29,4) Vásūnāṁ Rudrāṇāṁ Ādityānāṁ pannéjanīḥ stha, Vásavo

Rúdrā Ādityā etá voh pannéjanīḥ. There are more instances of such an extension

of a formula in TS., cf. IV.3.10.3 (VS. XIV.30, ŚB.VIII.4.3.16) dyávāprthiví

vyaitām ... Vásavo Rudrá Ādityā anuyáyāmis, tá evá 'dhipataya āsan, as against

MS. II.8.6 (110.19), KS. XVII.5 (349,4) Vásavo Rudrá anuvyáyan.

169 It may be noted that the five groups of gods fighting against the Asuras can

mythologically be paralleled to the five Pāṇḍavas (four of whom go to the four

quarters during the digvijaya, Mbh. II.23–29) waging war against the countless

Kauravas (G. J. Held, The Mahābhārata, F. D. K. Bosch, The Golden Germ).

170 See Arbman, Rudra, p. 14, n. 3. The name Rudras does not occur in books IV

and VI of the Rigveda.

171 MS. IV.1.10 (13,14) ásurānāṁ vō iyám prthivy ásiṭ. té devá abruvan : dattá no

'syāḥ prthivyá íti. té vai svayáṁ evá 'brūdhvam (read evā 'vruddhvam) íti. táto vai

Vásavah prácīm údīcayan, Rudráḥ dakṣiṇām, Ādityáḥ prátīcīm, Marúta ūdícīm.

172 MS. II.2.6 (19,13), III.7.10 (90,1) devá anyó 'nyásya śraiṣṭhyé 'tiṣṭhamānā

(emendation of śriṣṭe, skaisthye tiṣṭhamānā, von Schroeder, edition Aundh) vyúdakrāman, Agnír Vásubhiḥ, Sómo Rudraíḥ, Índro Marúdbhiḥ, Váruṇa Ādityaiḥ.

tán vā eténa Brhaspátir áyājayat ... táto devā ábhavan, párā 'surāḥ, KS. XXIV.9

(100,3), KKS. XXVIII.2 (206,1/240,10) devā́ evā 'nyo'nyásraiṣṭhyāya (KS. anyo-

'nyasya) nātiṣṭhanta, te caturdhā́ vyudakrāmann, Agnír Vásubhiḥ, Somo Rudraíḥ,

Índro Marúdbhiḥ, Váruṇa Ādityaiḥ. te devā́ pāpīyāṁso vai tye bhavanti, vasiyāṁso 'surāḥ.

173 TS. VI.2.2.1 devā́surāḥ sáṁyattā āsan, té devā́ míthó vipriyā āsan. té 'nyo'nyásmai

jyaiṣṭhyāyá 'tiṣṭhamānāḥ paicadhasthá vy ákramann, Agnír Vásubhiḥ, Sómo Rudraíḥ

Índro Marúdbhiḥ, Váruṇa Ādityair Brhaspátir Viśvair Devaiḥ ..., ŚB. VIII.6.3.3

(about the Gārhapatya) tám íha 'hútyo 'pādadhatā, túsmin vyávadanta, Vásavah

purástāt,Rudrá dakṣiṇatá, Ādityáḥ pascán, Marúta uttarátó, Viśve Devā upáriṣṭāt ....,

Gop. Br. II.2.2 pañcadhasthá vai devā́ vyudakrāmann, Agnír Vásubhiḥ, Somo Rudraíḥ,

Índro Marúdbhiḥ, Váruṇa Ādityair, Brhaspátir Viśvair Devaiḥ. te devā́ abruvann

asurebhyo vā idaḿ bhrā́tṛvyebhyo radhyāmo yan mítho vipriyā smo ...

the zenith versus the nadir, corresponding with the top and the bottom

of the cosmic axis, and with the upper versus the nether world respectively.

When this distinction is made, Bṛhaspati is often located “above”

(upáriṣṭāt), that is, in the zenith 174. This place he shares with Savitṛ,

whom he seems to have ousted in the course of time 175. In later texts,

such as the Gŏbhila Gṛhyasūtra and the Purāṇas, this place is attributed

to Brahmā 176. By the side of the ŚB. passage mentioned above there

is another one in the same text which polemizes against the introduction

of the Viśve Devāḥ. This passage is mainly interesting because it seems

to show that its author no longer understood the idea of a totality which

is more than the sum total of its components 177.

In connection with Bṛhaspati's place in the zenith a brief digression

may here be inserted. In the Brāhmaṇas, just as well as in the Saṁhitās,

it is characteristic of the Ādityas, and particularly of their leader Varuṇa,

that they only very rarely take part in the fight of the other Devas against

the Asuras 178. One of the exceptions is a passage in the Jaiminīya Brāhmaṇa

which tells that the Asuras fled with the essence and power of the savanas

(pressings) into the darkness and that Indra won back one third with

help of Varuṇa, one third with Bṛhaspati, and one third with Viṣṇu 179.

174 ŚB. V.1.1.4 egò 'rdhvā́ Bfhaspáter dìg ìty evò 'huh “They say that the upper

region belongs to Bṛhaspati.” From JB II.25,3, where Bṛhaspati is mentioned

conjointly with Prajāpati in the description of the rājāsandī, it does not follow that

Bṛhaspati had some traits in common with Prajāpati. Cf. AB. VII.12.3 and 17.2,

where Bṛhaspati is associated with Savitṛ.

175 See Hanns-Peter Schmidt, Bṛhaspati und Indra, p. 88f.

176 Bṛhaspati in the zenith, as opposed to Viṣṇu in the nadir, is already found in

the Atharvaveda, see below, p. 56. That this may be interpreted as upper world

versus nether world may be inferred from AS. IV.40.5-7, where in the centre are

located Earth, Vāyu and Sūrya. See in general for the system of classification

E. W. Hopkins, Epic Mythology, p. 149ff., Mbhṇ. VII.30.76-77, V.106-110,

XIII.100.11, Rām. II.14,321 *, etc., A. K. Coomaraswamy, Yaksas II, p. 31, Thérèse

de Mallmann, L'iconographie de l'Agnipurāṇa, p. 124ff. (table on p. 127), whose

material, in so far as it has been taken from J. N. Banerjea, The Development of

Hindu Iconography, p. 520 n. 1, is useful but not always reliable: the latter's re-

construction of the classification on the basis of Manu V.96 is incorrect. Banerjea's

book itself was not accessible to me. For Brahmā in the zenith, as opposed to

Ananta (= Śeṣa, Viṣṇu's snake) or Vāsuki in the nadir, see de Mallmann, pp. 56, 96

(who refers to Gŏbhila Gṛhyasūtra and the Purāṇas). See n. 195.

177 ŚB. III.4.2.1 àtithyèna vai devā́ iṣṭvā́, tàṁ samād avindant tàṁ caturdhā́ vyàdravann

anyò'nyásya śriyè átiṣṭhamānā, Agnír Vásubhiḥ, Somo Rudráir. Váruna Ādityair,

Indro Marúdbhír Bṛhaspáti Vísvān Devā́n itv ù haika áhur ètá há tv èvá té Vísve

Devā́ yé té caturdhā́ vyàdravams. tàṁ vidrutā́n asurarakṣásāny anuvyā́yeyuh (2) tè

'viduh pápīyāṁso vái bhávāmo, asurarakṣásāni vái no 'nuvyā́guh.

178 For the view that the Ādityas were fighting the Dānavas (Jean Varenne, Di

alcuni miti cosmogonici del Ṛg-Veda, Torino 1969, p. 12) I find little support in the

texts (see n. 223). MS. II.2.1 (15,5) Ādityā́ vái trātára, Ādityā́ aparoddhárah, KS. XI.6 (151,4)

Ādityā́ vái trātára, Ādityā́ aparoddhárah only conceal the existing fear of Varuṇa's

wrath. But see also p. 76 and n. 223.

179 AB. III.49.1, 50.1-3 agniṣṭomam̀ vái devā́ aṣṛayanto, 'kthāny asurā́n, te samāvad-

viryā̀ evā́ 'san . . . te vā́ asurā́ Maitrāvaruṇasyo 'ktham aṣṛayanta. so 'bravīd Indraḥ:

kaś cā 'haṁ ce 'mān ito 'surān notsyāvahā́ ity. ahaṁ ce 'ty abravīd Varuṇaḥ . . .,

In view of the characteristic association of Viṣṇu with the third part (as in Jaim. Brāhmaṇa II.243, where Indra is said to get two parts of Vṛtra's indriyam viryam, but Viṣṇu the third), it has been argued that, if Viṣṇu here represents the totality of the cosmos, Varuṇa is likely to stand for the nether world and Brhaspati for the upper world 180. The reason for choosing Brhaspati as the representative of the upper world, in spite of his later ambiguous character (n. 361), may have been his place in the system of classification 181.

In spite of the criticism by one author of the Śatapatha Brāhmaṇa there can be no doubt that a system which divided the universe into five regions belonged to the old tradition of the Yajurveda. In a list which gives for each region its mythical name, the name of its "overlord" (ādhipati) and that of the "arrester of the arrows" (hetindḿ pratidhartṛ), we find in MS. II.8.9, KS. XVII.8, KKS. XXVI.7 and TS. IV.4.2.1ff.:

East: name Rājñī, overlords: Vásus, arrester: Agní.

South: resp. Virāj, Rudrás, Īndra (KKS.: Sóma).

West: resp. Samrāj, Ādityás, Sóma [!] (KKS. Īndra).

North : resp. Svarāj, Marúts (TS.: Viśve Devāḥ), Váruṇa [!].

Zenith: resp. Ādhipatní, Viśve Devâḥ (TS.: Marúts), Bṛhaspáti.

The Taittirīya Saṃhitā is consistent in associating the Viśve Devas with the North (see n. 168), although the combination of them with Varuṇa (as in TS. IV.3.3.2) is sufficient to prove the secondary character of this system. All the texts of the Black Yajurveda, however, concur in locating Varuṇa in the North (but cf. MS. II.7.20: 105, 12). Here the White Yajurveda has a more logical (although not necessarily textually older) division. Cf. VS. XV.10-14, ŚB. VIII.6.1,5ff.:

JB. I. 179-180 devāsurā yajñe 'spardhanta. te devā asurān yajñān niravāghanan. te 'surās trayānām savandānām rasaṃ viryam pravṛhyā 'ndham tamaḥ prāviśan. yad dha vai kim ca parācinam agniṣṭomāt tad andham tamaḥ. yat parācinaṃ saṃvatsarāt dvādaśa vā agniṣṭomasya stotrāṇi, dvādaśa māsāḥ saṃvatsaraḥ. dvādaśa stotrāṇi, dvādaśa saṣtrāṇi. dvādaśa pūrvapakṣā dvādaśa parapakṣāḥ. atha yat tataḥ parācinam tad andham tamaḥ. tat prāviśat. te devā asurān anvabhyavāyan. tān asurān etair evo 'kthāḥ pratyudatiṣṭhan. yad ukthāḥ pratyudatiṣṭhāṃs tad ukthānām ukthātvam. tāṃs tato nir evā 'vāghnaṛ eva. te 'ndham tamaḥ prāviśan, rātrim eva. sa Indro 'bravīt kaś cā 'ham dam anvabhyavāvṛṣyā iti. [180] aham ce 'ty abravīd Varuṇaḥ. aham vāṃ jyotir dhārayiṣyāmi ty Agnir abravīt... Cf. also JB. III.104, where Varuṇa helps the Devas.

180 See Indological Studies-W. Norman Brown, pp. 145, 150. Varuna was there explained as standing for the Asuras, which is inaccurate (p. 150). He acts as a devá ásura (= Āditya) of the dualistic cosmos, the Asuras themselves standing outside the world of order. Varuṇa, although associated with the West in the classification system, was thought of as residing below (adhástāt). Cf. TS. V.5.9.5, as interpreted below, n. 183, and IIJ.8, p. 108. For the general problem of the difference between function and place in the classification see IIJ. 13, p. 283.

181 The role of this system appears in Gopatha Brāhmaṇa II.4.11, where Agni, Varuṇa, Bṛhaspati and Viṣṇu may stand for the main opposition in the horizontal plane (Agni: East, life, light, etc. versus Varuṇa: West, death, darkness, etc.) and for the vertical opposition (Bṛhaspati: zenith, Viṣṇu: nadir) but Indra's place in between remains to be explained.

East: name: Rájñí, adhip.: Vásus, pratidh.: Agní.

South: resp. Viráj, Rudrás, Indra.

West: resp. Samráj, Adityás, Varuṇa.

North: resp. Svaráj, Marúts, Soma.

Zenith: resp. Adhipatni, Viśve Deváh, Bṛhaspáti.

As an illustration of the many variations that occur in details the

following table gives a synopsis of the data of two Atharvavedic passages

(III.27.1–6, XII.3.55–60), compared with two Yajurvedic texts (MS.

II.13.21, TS. V.5.10.1–2). They give for each region its mythical name,

its overlord (adhipati), its “defender” (rakṣitṛ) and its arrows (only

in the AS.):

East :

South :

West :

North :

AS. name

Prācī

Dakṣiṇā

Pratīcī

Ūdicī

overlord

Agní 182

Indra

Varuṇa

Sóma

defender

Asitā

Tiráścarāji

Pṛdāku

Svajā

arrows

Ādityáh [!]

(but XII : Ādityáh)

Pitáras 183

ánnam

āśani

MS. name

Samcī

Ojasyă

Prācī [!]

Suṣádā

overlord

Agní

Indra

Sóma [!]

Varuṇa [!]

defender

Asitā

Tiraścinarāji

Svajā

Sṛdāgu, v.l. Sṛdāku

TS. name

Samcī

Ojasvinī

Prācī [!]

Avasthāvan 184

overlord

Agní

Indra

Sóma [!]

Varuṇa [!]

defender

Asitā

Pṛdāku [!]

Svajā

Tiráścarāji

Zenith

Nadir

AS. name

Ūrdhvā

Dhruvā

overlord

Bṛhaspáti

Viṣṇu

defender

Śvitré

Kalmāṣagriva

arrows

varṣám

virūdhah 185

182 But AS. XVIII.3.25: East: Indra and Maruts, South: Dhātṛ, West: Aditi, North:

Soma.

183 Taboo may have been the main reason why the Pitáras are but seldom mentioned

in the system of classification in connection with the region that is characteristically

theirs. They belong, indeed, to the cosmos (see above, p. 12). When they are included

in the system, the Rudras are shifted to the East, cf. TS. V.5.9.4 Vásavas tvā Rudráḥ

purástāt pāntu, pitáras tvā yamárājānaḥ pitṛbhir dakṣiṇatáh pāntv, Ādityás tvā

Viśvair Devaíḥ paścát pāntu, Dyutānás tvā mārutó uttaraláh pāntu (5) devā́s

twé 'nrajyeṣṭhā várunarájāno 'dhástāc copáriṣṭāc ca pāntu. Cf. VS. XXXVIII.9

(below, p. 61). The term devā́h here seems to denote the Viśve Devā́h. For a parallel

see above, n. 158. The phrase anrajyeṣṭhā várunarájānaḥ refers to the cosmic

dualism, since Indra and Varuṇa are regularly opposed to each other. Moreover,

in RS. X.66.2 similar epithets are used for the Viśve Devā́h, which may lead to the

conclusion that they transcend the dualistic contrasts of this world (n. 162). It is

tempting, then, to interpret this TS. passage in the same manner, the more so as

here the contrast is indicated expressis verbis by the words adhástāt and upáriṣṭāt,

denoting the nadir and the zenith. In this vertical opposition, corresponding to top

and bottom of the world axis, Varuṇa belongs, indeed, to the lower end of the axis,

the roots of the world tree. See n. 180

184 Cf. ASPaipp. II.56.4, MS. II.13.21 (167,8) avasthā́.

185 Thus also the Paippalāda version of XII.3, where the Śaunaka version has

oṣadhī́h.

56

MS. name Ādhipatnī Avasthā

overlord Bṛhaspáti Viṣṇu

defender Citrá 186 Kalmāṣagrīva

TS. name Ādhipatnī Vaśinī

overlord Bṛhaspáti Yamá

defender Śvitrá Kalmāṣagrīva

The normal term for the zenith in the Yajurveda is brhatī dík "the high region" 187, perhaps with assonance of Bṛhaspati. Only MS. has ūrdhvā dík. The nadir is sometimes called dhruvā dík "the fixed region" (TS. iyám dík), the term for the centre in general.

In addition to the VS.-passage quoted on p. 56 another may be mentioned where the White Yajurveda seems also to have preserved (this time in agreement with the Maitr. version of the Black Yajurveda) the old combination of three groups of gods. This is the passage dealing with the partition of the vertebrae of the sacrificial horse 188. The parallel versions of KS. and TS. here add the Aṅgirasas, Sādhyas and Viśve Devas, which is no doubt a secondary extension of the original grouping 189.

A very elaborate form of the Vedic classification, which may be mentioned in conclusion, occurs in JB. II.142. After Indra has slain all his enemies, the gods (sarve devāḥ) sit down around him 190 in a ceremonial manner, which is called daivī sabhā or daivī saṃsad (TS. VII.4.2.1f.) and reproduces the system of classification, viz. in the East the Vasus, in the South the Rudras, in the West the Ādityas, in the North the Maruts, in the zenith (upariṣṭāt) the Viśve Devas, in the nadir (adhastāt) the Sādhyas 191 and Āptyas, and round about (abhitah) the Aṅgirasas.

9. THE ĀDITYAS V: LATER TEXTS

In the "ideal" ceremonial grouping of the gods in the passage just quoted the Aṅgirasas stand more or less outside the system. They were,

186 Probably a misreading, according to Whitney ad AS. III.27.6. The new Indian edition (by Viśvabandhu) reads śvitró, but Papp. III.24,6 has citró.

187 brhatī dík does not mean "the great quarter" (Keith, H.O.S. 19, p. 449).

188 MS. III.15.6 (179,7), VS. XXV.6 Viśveṣāṃ Devānāṃ prathamā kīkasā, Rudrāṇāṃ dvitīyā, 'dityānāṃ tritīyā.

189 KSA. XIII.7, TS. V.7.17 Vāsūnāṃ prathamā kīkasā, Rudrāṇāṃ dvitīyā, 'dityānāṃ tritīyā, 'ṅgirasāṃ caturthī, Sādhyānāṃ pañcamī, Viśveṣāṃ Devānāṃ ṣaṣṭhī. The curious insertion of the Aṅgirasas (after the Ādityas) may possibly be due to the close connection between these groups, to be discussed below, section 10.

190 tam் vijigyānasm் sarve devā abhitaḥ samantam் paryaviśan. For this ceremony cf., e.g., KS. X.7 (132,16), TS. II.4.1.2 samantāṃ devān páriy avisan, TB. II.2.10.5 tam் devāḥ samantāṃ páryaviśan (114,11), KS. XVII.8 (251,19), KKS. XXVI.7 (110,1/128,3), TS. IV.4.2.3, where a classificatory enumeration is concluded with the words té tvā sárve saṃvidānāḥ nākasyá prṣṭhé suvargé loké yajamānāṃ ca sādayantu. Cf. Keith, Religion and Philosophy p. 213 n. 2, Renou, EVP. 4, p. 3.

191 The proper place of the Sādhyas was upáriṣṭāt (see n. 192), in contrast with the Āptyas, who were adháṣṭāt. They were always located in the centre. See further p. 242ff.

indeed, mythical forefathers but, properly speaking, no gods, and in the system of the Jaiminīya Brāhmaṇa they have been introduced as a kind of appendix, at the end of the list of gods.

In this way they are also added when groups of gods are enumerated without a direct reference to the system of classification. In those cases the authors were freer and could add the Aṅgirasas as a last member of the series, without necessarily implying that the Aṅgirasas belonged to the cosmic centre 192. The Maruts, too, are sometimes mentioned after (instead of before) the Viśve Devas 193 but since these two groups were both used to complement the older system of three groups, variation in their respective order is not unnatural. However, the variation found in the Aitareya Brāhmaṇa, where not only the Aṅgirasas but also the Maruts occur at the very end of the list, after the Sādhyas and Āptyas 194, is rather uncommon. The result is that the Viśve Devas are located in the North, but the Maruts and Aṅgirasas in the zenith 195. Only a systematic study of all classifications occurring in the brāhmaṇas could show to what extent particular variations are peculiarities of a single work. Thus the rare occurrence of the Viśve Devas in the classifications of the Atharvaveda (only XVIII.3.25-29?) 196 would seem an individual feature of that Saṃhitā.

192 As an instance may be quoted TS. VII.1.18 Vásubhir devébhir devátayā . . . Rudrébhir devébhir devátayā . . . Ādityébhir devébhir devátayā . . . Viśvebhir Devébhir devátayā . . . Aṅgirobhir devébhir devátayā. A difficulty in any attempt to explain such variations is the impossibility of determining to what extent they may simply have been due to the personal idiosyncrasies of an individual priest or to mere differentiations between various schools, and how far they are essential, as expressions of divergent theological speculations. The Aṅgirasas are sometimes associated with the nadir (cf. Vādhūla Sūtra, Acta Or. 6, p. 130 Aṅgiraso 'dhastāt, Sādhyā upariṣṭāt), but more often with the zenith. See n. 215.

193 JB. II.25, lines 5-6, III.152 (in connection with the "King's throne" (rājāsandī) : Vasavo rājyāya, Rudrā vairājyāya, 'dityās svārājyāya (sic !), Viśve Devās samrājyāya, Marutas sārvavaśyāya, Sādhyās ca 'pṭyās ca pārameṣṭhya.

194 AB. VIII.12.4 Vasavas . . . samrājyāya. Rudrās . . . bhavajyāyā, 'dityās . . . svārājyāya, Viśve . . . Devā . . . vairājyāya, Sādhyās ca tvā 'pṭyās ca devāḥ . . . rājyāya, Marutaś ca . . . Aṅgirasas ca devā . . . pārameṣṭhyāya māhārājyāyā 'dhipatyāya svāvaśyāyā 'tiṣṭhāya . . .

195 AB. VIII.14.1-4 East: Vasus, South: Rudras, West: Ādityas, North: Viśve Devāḥ, asyām dhruvāyām madhyamāyām pratiṣṭhāyām diśi : Sādyas and Āptyas, ūrdhvāyām diśi: Maruts and Aṅgiraso devāḥ. For the Aṅgirasas this location is not unparalleled: cf. Maitrī Upaniṣad VII.1-5, where the zenith is assigned to Mitra-Varuṇa and the Aṅgirasas, which conception cannot be old, the less so, as the North is assigned to Varuṇa alone. The Maitr. Up. is interesting in that it locates Śani with the snakes in the nadir. See n. 176.

196 East: Indra marutvat, South: Dhātr, West: Aditi with the Ādityas, North: Soma with the Viśve Devāḥ; further Dhṛtar and Savitr, apparently in the nadir and the zenith respectively (see above, p. 54 for Savitr). As for Dhṛtar who, in that case, is at the bottom of the world axis and supports it, see H.-P. Schmidt, Brhaspati und Indra, pp. 88f., and cf. IIJ. 13, p. 283. Some places where the system seems to have been abandoned are AS. XIX.17.1-9, KS. XXXVI.15 (83,11), which gives Agni, Indra, Viśve Devāḥ and Prajāpati, and ĀpŚS. XVII.7.6.

In conclusion two passages from the two oldest Upaniṣads may be mentioned. The Chāndogya Upaniṣad (III.1-5) contains a series of speculations on the rasa that is gathered from the Vedas. It is then argued (III.5.4) that this rasa is identical with the nectar (amṛta), and that there are five forms of amṛta (III.6-10): first, that on which the Vasus live, using Agni as their mouth; second, that of the Rudras, who use Indra as their mouth; third, that of the Ādityas, whose mouth is Varuṇa; fourth, that of the Maruts, with Soma as their mouth; and fifth, that of the Sādhyas, who use Brahma as their mouth. This Brahma is the pitā Brahma (not Brahmā!) of III.11.4-5, the predecessor of god Brahmā, who was to take Bṛhaspati's place in the zenith.

While in this classification the Viśve Devas are absent, they occur in the Bṛhad Āraṇyaka Upaniṣad as the fourth group, the fifth being the Maruts 197.

From the preceding sketch it will be clear that, whatever the complications and contradictions of the system of classification, the world of the Devas consisted basically of three groups, and that the Ādityas belonged from the beginning (that is, from the mythic beginning of this world) 198 to this main body of gods, which was distinct from "the former gods", the Sādhyas and Āptyas 199.

This is the more remarkable because even the Ṛigvedic poets, though in a lesser degree than the later theologians, at times hint at their ambiguous character. In X.63.4 they are called "the gods who have the magic power of the serpent(s)" 300, in II.27.16, 29.5 201 there is a reference to their snares (pāsāh) and VII.61.5 mentions the "attacks" or "deceits" (druhah) of Varuṇa and Mitra, a notion more familiar in the brāhmaṇas 202. Varuṇa's dhṛti "malice, viciousness" is also referred to 203. That the Ādityas were Devas of a special kind, different from the other groups, has, indeed, never been questioned 204.

The general conclusion must be that the Ādityas were originally Asuras, who had been received (a mythological conception, not a historical process !) in the world of the gods. However, although fully incorporated in the Devas, they still retained some ambiguous traits, being both beneficent and demoniacally destructive. Most illustrative in this respect is the contrast between the two aspects of the dual deity Mitrā-Varuṇā. In the organized cosmos the Ādityas represent the under world aspect. This is

197 BĀU. I.4.12 (=ŚBM. XIV.4.2.24) sa viśam asṛjata, yāny etāni devajātāni gaṇaśa ākhyāyante : Vasavo Rudrā Ādityā Viśve Devā Maruta iti.

198 TS. VII.1.5.1 (after the earth has arisen in the waters) tásyām aśrāmyat Prajāpatih, sá devân asṛjata, Vásūn Rudrân Âdityân. té devâḥ Prajâpatim abruvan.

199 On the Sādhyas see Excursus.

200 deváso . . . dhīmāyāḥ, but the epithet may have been taken from VI.52.15, where it characterizes the Viśve Devāḥ. See n. 205.

201 See the ample discussion by Renou, EVP. 7, p. 58.

202 Cf., e.g., KS. XVIII.19 (264,10) yas te rājan Varuṇa druhah pāśo.

203 E.g., I.128.7, VIII.27.15 and cf. Geldner, Ved. Studien I, p. 142.

204 See Bergaigne, Rel. véd. III, pp. 88ff., 256ff. and passim, B. Geiger, Die Aməša Spəntas, pp. 157, 220ff., etc.

especially true of Varuṇa who in the day-time resides in the cosmic waters beneath the earth, at the lower end of the world axis. In the dualistic structure of the cosmos the Ādityas are the opponents of the gods of the upper world (Vasus, or Vasus and Rudras?). All that can at present be said about their mutual relations, however, has a provisory character. There does not seem to be any open hostility between the Ādityas and the other gods except at the end of the year. It must be assumed that during the critical period of the transition to the New Year the Ādityas again sided with the Asuras, who returned from outside the cosmos to resume their contest with Indra for the supremacy in the world. Only this assumption would seem to make the Rigvedic theology intelligible as a coherent system. If, however, we compare the scarce Rigvedic data about the verbal contest between Varuṇa and Indra and the former's transition to the world of the gods (IV.42, X.124) with the epic tale of the Churning of the Ocean, where Varuna stands apart from (but not above) the contending parties (see p. 108), we must come to the conclusion that even during the annual strife between Asuras and Devas Varuṇa does not take part in the fight. Others have studied his inactive role in connection with the Indian conception of kingship. The conclusions here summarized are strictly limited to the theological system, without reference to the reality of Indian life in which these notions functioned.

Two references in the tenth book to Ādityas born on earth, in heaven and in the waters are here left out of account, because the same is said in earlier books of the Viśve Devāh205, from whom this idea may have been transferred to the Ādityas.

10. ĀDITYAS AND AṄGIRASAS

The relation between the Ādityas and the Aṅgirasas is particularly interesting206. The nature of the latter, too, was ambiguous. As the mythological ancestors of man they were, on the one hand, the first performers of the rites and, as magical singers, they assisted (in the Vala-version of the cosmogonical myth) the god Indra Brhaspati in liberating the cows207. Their role was so important that the god was also

205 Cf. X.65.9 yé pŕthiváso divyáso apsú yé, X.63.2 yé sthā jātā Áditer adbhyā́ pári yé prthivyā́s and, with reference to the Viśve Devāh, VI.52.15 yé ké ca jmā́ mahíno áhimāyā́ divó jajñiré apā́m sadhasthé. Since X.63.4 devā́so ... áhimāyā́h may be an imitation of VI.52.15, the same may be true for stanza 2. Cf. also VI.50.11 dasasyánto divyā́h pŕthivā́so gótā āpā́ mrlātā ca devā́h (Viśve Devāh).

206 See Sylvain Lévi, La doctrine du sacrifice dans les brāhmaṇas, p. 63ff.

207 For the relation of the "singers" to Indra see Hanns-Peter Schmidt, Brhaspati und Indra, p. 238. Indra is aṅgirastama (but this is more often an epithet of Agni and once, in book X, of Soma) and aṅgirasvat (II.11.20, VI.17.6, also once of the Aśvins). Brhaspati is aṅgirasá (RS., GB II.2.14,15, Bauddh. ŚS. XVIII.47: 403,9, 404,3,11) or simply an aṅgiras, e.g. JB. III.188, line 4, where Brhaspati functions as the udgātŕ of the Aṅgirasas ("Die als Opferpriester erforderten sind alle Angirasen", Caland, Jaim. Brāhmaṇa in Auswahl, p. 161 n. 8). This is, no doubt, an exception: the only parallel is KB. XXX.6, where Brhaspati is a brahmán, but the other versions (ŚB. III.5.1.15, AB. VI.34.2, GB. II.6.14 (266,11) only mention

60

said to be an angiras. On the other hand, they were also, as the mythological prototype of man, the first to follow the primeval man Yama to the realm of the dead, which accounts for their close connection with Yama and the Fathers. Their character was, accordingly, intrinsically ambiguous, beneficent and inauspicious. As benefactors they are sometimes, even in the Rigveda208, treated as gods209. As for their connection with the world of the dead cf. VS. XXXVIII.9 "(I offer) thee to Yama, who is accompanied by the Angirasas and the Pitaras'210, and especially in the funeral rites, cf. RS. X.14.3-6 (3) "Mātalī with the kavyas, Yama with the Angirasas, Brhaspati, growing, with the singers, revel, some svadhā, others svadhāyā (in the way corresponding to the funeral ceremonies)", (4ab) "Sit down on this prastarā, O Yama, uniting with the Angirasas, the Pitaras . . .", (5ab) "Come thou here, with the Angirasas who share in the sacrifice, O Yamā, revel with the descendants of the Virūpas . . ." (6ab) "The Angirasas, our Fathers, the Navagvas, the Atharvans, the soma-spending Brgus, [may we be in their favour] . . .211.

The Angirasas are called virūpāsah212, which has been interpreted as referring to their ambiguous nature, since they belong to the opposed parts of the cosmos. The vairūpās, "sons of the virūpa(s)", were probably not different from the Angirasas213.

It thus appears that in many passages where the Angirasas are not

Agni as the hotr. As for the Jaiminīya Brāhmaṇa itself, in I.125 Brhaspati acts as the puroḥita of the Devas, as against Uśanā Kāvya, who is the puroḥita of the Asuras. In the epic, again, Brhaspati is Angiras muni (I.7.1.6), Angirasāṃ varisṭha (V.16.27), but also an Āditya (I.60.38 Brhaspatiś ca bhagavān Ādityeṣv eva ganyate).

208 Cf. VII.44.4, where they take the place of the Rudras (Ādityébhir Vāsubhír Angirobhib) and VII.10.4, where Brhaspátim ṛkvabhir viśvāvāram is mentioned by the side of Indra with the Vasus, etc. Otto Strauss, Brhaspati im Veda, p. 40, characterized them as "die priesterlichen Heroen". They are yajñíya (X.14.5).

209 Later they are sometimes actually called Angiraso devāḥ, e.g., AB. VIII.14.4 Marutaś ca tvā 'ngirasaś ca devāḥ, TS. VII.1.18 Angirobhir devébhir devátayā.

210 Yamáya tvā 'ngirasvate pītṛmáte svāhā.

211 X.14.3 Mātalī kavayair Yamó Angirobhir Brhaspátir ṛkvabhir vāṛdhānāḥ, . . . (4) imáṃ Yama prastarám ā hí sāda 'ngirobhibh pítṛbhis sadhamádāḥ, . . . (5) Angirobhir návagvā Atharvāṇo Brhágavaḥ somyāsaḥ . . . See H.-P. Schmidt, op. c., p. 56f.

212 RS. X.62.5-6. See Schmidt, p. 40 and IIJ. 13, p. 282. Macdonell, Ved. Myth. 143, it is true, takes virūpa as a proper name; similarly Geldner in his translation of X.62.5-6, 78.5 (as against III.53.7).

213 Whereas in VI.10.4 the singers (ṛkvan), who must be identical with the Angirasas, are the companions of Brhaspati, they have here been doubled because the Angirasas occur under their proper name as companions of Yama. This is a case where the question arises what influence the poetical technique of the Rigvedic poets may have had on their theological phrasing. The poet who composed the pāda X.14.3b Brhaspátir ṛkvabhir vāṛdhānāḥ may well have had in mind VII.10.4d Brhaspátim ṛkvabhir viśvāvāram. In that case (which cannot be proved to be true since no other possible reminiscences of VII.10 can be pointed out) the main problem would be why the poet introduced Brhaspati (see Hanns-Peter Schmidt, l.c.). The doubling itself may either be due to purely technical considerations of versification or to a theological necessity of mentioning both Yama and Brhaspati.

61

fulfilling their cosmogonial task of opening the Vala, they are distinctly inauspicious. Their name is sometimes explained from their birth from an añga (member) of Death (Varuṇa Mrtyu !)214, which reminds us of their Rigvedic designation as sons of the (primordial father) Asura215. An aetiological myth related of a pact which they concluded with the Fathers216. These unmistakable connections with the world of the dead cannot but have affected the image that Vedic Indians formed of the Añgirasas and must have contributed to “die niedrige Bewertung von Atharvan und Añgiras”217. The same is true of Yama, the mythical first man. From AS. XIX.56.1, which says that Sleep originated in Yama's realm and was formed “in the womb of the Asura” (ásurasya yonau), von Bradke, p. 101, had already concluded that Yama had Asuric traits of character218. He had also rightly pointed to TS. II.1.4.3, where, in a curious contest between the Devas and Yama, the latter apparently stands for the entire group of Asuras219. Would it be right to say as a general rule that all deities who have gone over from the Asuras to the Devas still retain, in spite of their having become yajñíya,

214 GB. I.1.7 tam Varuṇam mrtyum abhyaśrāmyad, abhyatapat, samatapat. tásya śrāntasya taptya samtaptasya sarvebhyo 'ngebhyo raso 'kramat, so 'ngaraso 'bhavat. tam ye 'ngārā āsams, te 'ngiraso 'bhavan. Otherwise AB. III.34.1

ye 'ṅgārā āsam, te 'ṅgirasaṁ bhavantam Aṅgirā ācakṣate parokṣeṇa. Otherwise AB. III.34.1 ye 'ṅgārā āsams, te 'ṅgiraso 'bhavan.

215 Properly speaking RS. III.53.7, X.67.2 divás putráso ásurasya vīrāḥ only implies the existence of a pitṛ ásura (who is here identified with heaven, although there can be no doubt that the old father Asura belonged to the primeval world, see in general Bergaigne III, p. 85, and cf. above, p. 136f. Etudes Asiatiques 25, p. 97f.). As is well known, von Bradke, Dyāus Asura, pp. 44–60, thought that in such passages traces of an old god Dyāus Asura could be found. It is no longer necessary to dwell upon his theory and those of Hillebrandt (Varuṇa und Mitra, p. 108, Ved. Myth. II2, p. 63). The question remains to be answered, however, why the Asura, in whose realm of undifferentiated chaos heaven can only have existed as a potentiality, is here identified with that very heaven. As for the Aṅgirasas, who are said to have been born from the primordial world, a warning against euhemeristical explanations

may not be out of place. It may be correct to define the Aṅgirasas as deified ancestral priests, but this does not mean that they are a reminiscence of older times. They belonged entirely to mythology, as priests who in the beginning had cooperated with the Devas to create the cosmos. To the Indians, who transformed every historical process into a repetition of the mythical pattern, there can hardly have been a contradiction between “die priesterlichen Heroen” (see n. 208) and semigods, who even had their place in the system of classification among the gods. Was there some connection between the fact that they were “sons of the Asura Heaven” and the place in the zenith which is often assigned to them ?216

216 See Lévi, Doctrine, p. 68 and TB. II.1.1.1–2 Aṅgiraso vai sattrám āsata. téṣām Pŕśniḥ gharmadhúg āsīt. sá 'rjiṣéṇā 'jīvat. té 'bruvan, kásmai nú sattrám āsmahe yé 'syò 'sadhir nā janáyāma iti. té divó vrṣṭím asṛjanta. yáxcantaḥ stokā avāpadyanta, távatīr óṣadhayo 'jāyanta, iti. tá jātáḥ pitáro viśenā 'limpan. (2) tásām jagdhvà rúpyanty ait. té 'bruvan, ké idám itthám akar iti. vayám cā bhāgadhéyām icchámānā iti pitáro 'bruvan. For rupyaṇti cf. KS. XXV.4 (107,2) arupyat.

217 Paul Horsch, Die vedische Gāthā- und Śloka-Literatur, p. 238.

218 See above, p. 306 n. 2.

219 TS. II.1.4.3 deváś ca vai yamáś cā 'emiṁ loké 'spardhanta. sá Yamó devānām indriyáṁ vīryàm ayuvata. tád Yamásya yamatvám. 62

a dark, demonic aspect? This is certainly true of Agni and Soma, who

are sometimes called āsurá in the Rigveda and about whose “Varuṇic”

character at certain stages of the ritual (particularly when they are

upanaddha) the texts do not leave any doubt 220. This would also explain

why Svápna, the god of sleep, in spite of his transition from the Asuras

to the Devas (see p. 31) is still considered a harmful, inauspicious

deity 221. As such he can be paralleled with the Aṅgirasas. The question

as to just when the latter were considered ominous and when, apart from

the Vala-myth, they were godlike would require a special study 222.

The preceding remarks were essential for a better insight into one

of the most intriguing characteristics of the Ādityas. Within the limits

of the organized cosmos, that is, in terms of the cyclical cosmic process

as completed within a year, up to the New Year Festival, they are never,

as far as I can see, in conflict with the other groups of gods. On the other

hand, in the brāhmaṇas they are depicted as being traditionally in conflict

with the Aṅgirasas, in spite of the fact that the latter are not really gods

(see n. 208). While both parties have an ambiguous character, the Ādityas

here fight as true Devas 223. Their opponents, the Aṅgirasas, by the very

fact of this opposition, here become representatives of the Asuras. Just

as in the contest between the Devas and Yama mentioned above, the

fight between Ādityas and Aṅgirasas was interpreted, in accordance

with the fixed pattern of the cosmic contest, as a special instance of

the eternal struggle between Devas and Asuras. As a result of this

polarization the Aṅgirasas became simply Asuras. That this interpretation

gives a correct picture of the way the Indian priests looked at their

theological system is shown by the fact that in two brāhmaṇas (SB. and

JB.) the authors, in the middle of the tale, seem for a brief moment to

have forgotten who the real actors are. They substitute deváḥ and ásurāḥ

for the correct terms ādityáḥ and aṅgirasas, which reappear at the end

of the legend. This is, no doubt, primarily a problem of text tradition.

220 For agnír úpanaddhah see, e.g., Hillebrandt, Ved. Myth. II, p. 60 and cf. p. 21,

IIJ.12, p. 281, 15, p. 231f.

221 As for Svapna (Sleep), in spite of his changing sides and joining the gods (see

above, p. 31), the Atharvavedic poets do not rate him very high. Cf. XI.8.19

"Sleep, laziness, destruction, the deities called evils (pāpmán), old age, baldness

and greyness entered, one after the other, the body" (svápno vai tandrír nirṛtíḥ

pāpmáno nā́ma devatā́ḥ, jarā́ khālatyā́ṃ pā́lityā́ṃ śā́rīraṃ ánu prā́viśan).

222 In this connection the circumstance that in the hymns Agni is often referred to

as áṅgiras and even áṅgirastama may be meaningful (cf. also AB. VI.34.2 Aṅgirasā́ṃ

vā eko 'gnị́). Our insight into Vedic theology, however, is still far too imperfect to

explain the exact character of the relation between Agni and the Aṅgirasas.

223 In contrast with the normal practice of the Ādityas of keeping out of the contest

with the Asuras (with whom they remain connected by relationship, see p. 59),

there are some Yajurvedic passages where the Ādityas stand for the Devas as a

group in their fight against the Asuras. Cf. KS. XXVIII.6 (160,6), KKS. XLIV.6

(261,20/306,10) Ādityā́ vā asurā́n hatvā versus an immediately preceding passage

KS. XXIII.8 (83,10) XXVIII.2 (154,1), KKS. XXXVI.5 (191,14/223,5) XLIV.2

(256,15/299,17) devā́ vā asurā́n hatvā and the corresponding passage MS. IV.6.9 (92,14)

devā́ ásurān hatvá, etc.

63

It is possible that in some text two different versions of one and the same tale have been confused. Still, there is apparently more to it, and if there has been a confusion, this was made possible by the fact that the contrast between Ādityas and Angirasas had a structural character. This is shown by the first introductory words in ŚB. XX.2.2.9 "Now, the Ādityas and the Angiras, both of them sprung from Prajāpati 224, were contending together saying . . .". These words, and ŚB. III.5.1.13 "Now, in the beginning there were two kinds of beings here, the Ādityas and the Angirasas" explain what the author of AB. I.16.39 meant with his laconic words "The Ādityas and the Angirasas existed here" 225. From these passages it appears that the contrast between Ādityas and Angirasas was one of the ways in which the fundamental cosmic dualism was expressed. It is clear that this conclusion only reveals a basic pattern but does not explain why the Angirasas were the chosen adversaries of the Ādityas 226. It must also be repeated that, apart from the structural aspects, there is a textual problem that has to be solved by its specific method. In particular the strange contradiction in the version of the Jaiminīya Brāhmaṇa, where the Devas are, on the one hand, represented as identical with the Ādityas 227 and, on the other, as standing above the two parties may be due to a confusion, two different versions having been mixed.

The various passages where the Ādityas are depicted as contending with the Angirasas refer partly to the ritual of the sadyahkrī, when the Soma is bought and offered up on one and the same day 228, and partly

224 See above, n. 23. This discussion is a more elaborate version of some brief remarks made in IIJ. 13, p. 282.

225 ŚB. XII.2.2.9 athā́dityā́ś ca ha vā Angirasas ca, ubhāye prajāpatyā́ aspardhanta, vayā́m pūrve svargā́m lokā́m eṣyāmo vayā́m pūrva iti, III.5. 1.13 dvayyò ha vā idā́m agre prajā́ āsuh, Ādityā́ś caivā́ 'hā 'sann Angirasas ca, AB. I.16.39 Ādityā̄́s caive 'hā 'sann Angirasas ca.

226 The basic pattern could also be elaborated in other ways. Much better known is the contrast between Atharvans and Angirasas, which is a replica, on the level of non-divine groups, of the dualism. For ātharvaṇa "holy" versus āṅgirasa "magical" see Maurice Bloomfield, The Atharvaveda, p. 112, H.-P. Schmidt, Brhaspati und Indra, p. 39.

227 Already pointed out by Caland, Das Jaiminīya-Brāhmaṇa im Auswahl, p. 161 n. 11.

228 Cf. AB. VI.34.2 Ādityā́ś ca ha vā Angirasas ca svarge loke 'spardhanta : vayam pūrva eṣyāmo, vayam iti. te hā 'ṅgirasah pūrve svahsutyā́m svargasyā lokasyā dadrśus, te 'gnim prajighyur — parehy, Ādityebhyāḥ svahsutyā́m svargasyā lokasyā prabhṝṣṭi — . KB. XXX.5(6).2 Ādityā́ś ca ha vā Aṅgirasas cā 'spardhanta: vayam pūrve svargam lokam eṣyāma ity Ādityā, vayam ity Aṅgirasas. te 'ṅgirasa Ādityebhyah(-ān) prajighyuh: svahsutyā no, yājayata na iti. teṣā́m hā 'gnir dūtā āsa, ta Ādityā ūcur : athā 'smākam adyasutyā, teṣā́m nas tvam eva hotā 'si, Brhaspatir brahmā, 'yāsya udgātā, Ghora Āṅgirāso 'dhvaryur iti . . . GB. II.6.14 Ādityā́ś ca ha vā Aṅgirasas ca svarge loke 'spardhanta : vayam pūrva eṣyāmo, vayam pūrva iti. te hā 'ṅgirasah svahsutyā́m dadrśus. te hā 'gnim ūcuh : parehy, Ādityebhyāḥ svahsutyā́m svargasyā lokasyā prabhṝṣṭi . . . JB. III.187f. Ādityā́ś ca vā Aṅgirasas ca svarge loke 'spardhanta. yatare no yātārān yājayiṣyanti, te hāsyanta iti. te 'ṅgirasah pūrve yojñā́m samabharān svahsutyā́m, te 'gnim Aṅgirasam dūtám prāhiṇvann : imā́m nah svahsutyā́m Ādityebhyah prabhṝṣṭi 'ti. sa etyā 'bravid: Ādityā, Aṅgiraso vah

to the Abhiplava ṣadaha ritual229. The confusion is only found in

ŚB. III.5.1.13-22 and in the corresponding passages in JB. III.187-188

and II.115-116, which Caland has put together in his Auswah, pp. 158-161.

These two passages are here quoted in the translations by Eggeling and

Caland respectively. The italics are mine. Of the other passages only the

first lines are quoted in the foot-notes.

Śatapatha Brāhmana III.5.1.13 “Now, in the beginning there were two kinds

of beings here, the Ādityas and the Angiras. The Angiras then were the first

to prepare a sacrifice, and having prepared the sacrifice they said to Agni,

‘Announce thou to the Ādityas this our to-morrow's Soma-feast, saying,

‘Minister ye at this sacrifice of ours!’” (14) The Ādityas spake (to one another),

‘Contrive ye how the Angiras shall minister unto us, and not we unto the

Angiras!’ (15) They said, ‘Verily by nothing but sacrifice is there a way out

of this: let us undertake another Soma-feast!’ They brought together the

(materials for) sacrifice, and having made ready the sacrifice, they said, ‘Agni,

thou hast announced to us a Soma-feast for to-morrow; but we announce to

thee and the Angiras a Soma-feast even for to-day: it is for us that thou art

(to officiate as) Hotṛ!’ (16) They sent back some other (messenger) to the

Angiras . . . (17) The Angiras then officiated for the Ādityas in the sacrifice

with Soma brought (kri) on the same day (sadyas); whence this Sadyahkri.

śvahsutyāṁ prāhur iti . . . (188) . . . te 'bruvann : Agne, cirāṁ tad yac chva, imāṁ

eva vayāṁ tubhyam adyasutyāṁ prabruimas, teṣāṁ nas tvāṁ hotā 'si, Gaur Angiraso

'dhvaryur, Bhaspatir udgātā, 'yāsyo brahme 'ti sarvān avṛṇata . . . PB. XVI.12.1

Ādityāś cā 'ngirasaś cā 'dīkṣanta. te svarge loke 'spardhanta. te 'ngirasa ha vā idāṁ agre praiṣ āsṛh,

śvahsutyāṁ prābruvan . . . ŚB. III.5.1. 13ff. dvayyò ha vā idāṁ agre praiṣ āsṛh,

Ādityāś caivā 'ngirasaś ca, tátó 'ngirasaḥ pūrve yajñāṁ sám abharaiṁs. té yajñāṁ

sambhṛtyo 'cur Agním: imāṁ naḥ śvahsutyāṁ Ādityebhyah prá brūhy, anéna no

yajñéna yājayaté 'ti . . . BaudhŚS. XVIII.22 (369,11) athā 'dityāś ca ha vā Angirasaś

ca svarge loke pasprdhire. te 'ngirasaḥ te 'ngirasaḥ te 'ngirasaḥ te 'ngirasaḥ

ca svarge loke pasprdhire. te 'ngirasaḥ te 'ngirasaḥ te 'ngirasaḥ te 'ngirasaḥ

śvahsutyā na ity, athā 'smākam adyasutyā 'ti. (23) Ādityā procuḥ: teṣāṁ nas tvāṁ

hote 'ti.

229 Cf. AB. IV.17.5 Ādityāś ca ha vā Angirasaś ca svarge loke 'spardhanta: vayāṁ

pūrva eṣyāmo, vayāṁ iti. te hā 'dityāḥ pūrve svargaṁ lokaṁ jagmuh, paśce 'vā

'ngirasaḥ, saśtyāṁś vā vārṣeṣu . . . GB. I.4.23 Ādityāś ca ha vā Angirasaś ca svarge

loke 'spardhanta: vayāṁ pūrve svar eṣyāmo, vayāṁ pūrva iti . . . ŚB. XII.2.2.9

áthā 'dityāś ca ha vā Angirasas ca, ubhāye Prajāpatyā, aspardhanta: vayāṁ pūrve

svargāṁ lokāṁ eṣyāmo, vayāṁ pūrva iti . . . See also MS. III.4.2 (46,10) dvyuttarénā

vai stómenā 'dityāḥ svargāṁ lokāṁ āyamis, catturuttarénā 'ngiraso (as must be read

for stómenádityāḥ and aṅgirasau in the editions from Schroeder and Aundh) and

KS. IX.16 (119,10,11) Angirasas ca vā Ādityāś ca svarge loké 'spardhanta. tá Ādityā

etā́ṁ pañcahotrāṁ cakā́ṁs, tī́ṁ mānusā́ 'nūddṛtyā 'yuhams, tátā Ādityò svargā́ṁ

lokā́ṁ āyan, ápā 'ngiraso 'bhram்santa (as against IX.14: 116,19 pañcahotrā vai

devā asurāṁ prāṇudanta, MS. I.9.6: 138,3 pañcahotrā vai devā́ asurāṁ párābhāvayan,

. . . 138,6 pañcahotrā vai devā́h svār áyan, I.9.5: 136,17 pañcahotrā vai devā́h paścā́n

ásṛjanta, párā 'surān abhāvayan), TB. II.2.3.5f. Ādityāś ca 'ngirasaś ca svarge lokè

'spardhanta: vayā́ṁ pūrve suvargā́ṁ lokā́ṁ iyāma, vayā́ṁ párva iti (6) tá Ādityā́

etā́ṁ pañcahotrāṁ apasyan . . . tátó vai té pūrve suvargā́ṁ lokā́ṁ āyan, Cf. KS. XX.11

(31,19), KKS. XXXI.13 (161,22/188,20) ādityadhāmāno vā anye prāṇā, aṅgiro-

dhāmāno 'nye. ye prastā́t te ādityadhāmāno, ye paścā́t te 'ngirodhāmānaḥ, MS. III.2.9

(30,8) ādityādhāmāno vā úttare prāṇā́, aṅgirodhāmāno 'dhare, and GB. I.1.7

(Aṅgirasas from Varuṇa Mṛtyu).

(18) They brought Vāc (speech) to them for their sacrificial fee. They accepted her not . . . (21) Now Vāc was angry with them: 'In what respect, forsooth, is that one better than I—wherefore is it, that they should have accepted him and not me?' So saying she went away from them. Having become a lioness she went on seizing upon (everything) between those two contending parties, the gods and the Asuras. The gods called her to them, and so did the Asuras. Agni was the messenger of the gods, and one Saharakshas for the Asura-Rakshas. (22) Being willing to go over to the gods, she said, 'What would be mine, if I were to come over to you?' . . . So she went over to the gods".

Jaim. Brāhmaṇa III.1872 "Die Ādityas und Angirasen stritten sich um den (Besitz des) Himmelsraumes. Sie überlegten: "Welche von uns beiden für die anderen als Opferpriester auftreten werden, die werden (auf der Erde) zurückbleiben". Die Angirasen bereiteten (zur Erreichung des Himmels) zuerst ein Opfer vor, nl. eine sich über einen Tag erstreckende Soma-feier, welcher ein für die Weihe und den Somakauf bestimmter Tag vorangeht (eine śvahsutyā). Sie sandten den zu ihnen gehörigen Agni als Boten ab: "Sage diese unsere Śvahsutyā den Ādityas an". Der ging hin und sprach zu den Ādityas: "Die Angirasen sagen euch eine Śvahsutyā an" . . . (III.188) "Da bereiteten die Ādityas diese Soma-feier vor, bei welcher an demselben Tage der Soma gekauft (und dargeopfert) wird (eine adyasutyā). Da sagten sie: "Agni, bis morgen ist lang; hier sagen wir die Adyasutyā an. Du sollst unser Hotṛ, der Angirase Go unser Adhvaryu, Brhaspati unser Udgātr und Ayāsya unser Brahman sein" . . . (II.115) "Die (Angirasen, von den Ādityas als Opferpriester) erwählt, entzogen sich nicht. Sie gingen hin und verrichteten das Opfer für sie. Als Opferlohn führten sie (nl. die Ādityas) ihnen (den Angirasen) die Vāc in der Gestalt einer weissen mit einem Zügel angebundenen Stute herbei: 'Nehmt diese entgegen'. Sie sagten: "Diese ist stärker als wir, wir werden sie nicht aufheben". Diese (nl. Vāc), erzürnt auf Erwägung: "Sie haben mich nicht entgegengenommen", nahm die Gestalt einer Löwin230 mit zwei Mäulern (einem auf beiden Seiten) an, stieg (in die Luft) empor, und stellte sich zwischen Göttern und Asuras (d.h. zwischen Ādityas und Angirasen), angreifend wen sie von den Göttern und Asuras ergreifen konnte. Sowohl die Götter als die Asuras rief[en] sie zu sich heran231. Sich zu den Göttern hinwendend sprach sie . . . (II.116) "Nun führten die Ādityas ihnen (den Angirasen) die Sonne in der Gestalt eines weissen mit einem Zügel angebundenen Rosses herbei: "Nimmt diesen entgegen". Sie sagten: "Dieser aber ist stärker als wir; wir werden auch diesen nicht aufheben". Da kamen die Götter und Seher zusammen herbei und sagten: "Opfer ohne Opferlohn werden zunichte gehen. Nehmt (ihn) an".

230 For this lioness see below, n. 421.

231 For the text see n. 421. (For tám upaiva devā jāmantrāyanto 'pā 'surāḥ (Caland, Raghu Vira-Lokesh Chandra) the MSS. read amantrayantovāsurāḥ (ka), amantrayanto-pāsurāḥ (kā, gā). The reading āmantrayanta seems to be an emendation but is wrong: Viśvabandhu Śāstrī gives for the brāhmaṇas (I, p. 278) some 25 occurrences of upa-mantrāyai, but only this passage for upa-ā-mantrāyati (I, p. 289). As for the Samhitās (Viśvabandhu, p. 954), it has long been noticed that KS. XXVII.8 (147,1) tā upāmantrayata abruvatām is corrupt for tā upāmantrayatā. tāv abruvatām. Cf. Caland, WZKM. 26 (1912), p. 125, Raghu Vira, Kapiṣṭhala-Kaṭha Samhitā, Introduction, p. 30 (2p. XXVIII), who however read tā for tāv, and TS. VI.4.10.1 té deváh sāndāmárkāv upāmantrayanta. tāv abrūtām . . . So there remains ASPaipp. XVII.28.5 athe'n drāgní tam upāmantrayante, which L. C. Barret emends to upāmantrayete (registered by Viśvabandhu) but for which the context rather requires an imperfect. See further n. 418.

11. VARUṆA AS A DEMONIACAL FIGURE AND AS THE GOD OF DEATH

In the context of this study of Varuṇa's relation to Asuras it is necessary to say a few words on the demoniacal aspect of Varuṇa in the Veda and the epic. Since much has been written about it232 and the later Vedic texts are explicit, there is no need for a circumstantial exposition of facts which are generally known.

When, however, it comes to interpreting the evidence one is again faced with the well-known basic difficulty that the main part of the Rigveda differs from book X and the Atharvaveda in that it contains but few references to the dark aspect.

First of all, it is necessary to stress the fact that Varuṇa was associated, not with Ṛta (the cosmic order) only, but with the pair of opposites Ṛta and Druh. The last word denotes all that is not in harmony with the divine laws of the Universe. Such translations as "malice" or "falsehood" evoke too exclusively the idea of ethical relations between men. To avoid this, it will here be translated by "wrongness", which, although admittedly not very satisfactory, has the advantage of not directly evoking ethical associations. This pair of notions originated in the Proto-Indo-Iranian religion233. Varuṇa's relations to both entities (an inheritance, it seems, from the undifferentiated primordial world) constitute his basically ambiguous character. On the one hand, it is true, Varuṇa and Mitra are guardians of the Ṛta, they are rtāvr̥dhā (VII.66.19) and are said to be free from drúh234. On the other hand, Varuṇa not only discovers every act against the Ṛta235 but he is also invoked to

232 See, e.g., (1877) Hillebrandt, Varuṇa und Mitra, pp. 53–63 "Varuṇa's Fesseln als Ausdrück für die Nacht (Ungemach, Tod)", pp. 63–70 "Varuṇa's Fessel ein Ausdrück für die Wassersucht"; (1883) A. Bergaigne, Religion védique III, pp. 84 ("caractère équivoque"), 115, 128–129 ("caractère quasi démoniaque"), 139, 147–148, 156; (1893) K. Bohnenberger, Der altindische Gott Varuṇa nach den Liedern des Ṛgveda, pp. 56–61, E. Hardy, Die Vedisch-brahmanische Periode der Religion des alten Indiens, p. 58; (1897), A. A. Macdonell, Vedic Mythology, p. 27; (1898) S. Lévi, La doctrine du sacrifice dans les brâhmaṇas, pp. 96, 152–171; (1899) H. Oldenberg, Aus Indien und Iran, p. 179; (1905) Otto Strauss, Br̥haspati im Veda, p. 19f. ("er wird als eine zu überwindende dämonische Kraft betrachtet"); (1916) B. Geiger, Die Ameṣ̌a Spəntas (Sitzungsber. Kais. Akad. Wiss. Wien, 176 Bd, 7. Abh.), p. 178; (1917) K. F. Johansson, Über die altindische Göttin Dhiṣaṇā und Verwandtes, Upsala, p. 128ff.; (1923) H. Oldenberg, Religion des Veda, 3. und 4. Aufl., pp. 183, 293, 299–304; (1929) A. Hillebrandt, Ved. Mythologie II2, pp. 20, 22, 31f., 40, 127 ("mit der doch im ganzen unheilvollen Natur Varuṇa's"), 420 n. 1, Festgabe Garbe, p. 19f.; (1937), J. J. Meyer, Trilogie altindischer Mächte und Feste der Vegetation III, pp. 204ff., 209, 214, 265; (1951), H. Lüders, Varuṇa, p. 676; (1960) J. Gonda, Die Religionen Indiens I, p. 76f., Renou, EVP. 7, pp. 4 ("un artisan du mal"), 20, 58, Festgabe H. Lommel [=Paideuma VII, 4–6], pp. 122–128, especially p. 124; cf. IIJ. 5 (1961), pp. 51–53; 8 (1965), p. 108 n. 63; (1972), J. Gonda, The Vedic god Mitra, pp. 40, 75.

233 B. Geiger, Die Ameṣ̌a Spəntas, p. 178ff.

234 RS. V.68.4, VII.66.18 adruhā, V.70.2 adruhvanā.

235 AS. I.10.2 viśvan hy ugra nicikēṣi drugdhám.

deliver from his noose 236 , which is an indirect indication of his power over druh 237 . Most earlier studies on Varuṇa, indeed, omit to mention that Varuṇa's dark aspect is identical with the antithesis of Ṛta, namely Druh. It has been argued elsewhere that the other aspect, the one which delivers from evil, is identical with Mitra, when and in so far as there is a reference to the dual godhead Mitrá-Varuṇā 238 , the "dieu unique à deux noms" 239 .

An indication that there has, indeed, been a euphemistic reticence with regard to Varuṇa is the occurrence of taboo substitutes for his name. It has been suggested that the Old Iranian name Ahura Mazdā may have been such a substitute for the name Varuṇa 240 , which is attested as early as the treaty of the Mitanni king Šattiwaša of c. 1380 B.C. As B. Geiger has pointed out 241 , similar indirect denotations are found in the Rigveda (VIII.41.1 ásuro viśvávedāḥ, I.24.14 asura pracetāḥ). It might be objected that the Rigvedic terms are not specific (as Geiger himself remarks), prá cetas being much more commonly used as an epithet of Agni, but in the Yajurveda the words prá cetās tvā Rudrāḥ paścát pātu can hardly refer to any other god but Varuṇa (as it is, indeed, interpreted by Sāyaṇa, Uvaṭa and Mahīdhara), and in the classical literature Pracetas is, from Kālidāsa and Amara onwards, used as a name of Varuṇa 242 . Cf., e.g., Bāṇa, Kādambarī, p. 217, l. 2 Peterson (as against pp. 37,13, 79,21, 119,8, 123,15, where the common name is used).

The question as to why the Rigvedic poets seldom refer to this dark aspect has rarely been given the attention it obviously deserves. It has often struck scholars that Death, for instance, is rarely mentioned in the old family collections 243 . It may be considered significant that in the

236 In support of this thesis may be quoted, e.g., TS. II.3.13.2 "Varuṇa seizes him with the Varuṇa-noose" (Váruṇa enaṁ varuṇapāśéna gṛhṇāti) versus "Varuṇa frees him from the Varuṇa-noose" (e.g. II.3.12.2 sá evaí 'nam varuṇapāśán muñcati. See further below, n. 252, etc.

237 RS. VII.86.5 áva drugdháni pītryā srjā nó 'va yá vayám cakrmá tanúbhih, I.24.14 rá jann émási sísrathah kṛtáni, etc.

238 See IIJ. 5, p. 46ff. and Bergaigne III, p. 128f., who opposes his dark aspect as belonging to the natural order to his lighter one, as connected with the religious order, which is an illegitimate intrusion of Roman-Catholic theology into the domain of Vedic religion. Nevertheless Bergaigne has, more sharply than most other scholars, recognized the basic problem.

239 Renou, EVP. 7, p. 3. Similarly Johansson, Die altindische Göttin Dhiṣaṇā und Verwandtes, p. 134.

240 See IIJ. 5, p. 55, 8, p. 109 n. 68. Others may already have expressed the same opinion previously but, if so, they have escaped my notice.

241 Die Aṃśa Spentas, p. 213.

242 For the Yajurvedic formula see KS. II.9 (14,20), KKS. II.3 (16,18/19,22), TS. I.2.12.2, VS. V.11, ŚB. III.5.2.5, cf. TĀ. I.20.1 and TS. VI.2.7.5. For the classical literature see PW., and cf. Bāṇa, Kādambarī p. 217, 2 Peterson (as against pp. 37,13, 79,21, 119,8 and 123,15, where the common name is used). See for the mythological background IIJ.18, p. 35.

243 Macdonell, Ved. Mythology, p. 165 "In the Vedic hymns there is little reference to death".

tenth book of the Rigveda there are fifteen occurrences of the word mrtyú,

whereas in the other books it does not occur at all, except in one of the

latest interpolations inserted after the composition of the Padapāṭha

(VII.59.12 mrtyór mukṣiya, see Oldenberg, Prolegomena, p. 511). Saying

that "the thoughts of the poets of the RV., intent on the happiness of

this earth, appear to have rarely dwelt on the joys of the next life" 244

does not provide an explanation for this fact, nor can the characterization

of their spirit as diesseitig or positive 245 be regarded as such. One is

driven to the conclusion that there was an intentional euphemistic

reticence 246. The only explanation so far proposed for this reticence is

the theory that the Rigvedic hymns differed from those of the later

Samhitās in that they had been composed for a specific seasonal festival,

during which Varuṇa was particularly dreaded as he had probably again

become an Asura for a short while 247.

It will be sufficient to summarize a few details. First, there is, in spite

of Oldenberg's rejection 248, an undeniable relation between Varuṇa and

Vr̥tra 249. Second, there is an unmistakable relation between Varuṇa and

244 Op. c., p. 169.

245 So recently Ogibenin, see the next note.

246 Cf. Bergaigne III, pp. 76 ("atténuation de l'idée primitive d'une défaite"), 115

(Varuṇa is "une forme de Vr̥tra élevée à la dignité divine"), 144 ("allusion plus

ou moins vague à certains traits de ressemblance entre le personnage divin de

Varuṇa et le personnage démoniaque de Vritra"), 143 ("atténuations"), S. Lévi,

Doctrine du sacrifice, p. 167: "Le caractère de Varuṇa ne doit donc pas faire illusion;

si parfois il apparaît comme le gardien de la morale, c'est simplement en vertu

de sa nature rituelle. Il est . . . divinité chatouilleuse, mal commode, peu maniable,

prompte à se retourner contre qui la manie, dangereuse à heurter, dangereuse à

invoquer. Un nom si périlleux, susceptible par sa seule énergie de déchaîner tant

de maux, appelait un palliatif. Comme le farouche Rudra, évocateur des larmes,

a reçu des noms de propitiation et de paix pour annuler l'effet de ses noms sauvages

et violents et s'est dédoublé en Rudra-Çiva, Varuṇa a été associé à Mitra, divinité

très vague . . .", J. J. Meyer, Trilogie III, p. 254 n., Kuiper, IIJ. 5, p. 52 ("the

euphemistic reticence of India's oldest text"), B. L. Ogibenin, Struktura mifo-

logičeskix tekstov "Rigvedy" (Moskva, Akad. Nauk, 1968), p. 77 n. 1 ('sledvet

ukazat', čto vedijskie teksty otražajut liš' položitel'nye (pravye) členy protivo-

postavlenij; strogo govorja, to, čto oboznačajutsja otricatel'nymi (levymi) členami,

ne predstavleno v 'Rigvede'—eto svjazano s odnostoronne pozitivnym xarakterom

pamjatnika." That the character of the text may seem different from the other

Vedas is not so much due to its different theological position as to a taboo caused

by the period of contest. A similar euphemism can be found in the reticence of

the Rigvedic poets with regard to the slaying of Soma; see H. Lommel, Symbolon 4

p. 158f.

247 Cf. IIJ. 5, p. 53 and above, pp. 23, 41f.

248 Oldenberg, Textkritische und exegetische Noten II, p. 343.

249 See above, n. 246 and Bergaigne II, p. 202, III, pp. 115, 128, 144, 148, S. Lévi,

Doctrine du sacrifice, p. 168, Geldner, Vedische Studien II, p. 292ff., Übersetzung III,

p. 353, Bloomfield, The Atharvaveda, p. 107, SBE. 42, p. 370, Keith, Religion

and Philosophy, p. 100, Charpentier, Die Suparnasage, p. 118, Hillebrandt, Ved.

Myth. II2, pp. 33, 40, Renou, Festgabe Lommel, p. 125 n. 12, EVP. 7, p. 82, Eliade,

Eranos Jahrbuch 27 (1958), p. 208ff., Traité d'histoire des religions p. 365f., Images

et symboles p. 128f., Agrawala, Indological Studies-W. Norman Brown, p. 1 and

cf. IIJ. 5, p. 52 n. 12.

69

drúh, as pointed out above. Renou aptly remarks: “Vr [= Varuṇa] est celui qui délivre et en même temps ce dont on se délivre, c'est un personnage ambivalent: les drúh n'étaient-elles pas à son service RV. 7.61,5 (et au service de Mitra)?”250. Later Vedic texts were more outspoken about this aspect (which is only vaguely indicated in the Rigveda, cf. II.27.3, where the Ādityas are characterized as dipsantaḥ “intending to injure” and v. 16 yád vo māyád abhidruhé). They did not tend to conceal euphemistically the god's real character, as was done in the Rigveda, because as theological treatises they were less dangerous than the mantras with their “magical” potency. There is in the brāhmaṇas a constant association of the notions of noose (pāśa) and drúh. Vedic man prayed to be delivered from Varuṇa's noose, or from “wrongness” and the noose, or from the noose of wrongness. Cf. “With the words 'I have been delivered from Varuṇa's noose' he is delivered from the Varuṇa-noose”251, “With the words 'May I be delivered from wrongness, from Varuṇa's noose' he delivers him from wrongness, from the Varuṇa-noose”252, “That snare of wrongness of thine, O king Varuṇa, that, consisting of the metre Gāyatrī, has entered the earth and has the brahman for its support, that of thine I hereby avert by sacrifice, svāhā to it!”253 The phrase “snares of wrongness” occurs once in the Rigveda: “The furious man, O Maruts, who is fain to kill us, even when we do not expect it, O Vasus, may he put on himself the snares of wrongness”254. Cf. also in the Atharvaveda: “Thus I deliver thee from the afterbirth, Nirṛti, the curse that comes from thy kin, from wrongness, from Varuṇa's noose”255 and “the fetters of wrongness that does not release”256.

250 Renou, Festgabe Lommel, p. 124. Cf. also EVP. 7 (1960), p. 58: “Il semble bien qu'on ne puisse échapper à la notion de drúh qui sont au service de VrMi . . . on en arrive à associer Vr à l'idée du mal AV. 1.10,3 . . .” (where, again, the evolutionistic scheme emerges).

251 MS. III.9.1 (113,17) etc. nír Várunasyā pāśād amukṣí 'ti varuṇapāśād evá nirmucyata titi.

252 KS. XXVI.2 (124,4), KKS. XL.5 (230,5/269,14) nír druho nír Varuṇasya pāśān mukṣíye 'ti drúha evái 'namí varuṇapāśān muñcati. Cf. KS. III.1 (23,8), KKS. II.8 (19,13/23,9), the Vedic Concordance under nír varuṇasyā and ŚB. V.2.5.16 átha yád vāruṇó yavamáyas carúr bhávati, tát sárvasmād evátād varuṇapāśād sárvasmād varuṇyàt prajáy prá muñcati tā́ asyā́ 'nāmivá akilbiṣáh prajáy prá jāyante . . .

253 KS. XVII.19 (264,10) yas te, rājan Varuṇa, druháh pāso gāyatrachandáh prthivím anvāviśa brahmaṇí pratitiṣṭhati, tám ta etád vayaje, tasmái svāhā, etc. Cf. XII.6 (168,19) yas te, rājan Varuṇa, gāyatrachandáh pāśas ta(m) ta etád avayaje tasmai svāhā, etc., MS. II.3.3 (30,10) yás te, rājan Varuṇa, gāyatráchandáh pāso bráhman prátiṣṭhitas tám ta eténa 'vayaje, Vedic Concordance under yas te rājan.

254 RS. VII.59.8 yó no Maruto abhí durhrṇāyús tirás cittāní Vasavo jíghāṃsati, druháh pāśān práti sā mucista . . . Geldner translates: “so soll er sich die Schlinge des Trugs (selbst) umlegen”, Keith (transl. of TS. IV.3.13.4) “in the noose [pāśam] of destruction may he be caught.”

255 AS. II.10.2-8 evá 'hám tvám kṣetriyán nirṛtyā jāmísaṃsád druhó muñcāmi Várunasyā pāśāt. Cf. TB. II.5.6.1 and Heesterman, The Ancient Indian Royal Consecration, p. 18.

256 AS. XVI.6.10 ámucyā druháh pāśān.

In general the Atharvaveda is much less reticent about the inauspicious character of Varuṇa than the Rigveda and we find there the first passages which openly state that man has to be delivered from Varuṇa: "As thou hast spoken untruth with thy tongue, much wrong, I release thee from king Varuṇa whose laws do not fail" 257. In the Rigveda similar passages are only met with in the tenth book, cf. "May they (viz. the plants) deliver me from the (evil resulting from a) curse and from what comes from Varuṇa, and also from Yama's fetter and from every sin against the gods" 258. This so-called "emergence" of Varuṇa's krūra aspect has often been explained in terms of a historical evolution and as the result of a biased deformation of the god. As Renou put it in one of the most recent discussions of the problem: "Dans le domaine de la description mythique, l'AV. est un témoin peut-être moins sincère que le Rigveda: les faits sont biaisés selon une orientation tantôt magique, tantôt spéculative... ." 259. It is, however, rather the Rigvedic evidence that may be called biased. See n. 82, etc.

It need hardly be said that whatever is well-offered is seized by Mitra but that the ill-offered part of the sacrifice is seized by Varuṇa 260. Varuṇa's krūra character is most clearly shown by the fact that he is identical with Death. The Gopatha Brāhmaṇa even uses the term Varuṇa Mrtyu 261. In the funeral hymns of the Rigveda it is said that the deceased will see king Varuṇa and Yama in the next world 262 and in one of the Vasiṣṭha hymns, which testify to a close intimacy with Varuṇa, the poet prays the gods that he may not have to go into the "clay house" 263. In spite of Geldner's different interpretation 264 these words probably refer to the grave 265.

257 AS. I.10.3 yád vákthā 'nrtaṃ jihvāyā vrjinám bahú, rájñas tvā satyádharmāṇo muñcāmi Várunād ahám. Cf. verse 1, where "this man" is said to be led "upwards" (ud-nī-) from the wrath of the mighty one, and verse 2, where homage is paid to Varuṇa's wrath. In ud-nī- an indication may be seen of Varuṇa's abode being in the nether world; cf. India Maior (Congratulatory Vol. J. Gonda), p. 147.

258 RS. X.97.16 muñcántu mā sapathyād átho varuṇyād utá, átho Yamásya pádbīśāt sárvasmād devakīlbiṣāt. For varunyà see Eggeling, SBE. 41, p. 57 n. 2.

259 Festgabe für Herman Lommel (1960) [=Paideuma VII, 4-6], p. 122.

260 ŚB. IV.5.1.6 yád vā íjānásya sviṣṭám bhávati, Mitró 'sya tád grhṇāti, yád v asya dúriṣṭaṃ bhávati, Váruṇo 'sya tád grhṇāti, TB. I.2.5.3 Mitréṇa 'vá yajñásya duriṣṭaṃ, PB. XIII.2.4, XV.1.3; 2.4; 7.7 yad vai yajñásya duriṣṭaṃ, tad Varuṇo grhṇāti, tad eva tad avayajati. See below, p. 207f. For śamayanti Varuṇena dúriṣṭam, cf. KS. XXVII.4 (142,12), KKS. XLII.4 (250,18/292,18) tá imáh prajá Mitréṇa śántā, Varuṇena vidhrtáh ("restrained"). Therefore, AB. III.38.3, VII.5.4 Viṣṇur vai yajñásya duriṣṭam páti, Varuṇah sviṣṭaṃ, tayor ubhayor eva śántyai cannot be explained from what we know of Vedic mythological thought.

261 GB. I.1.7.

262 RS. X.14.7 préhi préhi pathíbhiḥ pūrvyébhir yātrā naḥ párve pitáraḥ pareyúḥ, ubhá rá̄janā svadháyā mádantā Yamáṃ paśyāsi Várunam ca devám.

263 RS. VII.89.1 mó ṣú Varuṇa mrṇmáyam grhám rājann ahám gamam.

264 Geldner, Kommentar, p. 115: "die irdeme Gruft für die Gebeine, das Beinhaus"; Übersetzung: "die Urne, in der die Gebeine beigesetzt wurden."

265 Zimmer, Altindisches Leben, p. 407, pointed to AS. XVIII.2.50 mātá putrám yáthā sicó 'bhy ènam bhūma ūrṇuhi, 51 jáyā pátiṃ iva vā́saṁ 'bhy ènaṃ bhūma ūrṇuhi,

71

Varuna's connection with Death is most explicitly stated in the practice of black magic (abhīcāra): "When practising witchcraft one should sacrifice a black ram, which belongs to Varuna. For Varuna is Death. So he causes Death to seize upon him. Verily, this [ram] is symbolic of pāpman. Pāpman is black, as it were. In that it [the ram] is black, he endows him [the victim] with pāpman, saying: "I tie this victim for King Varuna, a bull as a share for Indra, for it is his alone. May the gods enter his limbs, may Yama, may Destruction seize our enemies! These deities, verily, have power over man. Having bought him off from these - Yama is Death - he causes Death to seize upon him"266. The identification of Varuna with pāpmán (also MS. II.5.6: 55,3) should be noted in the light of the same identification of Nirrti (Destruction) in MS. II.5.5 (54,7): tád enam Nirr̥tir pāpmá 'grhṇāt. As is well-known, the black colour is generally associated with Varuna, e.g., ŚB. V.2.5.17 "A black cloth is [the fee] for the Vāruṇa [oblation], for what is black belongs to Varuna"267.

52 abhí tvo 'rṇomi prthivya máhīr vásrena bhadráyā, and to V.30.14 (má nú gān) má nú bhúmigrho bhuvat "Möge er nicht Bewohner des Hauses von Erde werden." Similarly Hillebrandt, Lieder des R̥gveda, p. 79 "Haus aus Lehm" and Renou, EVP. 5, p. 72; 7, p. 27. But see Caland, Die altind. Todten- und Bestattungs- gebräuche p. 166

266 Most explicit is KS. XIII.2 (181,7) vāruṇam kṛsnám vṛ́sim abhicaran álabheta, mrtyur vai Varuno. mrtyunai 'vai 'nam grāhayaty. etad vai pāpmano rūpám. kṛsná iva pāpmā. yat kṛsnah, pāpmanai 'vai 'nam abhiṣ̣uvati "paśúm badhnāmi Varunāya rājñe Indrāya bhāgám rśabham kevalo hi, gātrāṇi devā abhiṣ̣avikantu Yamo grhṇātu nirrtis sapatnān" ity etā vai devatāh puruṣasya 'sute, tābhyo evai 'nam adhi niskriyā (adhiniskrīya von Schroeder and ed. Aundh but cf. TS. III.4.3.1)- mrtyur Yamo- mrtyunai 'vai 'nam grāhayati. Cf. MS. II.5.6 (55,6) vāruṇám kṛsnám pétram stṛnute. kṛsnó bhavati. tāmo vai kṛsnám. mrtyunai 'vai 'nam grāhayaty etád vai pāpmáno rūpám yát kṛsnám. kṛsná iva hí pāpmā. pāpmanai 'vai 'nam abhiṣ̣uvati. tám íyunjíyat "paśúm badhnāmi..." ity etábhya evai 'nam devatábhyo niryácya-mrtyúr vai Yámo-mrtyúnai 'vai 'nam grāhayati ..., TS. II.1.8.2 Vāruṇenai 'vá bhrātr̥vyam grāhayitvá...

267 ŚB. V.2.5.17 támad dhí vārṇám yát kṛsnám. yádi kṛsnám ná vindéd, ápi yád evá kím ca vāsah syāt. For black sacrificial animals offered to Varuna, and for the black colour connected with Varuna, see Hillebrandt, Ved. Myth. II2, p. 21f., J. J. Meyer, Trilogie II, p. 43, III, pp. 208f, 230, 266. On p. 325 Meyer points out that at the Vājapeya sacrifice black and white animals were offered up to Prajāpati, which is based on Sāyaṇa's interpretation of śyāmā in ŚB. V.1.3.9 śárve śyāmádh, dvé vai śyāmásya rūpè, śuklám caíva lómnā kṛsnám ca, dvandvó vai mithunám, prajānanám Prajāpatíh. prajāpaty à etát. It is characteristic of Prajāpati that as a god of totality he comprises the light and dark aspects. In the difficult Varuṇa- hymn RS. VIII.41 the last stanza (10) yáh śvetám ádhinirnijah cakré kṛṇdám ánu vraté may mean "He who made the white wearers of garments black ones, in accordance with his vows". See Grassmann and Oldenberg, Noten for ádhinirnij. Since, however, it remains obscure to which "wearers of nirṇij" the verse refers, no conclusion can be drawn from these words. Other interpretations have been given by Geldner: "Der die weissen, die schwarzen (Farben) nach seinen Gesetzen gemacht hat" (implying a double relation which would only be explicable if Varuṇa here stands for the dual deity Mitrá-Varuṇā) and by Renou:

In the same manner Varuna's bond (bandhá) is used in the Atharvaveda: "King Varuna's bond art thou; do thou bind so-and-so, of such-and-such lineage, son of such-and-such mother, in food, in breath" (Whitney) 268.

The most prominent role of Varuna in the ritual is that of a "seizer". Instances of this feature are well-known to every reader of brāhmaṇas. Cf., e.g., "With what vehemence Varuṇa seized these creatures, with that vehemence..." 269 but the same expression already occurs in the Rigveda 270. It is true that sometimes other words are used, such as "injurer" 271, or that he is said to kill or harm 272, but they are much less characteristic. Seized by Varuṇa is the sinner as well as the sick man: Prajāpati, after creating the creatures, caused Varuṇa to seize them as they were offending (against his laws). They then had recourse to Prajāpati and "he saw which this one [was, which was] not seized by Varuṇa. He offered this one [viz., the black ram], which belonged to Varuṇa, in order to free them from Varuṇa. With it he freed them from Varuṇa. A sick man is seized by Varuṇa. Because (this black ram) belongs to Varuṇa, he frees him from Varuṇa" 273.

"Lui qui a mis sur soi les ornements blancs (du jour), noirs (de la nuit)", see EVP. 5, p. 73, 7, p. 31. These translations disregard the grammatical difficulty of ādhinirnij being a masculine noun (see Oldenberg).

268 AS. X.5.44 rájño Váruṇasya bandhò 'si, sò 'múm āmusyàyanám amúṣyàḥ putrám ánne prāṇé badhāna.

269 ŚB. V.4.5.12 só yénai 'vaú 'jase 'máh prajà Váruṇó 'grhṇāt, ténai 'vá... For ójas cf. MS. II.3.1 (27,15) yá vāíṃ Mitrāvaruṇā ojasyà tanús, tāyā vāíṃ Mitrāvaruṇā ojasyà sahasyà táye 'mám amúm muñcatam ámhasaḥ (and 27,19 yá vāíṃ Mitrāvaruṇā ojasyà sahasyà yātavyà rakṣasyà tanús, tāyā...), J. J. Meyer, Trilogie III, p. 205 n. 2). Similarly KS. XI.11 (158,11), but TS. II.3.13.1 reads yá vāíṃ Indrāvaruṇā sahasyà rakṣasyà tejasyà táye 'mám ámhaso mucatam.

270 RS. II.29.5 má má 'dhi putré víṃ iva grabhīṣṭa, X.12.5 kím svin no rájā jaghé kad asyà 'ti vratám cakṛmá, I.139.2 ādadāthe ánṛtam svéna manyúnā. See Bohnenberger, Der altindische Gott Varuṇa, p. 57.

271 ŚB. V.5.4.31 Váruṇo vá ārpayitá.

272 Cf., e.g., Śāṅkh. Ār. XII.5 (21) nainám divyo Varuṇo hanti bhītam, 6 (28) nainam pramattám Varuṇo hinasti and such well-known passages in the Rigveda as II.27.4 cāyamānā ṛtāní, VIII.60.5 imé cetáro ánṛtāya bhūrek, I.24.11 má na dyúb prá moṣīḥ, I.25.2 má no vadháḥ hatnáve jihīlānásya riradhaḥ, má hṛṇānásya manyáve.

273 KS. XIII.2 (180,15) varuṇíṃ kṛṣṇám petvám ekasitipād ám abhetā 'mayāvī jyogāmayāví. Prajāpatíḥ prajà asṛjata. tá enam atyacaram. tá aticarantair Varuṇenā 'grāhayat. tá jīhmáh pannáḥ aserata varuṇagṛhītáḥ. táḥ kṛṣṇáḥ petvó āpravata. tasyā 'bhīhāyā pādam agṛhṇāt. sa práṛhyata. sa ekasitipād abhavat. tā́h Prajāpatíḥ anāthantā,

so 'paśyad yo 'yam avaruṇagṛhītāḥ, tenai 'nā Varuṇám muñcānī 'ti tám vāruṇám ālabhata. tenai 'nā́ Varuṇād amuñcat... (see n. 275), MS. III.5.6 (54,17) Prajāpatíḥ prajà asṛjata. tá enam erṣá 'dīyamanyanta. tá atimányamānā Váruṇenā 'grāhayat. tá vāruṇagṛhītāḥ kṛṣṇáḥ pétvó 'dhyaskandat. tásyā 'nūhāyā pādam agṛhṇāt. tásyā 'prṛhyata. sā ekasitipād abhavat. tám acāyad: ayám vā́vā 'saú prajānā́m ávaruṇagrīto. 'néme 'máh prajà Váruṇān muñcānī 'ti. tám vāruṇám ālabhata. tátā imáḥ prajà Váruṇāt prámucyanta. tád varuṇapramocaníya evá 'sā́... (see further n. 275). Varuṇa even seizes the gods (as he seized Mitra, MS. IV.5.8: 76,7, KS. XXVII.4: 142,18, KKS. XLII.4: 251,2/293,5); cf. KS. XXVI.2 (123,7), KKS. XL.5 (229,7/268,10) sarvā́n vā etád Varuṇó grhṇāti yad dīkṣite, manuṣyā́n pitṛ́n devā́n... prajà eva varuṇapāśān muñcati. That deliverance from varuṇyà

A sick person is várunagrhīta, seized by Varuṇa 274. This was particularly true of those who suffered from a chronic disease (jyógāmayāvin). As we have seen above, in these circumstances Varuṇa was equated to pāpmán and mrtyú, e.g., "Who is suffering from a chronic disease, him he should cause to offer, for this man is seized by Varuṇa, by pāpmán, and he has been ill for a long time. In that it [the ram] belongs to Varuṇa, he thereby delivers him from Varuṇa" 275.

In the speculations of some brāhmaṇas, accordingly, Varuṇa is still associated with Death. In the epic his function is assigned to Yama, whose abode (yamasādana) is the "world of the Fathers" (pitṛloka), whereas Varuṇa's world has here become the abode of demons (see p. 90). As early as the Black Yajurveda, however, there is a tendency to connect the pitáras with the worlds of gods and men (p. 12). In 1898 Caland suggested the idea that the Yajurvedic custom of locating the pitṛs not in the South but in the South-East, that is, closer to the East, was due to an attempt to bring them in closer contact with the gods. He even took a further step and put forward the theory that originally the pitáras had been associated with the western quarter. In that case Varuṇa's ambiguous relation to the demons might explain why he had been unable to join the blessed fathers' trek from the western to the south-eastern quarter. Since, however, all this is purely hypothetical, there is no need for dwelling upon this theory (see Een idg. Lustratiegebruik, p. 279).

12. VARUṆA IN THE EPIC I

It is generally assumed that in post-Vedic times Varuṇa had sunk to the level of a mere “Lord of the Waters” (apāṁ patiḥ). This opinion

(see n. 258) is equivalent to deliverance from Varuṇa himself can be inferred from a comparison of, e.g., ŚB. V.2.5.16 átha yád vāruṇó yavamáyas carúr bhávati, tát sárvasmād evà 'tád varuṇapāśát, sárvasmād varuṇyàt prajā́ḥ prámuñcati with MS. II.1.2 (3,9) vā́ruṇam̀ yavamáyam̀ carúm̀ āmayā́vinā́m yājayed. vā́ruṇagrhīto vá eṣá yá āmayā́vī. Vāruṇād evà 'naṁ ténā muñcati, KS X.4 (128,14) vā́ruṇam̀ carati. yad vā́ruṇo, Varuṇād evai 'naṁ muñcati.

274 Cf. also in classical Sanskrit (lex.) abhyā́ṃta "sick, diseased". In RS. I.189.3 Ágne tvā́m asmā́d yuyodhy āmīvā́ ánagmitrā́ abhy à́manta krṣṭī́ḥ "O Agni, ward off the diseases from us. Let them seize tribes that are not protected by Agni" the verb abhi-am- is used with reference to the diseases themselves. It is certainly not accidental that only in the Yajurveda Varuṇa is said to be the 'seizer', cf. MS.III.12.1 (160,7), KSA. 4.4, TS. VII.4.15. VS, XXII.5, VSK. XXIV.1.5 tám abhy àmíṣi Vāruṇaḥ as against TS. II.2.6.2, etc. tám Vāruṇo grhṇāti.

275 MS. II.5.6 (58,2), IV.3.7 (46,18) yó jyógāmayāví syā́t, tám eténa yājayed. Vāruṇena hi vā eṣá pāpmánā grhītó. 'thai 'tásya jyóg āmayati. yād vāruṇo, Vāruṇād evai 'naṁ téna muñcati, KS. XIII.2 (181,1) pāpmanai 'ṣa grhīto ya āmayā́vī. krṣṇa iva pāpmā. yat krṣṇaḥ, pāpmānam evà 'pahate. yad ekākitipād, varuṇapāśam eva tena pramuñcate (181,5 Varuṇād evai 'naṁ muñcati), XXVII.4 (143,1) tena 'naṁ Varuṇād amuñcaṁ. yan maitrāvaruṇo grhyate, nirvaruṇatvāya, TS. II.3.11.1f. Vāruṇa enam varuṇapāśéna grhṇāti . . . yó jyógāmayāví syā́t, yó vā kāmá́yeta : sárvam á́yur iyām iti . . . vāruṇénai '276 Religion védique III, p. 114.

was already expressed in 1883 by Bergaigne in the words: "Cette fonction [viz. as binder] de Varuna est peut-être . . . celle sur laquelle les poètes védiques insistent le plus souvent, et l'attribut en est resté attaché à la représentation plastique de Varuna dans la mythologie brahmanique, alors qu'il était descendu du rang de Providence vengeresse à celui d'un simple dieu des eaux. Varuna est en effet toujours figuré avec une corde à la main" 277. Sylvain Lévi quoted these words 278 with the sceptical comment: "Reste à savoir s'il s'agit réellement d'une évolution chronologique du personnage". His reservation, dating from a time (1896-1897) when as a young teacher he gave a course on a subject that was still unfamiliar to him, testifies to a remarkably sharp insight into the basic difficulties that surround every study of a Vedic god. There is no denying that there has actually been an evolution in the religious ideas ever since Vedic times. The problem is, however, how far such a development can be reconstructed from what the texts say and omit to say with regard to such an ominous god as Varuna, who was apparently surrounded by taboos. Bergaigne's words "un simple dieu des eaux" show how even a scholar of his high rank could fall a victim to the seemingly innocent wording of the texts and to a wrong perspective in which the textual evidence was viewed. In fact, even in the epic, Varuna, far from being a simple god of water, still retains a good deal of the ambiguity and complexity that characterized him in the Vedic period. What had changed in the religious conception of the god was not so much his character itself as rather the importance of the cosmic dualism and, accordingly, of the annual contest between Varuna and Indra 279. The increasing emphasis laid on the gods of totality, who transcended this dualism, would seem to be the main reason why the importance of both Varuna and Indra, when considered in the context of the total mythological system, is no longer equal to what it was in the Vedic period. It is significant, however, that even in the epic, where Siva and Visnu have

277 At an early stage of Vedic studies it had already been pointed out that it was one of the characteristics of book X that here Varuna had "already" become a "Lord of the waters". See R. Roth, ZDMG. 6, p. 73, H. Oldenberg, Prolegomena, p. 267: "Aber Varuna und Mitrā-varunau verschwinden nahezu, ebenso Uśas . . .", H. Lommel, DLZ. 74 col. 403 (in a review of Lüders's Varuna I): 'Neben Feststellung der Tatsachen kommt Entwicklungsgeschichte bei I. wenig zu Wort, und die sich aufdrängende Annahme, dass die Gestalt des ursprünglich noch viel umfassenderen Gottes im Lauf der Entwicklung verarmt sei, so dass er nur mehr der "Herr der Gewässer" ist, bleibt unausgesprochen' [but see the second part: Varuna, p. 714]. An attempt to explain the supposed evolution in religious thought was made by R. N. Dandekar, ABORI. 21 (1940), p. 190: "The deterioration in Varuna's religious supremacy began with the ascendancy of Indra." It is hoped that the present study will show that the real development was different.

278 La doctrine du sacrifice, p. 153 n. 6.

279 This is also the main objection that must be raised to Hopkins's words (Epic Mythology, p. 117): "All this, however, is no indication of Varuna's real epic position. He is no longer a heavenly god, no longer a god rivalling Indra, or having stars as eyes. He is lord of water . . .". For the decline of the idea of a cosmic contest see above, p. 43.

become the principal gods, Varuṇa still appears to have preserved much of his original character.

We have seen above (n. 113) that in the Veda the Devas often had recourse to Prajāpati for help against the Asuras. The same situation is found in the epic, e.g. when the gods, before the Churning of the Ocean, approach Brahmā280, or when Prajāpati predicts the defeat of the Asuras in words which still echo the Vedic phraseology281.

Varuṇa is mentioned as the fifth or fourth of the Ādityas282, who in the epic are merely one of the classes of the gods and are seldom referred to individually. Just as in the Vedic contest with the Aṅgirasas, they sometimes stand for the gods in general and fight the demons283. Since elsewhere both parties are invoked to give their blessings284, the Ādityas apparently stand for the whole group of Devas, which does not imply that their name is here synonymous with Deva (see PW. and Sørensen, Index). As a curious detail, whatever its significance, it may be pointed out that in the great final single combat between Arjuna and Karṇa, which is full of cosmic symbolism285, the Ādityas side with Karṇa286 but the Devas (sic !) and Pitaras with Arjuna.

The demons in these passages are called Daityas, a name nearly or entirely synonymous with Dānavas and Asuras287. It has been pointed out above (p. 36) that traces can be found in the epic of the idea that the “demons”, no less than the gods, were sons of the World father. Hopkins refers, in illustration of his statement “for Brahman remains also their ‘father’ and gives them not only boons but good advice”288, to Mhbh. XIV.26.6f., where an “old tale” is related, according to which Prajāpati,

280 Mhbh. I.16.4 Viṣṇum āsinam abhyetya Brahmaṇam ce ‘dam abruvan.

281 Mhbh. V.126.42 parābhaviṣyanty Asurā Daiteyā Dānavaiḥ saha, Ādityā Vasavo Rudrā bhaviṣyanti divaukasasaḥ. Cf., e.g., MS. III.7.10 (90,6), III.8.5 (99,12) tāto devā ábhavan, pārá 'surāḥ abhavan, pārá 'surāḥ, KS. XII.3 (165,4) tato devā abhavan, parā 'surā abhavan.

282 He is “the oldest and best” of the Ādityas according to Sorensen, An Index to the Names in the Mahābhārata, p. 13b (but the reading of V.96.13 crit. ed. differs from V.98.13 Bomb. ed.).

283 Mhbh. I.58.26 Ādityair hi tadā Daityā bahuśo nirjitā yudhi, Rām. I.44.26 Āditer ātmajā vīra Diteḥ putrān nijaghniśre. See n. 178 !

284 Rām. II.22.569* tavā 'dityāś ca Daityāś ca bhavantu sukhadaḥ sadā. See Hopkins, Epic Myth., p. 47.

285 See also IIJ. 11, p. 154f. and cf., e.g., VIII.67.26 papāta Karṇasya śarīram ucchritam . . . girer yaṭhā vajrahatam śiras tathā.

286 See VIII.63.39 Ādityāḥ Karṇato ‘bhavan. The account is full of contradictions in details. Although Varuṇa is said to side with Arjuna (śloka 40), and Karṇa after his death goes to the world of Vasus and Maruts in heaven (śl. 56), the latter shoots his vāruṇam astram (VIII.65.13+, 1039*), and his arrows are like snakes (65.33 nirmuktasarpaṅpratimaś ca tīkṣṇais . . . saṅtyā nāācais, 37 śarāṃs tu pañca jvalitān ivoragān), or they are snakes that have sided with Takṣaka's son (65.39 te nyapatan prthivyām mahāhayas takṣakaputrapakṣāḥ).

287 Sørensen, Index to the Names, p. 98a, says that Asura “seems to have been totally synonymous with Dānava and Daitya”, whereas Hopkins, Epic Mythology, p. 47, thinks that the Dānavas are more “god-like” than the Daityas. If there was a distinction at all, it must have been very small.

288 Epic Mythology, p. 50.

by merely pronouncing the syllable "om" was teaching "happiness" (śreyas) to the Gods, the Ṛsis, the Nāgas and the Asuras. Much more frequent are the passages where gods of the totality are represented as protecting the surāsurāḥ, demons and gods. Thus Brahmā fulfils their wishes and is their guru289. Kaśyapa, whom Hopkins describes as "the sire of all demons"290, in fact stands for the cosmic totality: the epic calls him, who had married the two daughters of Prajāpati, Kadrū and Vinatā, "a husband similar to Prajāpati"291 and according to the Viṣṇupurāṇa he is the father of the Daityas (with Diti), the Nāgas (with Kadrū) and the Ādityas (with Aditi)292. In the epic Kaśyapa is said to be the father of gods and Asuras293. In the same manner Śiva and Hari/ Viṣṇu are called gurus of gods and demons294, although Viṣṇu is predominantly the destroyer of the demons295.

13. VARUṆA IN THE EPIC II: THE LORD OF THE WATER

It is not necessary to demonstrate in detail that in the Mahābhārata Varuṇa is still the god of the water (apāṁ patiḥ) and the ocean, who can stand for this ocean itself. His palace is on the bottom of the sea. That he resides and even conceals himself in the water is also apparent from the "trick of Varuṇa" (vāruṇa yoga), to which Kauṭalya 7.17.42 refers and which according to Kangle consists in "remaining submerged in water for a long time and coming up at a very distant place". The ancient idea of Varuṇa as the god of the subterranean samudra seems, however, gradually to fade away, at least as far as the term is involved. The Rāmāyaṇa, in locating Pannagas and Dānavas in Pātāla in the ocean (VI.21.27, 30; 22.4) apparently mixes up the two notions of the subterranean abode and the sea (see p. 85). When, however, different terms are used, Varuṇa remains the god who provides water in wells and tanks.

In the way the poets represented the relation between the god and his element there is, understandably enough, some variation. It is natural

289 Mhbh. XII.326.3-59-60 tvam caiva varado brahman varepsundām bhavisyasi, surāśuragaṇānām ca ṛṣīnām ca tapodhāna (60) pitṛnām ca mahābhāga satatam samśitavrata, vividānām ca bhūtānām tvam upāyo bhavisyasi and XIII.151.3 devāsuragurur devo devāsuranamaskṛtaḥ (cf. XIII.17.141).

290 Hopkins, Epic Mythology, p. 47.

291 Mbh. I.14.6 prajāpatisamaḥ patiḥ. In some versions Kaśyapa is even incidentally identified with Prajāpati, cf. I.20.16 Bombay ed.=fragment 285*, line 10 crit. ed.

292 Viṣṇupurāṇa I.15.141-21.3, I.21.19-22 and I.15.126-134 respectively. This is, of course, different from the older account of the Suparṇākhyāna, where Kadrū and Aditi are synonymous, whereas here Aditi, as the mother of the Ādityas, may possibly have been imagined as standing for the upperworld (=Suparṇī in the Yajurveda, Vinatā in the epic).

293 I.60.33 Mariceḥ Kaśyapaḥ putraḥ Kaśyapasya surāsurāḥ, jajñire nṛpaśārdūla lokānām prabhavas tu saḥ. Cf. XIII.12.26 rājyahetor vivaditāḥ Kaśyapasya surāsurāḥ.

294 See respectively XIII.17.141 devāsuragurur devo devāsuranamaskṛtaḥ, and XII.324.20 surāsuraguruṃ Harim.

295 XII.202.29 nihatya dānavapaṭan, XII.334.14 v. 1. asuravadhaḳaraḥ for tapasāṃ nidhiḥ. Cf. VIII.35.34 (tān . . .) pothayāṃ āsa Bhīmo, Viṣṇur ivā 'surān.

77

that they had some difficulty in visualizing what in the brāhmaṇas was

a mere mythical identity (samudró vai Várunah) and as such a purely

abstract notion. In the epic version of the Churning of the Ocean the

gods announce the churning to the sea, whereupon Varuṇa answers

(p. 108). The personification of the sea, however, is not always necessarily

the god. Cf., e.g., Bhāg. Pur. IX.10.13 sindhuh … rūpī, which does not

denote Varuṇa. When Rāma, approaching Lañkā, sees the sea which

separates the mainland from his enemy's island, he simply says "Give

orders to the sea" (Rām. VI.13.22 samudro 'yaṃ niyujyatām), although

the poet here often refers to this ocean as "Varuṇa's abode" (varuṇālaya:

VI.13.11, 12, 21; 14.13; 15.11). The fact that such other epithets as "abode

of the Dānavas" (App. I.11.8) and "abode of the makaras" (VI.14.8, 11)

occur side by side with it shows that samudra had come to be taken in

entirely the same sense as sāgara.

The same episode is found in Bhāsa's Abhiṣekanāṭaka, act IV. Rāma here

describes his disappointment at seeing his way barred by the ocean after

all the hardships he has overcome. Lakṣmaṇa says, pointing to the sea,

"Here is Lord Varuṇa" but then describes the ocean (saritpatih). While

Rāma again expresses his annoyance, Rāvaṇa's brother Vibhīṣaṇa arrives

through the air and is led before Rāma, who receives him with the

words (IV.11,3) "O Vibhīṣaṇa, the mere fact of thine arrival guarantees

the success of our enterprise!" (Vibhīṣaṇa, tvadāgamānād eva siddham

asmatkāryam). These words, which Bhāsa did not find in Rām. VI.19.7,

Bomb., 14.10 crit. ed., prove that he was still aware of the importance

of the underlying mythical pattern. When Rāma observes that there is

no possibility of crossing the ocean, Vibhīṣaṇa says (IV.11,5) "If the

ocean does not grant you passage, your majesty should throw his divine

missile on it" (yadi mārgaṃ na dadāti, samudre divyam astraṃ tāvad

visraṣṭum arhati devah). When Rāma threatens to dry up the ocean with

his arrows (IV.12), Varuṇa enters the stage in confusion. He confesses

that he has committed an offense against Nārāyaṇa and says that he

will seek protection with the latter. In this passage there is not only

the double aspect of the god and his sea but also of Rāma as an avatāra

of Viṣṇu and as the human hero of the saga. In his first aspect he is

worshipped by Varuṇa and respectfully referred to as bhagavat, whereas

in his second aspect (that is, as rājaputra 15,3) he is addressed with

bhavat. When Varuṇa has granted him passage (16,2 esa mārgaḥ. prayātu

bhavān) and while Rāma's troops march through the passage formed by

the sea, Rāma describes the beauty of the sea in the traditional way,

well-known from such texts as the Caturbhānī and the Mṛcchakaṭikā

(16,13 aho vicitratā sāgarasya. iha hi…). Here and in the final words

"Through the favour of the holy god we have crossed the sea" (17,1

bhagavato prasādād atītaḥ sāgarah) the distinction between the god and

his element is stressed. In Rām. VI.15.1, on the other hand, it is the

Ocean himself who appears after Rāma's menaces. In the same verse he

is referred to as Sāgara, arising from the middle of the Samudra, and

as Samudra. The first word is more common for the non-personified

78

element (VI.14.12; 15.15, 17, 22, 24, 25, 26, 30, 33; 16.1, 2, 3, 6), but

cf. samudra (VI.14.10, 16, 17; 15.27).

While the Vedic idea of a dualism which splits the world of the Devas

into two parties has faded in the epic, this does not mean that the

Ādityas, and in particular Varuṇa, have lost their ambiguous character

in this period. On the contrary, more openly than in any older text it

is now said that Varuṇa maintains secret relations with the dispelled

Asuras, who continue to exist outside or on the fringe of the ordered cosmos.

Before considering, however, the relations between the Cosmos and

the powers of Chaos on its border-lines, it must first be stated that the

Vedic conception of Varuṇa as a rājā and samrāṭ, which incidentally evoked

the idea of Varuṇa's consecration by Prajāpati and the gods 296, survives

in the epic. Here several times mention is made of Varuṇa's abhiṣeka,

e.g. by Brahmā 297, who stands for (or rather, is) the Vedic Prajāpati.

One of these passages, viz. Mbbh. IX.46.5ff. is here given in full:

(5) ādau kṛtayuge tasmin vartamāne yathāvidhi

Varuṇam devarāj sarvāḥ sametye 'dam athā 'bruvan

(6) yathā 'smān surarāṭ Śakro bhayebhyoḥ pāti sarvadā

tathā tuvam api sarvāsām saritām vai patir bhava 298.

(7) vāsaś ca te sadā deva sāgare makaralaye

samudro 'yam tava vāse bhaviṣyati nadīpatih.

(8) somena sārdham ca tava hānivrddhī bhaviṣyakah 299

evam astv iti tān devān Varuṇo vākyam abravīt.

296 ŚB V.4.3.2 Vāruṇaṁ vā abhiṣiṣicāndāt, indriyāṁ víryàm ápa cakrāma (in the

description of the Rājasūya ritual) for "the Rājasūya is Varuṇa's consecration"

(varuṇasavó vá éṣá yád rājasūyam). In the Jaiminiya Brāhmaṇa there is a tale

of how Varuṇa wishes to become king over all the gods and practises asceticism

for a hundred years. Prajāpati then teaches him a sāman which makes all the gods

stand up in his presence because they see in him a form (rūpa) of Prajāpati. He

is then consecrated by Prajāpati and the gods: JB. III.152 atha varuṇasāma.

Varuṇo vai rājā sadhamādām ivā 'nyābhir devarādhir āsit. so 'kāmayata: sarveṣām

devānāṁ rājyayā sūyeye 'ti. sa Prajāpatau śataṁ varṣāṇi brahmacaryam avasat.

tasmā etat sāma 'bravīt: etad vai me rājyàm rūpam. gaccha. rājānàṁ tvā devāḥ

kariṣyanta iti. so devān abhyait. tam āyantamàṁ drṣṭvā devāḥ pratyavarohan . . . yad

vai naḥ pituḥ Prajāpate rūpāmiṅ, tad idam் tvayi paśyāma iti. Cf. further JB. II.25-26:

the abhiṣeka of Āditya at the ritual of the mahāvrata.

297 Mbbh. IX.44.20 purā yathā mahārāja Varuṇam vai jakṣivaram, tathā 'bhyaṣiñcad

bhagavān Brahmā lokapitāmahah, Kaśyapas ca mahātejā ye cā 'nye nā 'nukirtitāḥ

and Viṣṇu Purāṇa I.22.1 and 3. Mbbh. IX.45.92 Aujasaṁ nūma tat tīrthaṁ yatra

pūrvam apāṁ patih, abhiṣiktah suraganaih.

298 Lack of symmetry ("rupture d'équilibre", Renou, Festgabe für Herman Lommel,

p. 123), as in RS. VII.84.2 "May Varuṇa's wrath spare us, may Indra give us room"

(Bergaigne III, p. 142) and TS. I.8.16.2 (mantra for the adoration of the king)

"Thou art Mitra, the kindly . . . thou art Varuṇa, of true rule" (Mitró 'si susévo . . .

Váruṇo 'si satyádharmā) where Sāyaṇa stressed the ambiguity of the god (p. 988

line 25): Varuṇo 'pi kracid aniṣṭakāritvād anṛtarūpah, kvacit tannivartakatvāt

satyartūpah. Cf. also the commentary on TB. I.7.10.1 tvain Varuṇo 'si śaucādāv

aniṣṭavārakatvāt.

299 Soma is the moon (cf. I.16.38 prasannabhāḥ samutpanṇaḥ Somaḥ śitāṁśour

(9) samāgamya tataḥ sarve Varuṇam sāgarālayam

apām patim pracakrur hi vidhiḍṛṣṭena karmaṇā.

(10) abhiṣicya tato devā Varuṇam yādasām patim

jagmuḥ svāny eva sthānāni pūjayitvā jaleśvaram.

(11) abhiṣiktas tato devair Varuṇo 'pi mahāyaśāḥ

saritah sāgarāṃś caiva nadāṃś caiva sarāṃśi ca.

(12) pālayām āsa vidhinā yathā devān Śatakratuh.

Unlike Indra, who “always protects against dangers”, Varuṇa, although consecrated, is not expected to perform beneficial actions. There is, however, no trace of a specific degradation of Varuṇa in comparison with his status in the Veda. In one passage it is Indra who consecrates Varuṇa, viz. V.16.33-34: “Having thus pondered the Lord Mahendra Pākaśāsana, (consecrated) Kubera as the lord of all Yakṣas and of riches, Vivasvat's son as lord of the Pitaras, and also Varuṇa as lord of the water. Śakra, the grantor of wishes, thus gave him respectfully the overlordship”300. Although no direct connection need exist with the Rigvedic passage X.124.5, where Indra offers Varuṇa the (or, an) ādhipatya of his kingdom (see p. 29), the situation is remarkably similar. What has changed in the centuries that separated the epic period from the Vedic age is not so much Varuṇa's cosmic function as lord of the waters as rather man's attitude towards the cosmic law and its general status. Whereas Yama as dharmarāja in a way continues Varuṇa's function as ṛtásya gopáh, dhármapatiḥ, the general shift of emphasis from the

ujjvalah). In the classification the moon is associated with Varuṇa, as the sun is with Mitra. See Bijdragen tot de Taal-, Land- en Volkenkunde 107, p. 82 and cf. Mbhb. V.97.4 (in Varuṇa's abode) ataḥ Somasya hāniś ca vrddhiś caiva pradrśyate (for a different version see IX.34.40ff., where the moon is said to bathe in Prabhāsatīrtha). The moon is a vessel which is periodically refilled. In the same way, the ocean is replenished, for tava samudrasya implies that Varuṇa and the sea are identical (just as in I.16.7-8, see p. 108). There seems to be a reference to high and low tide, the former being considered a periodical replenishment of the sea by the rivers. Hopkins, who recognized the importance of this passage, wrote (Epic Mythology, p. 117): “Varuṇa is formally consecrated by the gods as lord of rivers and waters (9,45,22; 46,105) and told that his home shall be in ocean, the home of makaras; that Ocean, the Lord of Rivers, shall be under his will, and that his own decline and growth shall agree with the waning and waxing of Soma. There seems to be actually no difference felt here (though expressed) between Ocean and Varuṇa.” It should be added that elsewhere a distinction is made between the apām rājyam and mahodadhīḥ. Cf. XII.122.26ff., where Śiva consecrates both Indra and Varuṇa:

(26) bhūyah sa Bhagavān dhyātvā ciram śālavarāyudhaḥ, tasya tasya nikāyasya cakārai

'kaikam īśvaram (27) devānām īśvaram cakre devam daśātekṣaṇam, Yamam vai

vasavataṃ cāpi pitṛṇām akarot patim meruṃ saritāṃ ca mahodadhim (29) apām rājye

surānāṃ ca vidadhe Varuṇam prabhum, etc. Cf. also Bhāsa, Abhiṣekanāṭaka IV, discussed on p. 78. For surānām see below, n. 349.

300 The text has been quoted above, p. 29. Since the syntactic structure is rather loose, no literal translation is possible.

gods of the organized cosmos to those gods who transcended the wider all-encompassing opposition of sat and asat, of Cosmos and Chaos, also altered the character of sin. After the seasonal awakening of Visnu as the god of the upper and the nether world had been projected to the gigantic dimensions of the god who periodically comes to this world to restore the dharma, and who even survives the total destruction of the world, sleeping on the Oceanic waters of Chaos, sin was no longer primarily considered an offense against Varuna. But this degradation was a consequence of the fading of the ancient notion of the ordered cosmos as based upon a contest of opposed powers.

It was only part of the general degradation of all those Vedic Devas who did not stand for the totality 301. Once the religious consciousness had widened so as to include even the asat of the Veda, the conception of the cosmic drama had so much changed as to deprive most Devas of their importance.

14. VARUNA IN THE EPIC III: THE NETHER WORLD

In the epic Varuna is, as Lord of the Waters, one of the Devas 302, but, just as in the Veda, his most characteristic feature is his being the principal god of the nether world. It is significant that none of his epithets refers to this side of his nature. His is the world where the sun sets: "That place where the sun, O king, sets in accordance with truth, that men call the king of mountains Asta. King Varuna, living in that king of mountains and in the great ocean, protects the beings (bhūta)" 303. That this is an old genuine trait of Varuna is apparent from the Vedic evidence 304.

301 A reservation must again be made for Siva. If he, as Rudra, originally stood outside the sacrificial year, he may have been closely connected with Chaos. This problem, however, will not be considered here.

302 E.g., in the Nala-episode, Mhbh. III.51.22, 52.4-5, 54.31, and in XII.29.16 yasya sendrā savarunā bhaspatipurogamāh. derāh, a reminiscence, it seems, of the older position of Indra and of the opposition of Brhaspati and Varuna in the brāhmanas. See above, p. 55 with n. 180.

303 Mhbh. III.160.10 yain prāpya Sacitā rājan satyena pratitisṭhati, astain parvata-rājanain etain āhur manisinah (11) etain parvatarājānain samudrain ca mahodadhim, āvasan Varuno rājā bhūtāni parirakṣati. Cf. Oldenberg, Kleine Schriften, p. 702 n. 4 against Pischel, EGA. 1895, p. 449.

304 KB. XVIII.6.10 (9 Lindner) sa vā eṣo 'pah pravisya Varuno bhavati, AS. XIII.3.13 sa Varunah sāyám Agnir bharati, RS. V.62.1, where Sūrya's horses are said to be unharnessed in Mitra's and Varuna's sacred abode. See Bergaigne III, p. 117, Hillebrandt, Liedes des Rgveda, p. 80 n. 2, Geldner, Religionsgeschichtliches Lesebuch, Heft 9, 2nd ed., p. 42, Renou, EVP. 5, p. 78, 7, pp. 40, 29 and JB. II.25, 13-14. where Varuna's region is associated with Rta and night: mām abhy astamayye 'ti pratici dik prayacchad rtain ca rātrim ca. Cf. History of Religions 10, p. 97. On the other hand, Thieme, Mitra and Aryaman, p. 71 and Renou, Festgabe für Herman Lommel, p. 125, explain the AS. passage as due to a secondary development. Gonda, The Vedic God Mitra, p. 42, only remarks that the KB. passage does not prove 'a more or less exclusive, original or fundamental association of Varuna with night or the nocturnal heaven". This is correct but no-one, as far as I can see, has ever claimed this. See ILJ. 15, p. 226.

and Nārada’s pilgrimage to Varuṇa's realm. Nārada here says to Mātali:

“I will explain everything to you, while showing you the nether world.

Then Mātali and Nārada, the magnanimous ones, both descending into

the earth, saw the guardian of the region, the lord of the waters. Nārada

explained to the charioteer all beings without exception, who lived inside

the earth, as he knew them . . .”. They then visit the world of the nāgas:

“Taking leave of Varuṇa they roamed about in the nāgaloka”. It is not

said that this nāgaloka is different from Varuṇa's world. Cf. also V.106.12

“Here, living in Pātāla, Varuṇa got Śrī (good fortune)” 305.

Mythologically the (western) ocean and the subterranean water, which

the Rigveda calls samudrá, are identical. They are also identified in

Mhhh. I.19.3 “There they then saw the ocean (samudra), the receptacle

of water, the mine of all gems and the abode of Varuṇa, the lovely abode

of the nāgas, the highest lord of the rivers, the habitation of the Pātāla-fire

and the jail of the Asuras, and the frightener of beings, the sea, receptacle

of water”306. In those passages where a distinction is made, “Varuṇa's

abode”, which is identical with the “World of Nāgas”, is reached by a

hole in the ground. In the tale of the sons of Sagara, who are sent out

to find the sacrificial horse, we read: “Obeying this order, the sons of

Sagara began again to search the whole earth. Then the heroes saw

(a place where) the earth had been excavated. Having reached that hole

and digging it, the sons of Sagara dug out the ocean with hoes and spades.

This abode of Varuṇa (varuṇālaya), being dug out by all Sāgaras, became

very much afflicted as it was excavated on all sides. Asuras, snakes

(uraga), Rākṣasas and all sorts of beings (sattva) gave cries of distress

as they were killed by the sons of Sagara... While they thus dug out

the ocean, the abode of makaras, a long time passed away but the horse

did not appear. Thereupon the sons of Sagara, O king, enraged dug out

Pātāla in the north-eastern region and there they saw the horse roaming

about on the soil (lit. earth)” 307. Just as in V.96.8 quoted above, where

305 Mhhh. V.96.5 aham te eṣaṁ ākhyāṁse darśayan vasudhātalaṁ (6) avagāhya

tato bhūmiṁ ubhau Mātalināradaṁ, dadrśāte mahātmānau lokapālam apāṁ patiṁ

(9) Nāradah sarvabhūtānām antarbhūminivāsīnām, jānams cakāra vyākhyānāṁ

yantunḥ sarvaṁ aśeṣataḥ. Cf. also above n. 24 in fine, n. 61 and V.106.12 atra pātālam

āśritya Varuṇaḥ śriyam āpa ca. Just as in the epic Pātāla is the capital of the

nāgaloka (V.97.1), it is in classical literature an abode of snakes, cf. Kādambarī

p. 92, lines 2–3 pātālam iva mahākāñcukīhasahasrādhyāsitam.

306 Mhhh. V.96.8 Varuṇenā 'bhyanujñātaṁ nāgalokaṁ viceratuḥ and, e.g., I.19.3

dadrśāte tadā tatra samudraṁ nidhim ambhasāṁ (5) ākarāṇāṁ sarvaratnānāṁ ālayaṁ

Varuṇasya ca, māgānāṁ ālayaṁ ramyam uttamaṁ saritāṁ patiṁ (6) pātālajvalana-

nāvāsam Asurāṇāṁ ca bandhanam, bhayaṁkaram ca sattvānāṁ payasāṁ nidhim

arnavam, etc. See further Hopkins, Epic Mythology, p. 118f. On the

other hand in III.101.8 the udadhi as Varuṇa's abode is apparently considered

to be different from the Pātāla under the earth in III.103.12. The same notion

that serpents and Dānavas dwell in the ocean in Pātāla (that is, accordingly, in

the subterranean samudra) is met with in, e.g., Rām. VI.14.19.

307 Mhhh. III.105.18 pratigrhya tu saṁdeśaṁ tatas te sagarātmajāḥ, bhūya eva mahīṁ

kṛtsnāṁ vicetum upacakramuḥ (19) athā 'paśyanta te vīraḥ prthivīṁ avadāritām,

samāsādya bilaṁ tac ca khanantah sagarātmajāḥ, kuddālaiḥ hreṣukaiś caiva samudram

no clear distinction is made between Varuna's world and that of the serpents, this varunālaya is apparently identical with the nāgaloka, which has a similar entrance.

An interesting parallel can be found in the well-known Uttanka-episode of the epic 308. In the former of the two relevant passages the entrance to the nāgaloka is a "wide, big hole" (vivrtam mahābilam), in the latter, the serpent, having taken the two ear-rings in its mouth 309, disappears in an ant-hill, which is the entrance to the nether world. Since the abode of the Asuras under the earth is identical with that of the nāgas (as we shall see below), the vivrtam mahābilam (Mhbh. I.3.137) and the asuravivara, a term used in classical literature for the entrance to the nether world, are synonyms. The latter word occurs at Daśakumāracarita, p. 41,1 ed. Agashe, whereas the much later Introduction (Pūrvapīṭhikā, pp. 14,3; 15,12, etc.) uses the uncharacteristic terms bila, bilāpatha. The proper meaning of asuravivara was misunderstood by the 18th century commentator who glossed it with pātāla. In the Harṣacarita the word means "treasure cave" according to Thomas, JRAS. 31 (1899), p. 489. Cf. Kāṭh. 227,2.

The ant-hill in the Uttanka-episode reminds us of the fact that there is an ancient connection between ants and Asuras: in the Atharvaveda the ant upajikā is said to be a daughter of the Asuras (and a sister of the Devas !) 310. Because the Asuras have fled under the earth, they have at an early date become the prototype of diggers 311, who bury a remedy (AS. VI.109.3) or a charm (ŚB. III.5.4.3 krtyām . . . ní cakhnur, cf. AS. V.31.8). In the later belief that a digger of pools attains Varuna's

akhanams tadā (20) sa khanyamānāḥ sahitaiḥ sāgarair Varunālayah, agacchat paramām ārtim dāryamānāḥ samantataḥ (21) asuroragarakṣāmsi sattvāni vividāni ca, ārtanādam akurvanta vadhya­mānāni sāgaraiḥ . . . (23) evam hi khanatām teṣām samudram makalāyam, ryatiḥ sumahān kālo na cā 'sthā samadrśyata (24) tataḥ pūrvottare dese samudrasya mahipate, vidārya pātālam atha saṁkruddhāḥ sagarātmajāḥ, apasyanta hayam tatra vicarantam mahītala.

308 Cf. Mhbh. I.3.137 sa (viz. śramaṇaḥ) tad rūpam vihāya takṣakarūpam krteā sahasā dharanyām vivṛtam mahābilam vivesa (138) praviśya ca nāgalokam svabhavanam agacchat. tam Uttanko 'nādṛśeva tenaiva bilena; (158) tābhir nāgaloko dhūpitah, XIV.57.21 apasyad bhujagah kascit te tatra manikundale (22) Airāvata­kulotpan­naḥ sīghrro bhūtvā tad eva ca, vīdaśyā 'śvena valmikam vivesa 'tha sa kuṇḍale tataḥ khanata evātha viproarṣer dharaṇītalam, nāgalokasya panthānam kartukāmasya niścayāt (27) rathena hariyuktena tam deśam upajagmivān, vajrapāṇir mahātejā dadrśa ca dviśottamam (28), . . . Uttankam abhāṣit: 'tāta nītaḥ khanuṃ trave 'ti vai (28) tato vajraprahārais tair dāryamānā nāgaloko vai yojanāni sahasraśah . . . (32) tato vajraprahārais tair dāryamānā vasumdharā, nāgalokasya panthānam akaroṭ Janamejaya (33) sa tena mārgena tadā nāgalokam vivesa ha . . . For Taksaka see also n. 339.

309 The reading sa kuṇḍale as the object of vīdaśya. Otherwise PW. pw., whose interpretation of this passage is based on the different reading of the Calcutta edition.

310 AS. VI.100.3 Āsurāṇām duhitd 'ei sá devānām asi svásā, divás pṛthivyáh sáṃbhūtā śā cakarthā 'rasām viśām.

311 AS. II.3.3 nicaih khananty Āsurāḥ, VI.109.3 Āsurās tvā nyàkkhanan devās tvó 'davapan púnah.

world312 we find, again, a connection between Varuna's world and the

Asuras, which will be discussed in the next section. In this context it is

interesting to note that the "ants are endowed with the faculty of producing

water"313, which must be due to the fact that as "diggers" they were

associated with the nether world and Varuna's subterranean water. Hence

in a charm undertaken with springwater (AS. II.3.4) it is said that the

upajikā-ants bring up the remedy from the "sea" (úd bharanti samudŕd

ādhi), where, "deep down", the Asuras bury it (v. 3 nícaíh khananti).

This side of Varuna's character is still seen in the Purānas, where

Varuṇa is the king of the waters "uniquement lorsqu'il s'agit de rites de

consécration des puits, mares et réservoirs"314. It has been pointed out

elsewhere that the name Varuṇadeva, used in medieval times for stone

slabs of wells in Chamba State315 reflects the same conception of Varuṇa

as a god residing in the nether world. This idea can be traced back to

RS. VIII.69.12 "Thou art a good god, O Varuṇa, as thy seven rivers

stream through (along) the hollow of thy mouth as through a hollow

pipe"316. Just because Varuṇa could obstruct the water (cf. Vŕtra), he

is denoted by the auspicious term sudevá "a good, auspicious god" - an

instance of word-magic317. While allowing the water to spring forth from

312 Viṣnusmŕti 41.2. Such a digger may have been supposed to be digging "beyond

the earth", as the Jaiminīya Brāhmaṇa puts it (I.237 parena prthivīm . . . anātkhāya).

Cf. TS. VI.3.4.2 pŕthivatyàm hy ètád yán nikhā́tam.

313 Bloomfield, SBE. 42, p. 278 (referring to Kauśika Sūtra 25).

314 See M.-Th. de Mallmann, Les enseignements iconographiques de l'Agni-Purāṇa,

p. 132, no refers to Agni-Purāṇa 64 (kúpādi), and cf. History of Religions 10, p. 98.

315 See above, p. 87.

316 See above, p. 27. The current interpretation of sudevó asi varuṇa yásya te

saptá sindhavah, anukṣáranti kākudáṃ sūrmyàṃ suṣirā́m iva is hardly correct.

Geldner translates (like Ludwig and Grassmann) "in dessen Schlund die sieben

Ströme fliessen wie in eine hohle Röhre", which disregards the specific meaning

of ánu. Hillebrandt, Ved. Myth. II2, p. 19, thinks that this stanza has no mythological

importance at all, whereas Lüders, Varuṇa, p. 53, interprets it as saying that

Varuṇa drinks much Soma. Actually, the stanza should be paralleled with VIII.41.2,

where Varuṇa is said to be dwelling "at the origin of the rivers, with seven sisters,

(himself) amidst (them)" (yád sindhūnám úpo 'dayé saptávasā sá madhyamó

see Lüders, p. 412) and the other passages quoted in India Maior (Congratulatory

Volume J. Gonda), p. 151. It is interesting that MS. IV.7.8 (104,9) quotes this stanza

in support of samudró vái Varuṇaḥ "Varuṇa is the (subterranean) ocean". Cf. also

Thieme, Kleine Schriften, p. 620 and German Scholars on India, p. 334.

317 Not "ein rechter Gott" (Lüders, Varuṇa, p. 54, who interprets the word as

referring to Varuṇa's drinking much Soma). It should be noted that the stanza

is recited in order to be delivered from Varuṇa (MS. IV.7.8 nírvarunatắya).

Elsewhere the word is mostly a bahuvrīhi but cf. ŚB. VII.5.2.52 sudevó adyá tád

vidyād yátra nírāpaṇam dadhúḥ. The idea that Varuṇa releases the subterranean

water is indissolubly linked up with the certainty that he also could detain it.

Cf. in this connection RS. X.124.7 "without being compelled Varuṇa released the

waters" (áprabhūtī Váruno nír apáh srjat). See Bergaigne, III, pp. 127, 148, Renou,

EVP. 7, p. 5 (otherwise H. Güntert, Der arische Weltkönig, p. 287). However, I

now have my doubts about the explanation suggested in IIJ. 5, p. 52 for the name

Varuṇadeva, which in medieval times was given to the stone slabs of wells in

Chamba State. They need not necessarily owe their name to the obstructive

character of Varuṇa.

84

the earth he represents the Mitra-aspect of the dual deity 318, which also explains the use of sudānavah with reference to the three Ādityas (including Mitra !). Since dānu specifically denoted the primordial and the subterranean water, the epithet meant either "whose water is auspicious" or, perhaps "well (bounteously) giving water" 319. It has been one of Lüders's important observations that only conjointly with Mitra Varuna is said to give rain 320. This confirms the conclusion that the specific Varuna-aspect was the one which obstructed the water. Mythologically, this water was both the spring-water and the rain-water falling down during the night. While it is clear that drawing water from a well must have meant drawing it from Varuna's dwelling-place, the taboo which forbade drawing water at night 321 is not immediately clear. The texts say this was prohibited because at night the spring-water was "in Varuna". Whatever they may have meant by these words, it may be suggested that the original motive was that at night Varuna and the subterranean water were in the sky so that what was drawn up from a well at that time was not Varuna's water.

From the passages quoted it may be inferred that the poets, unless they expressly associated Varuna's palace (sabhā) with the western ocean (in accordance with the system of classification) 322, located the god in the nāgaloka at the lower end of the world axis. However, the association of Varuna with the quarter of the setting sun, although understandable enough, could easily overshadow the basic notion that Varuna's true place in the cosmos was in the nadir, at the roots of the world tree. Some later descriptions such as in Rāmāyaṇa VI.14-15 cr. ed., may give the impression of being mixtures of two different pictures, the varuṇālaya being identified with the makarālaya and the sea 323. It would require a separate study to determine if, and in how far, they are mixed up but

318 See IIJ. 5, pp. 51–53.

319 For sudānavah see Lüders, p. 716, for dānu cf. Hist. of Rel. 10, p. 122.

320 Varuṇa, p. 716: "Meistens ist aber Varuṇa, wenn er Regen spendet, mit Mitra verbunden", and p. 719: "Oft wird Varuṇa auch um Regen angerufen, gewöhnlich zusammen mit Mitra".

321 See J. J. Meyer, Trilogie altindischer Mächte und Feste der Vegetation III, p. 207 n.1. A simpler explanation would be that, since Varuṇa is a god of the night and of the water, his power over the waters is at night still more dreaded. For the rain falling during night see India Major, p. 150f.

322 The inevitable result of assigning every deity to one of the five or seven quarters was a potential conflict between the place of a god in the system of classification and the place which he naturally occupied (see IIJ. 13, p. 283). The identification of Varuṇa's home with the (western) ocean must have been due to the system (and his special relation to the setting sun) but theoretically Varuṇa was the lord of all the four oceans which in the mythical cosmology surrounded the earth.

323 Cf. Rām. VI.14.13, 278*, 280* varuṇālaya beside VI.14.3 and 11 makarālaya. The same notions are connected in Mhbh. III.105.20 versus 23, see n. 307; cf. also Mhbh. III.101.8-9, where varuṇālaya=udadhi ghora. In the Rāmāyaṇa the Pātāla is said to be in the sāgara (VI.14.19, App. I.11.2), which is the dwelling-place of the Dānavas (dānavālaya VI, App. I.11.8). The rasātala is also located there (VII.23.3).

85

this much would seem certain that in the older literature the samudrá

in Varuna's world was, ever since RS. VII.88.3, primarily (and perhaps

exclusively) the subterranean water. See p. 26f.).

We have seen above that in the Mahābhārata Varuna's abode

(varunālaya) is identical with the world of the serpents (nāgaloka), in

which the town Pātāla, situated in the very cosmic centre 324, is Varuna's

dwelling-place 325. Here are “Asuric” forms of Agni and Soma and here

is the moon 326. The latter belongs here as a consequence of the classification

upon which the mythic cosmology is based, viz. sun : day-time sky :

upper world versus moon : night-time sky : nether world. This led to

identifying the water under the earth and the underworld jar with nocturnal

sky and the moon 327. Classificatorily the second group consists of different

aspects or manifestations of Varuna's world. While every morning Varuna

324 Mhbh. V.97.1 etat tu nāgalokasya nābhisthāne sthitam puram, Pātālam iti

vikhyātam daityadānavasevitam.

325 See notes 305-307.

326 See Mhbh. V.97.3-4 (quoted in n. 68) and IX.46.8 (above, p. 79). The idea is

that the gods after winning the amŕta at the Churning of the Ocean and drinking

from it gave it to guardians of the nether world to guard it there. The same idea

is also found in the tale of the Churning of the Ocean, cf. I.17.30 dadau rakṣitum.

This is a detail that has arisen in the Vedic period among authors of Yajurvedic

texts. Its genesis can be reconstructed (see Études Asiatiques 25, p. 90) but it is

only a secondary variant of the original idea that the soma/amŕta had to be won

from the primordial world. See p. 17 and e.g., IIJ. 15, p. 231. For the āsura agni

see Mhbh. I.19.6 pātālajvalana and RS. III.29.11 gárbha ásuro (n. 68), and cf. IIJ. 8,

p. 111, above p. 119ff. For the waxing and waning moon (Soma) see above, n. 299.

327 For the vāruṇa kumbha (Agni-Purāṇa 64.4) cf. Varuṇa represented as sitting

beside a kumbha on a relief of Badami (see p. 146f.). The symbolism of the cosmic tree

rooted in the kumbha was, it seems, more common in Khmer and Indonesian art,

see J. Auboyer, Le symbolisme du trône, p. 97, F. D. K. Bosch, The Golden Germ,

An Introduction to Indian Symbolism (Indo-Iranian Monographs, vol. II), 1960,

pp. 110–113, 156f. In India the bowl is usually filled with lotus leaves, which can

be extended upwards, “especially when narrow vertical spaces are available,”

“to a considerable height, either as a conventional candelabra-like tree, or as a

long spray of lotus” (A. K. Coomaraswamy, Yakṣas II (1931), p. 62). The tree

seems to be comparatively rare (Bosch gives only one example on plate 83, a painting

in the Joganmohan Palace, Mysore). As a rule the symbol of the world axis is

represented not as a tree but as the cosmic pillar (stambha). Some illustrations

are the stone pillar crowned by the solar wheel depicted on a pilaster in Sarnath

and the pillar cum foliage arising from a bowl in a relief from the Hazāra Ramachandra

temple, Hampi, see reproduction in Prithvi K. Agrawala, Pūrṇa Kalaśa, the

Vase of Plenty (Varanasi, 1965), plate XX1 and figure 19. E. B. Havell, A Handbook

of Indian Art (1920/1927), p. 43, had rightly recognized that “The vase forming

the base of the pillar stood for the cosmic waters.” James Burgess, Indian

Antiquary 12 (1883), p. 321, reported about a South-Indian ritual in which

“a decorated kumbha represents king Varuṇa” (quoted by Coomaraswamy,

Yakṣas II, p. 62 n. 2). Cf. the mantra used in pouring water into a vessel: astu rājā

Varuṇo revatībhiḥ (Āśv. GS. II.9.5). That the bell-shaped capital represents the

inverted cask on top of the world axis in the nocturnal sky is now contested by

John Irwin, Burlington Magazine 117 (1975), p. 636f. For further references

concerning the pūrṇaghaṭa see M.-Th. de Mallmann, Les enseignements icono-

graphiques de l'Agni-purāṇa p. 242f. and below, ch. II n. 232.

86

and Mitra liberate the sun after its “Nachtweg” through their world,

the moon, whether visible or not, permanently belongs to Varuṇa's realm.

From the latter's close connection with the serpents it can be understood

that while in the Rigveda Varuṇa must have been thought to hold the

roots of the world tree and thus to support the universe, in the epic it

is the world serpent Śeṣa who has taken over this function 328.

The location of Pātāla as the dwelling-place of the Daityas-Dānavas

in the nadir (that is, at the lower end of the world axis) in Mhbh. V.97.1

is also met with in Sāyaṇa's commentary on ŚB. I.4.1.34. The brāhmaṇa

here tells how the Devas and Asuras were contending to win Gāyatrī,

while the latter was standing between the two parties 329. The author of

the brāhmaṇa equates Gāyatrī with the earth, which can, indeed, be

considered to be between heaven (Devas) and the nether world (Asuras).

In this connection the commentator has the following remark: “For on

the top of the (cosmic mountain) Meru is the town named Amarāvatī;

for in it the gods are dwelling. And so, the town that is situated under

the Meru is named Irāmukha. In it, indeed, the Asuras are dwelling.

Between these two there is the earth . . .” 330. Cf. Mhbh. V.108.9, where

Varuṇa's realm is said to be the foundation (mūla) of the Himālaya.

The common name Pātāla is again found in the Paisācī-passage of the

Kuvalayamālā (§ 139), which runs as follows: “More delightful is the

King of Mountains, the Mountain of the thirty(-three Gods), where the

divine damsels wandering at their sweet lovers' lineages, where even the Pātāla drips with the sweat of their joy,

328 See IIJ. 8, p. 108 and cf. Mhbh. V.101.2 eṣa Śeṣah sthito nāgo yene 'yam dhāryate

sadā . . . mahī, VII.69.48 adhastād dharaṇīm yo 'sau sadā dhārayate nrpa, sa Śeṣah

pannagaśreṣṭhah. For adhastāt see above, n. 183 on TS. V.59.4 devás tvé 'nrajyeṣṭhā

váruṇarājāno 'dhisthāc co 'pāriṣṭāc ca pāntu. For Varuṇa's connection with the world

axis in the Rigveda cf. VIII.41.10bc yó skambhéna ví ródasī ajó ná dyâm ádhārayat

(like the god Dhartṛ in the centre AS XVIII.3.29). The later replacement of Varuṇa

in this function by Śeṣa, the underworld aspect of the god of cosmic totality, is

quite in keeping with the general evolution of Indian religion as sketched above,

p. 40. For Varuṇa's connection with the serpents see below, n. 332, Shendge

BDCRI. 9, p. 281, G. Johnsen, IIJ. 9, p. 260, M.-Th. de Mallmann, Les enseignements

iconographiques de l'Agni-Purāṇa, p. 132 n. 2 (Varuṇa represented with a snake

in his hand), ŚB. III.1.1.7 (in the classification system the snakes are in the West:

yá pratīcí sá sarpánām) and, o.g., Mhbh. II.9.8, where Vāsuki and Takṣaka attend

on Varuṇa in his sabhā). As for the world tree, which belonged to both the nether

world and the upper world, it was in a way both the samrāj Varuṇa and (during

the New Year festival, as Indradhvaja) Indra (see p. 138). This may have been

meant by AS. VII.86.3 samrāḍ asy āsurānām . . . devānām adhabhāg asi.

329 A different version occurs in MS. II.1.11 (13,8), where the Gāyatrī is identified

with the year (line 16 samvatsaró vat gāyatrī).

330 Sāyaṇa ad ŚB. I.4.1.34 Meror agrabhāge hi Amarāvatī nāma nagarī, tasyām hi

devāḥ. tathā Meror adhastādbhāge ca yad [yat, Weber] Irāmukham nāma nagaram.

tasmin hy Asurā nivasanti. The name Irāmukha for Pātāla is only found here

(PW. pw.). Other names, such as Hiraṇyapura (name of the capital of Pātāla),

occur in the Mahābhārata.

(see p. 74).336 The first trace of the transfer of Varuṇa's function to

Yama, however, can be found in the Rigveda, e.g., IX.113.8337.

The presence of demons in Varuṇa's home may at first sight seem

somewhat puzzling because the demons have been driven from "these

worlds" (plural, see n. 24), whereas Varuṇa's abode must have formed

part of "these worlds". The position of the Ādityas among the Devas

is in itself sufficient proof for their belonging to the ordered cosmos. So

the idea of demons dwelling in Varuṇa's world must have offered some

difficulties to the epic poets. The question naturally arises how they

visualized this presence of the demons. The answer is twofold.

On the one hand, one meets with the notion that Varuṇa in his sabhā

is not only surrounded by the Ādityas, but also respectfully worshipped

by Daityas and Dānavas, "adorned with brilliant ear-rings"338. He is

an object of worship for various bhūtas (see nn. 289, 303), which term

may comprise the spirits and "devils". It is not surprising that also

Vāsuki and Takṣaka, the kings of serpents, are there, along with other

Nāgas339, since this is the natural consequence of the identity of (or,

close relationship between) varuṇālaya and nāgaloka, and of Varuṇa's

lordship over the serpents.

Varuṇa's world appears to be the refuge for demons after they have

been slain (e.g. Rāvaṇa) or expelled from the earth: "The great Asuras,

afflicted by the gods, entered the earth and the sea of saltish water"340.

Lomaśa relates that the Kāleyas at first were Vṛtra's allies (III.98.3ff.).

After seeing Vṛtra slain by Indra, they fly into Varuṇa's abode, but

at night they come back to kill seers and to destroy "people" (or, "the

worlds"). Because the sea is their refuge, they are unassailable341. This

passage is instructive in that it openly depicts Varuṇa's abode as a source

of evil, the underlying idea apparently being that Varuṇa protects the

336 Epic Mythology, p. 119.

337 yátra rājā váiśvāvatō yátra 'varódhanam dívāḥ, yátra 'múḥ yahvátīr ápas tátra

mám áṃtamím krḍhi.

338 Mbh. II.9.7 Ādityās tatrá Varuṇám jaleśvaram upāsate, 9.15 daityadānavasamghāś

ca sarve rucirakuṇḍalāḥ . . . (17) te tasyām Varuṇam devam dharmapāsādharam sadā,

upāsate . . .

339 See Mbh. II.9.8 Vāsukis Takṣakas caiva nāgāś cai 'rāvatās tathā . . . (11) ete cā

'nye ca bahavaḥ sarpāḥ tasyām Yudhiṣṭhira, upāsate mahātmānam Varuṇam

vigataklamāḥ and III.42.6, where Varuṇa is surrounded nāgair nadair nadībhiś ca

Daityaiḥ Sādhyaiś ca daivataiḥ. The nāgaloka into which Takṣaka enters (see n. 308)

must, accordingly, have been Varuṇa's world. For the Sādhyas residing in the

nadir see the Excursus, e.g., p. 88 and n. 24 in fine.

340 Mbh. I.17.28 tato mahīm lavaṇajalām ca sāgaram mahāsurāḥ praviśiśur arditāḥ

suraiḥ.

341 Mbh. III.101.7ff.: Kāleya iti vikhyāto gaṇaḥ paramādarunah, taiś ca Vṛtram

samāśritya jagat sarvam prabādhitam (8) te Vṛtram nihatam dṛṣvā sahasrākṣeṇa

dhīmata, jīvitam parirakṣantaḥ praviśṭā varuṇālayam (9) te praviśyo 'dadhiṃ ghoram

nakragrahāsamākulam, utsādanārtham lokānām rātrau ghnanti muniṃ iha (10) na tu

śakyāḥ keṣayiṃ netuin samudrāśrayagā hi te, samudrasyā kaye buddhir bhavadbhiḥ

saṃpradhāryatām. Cf. also III.99.21 samāśrittya mahorimantam ratnākaram

Varuṇasyā 'layam ema and 100.1 samudram te samāśrittya vāruṇaṃ nidhiṃ ambhasām.

90

evil-doers, who at night attack the world of order. Cf. also Mhbh. III.160.11 "he protects the bhūtas" (bhūtāni parirakṣati) and Rām. VII.23.3, where Varuṇa is said to protect the Rasātala, dwelling-place of the Daityas and serpents (p. 88). It may be noted in passing that the notion of niśācaras "(demons) moving about by night" has its roots in the Vedic ideas about rakṣases342. After being defeated by the Devas those of the Asuras who survived the slaughter split up the earth and vanished in the Pātāla343. Their defeat seems to have changed their position fundamentally. They live on344 in the nether world as exiles (III.101.8, 103.12), no longer protected by Varuṇa. Similar accounts are found elsewhere, e.g. in the Matsyapurāṇa, where the Dānavas flee into the rasātala or pātāla (the two terms are here used optionally)345. In the Kāleya-episode of the Mahābhārata there is, accordingly, a difference between the spontaneous flight in the beginning, during which they are in Varuṇa's abode and their second flight into the Pātāla, where Varuṇa is not mentioned by name. However, whether the author of this episode recognized this or not, Pātāla is also Varuṇa's abode (V. 106.12) at the bottom of the world-tree. So some kind of relation must have continued to exist between the dispelled demons and Varuṇa.

Just as in the Veda (n. 24), the demons are driven away from the earth346. But the epic bards, like the Vedic theologians (p. 34), were confronted by the dilemma that the demons, though dispelled, cannot be definitively destroyed. As Mhbh. III.98.3-5 puts it, the Dānavas and Asuras in Hiranyapuram cannot be killed by Śakra, nor by Varuṇa, Yama or Dhanada (Kubera). Similarly Rām. VII.23.10. In XIII.140.3ff. the gods defeated by the Asuras wander about on earth till they see Agastya. At their request he burns the Dānavas with the fire of his tejas. By thousands they fall down from the atmosphere on earth but others "leaving the two worlds" go to the South (kāṣṭhāṁ dakṣiṇām), where Bali is just performing the horse-sacrifice under(?) the earth (mahīṁ gataḥ)347. The Asuras who are under and in the earth (mahīstha) are not burnt. When the gods ask Agastya to kill them, too (bhūmiṣṭhān asurān jahi), he declares himself unable to burn the mahīgata Asuras.

342 KS. X.5 (129,19) āmāvasyāṁ vai rātriṁ niśi rākṣāṁsi prérate (Caland: prerate [?]), TS. II.2.3 niśiṁtāyāṁ hi rākṣāṁsi prérate. Cf. II.4.1.1 and Geldner, Ved. Studien II, p. 167 n. 1.

343 Mhbh. III.103.12–13 hatakaseṣas tataḥ kecit Kāleyā manusottama, vidārya vasudhāṁ deviṁ Pātālatalaṁ āśritāḥ (13) nihatān Dānavān dṛṣṭvā tridaśā munipumgavam, tuṣṭuvur vividhair vākyair idam caivā 'bruvan vacah.

344 Incorrect Hopkins, Epic Mythology, p. 119.

345 E.g., Matsyapurāṇa 47.63, 212, 233 (viviśus te rasātalaṁ, viz. Dānavāḥ) and 47.68, 215 Pātālam.

346 Cf. Mhbh. I.58.25ff. (so tasmāl lokād iha cyutāḥ).

347 The crit. ed. (śl. 12) reads mahīgatān. Sørensen, Index, p. 17a, rightly translates mahīṁ gataḥ by "in the nether regions". Neither mahīgata nor mahīstha (mahīstha) is recorded in the Petersburg dictionaries, but cf. PW. bhūmiṣṭha "in the earth" and bhūtā bhūmiṣamsthitiḥ, IIJ. 15, p. 177.

91

Varuṇa's character is accordingly basically ambiguous. Although he never ceases to be a Deva nor is called an Asura or king of the Asuras 348, some variant readings testify to an uncertainty, at least among some copyists, about the god's nature. In VIII.30.77 the critical edition reads praticim Varuṇah pāti pālayānaḥ surān bali "Varuṇa the mighty one, guards the western region, protecting the gods" but the Bombay edition (VIII.45.32) has here a different reading pālayann asurān bali "protecting the Asuras". A similar variant, which is mythologically the lectio difficilior, occurs in the Calcutta edition XII.4497 apāṁ rājye 'surānāṁ ca vidadhe Varuṇaṁ prabhum "(Viṣṇu) (installed him) as king over the water and made him lord of the Asuras" 349.

While in the passages here referred to the ambiguity of Varuṇa's nature was more or less openly acknowledged and left unsolved - and insoluble it was, indeed - there are other passages where Varuṇa is depicted as a warder of the imprisoned demons. Mythologically this must, in the light of the preceding observations, be considered a rationalistic attempt to explain away the dark aspects in Varuṇa's relation to the banished demons. Such a passage is Mhbh. V.126.44ff., where the highest god Prajāpati orders god Dharma to fetter the Daityas and Dānavas and to hand them over to Varuṇa. The latter fetters them also with his own nooses (pāśa) and guards them henceforth in the sea 350. Less clear is the situation in III.42.6, where Varuṇa, surrounded by Nāgas, rivers, Daityas and Sādhyas is coming to Arjuna, but then tells him that he has fettered thousands of Daiteyas by means of his nooses 351. The ocean which Kadrū and Vinatā are watching (I.19.6) is described in the words

348 Kings of the Asuras and Dānavas are, e.g., Bali, Śambara and Vṛtra, the last of whom is often called Asurapravīra, e.g. XII.271.44. This function of Vṛtra is already met with in the Veda, see MS. IV.3.4 (42,14) and cf. Geldner, Vedische Studien II, p. 299 n. 3.

349 The text-critical problem has, of course, to be judged by its own criteria. In XII.122.29 the critical edition reads apāṁ rājye surānāṁ ca vidadhe Varuṇaṁ prabhum (see n. 299) and records only one variant reading 'marāndāṁ in a single manuscript. But the writing or omitting of the avagraha was largely left to the discretion of the copyists, who besides used it in a way different from what has become usual in modern printed texts. So reading 'surānāṁ for surānām can hardly be called an emendation. Moreover Varuṇa is not, as a rule, said to be a protector of the gods. In IX.46.11-12 (see p. 80) he protects "the rivers, seas and lakes just as Indra protects the gods." The reading of the Calcutta edition 'surānāṁ . . . prabhum "lord of the Asuras" certainly makes better sense.

350 Mhbh. V.126.44ff. iti matā 'bravid Dharmam parameṣṭhī Prajāpatih, Varuṇāya prayacchai 'tān baddhvā Daiteyadānavān (45) evam uktas tato Dharmo niyogāt parameṣṭhinah, Varuṇāya dadau sarvān baddhvā Daiteyadānarān (47) tān baddhrā dharmapāśaiś ca svaiś ca pāśair jaleṣvarah, Varuṇaḥ sāgare yatto nityam rakṣati Dānavān.

351 Mhbh. III.42.5f. tato vaiḍūryavarnābho bhāsayan sarvato diśah, yādoganavṛtaḥ śimān ajagāma jaleśvarah (6) nāgair nadair nadībhiś ca Daityaiḥ Sādhyaiś ca daivataih, Varuṇo Yādavāṁ bhartā vasi tām deśam āgamat and vv. 27-28, where he says to Arjuna: (27) mayā samudiyatāṁ pāśān varuṇān amīśvarān, pratigrhṇīṣva Kaunteya sarahasyaivartanān (28) ebhis tadā mayā vira saṅgrāme tārakāmaye, Daiteyānāṁ sahasrāṇi samyatalāni mahātmanām.

92

'abode of the Pātāla-Agni (= āsuro 'gnih, n. 326) and prison of the Asuras'

(pātālajvalanāvāsam asurāṇām ca bandhanam). Here again there are some

significant variant readings, which suggest the idea that the notion of

captive Asuras was not generally accepted 352.

The only conclusion that can be drawn from these data is that also

in the epic Varuṇa, far from being "un simple dieu des eaux" (p. 75),

was still essentially the same ambiguous and ominous god as in the Veda.

Although the epic is less reticent than the older texts, the traditional

taboo is still prevalent. Judging by the indirect indications one is bound

to conclude that as a rule it prevented also the epic bards from openly

referring to the god's dark side. What is said about him explicitly only

relates to his opposite favourable aspect. Typologically the epic Varuṇa

belongs to the category of such two-sided figures as Viśvarūpa, who

publicly was the purohita of the gods but secretly of the Asuras, or

Vidura 353.

Hopkins incidentally touches upon the general problem with which

Varuṇa's relation to the demons confronts us but his explanation does

not go to the root of the matter. He writes: "Ordinarily the Devas exclude

the demons; they are as light to darkness, but (as shown below) all spiritual

beings are sons of the Father-god and so all are divine. It is rather the

nature of the individual which determines whether he is "god" or "demon",

than the class to which he is assigned" 354. In Varuṇa the divine and

the demoniacal are linked in a suprarational unity, which eludes every

attempt at a logical formulation 355. In contradistinction to Hopkins'

opinion this is not merely an isolated feature of the god alone but a

structural trait that can also be found in other mythological figures. One

of these will be studied more in detail in the next section.

16. Ambiguous Figures: Uśanā Kāvya and Viśvarūpa

A classical instance of the same ambiguity that has been pointed out

in Varuṇa is Uśanā Kāvya. (See also Dumézil, Mythe et épopée II,

pp. 133-228). The latter stemmed from the Proto-Indo-Iranian mythology

but in the Sanskrit epic (where he also bears the name Śukra) he had

become the father-in-law of Varuṇa, who married his daughter Devī 356.

Of old he bears the title kavi and in the epic he is the guru of gods and

Asuras 357. Of his relation to both parties that to the Asuras is the more

352 For asurāṇām ca bandhanam some manuscripts read asurāṇām nivesanam

"dwelling-place of the Asuras," asurāṇām tathā 'layam "and also the abode of the

Asuras," asurāṇām ca bāndhavam "and cognate of the Asuras."

353 For Viśvarūpa see also p. 101.

354 Epic Mythology, p. 3 n. 1.

355 Cf. Renou, EVP. 7, p. 4: "autour duquel on aurait voulu laisser planer une

sorte d'ambiguïté fondamentale." The words "on aurait voulu" reflect a different

view of the relation of Vedic man to his myth from the one here taken.

356 Mbh. I.60.51 Varuṇasya bhāryā jyeṣṭhā tu Śukrād Devī vyajāyāta.

357 Mbh. I.60.40 Bhṛgvoḥ putraḥ kavir vidvāñ Śukrah kavisuto grahaḥ . . . yogācāryo

mahābuddhir Daityānām abhavad guruḥ, surāṇām cāpi medhāṁī brahmacārī yataivrataḥ.

Cf. Sørensen, Index, p. 312b.

93

significant one for he is said to be their teacher (upādhyāya) and priest (purohita)358. As such he is the adversary of Bṛhaspati, whom the gods have chosen as their priest. In the last-quoted passage Kavi Uśanā has a special power (māyā), which is called the vidyā samjīvanī "science of revivification" and which at first ensures the superiority of the Asuras. Kavi Uśanā is the prototype of the kavis359. All this is rooted in ancient Vedic and pre-Vedic tradition360, according to which Uśanā Kāvya was originally the purohita of the Asuras, just as Bṛhaspati was of the gods. How essential Uśanā's relation to the Asuras was is shown by the fact that sometimes he is represented as the messenger of the Asuras, in which function his divine counterpart is Agni361. His connection with the Asuras was, therefore, more essential than his priesthood, or the opposition between him and Bṛhaspati.

As far as these details are concerned the epic tale (Mhbh. I.71.5-6) corresponds with the older sources. In some of the latter, however, important additional details are found. According to one version (occurring in the Jaiminīya Brāhmaṇa and in Baudhāyana) there was a Gandharva, who stood above the two contending parties of the Devas and Asuras and who was the only one to know the issue of the endless war. On account of his position it is probable that this Gandharva was conceived as still

358 Mhbh. I.71.5f. surāṇām asurāṇām ca samajāyata tai mithah, aiśvaryam prati samgharṣas trailokye sacarācare (6) jigiṣayā tato derā evaire 'ngirasam munim, paurohityena yājyārthe Kāvyam tū 'sanasam pare, brāhmaṇau tāv ubhau nityam anyonyaspardhinaṁ bhram. The Asuras are victorious until Uśanā goes over to the gods, cf. also Matsyapurāṇa 47.61 tato 'surān parityajya Śukro derān agacchata.

359 Mhbh. VI.32.37 (=Bhagavadgītā X.37) munīnām apy aham Vyāsaḥ, kavīnām Uśanā kaviḥ. Cf. RS. IV.26.1 kavir Uśanā.

360 See Geldner, Vedische Studien II (1897), p. 167ff., Caland, Über das rituelle Sūtra des Baudhāyana [Abhandlungen für die Kunde des Morgenlandes XII (1903)], p. 26, H. Oertel, JAOS. 28 (1907), pp. 81-88 (on Jaim. Br. I.125-127). J. Charpentier, Kleine Beiträge zur indoiranischen Mythologie [Uppsala Universitets Årsskrift 1911], p. 82, E. Sieg, Sonnenrennen im Rigveda [Nachr. Ges. Wiss. Göttingen. 1928], pp. 196-198, A. Hillebrandt, Vedische Mythologie II2 (1929), p. 423, H. Lommel, "Kārya Uśan", Mélanges Bally (1939), pp. 209-214.

361 PB. VII.5.20 Uśanā vai Kāryo 'surāṇām purohita āsit; JB. I.125 line 1 Devāsurās samyattā jyon na vyajayanta, Bṛhaspatir devānām purohita āsid, Uśanā Kāryo 'surāṇām, I.126 line 2 sa (viz. *triśiṣṭaḥ Gandharvo) hṛdaya imaṁ samam vidatur, Bṛhaspatir ayam deveṣu, 'śanā Kāryo 'sureṣu; Baudh.ŚS. XVIII.46 (401,101) Devāsurā ha yatara mahāsaṁgrāmaṁ samyetire, tad dhe 'māni bhūtāni dvedhai 'va vyapacakramur, devān eva 'nu anyān, asurān eva 'nu anyāni. Bṛhaspatir devānām purohita āsid, Uśanā Kāryo 'surāṇām; Śāṅkh. ŚS. XVI.27.1 Uśanā ha Kāryo 'surāṇām purohita āsa (commentary: Śukrah kaveḥ putraḥ, cf. Mhbh. I.60.40 Śukraḥ kavisuṭa). Lommel conjectured (Mélanges Bally, p. 211) that the epithet Kārya has been inherited from Proto-Indo-Iranian and is identical with Mlr. Kā(ū)s. For Uśanā and Agni as the messengers of the Asuras and the Devas respectively see TS. II.5.8.5 Agnir devānām dūtá āsid, Uśanā Kāryo 'surāṇām, tau Prajāpatim praśnám aitám (but in ŚB. III.5.1.21 Sahāraksas is thus opposed to Agni). In Maitrī Up. VII.9 Bṛhaspatir, disguised as Uśanā, gives the Asuras a wrong instruction: Bṛhaspatir vai Śukro bhūtvā 'ndrasya 'bhayāyā 'surebhyah kayāye 'mām avidyām asṛjat. Cf. van Buitenen, Maitr. Up., p. 88 "Bṛhaspati who has the (late) reputation of being a false teacher." This motif has been elaborated in, e.g., Matsya Pur. 47.183ff.

94

standing quite close to the primordial unity of the undifferentiated Chaos. A confirmation of this suggestion can be seen in the Rigvedic myth of Yama and Yami, where the Gandharva as their ancestor occupies a similar position. The interesting detail then is that on the part of the undivided primordial world there is a knowledge of things to come (or, perhaps, of the deeper nature of all that is) that has gone lost in the two moieties of the dualistic Cosmos. The Gandharva knows that it will be impossible for any of the two parties to gain the victory, unless it can persuade the purohita of the other party to go over 382. In other words what is found here is another instance of the evocatio deorum 383, already met with above where Indra exhorts Soma to come outside with the words niréhi soma 384. In the same way, Indra here, after having learnt by a ruse the ‘truth’ of the Gandharva, goes to Uśanā Kāvya and tries, in his quality of protagonist of the Devas, to induce him to go over 385.

The myth of Kavi Uśanā, accordingly, shows the same pattern that was seen in Varuṇa’s going over to the Devas (RS. X.124.4). In the light of this conclusion, however, the current interpretation of the Rigvedic references to Uśanā Kāvyá cannot be accepted as correct 386. When interpreted in the light of the later Vedic evidence, however, they appear to present difficulties which have escaped notice so far.

It must first be stressed that Kāvya Uśanā is here clearly a man, who refers to himself as a mártya. Cf. X.22.6 ‘As ye both [viz. Indra and Kutsa] arrive, Uśanā asks you: ‘For what purpose (do ye enter) our house? Ye have come from afar, from the heaven and the earth, to a mortal man’ ” Indra seeks spaciousness for him (VI.20.11): “Thou, O Indra, wert the first of the helpers, as thou wert looking for a way out for Uśanā Kāvya”. Cf. X.40.7: “Ye, O Aśvins, have aided Bhujyu, ye Vāsa, Siñjāra and Uśanā” 367. The persons here named were historical or legendary men:

382 JB. I.126 line 5 tayor yataro yatarān upasamesyati te jesyantī ‘ti. 383 The phrase evocatio deorum ex urbibus obsessis, used by Macrobius (Saturnalia 3.9) has here been adopted as a succinct term for this mythological process. The corresponding Sanskrit term is upamantrana. Cf. Livy 1.55.5 and 5.21.5 (deos) alios ex urbe sua evocatos. 384 RS. X.124.5. See above, pp. 17–20 and cf. X.60.7 éhi nír ihi. 385 JB. I.126, lines 7–8 sa ho ‘kanasam̆ Kāvyam ājagāmā ‘suresu. tam̆ hovāca : rse, kam̆ imam̆ janam̆ (“foreign people”) vardhayasy, asmakam̆ vai tvam asi, vayam̆ vā tavā ,smān abhyupāvartasve ‘ti, Baudh. XVIII.46 (403,2) sa ha gatvai ‘vo ‘sanasam̆ Kāvyam upamantrāyam̆ cakre, Jayantyās ca duhitrā catrasbhīs ca kāmadughābhīs. sa hā ‘ñanto [!] ‘surebhyo ‘dhi vapasamiyāva. tato ha vā etad devā asurān mahāsam̆grāmam̆ jigyuḥ. The situation as is described in JB. has a close parallel in ŚB. I.6.3.13 (see n. 49). For jana “foreign people” see Minoru Hara, Pratidānam, p. 256ff. (with references), for upamantrāyati see nn. 231, 418. 386 See Bergaigne II, pp. 338–341, Macdonell, Vedic Mythology p. 147. The seventeen passages where Uśanā (Kāvyá) occurs are I.51.10, 11, 83.5, 121.12, 130.9, IV.16.2 26.1, V.29.9, 31.8, 34.2, VI.20.11, VIII.7.26, 23.17, IX.87.3, 97.7, X.22.6, 40.7. 367 RS. X.22.6 ádha gmánto ‘sánā prchate vām̆ kádarthā na á grhám, á jagmathuh parākád divás ca gmás ca mártyam, VI.20.11 tvám̆ vrdhá indra pūrvó bhūr varivasyánn Uśáne Kāvyáya, X.40.7 yuvám̆ ha Bhujyúm̆ yuvám̆ Aśvinā Vásam̆ yuvám̆ Siñjāram Uśánām̆ úpārathuh.

Bhujyu is elsewhere said to be the son of Tugra (I.117.15), and Vāśa was the purohita of king Pṛthuśravas (I.116.21, etc.). Only about Śiñjāra (whose name points to a non-Āryan background) further data are lacking.

As a human being, Uśanā was a sacrificer, who invigorated Indra with Soma before the god's fight against Śuṣṇa (I.51.10-11): “As Uśanā with force fashions thy force, (thy) violence thrusts asunder the two worlds with their greatness” (11) “Indra, when revelling at Uśanā's, mounts the (chariot with the) horses, which run ever faster. The rigorous one (mounts) the (chariot which) hastens along. He effused the water in streams. He broke open the firm strongholds of Śuṣṇa” 368. The “force” (sáhas) which Uśanā here “fashions” for Indra is only a metaphor for the Soma. In the same way the Soma is called a vá̄jra in I.121.12 “The intoxicating (beverage), which Kāvya Uśanā gives to thee, which he fashioned as a resistance-breaking vá̄jra that belongs to new year's day (?)” 369 or indirectly denoted as such in V.34.2 “He, the liberal one, who filled his belly with Soma and revelled in the (beverage of the) sweet plant, when Uśanā, who has great weapons, gave him the thousand-pointed” weapon to kill the animal” 370.

When Uśanā appoints Agni as his hótr 371, or is said to be supported by Indra and the Aśvins 372, this still agrees with the current picture of the Vedic sacrificer. Things become different when his quality of kavi is stressed. This was the most marked characteristic of Uśanā, who is the only Proto-Indo-Iranian kavi whose name has been handed down to us. Uśanā is introduced as saying “I am the kavi Uśanā, look at me” 373. Kavi was the specific term to denote an initiate who, as a devotee of Varuṇa, had received his initiation and knowledge of the cosmic mysteries (medhá̄) in the nether world 374. This character of the initiation explains how the mortal Uśanā could become the priest of the Asuras.

On the other hand, since Uśanā as the kavi par excellence must have stood in close relation to Varuṇa, the ancient Vedic tradition about his being the purohita of the Asuras may be considered to prove indirectly

368 RS. I.51.10 tákṣad yát ta Uśanā sáhasā sáho vi ródasī majmánā bádhate sávah (11) mándișṭa yád Usáne Kávye sácām Indro vaṅkú vaṅkutárd 'dhi tiṣṭhati, ugró yayíṁ nír apáh sr̥tásā 'srjad ví Súṅnasya dr̥mhítā áirayat pírah.

369 I.121.12 yáṁ te Kāvya Uśanā mandinám dad vr̥traháṇam páryam̀ tatakṣa vá̄jram.

This translation is different from Geldner's. For dád see see K. Hoffmann, Injunktiv p. 191 n. 157: 'genereller Injunktiv oder prospektiver Konjunktiv'; for párya see IIJ. 5, p. 181; for vá̄jra see Bergaigne, II, p. 253, Oldenberg, Noten, and Hillebrandt, Ved. Mythologie I, p. 336 n. 2.

370 RS. V.34.2 á yáh sómena jatháram ápiparlá 'mandata magháravā mádhvo ándhasah, yád ím mr̥gáya hántave mahávadhah sáhasrabhṛṣṭim Uśanā vadhaṁ yámat. The soma acted for Indra as a vá̄jra. This explains the use of the adj., four times an epithet of the vá̄jra, as an epithet of Soma (twice). Cf. IIJ. 12, p. 283.

371 RS. VIII.23.17 Uśanā Kāvyaś tvā ní hótrāram asādayat, áyajtiṁ teá Mánave jātávedasam.

372 RS. VI.20.11 tváṁ vr̥dhá Indra pūrvayó bhūr varivasyám Uśáne Kávyāya, X.40.7 yuváṁ ha Bhujyúṁ yuváṁ Aśvinā Vāśám yuváṁ Siñjáraṁ Uśánām úpārathuḥ.

373 RS. IV.26.1 aháṁ kavír Uśanā páśyatā mā.

374 Cf. IIJ. 4, p. 187, 8, pp. 110, 127, and especially VIII.6.10 medháṁ rtásya.

that also in Vedic belief the devá ásura Varuṇa had a secret connection with the ásurā adevdh. It has been suggested above375 that at the end of the year Varuṇa was, again, temporarily one of them. As for Uśanā, it is characteristic of him that he himself is never called an Asura. The way in which Bauddhāyana accounts for his curious position in terms of the cosmic dualism is instructive: “When the Devas and Asuras were contending in the great war (mahāsamgrāma), then all these beings departed into two directions: some of them joined the Devas, others the Asuras. Bṛhaspati was the purohita of the Devas, Uśanā Kāvya of the Asuras”376.

That Uśanā, as a kavi, actually had a knowledge of the cosmic mysteries can indirectly be inferred from what is said of Soma in IX.87.3: “The inspired Ṛṣi, the leader of men, who is a skilful Ṛbhu and, owing to his prophetic inspiration, is Uśanā, found what was concealed of them [the cows], the hidden secret name of the cows”377. Since Soma is also a kavi378, it was natural for him to be likened to Uśanā. The same knowledge of cosmic secrets, however, is also attributed to Varuṇa himself as the divine kavi379, who is asked to reveal it to the poet380. This is the well-known pattern of the initiation of the poet-seer by Varuṇa381. Soma could be likened to Uśanā owing to his kāvya because the ancient Aryan *kavi Uśan(ā), being a mortal, had become the prototype of the Aryan seer in general. In a similar way Soma is likened to him in IX.97.7 “Openly professing his kavi-ship like Uśanā the god names the births of the gods”382, which may be paralleled with IV.16.2 “The (hótr̥) must recite a hymn of praise, a poem, he who is a vedhás like Uśanā, to thee, O Asuric one, who hast the insight”383.

On account of his being a kavi, Uśanā is also mentioned jointly with Atharvan and performs acts which have almost a cosmogonical character. Cf. I.83.5 “Atharvan was the first to clear the paths by means of sacrifices;

375 See above, p. 42.

376 See n. 361.

377 RS. IX.87.3 tṣir vipraḥ puraetā jānānām Ṛbhúr dhīra Ukánā kāvyena, sá cid viveda níhitam yád āsām apīcyaḿ gúhyam nāma gónām. Lüders, p. 521f., has convincingly demonstrated that this secret name is Uṣas. Cf. I. Renou, EVP. 9, p. 100.

378 Hillebrandt, Ved. Myth. I2, p. 370ff.

379 RS. VIII. 41.5 yó dhartā bhúvanānām yá usrāṇām apicyā védā nāmāni gúhyā, sá kavíḥ kāvyā purá rūpám dyaúr iva puṣyati . . . See Lüders, Varuṇa, p. 522.

380 RS. VII.87.4 uvóca me Váruno médhirāya tríḥ saptá nāmá ‘ghyā bíbharti, vidván padá́ sya gúhyā ná vocad . . .

381 Cf. IIJ. 8, p. 110ff. on RS. VII.88 (which is closely connected with the preceding hymn).

382 RS. IX.97.7 prá kāvyam Uśáneva bruvāṇó devó devā́nām jánimā vivakti. Ludwig II, p. 503 and Geldner translate jánimā as “Geschlechter”, Renou, EVP. 9, p. 46 renders “naissances”.

383 RS. IV.16.2 śaṁsā́tā ukthám Uśáneva vedhā́ś cikitúṣe asuryà̀ya mánma. The meaning of vedhás (“Meister” according to Geldner, “Meister in der Kunstdichtung” Lüders, p. 554) must have been close to that of kavi. See Renou, EVP. 16, p. 148 (also 4, p. 68, 7, p. 10 and 9, p. 91: “ce mot indéterminable”). Although cit- seems to have been an Indo-Iranian word for “vision”, its exact shade of meaning cannot be determined.

97

from these Sūrya, the seer, the guardian of vows, was born. At the same time (?sácā) Uśanā Kāvya drove the cows thither . . ."384

It is quite in keeping with Uśanā's function as purohita of the Asuras that the Rigvedic poets represent him as living "far away" or "coming from afar". This must have been a traditional notion since twice a pāda begins with the words Uśánā yát parāvátah, cf. I.130.9 (to Indra) "Since thou, O kavi, hast come from afar, together with Uśanā, for aid", V.31.8 (to Indra and Kutsa) "When the gods brought you into contact with Uśanā" and VIII.7.26 (to the Maruts) "When ye come from afar, with Uśanā, to Ukṣṇo Randhrā"385. On the other hand Indra and Kutsa are said to come from afar, from heaven and from the earth, to the mortal Uśanā386. The term "afar" (parāvát) is specifically used to denote the nether world and it is readily understood that Uśanā, living with the Asuras, had to come "from afar"387. When, therefore, it is said in X.22.6 that Indra (from heaven) and Kutsa (from the earth) had to come "from afar" (parākd) to meet the mortal Uśanā, this may be interpreted as implying that Uśanā was "far away", in the dwelling-place of the Asuras. In this connection it is particularly interesting that Uśanā asks them "For what purpose have ye come to our house?"388 As far as I can see the word "our" has been disregarded by all translators. It can only be fully understood, when it is realized that Indra and Kutsa, in order to meet Uśanā, had to go all the way from heaven and the earth to the nether world and the Asuras.

According to Geldner389 Indra and Kutsa came to ask Uśanā's advice. This opinion, which is characteristically based upon Uśanā's fame in the classical period as the author of a nītiśāstra, is hardly correct. Indra needed material support, not advice: the next stanza (X.22.7) refers to an incantation (bráhmó 'dyatam) for the benefit of Indra390. What the poet actually had in mind may be inferred from V.29.9 (to Indra and Kutsa): "When ye, O Indra, together with Uśanā, came to the house, with the powerful swift horses, thou camest here victorious, on the same chariot (with Kutsa); with (the help of) Kutsa and the Devas thou hadst

384 RS. I.83.5 yajñair átharvä prathamó pathás tate tátah sāryo vratapá vená ájani, á gā ájad Uśánā Kāvyáh sácā . . .

385 RS. I.130.9 Uśánā yát parāvátó 'jagann úláye kave (cf. Hoffmann, Injunktiv, p. 191 n. 155) V.31.8 sám ha yád vám Uśánā 'ranta devā́h, VIII.7.26 Uśánā yát parāvátā Ukṣnó rándhram āyátana (dyáur ná cakradad bhiyád). An accusative Uśánām (for Uśánām X.40.7), as assumed by Geldner ad I.130.9, V.29.9 and VIII.7.26, is impossible.

386 See n. 367.

387 It is certainly due to an assimilation of the well-known type that in I.130.9 Indra, addressed as kavi, is said to have come with Uśanā "from afar". On Indra's kavi-ship see H.-P. Schmidt, Brhaspati und Indra, pp. 151, 157f. and cf. IIJ. 13, p. 284. The same must be true of the Maruts in VIII.7.26.

388 X.22.6 kādarthā na á grhám, á jagmathuh . . .

389 Note ad X.22.4 and Vedische Studien II, p. 168 (for Uśanā's mantras), p. 169 (for his nīti !).

390 See J. Narten, Die sigmatischen Aoriste im Veda, p. 175.

98

overcome [or "then overcamest"]? Śuṣṇa"391. Whatever house may be meant here, Śuṣṇa's (before the fight) or Kutsa's (at the triumphal return after the fight, as is Sāyaṇa's interpretation), or perhaps even the poet's392, this much follows from the translation here given that Uśanā's presence was necessary for Indra to overcome the demon Śuṣṇa. Therefore, Uśanā's surprise at the unexpected arrival of Indra and Kutsa at the grhā́ of the Asuras ("our house" !) is quite understandable. Only a very serious purpose could have induced them to come to see him there, and this purpose can only have been the evocatio mentioned in later Vedic texts. The need for Uśanā's presence is clearly explained in the hymns: he is not yet (as Geldner thought) the wise counsellor, the author of a nītiśāstra which he was to become in the classical period, but the one who gave Indra the soma that became his vá́jra393. The "magical" strengthening by Uśanā was indispensable and the formulaic phrase "to come with Uśanā" expressed the certainty that the god would now be victorious. Since the Soma had to be won from the powers of the nether world (as the śyena-myth expresses in a mythical way, and the ritual of the somakrāyaṇa in a symbolical one) it is natural that Uśanā could act as a mediator between the two worlds. It is possible that according to one version of the myth Indra drank the Soma in Uśanā's house, that is, in the nether world394. Anyway, Uśanā had to side with Indra in order to enable the latter to fight Śuṣṇa and this involved the necessity of Uśanā's going over. The idea of the evocatio, as found in the brāhmaṇas, was also known to the Vedic poets (RS. X.124.6).

Leaving aside RS. I.130.9395 we may now turn to the epic, where Uśanā's "magical" power is described in greater detail as a vidyā́

391Geldner's principle that everywhere in the formulaic pāda-opening Uśánā yát there must be the same grammatical form and syntactice structure (Vedische Studien II [1897], p. 169) was quite sound, although the possibility of reinterpretation by later poets (in the sense of Manu Leumann's Homerische Wörter) or mere variation in the use of a traditional formula cannot be ruled out. He rightly, therefore, took Uśánā as an instrumental. Later, however, he changed his mind, taking it as a genitive in his Glossar (1907), and, apparently, as an accusative in his Übersetzung II, p. 26 (in spite of the reference to his Glossar). But his translation "Als ihr beide . . . zu Uśanas ins Haus kamt . . ." is grammatically impossible and presupposes a situation which it is difficult to explain. Ludwig II (1876), p. 109, rendered Uśánā as a vocative, which is impossible on account of the accent. This may have been the reason why in 1883 (vol. V, p. 95) he suggested that Uśánā here meant the vájra which Uśanā had given to Indra. Oldenberg took Uśánā as a nominative.

392Geldner, in a note on his translation of X.22.7-10, and Hoffmann, Injunktiv, p. 191, rightly point out that Indra's fight with the mrgá Śuṣṇa is a mythical exploit, which is represented as being repeated in the world of the poets. It is, therefore, of lasting importance.

393See nn. 368-370.

394Cf. I.51.11, quoted above, p. 96.

395RS. I.130.9 prapitvé vácam aruṇó muṣāyatí śáná á muṣāyati, Uśánā yát parāvátó 'jagann ūtá́ye kave. See Bergaigne II, p. 339 (Uśanā or Indra stole vāc?), Oldenberg, Noten I, p. 135, Hoffmann, Injunktiv, p. 191 ("ganz unklar"), etc.

99

samjivani (Mbh. I.71.9). Herman Lommel 396 has pointed out that this is a feature of Uśanā which can be traced back to the common Aryan mythology. In the Iranian tradition the well-known hero king Kay Kāūs 397 subdues the country of demons “und Kay Kāūs ist dadurch auch Gebieter über die Daemonen geworden. Diese errichten für ihn wunderbare Zauberschlösser im Elburzgebirge, wo ewiger Frühling und ewige Jugend herrscht, und gealterte Menschen verjüngt werden” (p. 212) 398. Since contact with demons and magic power (access to the source-wells of life) must of old have been characteristic of the Āryan kavi, the fight of Kay Kāūs in Middle Iranian sources is obviously a secondary aetiological myth and the Indian epic has here, as in other cases, preserved old features of the myth.

In this connection it is hardly accidental that in the Katha Upanisad I.1 Uśan(t)'s son Naciketas goes to Yama and the realm of the dead.

In the epic Uśanā has unquestionably an “underworldly” character: when Rudra is performing a dikṣā before his fight against Tripura, Uśanā cuts off some locks of hair. These then become serpents which wind about Rudra and thus make him Nilakantha 399. Just as in older texts, he is in the epic the purohita of the Asuras and, as such, the antagonist of Bṛhaspati 400. He possesses a knowledge which sometimes surpasses Bṛhaspati’s 401. He is known as a nītiśāstrakartṛ 402, his work having been revealed at the Auśanasa tīrtha 403, and is especially quoted as a military epigrammist 404, his instructions having the same value as those of Bṛhaspati 405.

396 See n. 360.

397 According to Lommel’s guess from Proto-Indo-Iranian *kari Kārya Uśan. See n. 361.

398 See A. Christensen, Les Kayanides, p. 73ff.

399 See Hopkins, Epic Mythology, pp. 180, 226, and cf. Mhbh. XII.329.15 (:342.26 Bomb. ed.) tripuravadhārtham dikṣām abhyupagatātasya Rudrasyo ‘sanasā śiraso jatā utkṛtya prayuktāḥ. tataḥ prādurbhūtā bhujagaḥ. tair asya bhujagaiḥ pīdyamānāḥ kanṭho nilatām upanītah. pūrve ca manvantare svāyambhuve nārāyanahastabandhagrahanān nilakanṭhatvam eva vā.

400 Mhbh. XV.3.13 Bṛhaspatir vā deveṣu Śukro vā 'pyASUREṣu yaḥ . . . and Bhāg. Pur. VIII.10.33, where Bṛhaspati, in the battle of the Amṛtamanthana, is fighting with Uśanā.

401 Cf. Matsyapurāṇa 47.81 mantrān icchāmy aham deva ye na santi Bṛhaspatau.

402 Thus Nilakaṇṭha in his commentary on V.39.30 Bomb. ed. Cf. Pañcatantra I śl. 185 (BBS, ed. Kielhorn) Uśanā veda yac chāstraṁ yac ca veda Bṛhaspatih, stribuddhyā na viśīyate tasmād rakṣyāḥ katham hi tāḥ (: Mhbh. XIII.39.7. tāḥ sma raksyāḥ katham naraih), Hitop. I śl. 114 (edd. von Schlegel and Lassen), I śl. 122 (ed. Nirṇaya Sāgara Press, Bombay 1947), not in the ed. Peterson (p. 27).

403 Mhbh. IX.38.6 tatra pūrvam tapas taplami Kāvyena sumahatmanā, yatrā 'sya nītir akhilā prādurbhūtā mahātmanah.

404 Mhbh. XII.136.127 buddhyā tvam Uśanāḥ sākṣād bale te adhikṛtā rayam, tvanmantrabalayukto hi vindeta jayam eva ha, XII.56.28 ślokau co 'śanasā gitau purā tātā maharṣinā, tau nibodhā, XV.12.15 Uśanā veda yac chāstraṁ tatraí 'tad vihitaṁ vibho, IX.57.12 api co 'śanasā gitāḥ śrūyate 'yaṁ purātanah, śokas tattvārthasahitas tan [v.l. taṁ] me nigadataḥ śṛṇu (13) punar āvartamānānāṁ bhagnānāṁ jivitaiṣiṇāṁ, bhetavyam ariśeṣāṇāṁ ekāyanāgataḥ hi te.

405 Mhbh. IX.61.48 Bomb. ed. (fragment 367 * after IX.60.46 crit. ed.) Bṛhaspater Uśanaso nopadeśaḥ śrutas tvayā.

Still more ambiguous is Viśvarūpa, son of Tvaṣṭr. Geldner has been

the first to point to "die eigentümliche Zwitterstellung des Tvaṣṭr im

ganzen RV." 406 . Since his patnī (or sapatnī) was an Asurī, the TS. II.5.1.1

says (in the translation by Keith) "Viśvarūpa, son of Tvaṣṭr, was the

domestic priest of the gods and the sister's son of the Asuras. He had

three heads, one which drank Soma, one Surā, and one which ate food.

He promised openly the share to the gods, secretly to the Asuras. Men

promise openly the share to every one; if they promise any one secretly,

his share is indeed promised. Therefore Indra was afraid (thinking), 'Such

an one is diverting the sovereignty (from me)'. He took his bolt and smote

off his heads" 407. The Mahābhārata, which relates the story in detail as

an itihāsa purāṇa in V.9.3–36, gives the version of the Taitt. Saṃhitā

(II.5.1.1) in XII.329.17–19 408, where the ambiguous behaviour of

Viśvarūpa is particularly stressed. In this light the role of Vidura in the

epic may also prove interesting.

17. THE EVOCATIO (UPAMANTRANA)

From the myths mentioned before there emerges a general picture,

which was briefly referred to on p. 20 but which it is now possible and

necessary to analyze in greater detail. The basic idea is that in the

primordial fight between Devas and Asuras, which is periodically repeated,

there is a balance of power so that neither of the two parties can gain the

victory. For this to become possible there must be a shift in the status quo,

which can be brought about in various ways:

A. Intervention of a god of the totality in favour of one of the parties.

Since these gods by their very nature transcend the dualism of the

406 See Vedische Studien II (1898), p. 168.

407 TS. II.5.1.1 Viśvárūpo vái tvaṣṭráḥ puróhito devánām āsīt, svasríyó 'surāṇām.

tásya tríni sī́rṣāṇy dísant, somapā́nam surāpā́nam annādánam. sá pratyákṣam devébhyo

bhāgám avadat, parókṣam ásurebhyah. sárvasmai vái pratyákṣam bhāgáṃ vadanti,

yásmā evá parókṣam vadánti, tásya bhāgó uditáh. tásmād Índro 'bīhed vái

rāṣṭrā́m vi paryāvartayatí 'ti. tásya vájram ádāya sī́rṣāṇy achinad . . . Cf. MS. II.4.1

(38,1) Viśvārūpo vái tvaṣṭrá āsīt trísīrṣá 'surāṇāṃ svasríyah. sá somam ekena sī́rṣṇā

'pibat, súrām ékenā, 'nnam ékena 'vayat. sá Índro 'manyutāḥ : 'yám vāvé 'dám bhavisyáti

'ti. téna sámalabhata. téna yugaśāráṃ aparat. sa tákṣāṇam tiṣṭhantam abravīd : ādhāve,

'māny asya sī́rṣāṇi chinddhī 'ti. tásya tákṣo 'paskadyā paraśúnā sī́rṣāṇy achinat.

tásmāt tákṣṇe śiro dhṛtám, KS. XII.10 (172,5) Viśvarūpo vái triśīrṣá 'sīt, tvaṣṭúḥ putro

'surāṇā́m svasríyas. sa somam ekena sī́rṣṇā 'pibat, annam ekenā 'vayat, súrām ekenā

'pibat. sá Índro 'manyatā : 'yam vā́ve 'dā́n bhaviṣyati 'ti. téna samalabdhata. téna

yugaśarám apadyata. sa tákṣāṇam tiṣṭhantam abravīd : ādhā́ve, 'sye 'mā́ni sī́rṣāṇi

cchinddhī 'ti. tásya tákṣo 'pādrutya paraśúnā sī́rṣāṇy acchinat. tásmāt tákṣṇas śiro

dhrtam (viz. at the sacrifice, cf. Mhbh. V.9.33 śiraḥ paśos te dāsyanti bhāgáṃ yajñéṣu

mānavā́ḥ). After Viśvarūpa's death Tvaṣṭr creates Indraśatru and Prajāpati dips

Indra's vájra, see n. 113.

408 In the epic the story is found in Mhbh. V.9.39ff. (esp. 42), VII.69.54, V.16.20

and 28 and (in a prose passage) XII.329.17, which literally quotes part of TS. II.5.1.1

(see Hopkins, Epic Mythology, p. 130f.). The Taittirīya version is not, however

the original one. Cf. MS. II.4.3 (40,14), where it is Tvaṣṭr himself who dips the

bolt, just as he fashions it in the Rigveda (V.31.4, etc.).

contending parties any step taken by them turns out to be decisive. It may be true that in the phenomenal world every victory is only temporary 409, but the myth describes the ideal structure of this process. According to the pattern of the myth the god may stand either between the two parties, or above them. In the former case it can be said that the god participates in either party, the clearest instance of which is Viṣṇu. Regardless of whether the belief of later Hinduism about his being eight months in the upper world, while sleeping 410 four months in the nether world, is in this form old or rather a later elaboration, it is in any case the clearest expression of Viṣṇu's nature. In this context it is evident that his taking three steps before Indra strikes at Vṛtra is deeply rooted in the constitution of the Universe as the Vedic Indians saw it 411.

A god of totality who stands above the parties is Prajāpati. He is best characterized as the god in whom the two moieties participate. His character of an impartial high god has been discussed above. Only rarely is he said to give direct support to the Devas 412. More often this is done in a subtle, indirect way 413. A clear illustration of the different positions of Viṣṇu and Prajāpati will be seen below in the epic account of the Churning of the Ocean.

B. An entity is surreptitiously stolen from one party and taken to the other. This is found in the myth of Mātariśvan's stealing Agni from the Asuras for the gods 414, and in that of the eagle (śyenā) who brings Soma to Indra. It is obvious that these two agents must originally have had a mythological status outside the two parties, otherwise they would not have been able to perform their acts. In the myths, however, no attempt is made to define their status more exactly.

C. One or more entities go over of their own free will:

1. as the result of their own decision, e.g., in the case of Agni who leaves the father Asura and prefers Indra (X.124.4) or in the epic Śrī, who during the Churning of the Ocean, leaves the primordial world and sides with the gods 415, or, having lived with the Asuras, prefers to live with Indra 416. Another instance is Vibhīṣaṇa in the Rāma-saga, who goes over from his brother Rāvaṇa to Rāma and is welcomed by the latter as the one who by his mere arrival guarantees Rāma's victory (p. 78).

2. as the result of a contest, the (mostly female) entity standing between the two parties. Although this could be considered a variant of 1),

409 See p. 34.

410 For the mythical meaning of sleep see p. 31.

411 Cf. Indological Studies - W. Norman Brown, p. 150 n. 88.

412 See n. 113.

413 For a characteristic instance from a later stage of Indian mythology, viz. Brahmā's position with respect to Devas and Asuras in the Nāṭyaśāstra, see above, p. 36 f.

414 Cf. Etudes Asiatiques 25, p. 95ff.

415 Mbhb. I.16.36 Śriḥ surā caiva Somaś ca turagaś ca manojavāḥ, yato devās tato jagmur ādityapathām aśritāḥ.

416 See p. 20.

the two additional elements of the contest and the intermediate position are so characteristic that it can better be considered a separate way of deciding the strife.

Some instances will be discussed below.

3. as the result of an evocatio 417.

While C 1-3 are different ways of looking at the same myth rather than different myths, B constituted a separate myth. The four possible variations can be illustrated by Soma being stolen by the śyená (passim in the Rigveda), Soma going over to the Devas (X.124.4c), his evocatio by Indra (X.124.6c niréhi Soma), and the contest in I.108.6 ayám Sómo ásurair no vihávyaḥ "We have to compete with the Asuras in calling Soma towards either side".

It has been seen above that the typical term for the 'evocatio' is upamantrayati, which does not mean "bestechen" 418 but simply "to call towards oneself" 419. If an entity is the stake of a contest, the Rigveda uses vi hvayete, vi hvayante, irrespective of whether a mythical or a real strife is referred to 420. Cf. vihavá "competitive invocation" (Wettruf, Wettanrufing II X2). In the Atharvaveda Agni is vihávya by the kings (II.6.4) and this is the technical name of AS. V.3 (cf. GB. II.2.34, perhaps also AS. VII.5.4, cf. V.3.11c). In the Yajurveda upamantrayante (MS. KS. TS. ŚB., also JB.) is the common term, icchante is only used in MS. 421, and nirhāya, corresponding to RS. niréhi, is a hapax in TS. II.5.2.4

417 See on this term n. 363.

418 Thus Geldner, Vedische Studien II, p. 167. For the form upa-mantráyati see also n. 231.

419 E.g., the gods try to persuade their mahimāṇah, which have fled to the Asuras, to come back. Cf. KS. XXIX.1 (167,17), KKS. XLV.2 (268,13/314,19) devāś ca vā asurāś ca samyattā asains. te devehyo mahimāno 'pakrāmanś. te 'surān agacchamś. tā́n 'māni cchandā́msi yā́ny ayajñavā́hāni. te devā abibhayur : ítham vā́va nas sarvā́n asurā apavatsyanti 'ti. tā́n y upāmantrayanta. te 'bruvan : vāryam̀ vrṇīmahā iti. Or the gods ask the sacrifice (their heritage from Prajāpati's bequest) to entice Vác away from their adversaries: ŚB. III.2.1.19 té devā́ yajñā́m abruvan, yóṣā vā́ íyám vā́c, úpa mantrasyasva, hvayisyáte vai tvé 'ti, etc.

In AS. VI.73.1 ("for supremacy") éhá yātu Várunah Somo Agnír Bháspátir Vásubhir éhá yātu, asyá śríyam upasámyātā sarva ugráseya cetíth sámmanasasah sajātáh

it is doubtful if a formal evocatio is meant. I know no passage where é há yātu is used in the sense of nir-ā-etu.

420 Cf., e.g., RS. X.80.5 agním ukthair ṛcayo vi hvayante, IV.24.3, 39,5, X.42.4, 112.7 vi hvayante (tám, Índram, tvā́, tvām), II.12.8 yám (Índram) víhvāyete. Only I.102.6 does not refer to Indra.

421 This may concern the téjo yajñíyam of the altar, or Gāyatrí, or Vāc. Cf. KS. XXV.6 (108, 17), KKS. XXXIX.3 (214,18/251,6) devāś ca vā asurāś ca samyattā āsains. tā́n yad asyā́s tejo yajñíyam āsīt, tat simhírūpam iva mahiṣírūpam iva bhūtvā 'ntarā 'tiṣṭhat. te 'vidur : yatarán vā iyam upāvartsyati ta idám bhaviṣyanti 'ti. tā́n devā upāmantrayanta. só 'bravíd: vāryam̀ vrṇai and MS. III.8.5 (99,6) yád vā asyā́ yajñítyam̀ yát sádevaṃ, tád ukrá́myā 'tiṣṭhad antarā devāsurā́n sāmyattā́nt simhírūpam iva bhūtvā́. té 'vidur: yatarán vā iyám upāvartsyáti tá idám bhaviṣyanti

'ti. tásyaṃ vā ubhā́ya aichanta. In the following passages Gāyatrí stands between the parties: KS. X.7 (133,8) devāś ca vā ásurāś ca sāmyattā āsains. tā́n gāyatry àntará 'tiṣṭhad ójo vīryàm annā́dyam parig̣hya. samvatsaró vāvaínān só 'ntarā 'tiṣṭhat . . .

103

(see p. 18), since AS. nirhvayāmi (Śaun. 2, Paipp. 4) is only used in magic. Similarly indram nirāhvayan (JB. 3.1695), whereas (indram) nirhvayante (GS 4) is only corrupt for nihvayante (AB., cf. Av. nizbayemi).

In a modern paraphrasis of these myths they ultimately to express the conviction that the Universe is based upon an asymmetric balance between life and death, the powers of life just turning the scale 422. The same idea apparently underlies the symbolism of the numbers 101 and 99, of which the latter is the typically demoniacal number in the Rigveda 423.

The "divine" number 101 has remained in use in its derived form (101 + 7 =) 108 424.

18. VARUNA AND THE CHURNING OF THE OCEAN

In conclusion of this study of Varuna's relation to the world of the shaped and the shapeless his position in the Amrtamanthana will be considered more closely. It has been seen above that he does not personally take part in the cosmogonical fight 425. The same appears to be still true

té 'vidur : yatarân vâ iyám upâvartsyáti tá idám bhárisyanti 'ti. tám ryahvayanta . . . sá devân upâvartata. táto devâ ábhavan, párâ 'surâ abhâvan ; MS. II.1.11 (13, 8) devâs ca vâ ásurâs ca 'spárdhanta. tám gāyatrī sárvam ānám parigṛhyā 'ntará 'tiṣṭhat. té 'vidur : yatarân iyám upâvartsyáti tá idám bhárisyanti 'ti. tásyām vâ ubhayâ

aichanta ; ŚB. I.4.1.34 devâs ca vâ ásurâs co 'bhâye prajāpatyòh pasprdhire. tám̂t spárdhamānām gāyatrá āntará tasthau. yá vai sá gāyatrá ásād, iyám vai sá prthivî, 'yám haivá tád antarā tasthau. tá ubháyâ eva vidám cakrur : yatarân vai na iyám

upâvartsyáti, té bhárisyanti, páré 'tare bhísyanti 'ti. tám ubháyâ eva 'pamantrayâm cakrire. However, in MS. III.7.3 (73,8)=KS. XXIV.1 (90,10), KKS. XXXVII.2 (195,20/228,4) vyahvayâmahâ iti is used. Similarly ŚB. III.2.4.4 ry âhrayanta. The last quotations refer to the contest between the gods and Gandharvas for supremacy.

In a similar context vyahvayetâm is used at PB. IX.2.22. Finally two quotations with Vâc standing between the parties: JB. II.115 line 3 sâ kruddâ . . . simhy ubhayatomukhî bhûtvá 'rdhvo 'dakrāmat. so 'bhayân devâsurân antarâ 'tiṣṭhad. yam devânām upámrad, yam asurânâm, tam [as is the reading of two manuscripts instead of tad] ādadānā. tám upaiva devā amantrayanto [see n. 231], 'pā 'surāh. sâ

devân abhyupâvartamānā 'bravît: kîm me tataḥ syād, yad vo 'bhyupâvarteya 'ti ; ŚB. III.5.1.21 tébhyo ha vâk cukrodha, kéna mâd éṣá sréyān bandhunâ3 kénâ3 yád etâm pratyágrahîṣṭa, ná mâm iti. sá haibhyó 'pacakrāme. sá 'bháyān ántarena devâsuránt sám̂yatânt sim̂hy bhûtvò 'dādānā cakāra. tám úpaiva devā amantrayantó,

'pā 'surā . . . (22) sá devân upâvartsyanty uvāca : yád vā upārártêya kim me tátah syād iti. Since the dvandva-compound devâsurdh is not, as a rule, used in the ŚB. (Delbrück, Altindische Syntax, p. 57) this passage may have been taken over from another brāhmaṇa. The specific terminology got out of use in the classical literature. Cf. Matsyapurāṇa 47.229 vaiṣṇavêto 'śrayadhāmas tau (viz. Śaṇḍamārkavau) as against TS. VI.4.10.1 té devâḥ Śaṇḍamārkāv upāmantrayanta. [Add ápa-âs- MS.4].

422 See M. Lietert Peerbolte, Psychocybernetica (Amsterdam 1968), p. 101.

423 Cf. the ninety-nine citadels of Śambara or the Dāsas, which is a mythological notion by which the reality of contemporary life could get a religious interpretation; and the ninety-nine Vṛtras (I.84.13f.). But the number 100 also occurs. Cf. Hillebrandt, Ved. Myth. II2, p. 245, Macdonell, Ved. Myth., pp. 159, 161.

424 On the number 108 see I. Scheftelowitz, "Die bedeutsungsvolle Zahl 108 im Hinduismus und Buddhismus", Studia Indo-Iranica (Ehrengabe für Wilhelm Geiger), Leipzig 1931, pp. 83-88.

425 See p. 54,

104

in the epic account of the Churning of the Ocean, which proves the

relatively archaic character of the epic in contrast with the Purāṇas 426.

Whatever the origin of this creation myth may be, its archaic character

cannot be questioned. Instead of Indra’s fight with Vṛtra resulting in

the latter’s death and Indra’s opening the primordial hill 427, we here find

a cooperation, which is at the same time a competition and fight

between the two groups of gods for the purpose of winning the life-elixir

(and other goods) directly from the primeval water. The role of

Mount Mandara is here limited to its functioning as the cosmic pivot,

used as a churning-stick. The two parties need each other to achieve their

purposes, which is symbolized by the technique of churning. This consists

in pulling a cord to and fro on two sides, thus making the churning-stick

rotate 428. Held, who discusses the symbolism of the churning in detail,

aptly remarks that this is the most concrete expression for the notion

of the cosmic pravṛtti and nivṛtti 429.

An important and significant detail of the epic tale is that three gods

do not take part in the competition.

Brahmā stands above the parties and, apart from making incidentally

a request to Viṣṇu, he confines himself to giving his permission to start

the contest.

Viṣṇu stands between the parties, in accordance with his fundamental

characteristic that he belongs to the two moieties of this world. Since

Gonda, Viṣṇuism and Śivaism p. 7, contests this, I may point not only

to Viṣṇu’s position, in Yajurvedic texts, in the middle of an arrow (ch. II, n. 275) and likewise of the jarjara (p. 151), but also to the

emphasis which our texts lay upon his position in the Churning of the

Ocean. In Mhbh. I.16.12-13 it is said that the Asuras took hold of the

head of the world serpent, and the Devas of its tail—which may be

connected with their character of older and younger brothers (Indol.

Studies - W. Norman Brown, p. 145 n. 51)—“but where the god,

Lord Ananta was [viz. wound as a rope around the churning stick,

Bhāg. P. VIII.7.1], there was Nārāyaṇa”. In this version Viṣṇu helps

the gods, in this intermediate position, by shaking the head of his serpent,

which then produces fire and smoke for the Asuras, but cooling rain

for the gods (more explicitly at Viṣṇupurāṇa I.9.86-7). His ambiguous

character becomes particularly apparent when the gods get tired and

ask Brahmā for Viṣṇu’s help, without which they are unable to complete

their work. When Brahmā then asks Viṣṇu to give the gods strength,

the latter pretends to be neutral (I.16.31) 430: “Strength I give to all who

426 E.g., Matsyapurāṇa 249–251, Bhāgavatapurāṇa VIII.5.17; 8.15.

427 Cf. History of Religions 10, p. 107ff.

428 Mhbh. I.16.12 (Viṣṇupurāṇa 9.84, Bhāg. Pur. VIII.7.1) netra (cf. navanīta

"fresh butter"); but Rām. I.44.17 and Mhbh. Kumbakonam ed.: yoktra.

429 G. J. Held, The Mahābhārata, pp. 146, 143 and for cooperation cum competition,

e.g., p. 171.

430 Cf. I.16.31 balam dadāmi sarveṣām.

105

have undertaken this work. Let all agitate the 'jar'. (For kalaśa see

India Maior, p. 149). Only later is he to help the gods by a trick, which

was apparently an essential part of the myth (I.16.39 māyām āsthito

mohinim) 431. In interpolated passages of the Rāmāyana version the tortoise

which carries Mount Mandara on its back, and which the Mahābhārata

refers to as “Akūpāra, king of the tortoises” (I.16.10), becomes a form

of Viṣṇu himself, while at the same time he also takes part in the churning

"amidst the deities" (App. I.8d, 9, v.l. mamantha devalānām ca madhye,

v.l. 8.25 devānām madhyatah sthitvā). Similarly Bhāg. P. VIII.7.8f. Most

elaborate is the Viṣṇupurāṇa (3rd or 5th cent. A.D.): “In the middle

of the Milk Ocean was Lord Hari himself in the shape of a tortoise, as

a pivot for the mountain which turned round as the churning-stick. In

one shape, bearing his discus and club, he pulled the king of serpents

among the gods, in another among the Daityas. But above, he was, as

Keśava, standing on the top of the mountain in still another, great shape

of his, which was invisible to gods and Asuras” (I.9.88–90). Similarly

Bhāg. P. VIII.7.11 (āsureṇa rūpeṇa) and 12. This version comes nearest

to the bas-reliefs of Angkor Vat where, in a representation of the Churning

of the Ocean, Viṣṇu is seen, standing in the middle between the contesting

parties and joining hands with Devas as well as with Asuras. Beneath

is the tortoise as another shape of the god and above the mountain he

appears in his most heavenly shape, pressing down with both hands the

mountain top. Cf. nagāgram piḍayām āsa vāmaḥastena and parvatāgram

tu lokātmā hastenā 'kramya Keśavaḥ in two interpolations after Rām. I.44.17

(App. I.8.7 and 8.24). In all these versions, accordingly, Viṣṇu is the

"connecting link" (Indol. Studies, p. 145). This position is so fundamental

that it cannot possibly have been a later accretion: as early as the

Śatapatha Brāhmaṇa Viṣṇu is defined as "that which is between the

two". It has been argued that in the Rigveda Viṣṇu could not fight the

Asuras because they, too, were part of his essence. Gonda objects that

even the Rigveda describes him as destroying demons (VII.99.4f.) and

states that Viṣṇu goes to war (I.155.6). I am afraid, he here misses the

point. In VII.99.4-5 Indra and Viṣṇu are conjointly addressed as having

slain the māyās of the Dāsa and the men of the Asura Varcin, which

does not prove that Viṣṇu was supposed to fight demons single-handedly

(see n. 69 !). What Viṣṇu's role was in their joint exploits is clearly enough

described in the hymns: he takes part in the fight by his three strides,

and in this sense he can also be said to "meet the challenge" (I.155.6

práty ety āhavām). Cf. MS. II.4.3, TS. II.4.12.3–5 (viṣṇvanuṣṭhita ?). His

character of connecting link between the parties would seem to stand

unchallenged.

Varuṇa's part is also ambiguous, but for other reasons. He is not only

the overlord or lord (ādhipati, pati) of the waters but in a way he is

himself the water 432. Before the emergence of the dual cosmos (the very

431 On Viṣṇu as the divine trickster see Held, p. 299 and Hopkins, Epic Myth., p. 215.

432 MS. IV.7.8 (104,9) samudró vai Várunah, KS. XIII.2 (180,21) ápo vai Varunah.

See p. 77ff. and Hillebrandt, Vedische Mythologie II², p. 18.

106

process which is, in a way, described, in the myth of the Amṛtamanthana) the "waters" are the primordial world, which comprises, as a potentiality, all good and evil. This double-sidedness the myth expresses by the notion that not only the life-elixir amṛta, but also the deadly poison kālakūṭa 433 is produced from this water. It is true that according to the critical edition of the Mahābhārata the episode relating the emergence of this poison is a later interpolation, the aim of the churning being the creation of life. This does not necessarily mean that it is also a later addition to, or elaboration of the original myth. However that may be, the idea that by creating the ordered dualistic cosmos out of the undifferentiated primordial world not only life but also its opposite, death, came into existence was necessarily implied in this conception of creation itself. In the ocean (and in the pītṛ asura who impersonated it) such opposites as day versus night, life versus death did not yet exist but creating the one necessarily involved creating the other 434. Later Purāṇic versions of this myth took the idea of churning literally and evolved the notion of a Milk-ocean 435.

The churning, as the epic describes it, has the character of a violent attack on the ocean, that is, on Varuṇa's world: "As the ocean was churned by means of Mount Mandara by the gods and Asuras, it produced a terrible sound like the thunder of big rain-clouds. All kinds of aquatic animals, squeezed out by the big mountain, died by hundreds in the saltish water. Various creatures belonging to Varuṇa (and) living in Pātāla were killed by the big mountain. And while the mountain was turned round, big trees, rubbing against each other, fell down from the top of the hill with (all their) birds" 436. This particular feature of "Varuṇic animals" is preserved even in the later Purāṇas 437.

In spite of its different presentation of the process this myth is characterized by the same aggressive attack on the primordial world as Indra's fight. Just as Indra slays Vṛtra, the churning results in the killing of hundreds of creatures. For the founding of the cosmos the cooperation of the Asuras is needed but as soon as the existence of the dual world is ensured, the defeated Mahāsuras disappear, in accordance with the ancient Vedic pattern 438, into the nether world and the sea 439, while

433 Mhbh. I.16, 274*, 3-7.

434 See also Held, p. 145 and p. 306, where he rightly remarks that the Asuras were not entirely destitute of the amṛta.

435 E.g., Matsyapurāṇa 249.14 mathyatām kṣīravāridhịḥ as against Mhbh. I.15.2 mathyatām kalasodadhiḥ. See India Maior, p. 149 n. 3.

436 Mhbh. I.16.18-21—Matsyapurāṇa 249.60ff. kakṣāvṛta 'tra mahāghoṣo mahāmeghāvopamaḥ, udadher mathyamānāsya Mandarena surāsuraiḥ (19) tatra nānājalacarā viniṣpisṭā mahādriṇā, vilayaṃ samupājagmuḥ śataso lavaṇāmbhasi, (20) vāruṇāni ca bhūtāni vividāni mahīdharaḥ, pātālalavāsīni vilayaṃ samupānayat (21) tasmīṃś ca bhrāmyamāṇe 'drau saṃghrṣyantaḥ parasparam, nyapatan patagopetāḥ parvatāgrān mahādrumāḥ.

437 Matsyapur. 249.25 tathe 'ti Mandaraḥ prāha yady ādhāro bhaven mama, yatra sthītvā bhramiṣyāmi mathiṣye varuṇālayam.

438 See n. 24. Cf. Viṣṇupur. I.9.111 pātālaṃ ca viveśa vai.

439 Mhbh. I.17.28 tato mahīṃ lavaṇajalaṃ ca sāgarạṃ mahāsurāḥ praviviṣur ardītāḥ suraiḥ.

other deities, such as Śrī and Soma, side with the gods (as they did in RS. X.124) 440.

Inconspicuous, but for that very reason most instructive, is Varuna's position in this contest. According to the older classical formula of Baudhäyana the 'great war' of the cosmogony was a process of polarization, in which all creatures sided with either of the two parties 441. Varuna, however, impersonates the ocean itself and must, therefore, stand apart. He does not resist the Devas, nor can he prevent the churning from taking place. The most significant lines in this respect are I.16.7–9. They describe how at Viṣnu's command the world-serpent Ananta (actually his own 'underworldly' aspect) lifts Mount Mandara, which is to serve as the churning-stick Then the gods go with Mandara to the ocean and announce that they will churn its water in order to obtain the amrta. The lord of the waters then stipulates that he shall also get a part of it. On that condition he is willing to endure the violent crushing by Mount Mandara 442.

It is characteristic that the text here does not make a clear distinction between the element and its god. The divine element is propitiated, like a victim before its immolation. Clearly, it cannot oppose the irresistible course of things. The myth seems to suggest the idea that the emergence of the ordered world was a cosmic necessity 443, to which even Varuna had to resign himself. On the other hand, the condition he makes also implies his going over to the gods.

When compared to the epic tale, which neatly distinguishes Varuna's position from that of the Devas, the authors of the Harivamśa and the Purāṇas seem to be no longer fully aware of the mythical background. This may, at least, be inferred from the fact that such an early Purāṇa as the Matsyapurāṇa, while preserving the idea that Mount Mandara 444 is to churn "Varuna's abode", at the same time says that the gods made Varuna their assistant 445. Since, however, Varuna's aloofness is here still presupposed, it is not easy to interpret exactly the author's views. He may have been thinking of Varuna's belonging, in a later stage of the cosmogonial myth, to the Devas. In any case he was still aware of the fact that Varuna, true to his nature, did not take part in the fight 446. Not before the late Bhāgavatapurāṇa 447 is Varuṇa represented as fighting

440 See n. 415.

441 See n. 361.

442 Mhbh. I.16.7–9 atha parvatarājānamin tam ananto mahābalah, ujjahāra balād brahmaṇ svanāḍ savanukasam (8) tada tena surāḥ sādhayām samudram upasthire, tam ūcur : amṛtārthāya nirmathisyāmahe jalam (9) apāmin patir atho 'cāca : mama 'py amśo bhavet tatah, sodhāsmi vipulam mardam mandarabramaṇād iti.

443 Cf. in this connection Lommel's remark about the necessity of Soma's being slain. See Paideuma III, 6–7 (June 1949), pp. 215–216.

444 Matsyapurāṇa 249.25 (see n. 437).

445 Matsyapurāṇa 249.14 sahāyam Varuṇam kṛtā.

446 See pp. 54, 104.

447 For the tenth century A.D. as the date of the Bhāgavatapurāṇa see Jean Filliozat, Indological Studies - W. Norman Brown (1963), p. 74 (= Laghu-Prabandhāḥ 1974, p. 75).

108

as a mere lokapāla on the side of the gods448. However, the exact chronology of this change in the conceptions about Varuṇa's position would require a more detailed study than is possible here. The circumstance that in sculptural art Varuṇa is at a much earlier date represented as fighting among the Devas449 shows that the evolution was not as simple as it might seem to be at first sight.

448 Bhāgavatapurāṇa VIII.10.26 lokapālāḥ saha gānaiḥ Vāyu-Agni-Varuṇādayah (and 11.47 Vāyu-Agni-Varuṇādayah), 10.28 Varuṇo Hetinā 'yudhyam Mitro rājan Prahetinā. For Varuṇa as a fighter cf. also Hariv. App. I.42B line 2143f.

449 In the basreliefs of Badami, Cave II, dating from the sixth century (or, according to Agrawala, Matsya Purāṇa, opposite p. 311, from the seventh). See R. D. Banerjea, Basreliefs of Badami, Memoirs Archaeological Survey of India No. 25 (1928), p. 24 and Plate XI, e.

109

CHAPTER II

THE VIDŪṢAKA

1. INTRODUCTORY REMARKS

After all that has been written about the various theories on the origin of the Sanskrit drama1 there is no need for a new critical survey. In recent times Renou and Thieme have briefly discussed the main theories2. The old idea of a Greek origin3 and that of the Greek-Roman pantomime being the source of the (North-)Indian drama4 have definitively been refuted.

The main theories which still hold the field are those which derive the drama from an Indian pantomime5, which may have been connected with religious ballads or songs and dances6, or with the recitation of the epic7. Although some scholars, in view of Patañjali's well-known comment

1 For references see S. N. Dasgupta, A History of Sanskrit Literature, Classical Period, Vol. I (Calcutta, 1962), pp. 630–656 and see further, e.g., Sylvain Lévi, Le théâtre indien (Paris, 1890), pp. 343–366, Sten Konow, Das indische Drama (Berlin and Leipzig, 1920), pp. 40–42, Winternitz, Geschichte der indischen Litteratur III (1922), p. 174, A. B. Keith, The Sanskrit Drama (Oxford, 1924), p. 57ff., Paul Horsch, Die vedische Gāthā- und Śloka-Literatur (Bern, 1966), pp. 341–343, M. M. Ghosh, The Nāṭyaśāstra, Text, vol. I (Calcutta, 1967), Introduction, p. LI, n. 137, M. Chr. Byrski, Concept of Ancient Indian Theatre, New Delhi, 1974, G. H. Tarlekar, Studies in the Nāṭyaśāstra, Delhi-Patna-Varanasi, 1975, and many Indian publications which were not accessible to me, such as S. N. Shastri, Laws and Practices of Sanskrit Drama, vol. I, Benares, 1961.

2 See Louis Renou, "La recherche sur le théâtre indien depuis 1890", Extrait de l'Annuaire 1963–1964, Ecole pratique des hautes Etudes, IVe section, sciences historiques et philologiques, pp. 27–40 [=Introduction, Le Théâtre indien par Sylvain Lévi, réimpression, 1963] and Paul Thieme, "Das indische Theater" in: Kindermann, Fernöstliches Theater, Stuttgart, 1966, pp. 26–51.

3 A. Weber, Indische Studien II, p. 148, etc., Windisch, Der griechische Einfluss im indischen Drama (1882), Geschichte der Sanskrit Philologie, 2 vols (1917–1920), pp. 398ff.

4 Hermann Reich, Der Mimus I (Berlin 1903). See Winternitz, p. 175, Keith, Sanskrit Drama, p. 67 n. 2, Dasgupta, p. 650ff.

5 See Dasgupta, p. 642 and particularly Thieme, p. 29f.

6 See, e.g., Winternitz, Geschichte der ind. Lit. III, p. 161 ("Balladendichtung"), p. 163 ("Tanz und Mimik"), p. 169 ("Einfluss des Volkstückes"), Macdonell, Skt. Literature (1900), p. 347: "From all this it seems likely that the Indian drama was developed in connection with the cult of Viṣṇu-Kṛṣṇa, and that the earliest acted representations were therefore, like the mysteries of the Christian Middle Ages, a kind of religious plays, in which scenes from the legend of the god were enacted mainly with the aid of song and dance, supplemented with prose dialogue improvised by the performers", M. M. Ghosh, Nāṭyaśāstra, Translation I2 (1967), p. xlix [I1, p. lv]: "it developed probably out of dances and songs in honour of a deity like Śiva who in later times came to be styled the great dancer-actor (naṭarāja)."

7 H. Lüders, Philologica Indica, p. 428, Sten Konow, Aufsätze zur Kultur- und Sprachgeschichte (1916), p. 113f.: "als noch der Mimus das eigentliche Drama war ... In denjenigen Dramen Bhāsa, die von der zweiten Wurzel des indischen Dramas,

110

on Pāṇini III.1.26, have argued that these epic tales must have been

Krishnaitic, it has rightly been pointed out that the classical drama as

we know it, is predominantly based upon the story of Rāma8. Other

theories that have found considerable support trace the origin of the

drama back to puppet or shadow plays9. On the other hand, the idea

that the Sanskrit drama has developed from Prakrit dramas10 has long

been abandonned. The not uncommon but fallacious idea that the simplest

form is always necessarily the oldest has led some scholars to assume

that the one-act bhāṇa (with a single actor !) is the prototype of the drama11.

dem halbdramatischen Vortrage epischer Stoffe, herstammen . . . ”, Das indische

Drama (1920), p. 47: “Aus einer Verschmelzung des durch Schattenbilder illustrierten

Vortrages epischer Sagen und der Kunst der alten Mimen ist somit das indische

Drama entstanden”, and Thieme, p. 33.

8 See S. Lévi, p. 316, Renou, p. 39 with n. 2 (otherwise Thieme, p. 108; cf. also

von Schroeder, Mysterium und Mimus im RV., p. 17ff. who points to the Śaiva cult).

R. V. Jagirdar [Adya Rangacharya], Drama in Sanskrit Literature2 (Bombay, 1967),

p. 36ff., holds the sūta and the kuśilavas to have been the first performers of the

drama, the sūta having been the direct precursor of the sūtradhāra, with whom

he is sometimes identified (Mhbh. I.47.15 crit. ed.). He compares BhNŚ. 2.27C

śuklam̐ sūtram̐ prasārayet. G. T. Deshpande (in: Indian Drama, Delhi 1956, p. 16)

accepts his theory. Cf. also Dasgupta, p. 647, on Cakkyars and sūtas.

9 For the puppet play see Pis̃hel, Die Heimat des Puppenspiels (Halle, 1900),

Shankar P. Pandit, Vikramorvaśī, Notes p. 4f., Sylvain Lévi, p. 324f., Winternitz III,

p. 178 n. 1, Konow, Das indische Drama, p. 46f., Keith, p. 52f., Dasgupta p. 641f.

Thieme, p. 32 (“Ob es das Puppenspiel im alten Indien gegeben hat, muss dahin-

stehen: wirklich beweisende Spuren oder Hinweise aus alter Zeit haben sich nicht

auftreiben lassen”). Pis̃hel had drawn attention to Mhbh. (crit. ed.) III.31.22

dārumayī yoṣā, V.39.1 aniśvaro 'yam̐ puruṣo bhāvābhave sūtraprotā dārumayī 'va

yoṣā, dhātrā tu digtasya vāse kīlā 'yam, which are not discussed by Thieme.

H.-O. Feistel, Das Vorspiel auf dem Theater, Ein Beitrag zur Frühgeschichte des

klassischen indischen Schauspiels (Dissertation Tübingen, 1969), p. 34, quotes an

interesting passage from the Vākyapadīya 3, Kālasamuddeśa 4, where the puppet

play is called a yantra and the puppet (in the commentary) yantrapuruṣa. In this

passage the sūtradhāra is the performer of the puppet-show, although Thieme, p. 42

explains the term as having originally denoted the performer of a shadow-play,

on the supposition that this technique was different from that of the modern

shadow-play in South India (cf. F. Seltmann, BTLV. 127 [1971], pp. 452–489, 128

[1972], pp. 458–490) and of the Javanese wajang. S. K. Chatterji in: Indian Drama,

Introduction, p. 6, takes the same view as Feistel, but sūtradhāra has also been

explained as “measurer” (thus, e.g. Adya Rangacharya, Drama in Sanskrit

Literature2, p. 36) or as “sūtra-holder” (Renou, p. 33 n. 4: “tenant de [la doctrine

énoncée en] sūtra's [les naṭasūtra's pāṇinīcens]?). Adya Rangacharya (l.c.) may be

right in saying that “the puppet plays were merely the substitute of the populace

for the dramatic luxury of the intellectuals”. As for the shadow-plays see H. Lüders,

SAW. 1916, p. 698ff. [=Philologica Indica p. 391ff.], S. Konow, Das indische

Drama, p. 45f., Thieme, pp. 30–34, 111–116, Renou, p. 33 n. 3 (“Rien là, en tout cas,

ne nous conduit nécessairement aux origines du théâtre”).

10 See S. Lévi, pp. 329f., 335, A. Barth, Revue critique 1886, p. 265 [=Œuvres III,

p. 474f.], J. Huizinga, De Vidūṣaka in het indisch tooneel (1897), p. 6, Renou, p. 36,

on the possibility of “un type de littérature demeuré relativement “populaire”,

un peu comme l'ont été le Pañcatantra et ses dérivés en regard des contes savants.”

11 D. R. Mankad, The Types of Sanskrit Drama (Karachi, 1938), p. 136ff. Cf.

G. T. Deshpande in: Indian Drama (Delhi 1956), p. 17.

111

Still important are the various attempts, from 1869 onwards, to find in Vedic literature and ritual the roots of the drama. Possible connections were found in the Rigvedic dialogue hymns 12, or in the mahāvrata ritual 13, in which a contest between an Ārya (Arya) and a Sūdra symbolically represented the fight for the sun between the Devas and Asuras (the Rigvedic svàrṣāti). In view of the ample discussion which in recent times Horsch has devoted to this problem 14 it is necessary to say a few words about his book, which gives a summary and synthesis of former research 15.

Horsch starts with the ākhyāna literature and first carefully investigates if traces of dramatic action can here be found. His conclusion is that they are entirely lacking in the recitation of tales 16. In the pāriplavam ākhyānam of the horse sacrifice (p. 321)–which is not old according to Eggeling, SBE. 44, p. 361 n. 4–he finds "germs" of epic narrative and rhapsodical poetry 17 but not of drama. His own picture of the historical development is as follows (p. 324): from the Vedic "ballads" must have developed "dance songs", which again must have been the prototype of the popular drama ("das volkstümliche Drama") 18. He then points to the well-known prohibition of singing and dancing by the orthodox brahmans (p. 325, cf. pp. 231, 423) and rightly argues that since it was

12 Max Müller, Rig-Veda-Sanhita translated and explained, Vol. I Hymns to the Maruts (1869), p. 172f., repeated in the rewritten edition SBE. 32 (1891), p. 183, adopted and elaborated by S. Lévi, Le théâtre indien (1890), p. 301ff., L. von Schroeder, Mysterium und Mimus im Rigveda (1908), passim, J. Hertel, "Der Ursprung des indischen Dramas", WZKM. 18 (1904), pp. 59–83, "Der Suparnādhyāya, ein vedisches Mysterium", ibidem 23 (1909), p. 273ff., 24, p. 117ff., M. Winternitz, "Dialog, Ākhyāna und Drama in der indischen Literatur", WZKM. 23 (1909), p. 102ff. Cf. also Keith, p. 14ff., Konow, p. 39 and Horsch, Die vedische Gāthā- und Śloka-Literatur (1966), pp. 321, 329ff.

13 A. Hillebrandt, "Über die Anfänge des indischen Dramas", Sb. kön. bay. Akad. Wiss. (München, 1914), 4. Abh., p. 22f., Konow, p. 42ff. (criticized by Keith, p. 25f., J. Gonda, Acta Orient. 19 (1943), p. 373f.)

14 See Paul Horsch, Die vedische Gāthā- und Śloka-Literatur (Bern, 1966).

15 Also Tsuji, IIJ. 12 (1969), pp. 27–34 mainly confines himself to giving a survey of its contents.

16 See p. 321: "Die Rezitation ohne kultische Handlung hat bei Itihāsa-Motiven die magische Wirkung des Zaubertrituals. Jede Anspielung auf Mimus, Tanz, Gesang und so weiter fehlt in diesem Falle".

17 See p. 323 "Keine epischer Erzählkunst und der Rhapsodendichtung".

18 See p. 324: "In Indien auf jeden Fall müsste der Tanz den melodischen Vortrag begleiten. Sicher ist die Annahme berechtigt, dass sich das volkstümliche Drama aus Tanzliedern entwickelt hat. Letztere wiederum haben ihre Hauptwurzeln in der altvedischen, religiösen Balladendichtung. Das Drama setzt deshalb mehrere Entwicklungsstufen voraus, die man nicht überspringen darf, indem man ein vollendetes Kultdrama in die Hymnenzeit projiziert . . . Unsere Behauptung wird mehr Überzeugungskraft besitzen, wenn wir zeigen können, weshalb der Übergang von der Ballade und der von Lautgesang begleiteten epischen Gāthā zur eigentlichen theatralischen Schau innerhalb der konservativ-brahmanischen Kreise erst spät vollzogen wurde. Der Hauptgrund ist der Widerstand von seiten der orthodoxen Priesterschaft".

112

prohibited it must have existed 19. In this connection the mahāvrata ritual

can be considered the missing link, because it comprised singing and

dancing 20. From the fact that the words nātaka and nātya "drama" are

of Prakrit origin, he further concludes that the oldest form of the drama

must have been popular plays, which were performed on a primitive stage.

That the subjects were taken from the mythic-religious tradition is in

his opinion not essential 21.

The so-called "popular origin" of the Sanskrit drama, which also in

Horsch's speculations plays such a prominent part, calls for some comment.

In support of the assumption of these "volkstümliche Spiele" Hillebrandt,

for instance, had referred to such features as the vidūṣaka (interpreted

as the clown of the popular play) 22, which would require a more careful

approach.

It is certainly true that the use of various Prakrits and the alternative

use of prose and poetry in the Sanskrit drama have not yet been

satisfactorily explained. However, does this use shed any light on the

problem of the origin? Others have objected that "Le fait sanskrit, ici

comme d'ailleurs, est d'abord un fait singulier" 23. The circumstance that

the actors were low-class people can to some extent account for the use

of Prakrit terms, such as nātya, but does not give any clue to the origin.

Particularly in this field the danger of unjustified generalizations is great.

There are at least three points which speak against a "popular" origin

and which will be discussed in detail: first, the performance of the

19 See p. 325: "Was verboten und bekämpft wurde, musste bekannt sein. Nur

gehörte es andern Kreisen an, in diesem Falle also jenen volkstumlichen Schichten,

deren Lebensfuhrung dem konservativen Klerus ein Dorn im Auge war."

20 See p. 326f.: "Das Mahāvrat ist ein religioses Volksfest gewesen. Das missing link

zwischen dem vedischen Kunstdrama und dem authentischen indischen Schauspiel

izt deshalb nicht im orthodoxen Brahmanentum, sondern im gemeinen Volke

zu suchen".

21 See p. 327: "Von Leuten, die einen Dialekt sprachen, wurden die volkstumlichen

Spiele, die auf einer wohl höchst primitiven Bühne aufgeführt wurden, geschaffen.

Dass sie dabei auf mythisch-religiose Themen zurückgriffen, ist nicht erstsaunlich".

He refers to Winternitz III, p. 165 and Hillebrandt (see n. 13), p. 28: "Ganz in

Übereinstimmung mit Winternitz führt Hillebrandt die wichtigsten Momente des

klassischen Sanskritdramas auf die autochthonen, in Dialekt aufgeführten Spiele

zurück".

22 See Hillebrandt, p. 28 (quoted by Horsch): "Das die Dramen einleitende scheinbar

improvisierte Zwiesgespräch zwischen Theaterdirektor und Schauspielerin, die

Anwendung verschiedener Dialekte, die Mischung von Prosa und Lied, die

Verbindung mit Musik und Tanz, die Einfachheit der indischen Bühne, schliesslich

die Figur des Vidūsaka, sichern dem indischen Drama seine Entwicklung aus dem

Volksstück, dem primitiven Mimus und niedern Schauspielerkreisen, die, in Indien

autochthon, im Lande umherzogen, tanzten, musizierten, mimten und ihre Frauen

zur Liebe hingaben". The idea that the vidūṣaka was a clown and as such had his

origin in "popular" plays, particularly the mime, is fairly current. See Thieme, p. 38

(and pp. 50, 53f.), Feistel, p. 119f. and Samskrita Ranga Annual VI (1972), p. 7

"the figure of the vidūṣaka having indubiously developed in the sphere of popular

entertainment".

23 See Renou, p. 30.

113

"Preliminaries" (pūrvarañga), far from being a mere entertainment, was considered equal to a sacrifice for the benefit of the king and his country;

secondly, the older situation as described by the Nāṭyaśāstra seems to have been that the sponsor and patron of a performance was mainly, if not exclusively, the king; thirdly, the oldest forms of the drama we know were based on the Vedic mythical picture of the world. In the light of these facts it seems highly questionable whether such terms as "Klerus" and "volkstümlich" are any help in understanding the driving forces behind the evolution which Horsch assumes, even though it may be correct that the Indian gāthā, in contrast to the sacred ṛc, got a profane character as a result of the orthodox ban on singing. The 19th century frame of reference with which Oldenberg at one time tried to explain certain aspects of Vedic religion is no longer adequate.

The basic fact is that for the Vedic period there are no other sources but an exclusively religious literature, clearly originating in circles which had both ideologically and sociologically a special character. Little is known of any other facet of the cultural and social life. Kings and kṣatriyas are only incidentally mentioned in so far as they come, as peripheral phenomena, within the range of vision of the poets or theologians. It is misleading, therefore, to call everything that is not mentioned in Vedic literature (and, accordingly, remains unknown) "volkstümlich", that is, belonging to "the people's culture". At best it might be termed "profane", if this is not taken as a category basically separated from and contrasting with "sacred". Hillebrandt has further introduced the autochthons into the discussion and Horsch (p. 327) follows him in this respect. Here, too, it is necessary to distinguish between facts and modern guess-work. Although direct knowledge about the aborigines is lacking, it can be stated that many of the names of Bharata's sons as given in the Nāṭyaśāstra (1.26-39) are obviously of foreign origin 24. Some Yajurvedic texts (VS., TB.) associate the śailūṣá (which word later denotes the actor) with singing and dancing (gítá and ṇṛttá): it only proves that in those circles which felt themselves bound by the rules of the brahmins, people outside the three classes of Āryas were used when, for religious or not strictly religious purposes, singers and dancers were needed. Nothing is known about how far the rules of the brahmins were considered valid. It is certainly not advisable to look upon the precepts of brahmanical literature as a faithful picture of the real social life. Since nothing can possibly be said about the performance of dramas in that society, it is not necessary to dwell on the speculations of Hillebrandt, Horsch (p. 326) and others.

A few words must only be said about Horsch's conclusion, based upon Pāṇini IV.3.110, that there existed a dramatic theory (which presupposes performances) in the fifth or sixth century B.C. but that the Vedic hymns and the ritual were not the immediate precursors of the classical drama

24 See Gedenkschrift-Brandenstein, p. 83ff.

114

(p. 328 n. 4, p. 329f.)25. The first conclusion is not justified by the evidence26. As for the second point, Horsh is right. Since only a part of the top of the iceberg of Vedic culture is visible in the priestly literature, there is no reason to expect that here the starting-point for the later evolution of the drama can be found.

In the case of the mahāvrata ceremony, which has more than once been referred to as the missing link, the ritual has preserved an isolated instance of a ceremonial contest that must have stood central in Ancient Aryan social life. It must originally have been a ritual in which the whole community took part, not an amusement for the “masses” (as opposed to a social élite). Characterizing it as “ein religiöses Volksfest” may, therefore, easily evoke wrong associations. It is true, traces of cultural change can be seen in the fact that more primitive features were abolished in the course of time. Since, however, as a symbolical representation of the svàrsāti, it must have been a seasonal ritual, it is hard to see why its origin should have been “im gemeinen Volke” (Horsch, p. 327). While speculations on “die volkstümlichen Spiele, die auf einer wohl höchst primitiven Bühne aufgeführt wurden” are contradicted by the historical evidence, which points to performances sponsored by kings, the mythic-religious subjects suggest the inference that the drama had, from the oldest (pre-literary) times onwards, a religious function. If this is correct, it raises the problem to a different level, the real question then no longer being that of “popular or non-popular” but of religious versus profane27.

Two more theories must finally be mentioned. Thirty years ago J. Gonda put forward ample ethnographical material on which he based his view that the magical effect attributed to the representation of such acts as dancing, laughing and eating was one of the main roots from which the Sanskrit drama has sprung. It is impossible to summarize this paper, which contains many interesting remarks and deals extensively with the

25 As against Sylvain Lévi, Hertel, von Schroeder, Hillebrandt, etc.

26 It is not necessary to repeat what has been written about Pānini's bhikṣuṇaṭa-sūtra­yoḥ, “ein Sūtra für Mimen” (Böhtlingk), rather than “règles des acteurs” (Renou). The uncertainty about what exactly was meant by naṭa at that time has been stressed by Lüders, Philologica Indica, p. 427: “wenn wir einmal etwas Bestimmteres über die Kunst der naṭas der älteren Zeit erfahren, hat es mit dem Drama gewöhnlich gar nichts zu tun”. Cf. Konow, p. 48: “kann ebensogut ein Mime wie ein wirklicher Schauspieler gewesen sein”, p. 43: “Es kann nach alledem nicht zweifelhaft sein, dass sie wirkliche Mimen waren . . .”.

27 A different point is that, whenever a cult act evolves into a piece of literary art, such as tragedy and comedy in Greece, there is always the possibility of secularization if in the context of the culture concerned the religious origin and character of the ritual is forgotten. In Greece Euripides dropped the chorus in his tragedies as an element that had become cumbersome and had lost its meaning. In India the pūrvaranga, the religious drōmenon par excellence, was shortened and stripped of those elements which had at one time been of the highest importance, and new themes of a non-religious nature, as found in the “bourgeois” drama (prakaraṇa), arose at an early date—so early, indeed, that it cannot even be proved on the basis of the dramas that have come down to us, that the prakaraṇa was the result of a later development of dramatic art. Its origin is still obscure, and probably will remain so as long as no fresh evidence comes to light. See below, p. 213.

115

general problem of the origin of the drama 28. An entirely different thesis has been defended by Indu Shekar, who argued that the drama was a product of a non-Aryan culture of India. The present study will show why I think that the evidence available points to a different conclusion 29.

Finally, two ideas must be mentioned which in my opinion contain an important element of truth but which must be incorporated in a more comprehensive approach. The first, put forward in 1909 by M. Haraprasad Shastri, emphasizes the fact that according to the native tradition of the Bhāratīya Nātyaśāstra the first legendary dramatic performance took place on the occasion of Indra's Banner Festival (Indradhvajamaha). The importance of this ceremony will be discussed below 30, where it will be argued that the half-mythic, half-legendary form in which this tradition is embedded is no reason for questioning its fundamental correctness. The second theory has, as far as I can see, only been proposed by G. K. Bhat and has generally been rejected. In his opinion the vidūṣaka of the drama is historically connected with the Asuras of the Veda. The way he conceived this connection is, it is true, unacceptable and the argument in support of his view very weak 31. In spite of these shortcomings, however, his theory deserves to be mentioned because in its essence it is, in my opinion, of greater importance than Bhat may have been aware of.

From this brief survey it emerges that there is such a chronological gap between the classical drama as we know it and its supposed origin that all attempts to bridge it must needs be conjectural 32. Renou's

28 See J. Gonda, "Zur Frage nach dem Ursprung und Wesen des indischen Dramas", Acta or. 19 (1943), pp. 329-453. Renou reacted to this approach in the following words (p. 30), the purport of which is not quite clear to me: "Éclairer l'évolution préhistorique du théâtre sanskrit par le folklore peut amener des comparaisons intéressantes: l'inconvénient est que l'originalité d'un art savant se trouve diluée dans l'anonymat des structures élémentaires. Le fait sanskrit, ici comme ailleurs, est d'abord un fait singulier".

29 Indu Shekar, Sanskrit Drama, Its Origin and Decline (Leiden, 1960). See Renou, p. 29 n. 4: "hypothèse souvent formulée pour d'autres secteurs de l'indianisme et qui semble dispenser d'explication interne rationnelle". An attempt to determine to what extent non-Aryan influence may have been at work, and what its limits were was made in Studien zur Sprachwissenschaft und Kulturkunde, Gedenkschrift für Wilhelm Brandenstein (1968), p. 82ff. It is true, influence of non-Aryan cultures has too often been invoked, without the slightest proof, as a deus ex machina to explain away difficult problems. If, however, there are specific (mostly linguistic) indications pointing to that conclusion, there is obviously no point in ignoring their existence, our task then being to try to understand what the role of the influence can have been in the whole context of Indian culture.

30 M. Haraprasad Shastri, J. and Proc. As. Soc. Bengal 5 (1909), pp. 351-361 (especially p. 361), Raghavan, J. Or. Res. (Madras) 7, p. 41. This festival will be discussed below.

31 G. K. Bhat, The Vidūṣaka (Ahmedabad, 1959), p. 45: "The evidence of the Nātyaśāstra and the Sanskrit literature shows that the earliest phase of the drama was marked by the Deva-Asura conflict, where the representation of Asura as a comic character provided the only possibility for humor", and passim. Indu Shekar (see fn. 29), p. 78, rejected this theory as "far-fetched and entirely conjectural".

32 See Renou, p. 30.

116

characterization as a “problème difficile, insoluble peut-être” (p. 29) correctly summarizes the present situation. There is still a lack of agreement on one of the basic questions, viz. whether the origin of the drama was religious or profane. While Konow and Lüders assume a double origin and Renou's position, as usual, is non-committal 33, Thieme has made an interesting attempt at a logical reconstruction of the genesis on the basis of three existing theories, which started from dance, mime and shadow-play respectively. In his synthesis the mime, being a popular play without dance and music and with profane subjects 34 and, on the other hand, dance and shadow-plays with religious themes 35 have led to the emergence of the drama. In accordance with current theories 36 he then concludes that the vidūṣaka's part must have been “die Rolle eines reinen Spass-machers” 37 and that he, consequently, must have originated in the mime 38. Since the Nāṭyaśāstra according to him nowhere refers to a specific part of the vidūṣaka in the drama. 39, Thieme further assumes that originally the vidūṣaka's role was limited to the so-called “prelude” (pūrvaranga). A difficulty is, however, that the oldest extant dramatic fragments, dating roughly from about 150 A.D., force the conclusion upon us “dass zu Aśvaghoṣas Zeit dem vidūṣaka die typische Rolle des Heldenfreundes durch feste Tradition bereits zugeteilt war” 40.

In view of the lacunary character of the tradition it can be understood that other scholars restrict themselves to the evidence of the classical drama—which, however, teaches us next to nothing about the prehistory and genesis of the drama. This position has succinctly been expressed in Renou's words of resignation: “la préhistoire du théâtre—il faut décider y renoncer” 41.

33 For Lüders and Konow see above, n. 7. Also Gawroński assumed a double source, religious and “bardic” (see Renou, p. 31 n. 2; not accessible to me). See further Renou, p. 33 n. 3 and Dasgupta, p. 642.

34 See Thieme, p. 38.

35 Ibidem, p. 41.

36 Konow, p. 14: “Es wird allgemein angenommen, dass der vidūṣaka eine alte Figur der Volksbühne ist, die das klassische Drama aus den Volksstücken der fahrenden Tänzer und Mimen übernommen hat”; p. 15: “Der vidūṣaka ist eben eine Figur der alten Volksstücke, in welchen die Sitten und das Benehmen der höheren Kasten, besonders der Brahmanen genau so, wie heutzutage, verspottet wurden”; p. 47: “die der Mimenbühne entnommene Gestalt des vidūṣaka”. Cf. also Horsch, Die vedische Gāthā- und Śloka-Literatur, p. 343, where the vidūṣaka and the viṭa are characterized as “populäre Gestalten”. On the viṭa see below, p. 233ff.

37 Thieme, p. 53; cf. p. 50.

38 Ibidem, pp. 40, 53. Cf. Feistel, p. 119 !

39 This is, as far as I can see, contradicted by NŚ. 19.17-18C vayasya rājann iti vā bhaved vācyo mahīpatih, vidūṣakena rājñā ca ceṭi ca bhavatī 'ty api [=17.79–80 KM²].

40 Thieme, p. 54. Similarly already Lüders, Philologica Indica, p. 208 and Konow, p. 50.

41 Renou, p. 34.

2. THE BHĀRATĪYA NĀṬYAŚĀSTRA: THE RELIGIOUS CHARACTER OF THE DRAMATIC PERFORMANCE

Since the evidence that can be found in the existing classical dramas is so scarce, it is somewhat surprising that comparatively little use has been made of those indications about the origin of the drama which are contained in the handbook of dramaturgy known as the Bhāratīya-nāṭyaśāstra. This is the more striking as this Nāṭyaśāstra is rightly referred to in every discussion of the Sanskrit drama and as there exists a general agreement about its being anterior to all dramas that have been handed down to us, at least as far as its core is concerned. The explanation for this oversight is no doubt to be sought in the mythic form in which the obviously old and interesting tradition has been clothed. It is not difficult to understand the sceptical attitude of judicious European and Indian scholars. Sylvain Lévi was among the first to deny the value of this tradition when in 1890 he wrote: “La fable racontée par le Nāṭya-çāstra peut suffire à la pieuse crédulité des Indiens; la critique européenne ne saurait s'en contenter”. Others have expressed similar opinions, e.g., S. K. De (1962): “The orthodox account of the origin of the Sanskrit drama, by describing it as a gift from heaven in the form of a developed art invented by the divine sage Bharata, enveloṕs it in an impenetrable mist of myth; while modern scholarship, professing to find the earliest manifestation of a ritual drama in the dialogue-hymns of the Ṛgveda and presuming a development of the dramatic from the religious after the manner of the Greek drama, shrouds the question of its origin in a still greater mist of speculation” 42.

42 See Sylvain Lévi, Le théâtre indien, p. 299, S. K. De in: S. N. Dasgupta, A History of Sanskrit Literature I (see n. 1), p. 43 and cf. Adya Rangacharya, Drama in Sanskrit Literature2 (1967), p. 31f. on the fifth Veda: “Every new school of thought in India has striven to claim and establish for itself the sanction of the Vedic texts. So a statement of the kind under question is more a tribute to the sanctity and hold of the Vedas than a reference to a fact”. The notion of a fifth Veda has its roots in Chānd. Up. VII.1.2, where, elaborating on the traditional enumeration ... sāma-vedo’tharāṅgirasa itihaśah purāṇam ... (BĀUp. II.4.10, IV.1.2, V.5.11) the text has ātharvaṇam caturtham itihāsapurāṇam pañcamam (vedānām ...), which the BhāgPur. I.4.20 paraphrases as itihāsapurāṇam ca pañcamo veda ucyate. The Rāmāyaṇa I.1.77 calls itself “equal to the Vedas” (redaiś ca saṃmitam) and it is said that the Mahābhārata is considered a fifth Veda (A. Barth, Œuvres III, p. 473, M. Krishnamachariar, History of class. Skt. Lit., p. 30) but except in a śloka found in the beginning and at the end (Mbh. I.1.205, 56.17, XVIII.57* kāṛṣṇam vedam imaṁ vidvān śrāvayitvā ‘riham aśnute), the late interpolation after VII.173.107 kāṛṇas tu pañcamo vedo yaṅ Mahābhāratam smṛtam (in Ś. K3. K3. D1 pp. 1059-1062) and some passages which are not explicit (I.1.208 catvāra ekataḥ vedāḥ Bhārataṁ caikataḥ, XVIII.36* vedaḥ sāṅgās tathaikatra Bhārataṁ caikataḥ sthitam), the epic refers to itself as an ākhyāna, being the fifth after the Vedas; cf. III.55.230* yo ‘dhīte caturo vedān sarvān ākhyānapaiñcamān, V.43.23 ākhyāna-pañcamair vedaih. The expression “equal to the Veda(s)” also occurs in the Mahābhārata but only once, as far as I can see, with reference to the epic itself, and that in a spurious śloka occurring in a passage where the text “is in a chaotic condition” (I.56.27ff.). The relevant lines run as follows (I, App. 33, lines 17-19 = I.62.48-49 Bomb. ed.) tathā Bhāratam ucyate [scil. ratnadhī], idam hi vedaiḥ

118

Only G. K. Bhat, as far as I know, has defended the (relative) value and importance of the data of the Nāṭyaśāstra with respect to the problem of the origin of the drama 43 . It will be the object of the present study to show that on closer investigation the mythical story of this work turns out to lend itself much more to a rational interpretation than may appear at first sight. On the other hand an analysis of a number of details of the pūrvaranga discloses the existence of an underlying structure, which throws a new light on the historical background of the drama.

Although the work, in its present somewhat chaotic state of different versions, no doubt contains a great deal of later accretions, it is almost generally accepted that the original Nāṭyaśāstra, which was the prototype of the first part of the various versions now extant, must have been earlier than all the dramas known at present. Estimations vary between 150 A.D. (or earlier) and about 350 A.D. 44 .

sam்mitam pavitram api cottamam, śravyam śrutisukham caiva... In the other passages it is mostly an epithet of purāṇa or itihāsa, cf. XII.326.106 tebhyas tac chrāvayām āsa purāṇam vedasaṁmitam, 335.7 kathayisyāmi te sarvam purāṇam vedasaṁmitam, jagau yad bhagavān Vyāso . . ., 335.72 etad Dhayaśiro rājann ākhyānam tava kīrtitam, purāṇam vedasaṁmitam yan māṁ tvam paripṛcchasi, XVIII.5.43 itihāsam imam punyam mahārtham vedasaṁmitam, śrāvayed yas tu . . ., VII.173.101 dhanyam yaśasyam āyuśyam punyam vedaiś ca saṁmitam (!) . . . vyākhyātam satarudriyam (cf. the variant reading in XIII.17.169 . . . stavam . . ., svargyam ārogyam āyuśyam dhanyam vedena saṁmitam). As a v. 1. it also occurs in XIII.61.33 vyāhrtim . . . vedasaṁmitām. In the same way the Nāṭyaśāstra describes in its introductory part (1.4) how the munis interrogated Bharata about the Nāṭyaveda "equal to the Vedas" which he had composed: yo 'yam bhagavatā samyag grathito vedasaṁmitah, nāṭyavedah...

43 See G. K. Bhat, The Vidūṣaka (1959), p. 19: "There is a tendency to disregard the evidence furnished by tradition, but this is unjustified. It is certainly necessary to sift all such evidence very carefully; but when such evidence is nearly all that we possess as a starting point of an inquiry it cannot be treated lightly, much less ignored, as it has been done. The Nāṭyaśāstra of Bharata and the actual dramas must lead our way, I think, to the solution of the present problem".

44 From the various guesses the following may be mentioned: Konow, Das indische Drama, p. 2 (with references) considered it to be only slightly earlier than Bhāsa (towards the end of the 2nd century A.D.? Konow, p. 51) and dated it to the 2nd century (p. 49). S. K. De, Sanskrit Poetics I 2 , p. 31 wrote: "The lower limit of the date of Bharata's work, therefore, can be provisionally shifted back to the fourth or fifth century A.D., while it is almost certain that it existed in its present shape in the 8th century A.D. The upper limit cannot be put too early" (cf. p. 29 and p. 18 with references). P. V. Kane, History of Sanskrit Poetics 3 (1961), p. 41 (with references) thinks "that before 300 A.D. there existed a work on nāṭyaśāstra ascribed to Bharata and dealing with the rasa theory and dramaturgy in general" (similarly p. 47). Cf. p. 20: "That most of the chapters now found were in existence from at least the 3rd or 4th century A.D. follows from several considerations"; p. 43: "[These details] however, make it probable that the present Nāṭyaśāstra is not much older than the beginning of the Christian era"; p. 21: "the present Nāṭyaśāstra existed at least a century or two earlier than 350–450 A.D., i.e. in the 3rd or 4th century, if not earlier". Thieme, p. 52 (cf. p. 35) considers it to be earlier than Aśvaghoṣa and Bhāsa, i.e. earlier than the 2nd century A.D. Similarly H.-O. Feistel, Das Vorspiel auf dem Theater, pp. 138, 136 (after the beginning of the 1st century A.D.), Samskrita Ranga Annual VI, p. 26 (the kernel of the NS. probably composed in

119

Since the tradition referred to above is mainly contained in the first chapter of the Nāṭyaśāstra, it is necessary, before proceeding to a close analysis of that chapter, to note that according to one scholar this is a late addition to the work. Since experience teaches us that the first and last portions of literary works have a greater chance of being additions than the main body, this opinion has to be taken seriously, the more so as it emanates from a well-known scholar. P. V. Kane, History of Sanskrit Poetics 3-4 (1961-71) assumes that chapter I, and probably the next four chapters, were added some centuries before the fifth century A.D. (p. 20) to raise the status of dramatic art (p. 22). They were, accordingly, "comparatively late additions" (p. 16, cf. p. 41). All this, it is true, is advanced as "a very tentative theory" (p. 18). I fail to see, however, any decisive arguments in favour of Kane's rather startling supposition 45. As will be shown below, the first chapter contains such Vedic reminiscences as to exclude a comparatively late origin. On account of its contents it is hard to imagine that it was originally composed as a separate work. It was obviously conceived as the introductory part of a work on dramaturgy. Since the sixth chapter immediately starts, without any introduction, with the theory of the rasas, this cannot possibly have been the original beginning of the Nāṭyaśāstra. If Kane were right, the first five chapters would have taken the place of the original introduction. Since, however, the pūrvarañga was of the highest importance because it comprised the ritual for the "consecration" of the stage, it is hard to believe that the hypothetical older version of the work had not contained a special chapter on this ritual. As a consequence of Kane's theory, the present fifth chapter, which deals extensively with the pürvarañga and which is full of interesting details of an archaic character, would have been substituted for an older version. I fail to see any internal evidence for such a substitution. Both the first and the fifth chapter must to all appearances have belonged to the oldest stratum of the work. This conclusion is based on the analysis of these chapters which will be given in the next sections. As for the other chapters, only a critical analysis, which is outside the scope of this study, could determine their relation to these two chapters. From the point of view of composition, the second chapter, for instance, differs so much from the first that it is hard to believe that it has been composed by the same author. Possibly it has been inserted at a later date.

the 1st century A.D.). So there is a certain agreement about the first or second century as the time of origin of the older parts. The only exception is Manomohan Ghosh, who at first placed the date between 100 B.C. and 200 A.D. (which "cannot be far from the truth", Kane, p. 41f.), but in the second editions of his text and translation of 1967 changed this into 500 B.C. See The Nāṭyaśāstra ascribed to Bharatmuni, Vol. I2, Introduction, p. lxxxii, Translation I2, Introduction, p. lxv. As for the date of Bhāsa it may be noted that only Pusalker and D. S. Pathak have considered him to be earlier than the Nāṭyaśāstra (see the references in Feistel, Das Vorspiel, p. 128 n. 5).

45 Quoted without comment by S. K. De, Sanskrit Poetics I2 (1960), p. 28 n. 1.

120

An analysis of the first chapter of the Nāṭyaśāstra leads to the conclusion that the drama, and in particular its performance as a drōmenon, had a religious character. In the introductory verses it is related how the munis asked Bharata to inform them about the origin, the aim and the performance (prayoga) of the nāṭyaveda, which is here called “equal to the (four) Vedas” (see n. 42). According to the story as Bharata then tells it, Indra and the other gods, at the beginning of the Tretayuga, asked Brahmā for a play (krīdanīyaka) which could be seen and heard 46. As they pointed out to him, knowledge of the four Vedas was not allowed to the Śūdras. There was a need, therefore, for a fifth Veda, to which all classes could be admitted (1.12 sārvarṇika). After being sunk in meditation Brahmā then creates, on the basis of the four Vedas, the Nāṭyaśāstra as a fifth Veda. The fiction that it consists of elements of the four Vedas, the dramatic art being a combination of recitation (pāṭhya) from the Rigveda, singing (gīta) from the Sāmaveda, dramatic action (abhinaya) from the Yajur, and sentiments (rasāḥ) from the Atharvaveda 47, can here be ignored.

The term nāṭyaveda, mostly used, it seems, in this connection, stresses the religious character of dramatic art as a Veda. As a karmadhāraya compound it means “the Veda that is the drama” and is almost synonymous with nāṭya alone 48. Thus it can be said that the Nāṭyaveda was performed 49.

It is clear that this legendary story is a fiction, and this has rightly been stated by Sylvain Lévi and others (above, p. 118). It is also evident

46 1.11 drśyam śroṭavyam ca yad bhavet.

47 1.17 Cf., e.g., Nāṭyaśāstrasamgraha 9 pāṭhyam abhinayān gītam rasān samgrhya Padmabhūḥ (10) yairīracat trayam idam dharmakāmārthamokṣadam . . . etc.

48 According to Feistel, p. 100, the word has been created in analogy to Ṛgveda, etc. As the name of the fifth Veda, it is certainly parallel to those of the other Vedas but it is a karmadhāraya compound “the Veda that consists of the drama”, cf. 1.20 nāṭyasaṃjñito hi vedaḥ. As such it is almost synonymous with nāṭya, except that it stresses its religious character as a creation of Brahmā. It is, therefore, mainly used, it seems, in the first chapter and where nāṭya is contrasted with the other Vedas, e.g., 1.4 yo 'yam Bhagavatā samyag gṛhito vedasaṃmitaḥ, nāṭyavedaḥ . . . (in the versification of the BhNŚ (just as in the Mbh.) samyak functions as a stop-gap at the end of the odd pādas, cf. 1.51, 4.12, 5.30, 32, 179), 7 śrūyatām, nāṭyavedasya sambhavo, 16 nāṭyavedam tatas cakre, 18 vedopadeaih sambaddho nāṭyavedo mahātmanaḥ, 19 utpādya nāṭyavedam tu, 102 yo 'yam Bhagavatā sṛṣṭo nāṭyavedaḥ surecayā (: 4.12 aho nāṭyam idam samyak tvayā sṛṣṭam), 22.22 nāṭyavedasamutpannā vāg-āṅgābhinayātmikā (: 23 Ṛgveda, etc.), 36.8 nāṭyavedasya niścayam, etc. For 1.6 Bharato muniḥ, pratyuvāca tato vākyam nāṭyavedakathām prati “about the tale of the Nāṭya-Veda”, cf. 4.329C bhūyaḥ kim kathyatām viprā nāṭyayogavidhīm prati “about the performance of the yoga that consists of the drama” (v. 1. anyan nāṭyavedavidhīm prati 4.320B 2), 5.179 bhūyaḥ kim kathyatām samyaḥ nāṭyavedavidhīm prati “about the performance of the Veda 'Drama' ” (not “das Wissen, das das Schauspiel ist”, Feistel, p. 100). Vidhi here means “Ausführung”, rather than “Vorschrift” (Feistel, l.c.), cf., e.g., 5.60 prayujya gītakavidhīm, 142 sarvam evam vidhīm kṛtvā, 163 ṛttagītavidhīm prati, 4.322 bhāṇd̄avādyavidhīm prati.

49 1.55 ayaṃ vedo nāṭyasaṃjñāḥ prayujyatām “This Veda which bears the name 'Drama' must now be performed”.

121

that, once the general pattern had been given, it could be elaborated, as has been done, e.g., in the last detail mentioned about the elements taken from the four Vedas 50. The question remains to be answered whether the fiction was a mere literary artifice or whether it had, in the last resort, a mythical character. At all events, the fact that the dramatic art is called a fifth Veda shows that the drama was considered to belong to the domain of religion, not to that of entertainment. Was this claim well-founded?

As is well known, in theory and practice a dramatic performance consisted of two parts, the pūrvaraṅga and the play proper. The importance of the pūrvaraṅga for the success of the performance and, accordingly, also for our insight into the original character of the drama, has not sufficiently been understood and translations such as “Vorspiel auf dem Theater”, inspired by Goethe's reviving the Indian tradition, can easily lead to wrong conclusions. When the character of the pūrvaraṅga is defined as "a ritual of worshipping the gods" 51 this is only a late and weak reflex of the ancient drōmenon, the performance of rites, which the pūrvaraṅga clearly was.

Fortunately the author of the Nāṭyaśāstra was still fully acquainted with these rites 52 so that here we stand on solid ground. In later times, when the real nature of these "preliminaries" was no longer understood, the pūrvaraṅga must gradually have dwindled down to a short formality, although we have no direct information about this development. The older meaning of the preliminaries, however, was to enact the banner festival and transform the stage into the sacred world. The Nāṭyaśāstra states expressly that the "worship of the gods" (in which formula the ritual of the consecration is often summed up) is equal to the Vedic sacrifice (yajña) 53. The verb yaj- is also used to denote the "worshipping of the stage" (raṅgapūjana) 54, another term for the same consecration, and the Nāṭyaśāstra says that "hearing recitation, song and music is equal to words which express the contents of a sacred text of the Veda" 55.

50 Similarly the story of Brahmā's asking Indra to perform with the other gods the itihāsa which he has composed (1.19) and Indra's refusal, after which Brahmā gives it to Bharata (1.22-25) and his hundred sons (1.40).

51 NŚ. 36.29 devatābhyarcanam, 32.483 devatāpūjādhikārah, Sāgaranandin, NLRK. 2159/250 (also quoted by Raṅganātha ad Vikram. I.1) pūrvaraṅgo bhaved eṣām (bhavet teṣām) ādau devārcanavidhiḥ, see n. 53.

52 Although sometimes he already defined the aims of a dramatical performance in terms of entertainment. See below, n. 75.

53 1.126, 3.96C yajñena sammitam hy etad raṅgadaivatapūjanam.

54 3.93C samyag iṣṭas tu raṅgo vai. Cf. 5.112C, where the nāndī ends with the words ijyayā cā 'nayā nityam priyantām devatā iti (v. 1. sarvadevathī in one MS.). The word ijyā here probably refers to the pūrvaraṅga, although it is possible to take it as referring to the whole performance (Feistel, p. 70).

55 In the last chapter (36.26C 1), which is probably a later addition: paṭhyam śrutvā tathā caiva gānam vāditram eva ca, vedamantrārthavacanaiḥ samam hy etad bhaviṣyati. The reading of 36.21 KM 2 differs considerably (vādyam nāṭyam tathā geyam citram vāditram eva ca, vedamantrārthavacanaiḥ...) but vādyam is here redundant: it sometimes occurs, it is true, in conjunction with gīta (e.g. 5.149C vādyagītapramodita,

While in such passages the drama, or the pūrvaranga, is equated to a sacrifice, others show that a nāt̨ya was sometimes performed simultaneously with a sacrifice, as in Harivamśa II.91.25 Bomb. ed. (No. 29F. 51-52 crit. ed.). "While the sacrifice there took place, a naṭa named Bhadranāman satisfied the Mahars̨is with a good performance" 56. It remains doubtful, it is true, what the word sunāt̨ya here exactly denotes. Keith rendered it by "his excellent power of representation" but a mere dance may have been meant. This is apparently the meaning of nāt̨ya in Kālidāsa's Mālavikāgnimitra, where dancing is described as follows: "The Rs̨is consider this to be a charming visual sacrifice for the gods ... dancing, although of manifold forms, still is one and the same amusement for people of different tastes" 57. As is apparent from the last quotations the meanings "dance" and "dramatic performance" cannot always easily be distinguished. What is said of the former, is also true of the latter: both were regarded as a special form of sacrifice.

This was not a simple metaphor. The Vedic sacrifice was performed at the expense and for the benefit of the yajamāna by a group of officiating priests, who released him from ritual impurity. Just so the drama can be looked upon as a dṛśyamenon accomplished by a group of actors at the patron's expense, who is denoted either by more neutral names such as preksākartṛ, preksāpati, arthapati, sabhāpati, or by terms such as bhartṛ and svāmin, which also denoted the king 58. In the light of this parallelism the reward given to the actors 59 can be compared to the dakṣiṇā of the priests.

In connection with this suggestion a point on which there is much difference of opinion must here briefly be touched upon. This is the question as to the place of the drama in social life. In spite of its being intended for all classes (1.12 sarvavarnika) the general impression one gets from the Nāt̨yaśāstra is that the author constantly had the king in mind as the common "patron" of performances, although it is nowhere stated that this patronage was a royal prerogative. The vagueness of the text in this respect is reflected by modern studies. As an illustrative instance

164 gite vāye ca nr̨tte ca), but so does pāthya (cf. 2.21 pāthyaṁ ca geyaṁ ca) and gītavāditre "song and instrumental music" is attested as early as Chānd. Up. VIII.2.8.

56 Harivamśa II.91.25 (8575) tatra yajñe vartamāne sunāt̨yena natas tadā, mahars̨ibhiḥ toṣayām āsa Bhadranāme 'ti nāmataḥ, quoted by PW. IV, col. 99 V, col. 194 (Bhadra), Hertel, WZKM. 24, p. 118, Keith, p. 48. The passage, however, occurs in the long episode Pārijāta-harana (II.65-97), which according to P. L. Vaidya (Crit. ed. I, p. xxxiii) must have been interpolated after 1050 A.D. and has, therefore, been excluded from the text of the critical edition (see App. I, No. 29F, line 51).

57 Kālidāsa, Mālavikāgnimitra I.4 devānām idam āmananti munayaḥ kāntaṁ kratuṁ cākṣusaṁ ... nāt̨yaṁ bhinnarucer janasyabahudhā 'py ekaṁ samārādhanaṃ. Quoted by Thieme, p. 37.

58 E.g., 5.111C prekṣākartṛ, Abhinavagupta I, p. 55 (ad 2.29 svāmin) prekṣāpati, 1.127 arthapati. For sabhāpati see Balbir, IIJ. 6 (1962), p. 44. For svāmin see the foot-notes 64 and 68; bhartṛ occurs, e.g., 27.98.

59 See M. M. Ghosh, note on his translation of 1.58-61.

123

may be quoted Keith's subchapter on "the Audience"60. After having

pointed out that the audience should be cultivated, critical and experienced,

he goes on to say (p. 370):

"The rules for placing the patron at whose bidding the drama is performed,

Sabhāpati, and his guests, are elaborate 61. He sites himself on the Lion Throne,

the equivalent of the royal box, with the ladies of his harem on the left, and

on the right the personages of highest importance, such as the vassal princes

of a great king like Harṣa. Behind the latter are the treasurer and other officers,

and near them the learned men of the court, civil and religious, including the

poets, and in their midst the astrologers and physicians. On the left again are

the ministers and other courtiers; all around are maidens of the court. In front

again of the king [!] are Brahmins, behind the bearers of fans. radiant in

youthful beauty. On the left in front are the reciters and panegyrists, eloquent

and wise. Guards are present to protect the sacred person of the sovereign [!].

How far the dramas were viewed by the public in general we cannot say: the

rules regarding the play-house contemplate the presence of Çūdras. but

that is a vague term, and may apply to a very restricted class of royal hangers-on.

We have the general rule that barbarians, ignorant people, heretics, and those

of low class should not be admitted, but such prescriptions mean very little".

It may be added that BhNŚ. 2.46–53 mentions four particular pillars

relating to the four social classes in a context where the cosmic meaning

of the playhouse is stressed.

This quotation illustrates how in these discussions the patron and

sponsor is almost automatically identified with the king. The latter is

supposed to attend the performance 62 and to decide to which actor the

patāk (prize) is assigned 63.

The way the ritual prescribed for the building of a playhouse and that

of the pūrvaranga were performed had a direct favourable or unfavourable

effect on the king and the whole country. In the discussion of the

prescriptions of how to measure out the future theatre with a "white

string", it is said "when the string is broken in half, the king will

certainly die; when it is broken into three pieces a political disorder will

arise in the country"64. This is reminiscent of the rules for erecting the

60 The Sanskrit Drama, p. 369ff.

61 Keith here refers to Sarngataratnākara 1327ff., S. Lévi I, p. 375ff. and

Kāvyamīmāmsā p. 54f. For the seats for Śūdras in the theatre see, e.g., Keith, p. 359.

62 See, e.g., 2.59 purohitam nrpam caiva bhojayen madhupāyasaih, 3.83 nrpater

nartakīnām ca kuryād dīptyabhivardhanam, abhidyotya sahā 'todyair nrpatim nartakis

tathā (84) mantraptena toyena punar abhyukṣya tām vadet: mahākule prasūtās ca

gṛhaṅghaik cā 'nu dalankṛtāḥ ! (85) yad vo janmagunopetam tad vo bhavatu nityasah.

evam uktvā tato vākyam nrpater bhūtaye budhah (86) nātyayogaprasiddhyartham

āśiṣah saṃprayojayet.

63 See 27.77 ghātā yasya tv alpā saṁkhyātāḥ siddhayaś ca bahudhā syuh, viditām krtvā

rājñe tasmai deyā patāk tu (78) . . . siddhyadhike tu patākā samasiddhau vā 'jñayā

nrpateḥ (79) atha narapatih samah syād ubhāv api tu lambhanīyau tau.

64 2.29 ardhacchinne bhavet sūtre svāmino maranam dhruvam, tribhāgacchinnayā rajjvā

rāstrakopo vidhīyate "If the string is broken into two pieces, the king will surely die;

when broken into three pieces, the string causes political disorder". Since the

context deals with disasters befalling the country, svāmin must refer to the king

as preṣāpati. It may be conjectured that in Abhinavagupta's time it was no longer

124

Indra-banner 65. The pillars of the playhouse must be auspicious for the king 66 and the jarjara is implored to give victory and prosperity to the king 67.

In view of these facts it is natural to interpret the word svāmin in 2.29 68 not only as the patron (preksāpati, as Abhinavagupta glosses it), but more specifically as the king. It is no doubt true that “der Spielgeber (sabhāpati) ist häufig ein Fürst oder eine andere hervorragende Persönlichkeit” 69. All such general statements, however, which are bound to be vague because our sources are so, do not answer the basic question, viz., Is it at all conceivable that in older times other persons than the sovereign were entitled to organize and sponsor a performance which had such an influence on the whole country and its king? Again and again the Nātyaśāstra says that the rites, e.g. those for the consecration of a new playhouse, are of immediate importance for the king, who will either be threatened by his enemies or conquer them 70. Also the ritual preparations on the stage must be duly performed so as to be auspicious for the king 71. Even in the last chapter, apparently of later date, the “prosperity of the king” is mentioned in connection with the nāndi (introductory prayer) 72. All this need not, it is true, imply that the theatre had necessarily to be built by a king but the first clear statement to the contrary that I know of, dates from about 1200–1300 A.D., when Śāradātanaya wrote that “the audience in this theatre should be only males, consisting, besides the king, of proprietors of other theatres

self-evident that the patron was the king. Indirectly his gloss would seem to confirm the conclusion which will be drawn below (n. 73) from Śāradātanaya (about 1200–1300 A.D.).

65 See VarBS. 42.58 acchinnarajjum . . . uthāpayel laksma.

66 NŚ. 2.61 abhimañtrya yathānyāyam stambham uthāpayec chucit : yathā 'calo girir Merur Himavāṁś ca yathā 'calaḥ (62) jayāvo narendraśya tathā tvam acalo bhava !

67 NŚ. 3.81 jayaṁ cā 'bhyudayaṁ caiva pārthivāya prayaccha naḥ.

68 For Lévi's and Konow's theory that svāmin is a Sanskrit translation of murunda, title of the Śaka kings, and the conclusions drawn from it see Keith's summary (p. 71). P. V. Kane, while objecting that in the Nātyaśāstra svāmin is prescribed as a term of address for the prince royal (19.12 yuvarāja, see History of Sanskrit Poetics 3, p. 41), disregards the fact that, e.g. at 3.93: 94 svāmin is used as synonym of nrpa (see n. 71). A similar case is 2.29 (see above, n. 64). This use as a term of address for the king is sanctioned by the Sāhityadarpana VI.144 (431), which, however, restricts its use to the servants (bhrtya).

69 Konow, p. 8.

70 NŚ. 3.87 homaṁ kṛtvā yathānyāyaṁ havir mantraputaskṛtam (88) bhū(n)dyāt kumbhaṁ tataś caiva nātyācāryaḥ prayatnataḥ, abhinne tu bhavet kumbhe svāminaḥ satrutaṁ kṣayaḥ (89) bhinne caiva tu vijñeyaḥ svāminaḥ satrusamkṣayaḥ. For the meaning of these verses see below, p. 162ff.

71 NŚ. 3.93 samyag iṣṭas (consecrated) tu raṅgo vai svāminaḥ śubham āvahet (94) purasyā 'bālavṛddhasya tathā janapadasya ca, duriṣṭas tu tathā raṅgo daivatair duradhiṣṭhitah (95) nātyavidhvaṁsanāṁ kuryāt nrpasya ca tathā 'śubham.

72 NŚ. 36.24C1 yāvat taṁ pūrayed deśaṁ dhvanir ātodyasāṁśrayāḥ, na sthāsyanti hi rakṣāṁsi na ca vighnavināyakāḥ (25) āvāhe ca vivāhe ca yajñe nrpatimangale, nāndīśabdam upāśrutya hiṁsrā naśyanti sarvadā. Cf. 5.110 praśastv imāṁ mahārājaḥ pṛthivīṁ ca sasāgarām.

(para-mandapika) and the chief citizens" 73. The lack of explicitness of

the older sources, however, leaves room for divergent opinions.

As for the occasions on which the drama was performed, they are

not conclusive as regards the personage of the patron. What the

Nāṭyaśāstra says of dancing may have been generally true of dramatic

performances: 'In general dancing is by nature liked by every one; and

dancing is declared to be auspicious on the occasion of marriage, child-birth,

reception of a son-in-law, joyous religious festivals, etc.; dancing is also

called a source of amusement" 74. Although some occasional expressions

such as "it creates beauty" or "giving amusement" 75 seem to refer to

a purely aesthetic appreciation, the indications pointing to an originally

ritual character of the performance are too strong to be questioned on

this ground. In this connection the following summary of some of

Schärer’s conclusions about the Ngaju Dayak should be noted: “ . . . among

the Ngaju Dayak the two months between the old and the new year,

when all the agricultural tasks are finished, are considered the most

proper time for contracting marriages. It is the time between the expiry

of one period and the beginning of another in the existence of the world:

the time of a new creation. This is also the period in which the community

celebrates its major religious feasts. The tree of life is then erected, later

to be destroyed again. All sorts of contests are organised and theatrical

performances held. The wedding is one of these major religious feasts,

for to be married means to enter a new stage of the sacred life. It is the

same kind of event as birth, initiation and death . . . ” (J. J. Ras, Bijdr.

Taal-, Land- en Volkenkunde 109 (1973), p. 453f. Italics his).

For determining the character of the ancient performances the number

of spectators is naturally important. The interpretation of the prescriptions

for the building of a theatre is, however, notoriously a moot point. The

wording of the text lends itself to very divergent explanations. M. M. Ghosh

arrives on the basis of his interpretation at the conclusion that the theatre

which the author had in mind can at best have accommodated about

four hundred spectators 76. He, therefore, assumes that “dramatic

spectacles meant for the common people were held in the open halls

73 Bhāvaprakāśana, p. 295, line 10. The Nāṭyaśāstra differs from Śāradātanaya in

that it refers to the presence of stribālamūrkhāh, see IIJ. 16, p. 243 n. 10.

74 NŚ. 4.269 prāyena sarvalokasya nrttam iṣṭam svabhāvatah, maṅgalyam iti kṛtvā

ca nrttam etat prakṛtilam (270) vivāha-prasavā-vāhāpramoda'bhyudayādiṣu, vinoda-

karanam caiva nrttam etat prakṛtilam.

75 Cf. 4.268 śobhām janayati, 1.120 vinodajananam loke (cf. 4.270 vinodakaranam,

quoted in the preceding foot-note, 5.165 rāgajanaka). See also 2.21 yasmāt pāthyam

ca geyam ca sukham śravyataraṃ bhavet, which, however, more concerns the technical

aspect of acoustics in the play-house.

76 Ghosh, Translation, Introd. p. li; Mankad, Ancient Indian Theatre (1950), p. 11,

considered two possible solutions, viz. 500–600 or 20.000 spectators for the smaller

type of theatre, and 1500 or 25.000 for the larger one. The possibility of a covered

building accomodating 20.000 or 25.000 spectators cannot seriously be considered.

The mere comparison with the open-air theatres of Greece is sufficient to show

126

called the Nāṭ-mandir (Nāṭya-mandira) in front of temples, or in a temporarily devised theatre under the cover of a canopy, as in the case of the modern Bengali Yātrās which seem to have unmistakable resemblance and connexion with the ancient Indian Nāṭya described in the NŚ" (p. li). As far as I can see, this is pure speculation, without any support of the Nāṭyaśāstra, which everywhere seems to refer to a walled building 77. Nowhere in this text can a confirmation be found for the existence of separate buildings "for the common people" - a distinction also made by Mankad 78 - in addition to the regular playhouses. Since the lack of precise information has naturally led to all kinds of speculations, it may be added in conclusion that also Keith's words to the effect that "At great festivals, when plays were given in the temples, there must have been admission for as many as could be crowded in" cannot be substantiated with the evidence at hand. The Nāṭyaśāstra, at least, does not mention, as far as I can see, any other place for performances than the playhouse especially built and consecrated for it. That temples have been used for this purpose is just possible but cannot be proved [cf. p. 241].

The final conclusion to be drawn from the facts mentioned is that the classical Sanskrit drama must have been composed and performed for an élite. As far back as 1890 Sylvain Lévi already pointed out what high requirements the spectators had to meet: "Tous les spectateurs ne sont pas aptes à goûter le rasa; c'est une sorte de récompense qu'il faut mériter par une étude assidue des poèmes et par des impressions saines et délicates accumulées pendant les existences antérieures" 79 and Winternitz rightly characterized the drama as "das wertvollste Erzeugnis der höfischen Kunstdichtung" 80. This has been the conclusion of most scholars. Only that the technical problems of such huge buildings could hardly have been mastered in ancient India but, even apart from this negative argument, the indications pointing to a comparatively small audience are sufficient to dismiss Mankad's alternative.

77 The Nāṭyaśāstra uses, without any perceptible distinction, the terms nāṭyāgāra (34.79C1=24.53 KM2), nāṭyaveśman (2.2, 4, etc.), veśman (2.10, 20), nāṭyamaṇḍapa (2.3, 6, 18, 25, cf. 14.1), maṇḍapa (2.8, 17, 19), nāṭyagraha (2.31, 3.1), prekṣāgṛha (2.7, 12, 21), geha (2.23) and, as a very general, non-specific, term bhavana "building" (2.11). All these terms denote covered buildings. Cf. also Lévi II, p. 62, Jones, JAOS. 93, p. 288.

78 See D. R. Mankad, Ancient Indian Theatre (1950), p. 23: "the Śilparatna tries to describe the Nāṭyamaṇḍapa which was usually attached to the Royal palace, while the Nāṭyaśāstra describes the usual theatres which were mostly meant for the ordinary people". Macdonell's supposition (in A History of Sanskrit Literature [1900], p. 352) that in the Indian Middle Ages plays were performed in the concert-room (saṅgītaśālā) of royal palaces was rejected by Haraprasad Shastri, J. and Proc. As. Soc. Bengal 5 (1909), p. 353, who also referred to the term prekṣāgṛha, Prakrit pekkhāgharāra. But see S. Lévi II, p. 62.

79 See Le théâtre indien, p. 258.

80 See Geschichte der indischen Litteratur III, p. 160. Similarly J. K. Balbir, IIJ. 6 (1962), p. 44: "it is obvious that Sanskrit drama was intended to be a drama of the élite, written by master-artists, played on special occasions by finished actors, and enjoyed by qualified persons. It was a refined product religiously presented as an offering before a discriminating audience".

Konow, after stating “In dem klassischen Drama wird es überhaupt vorausgesetzt, dass das Publikum ein gebildetes ist”, then adds the words

“Auch das grosse Publikum hat aber ohne Zweifel eine gewisse Kenntnis von den technischen Mitteln der Schauspieler gehabt” 81. Whoever has

read the definition of the preksaka in BhN§. 27.49ff. and considered it

in connection with the other data will question the correctness of this

speculation. In its origin the drama was rather a ritual performed on

the initiative, in the presence, and for the benefit of the king.

3. BHĀRATĪYA-NĀTYAŚĀSTRA CH. I: DATE AND PARTICULARS OF THE

LEGENDARY FIRST PERFORMANCE

After these general introductory remarks our first task will be to

consider more closely the legendary story as related in chapter I of the

Nātyaśāstra. It runs as follows: At the request of the gods, Brahmā

composes the Nātyaveda as a fifth Veda, which contains the quintessence

of all four Vedas and is open to all social classes. He then asks Indra

to give, with the gods, a performance of the legend (itihāsa) he has composed

(1.19-23) but Indra refuses because the gods are, according to him,

incapable (aśakta) and unfit (ayogya) to learn and perform a drama.

Indra says that this is rather a task for the “wise men who know the

secrets of the Veda” (vedaguhyajñā munayah). Brahmā then orders

Bharata to rehearse the drama with his hundred sons to the accompaniment

of song and instrumental music (1.24ff.). The implication of this part of

the story is obviously that in the nātya human beings enact rôles which

were designed for gods: the drama is a “divine comedy”, in the sense

that cosmic events are presented on a human scale.

When Bharata informs Brahmā that the rôles have been studied and

asks for further instructions, Brahmā says: “ ‘Just now an important

moment has arrived for the performance. The auspicious banner festival

of the great Indra has just begun. On this occasion now perform this

Veda which is called Drama!’ Then, at that banner festival (celebrated)

on the occasion of the slaying of the Asuras and Dānavas, at that festival

of the victory of the great Indra, where the gods were assembled in joy,

I first uttered a Benediction (nāndī) with blessings, which was full of

variation as it consisted of words of the eight categories and which were

approved of by the gods. Thereafter I arranged an imitation of how the

Daityas were defeated by the gods, which was full of altercation, tumult,

mutual cutting off and piercing (of limbs)” 82.

From these lines there emerges a picture which, although it is a

projection of actual practices backwards into a primordial time, still bears

testimony to the preservation of recollections of earlier stages of the

Sanskrit drama. To begin with, this first performance was a re-enactment of

the cosmogonical strife between Indra and the Asuras, and by repeating the

defeat of this group the “creation” of the world was again actualized.

81 Konow, p. 8.

82 See n. 145.

128

This performance was baddha, which must mean “arranged”, since “composed” or “devised” (Ghosh) would seem out of the question here (see n. 144). If so, however, the question as to what exactly had been rehearsed by Bharata remains unanswered. It seems that the author of these lines was primarily thinking of the art of acting, rather than of memorizing a text. Anyway, the presentation of the cosmogonial event can only have been meant as a reiteration of the act of creation, that is, as a renovation of the world. It was, accordingly, a religious act par excellence, a drōmenon in the proper sense of the word.

This first performance is here connected with a religious ceremony, Indra’s banner festival, which is here represented as a celebration of Indra’s victory over the Asuras. This is confirmed by some of the oldest texts which deal with this festival and trace the origin of its ritual back to a primordial fight between Indra and the Asuras. Thus Varāhamihira (c. 490–587 A.D.) in his Bṛhatsaṃhitā 43 (42).1-7 relates how the gods at one time, unable to conquer the Asuras, appealed to Brahmā, who then advised them to ask Viṣṇu for the dhvaja. When Viṣṇu gave this banner to the gods, Indra placed it upon his chariot and with the help of the dhvaja he defeated his foes 83. The mythical situation is the same as in the Churning of the Ocean (Mbh. I.16.28-31), where Brahmā stands above the parties but Viṣṇu, at Brahmā’s request, gives the gods strength (although he says that he will give it to all who participate in the battle, an ambiguity which can hardly have been unintentional, see p. 105).

A fairly similar version (stemming from the same source, Garga) occurs in the Viṣṇudharmottarapurāṇa II.154.1-17, where Indra after his victory worships the banner himself (v. 17).

On the basic meaning of the banner festival, which is in some texts represented “as a royal rite for the benefit of the king’s welfare and superiority in war” 84 opinions differ 85. The character of the “banner”,

83 Cf. Abhinavagupta ad NŚ. 1.55: dhvajamahāhasya viśeṣanadvāreṇa sambhavam darśayati ‘nihatāsuradānava’ ityādinā “By means of the attribute ‘after the Asuras and Dānavas had been killed, etc.’ he indicates the origin (occasion) of the Banner Festival”. The importance of the detail in the version of the Bṛhatsaṃhitā that Indra places the banner upon his chariot seems never to have been noticed. Cf. 43.5 taiḥ samṛstutaḥ sa dewas tutoṣa Nārāyaṇo dadau caiṣām, dhvajam asurasuravadhūmu-khakamālavanatṛṣitakṣāmaṃśum (6) tam Viṣṇutejobhavam aṣṭcakre rathe sthitam bhāsati ratnacitre, dedīpyamānāṃ śaradī ‘va sūryaṃ dhvajam samāsādya mumoda Śakraḥ (7) sakṛnniḥalāparikṛtena srakchattraghantāpīṭakāṃvītena, samucchritenā ‘mararād dhvajenā vinatāṃ samare ‘risainyam. It has been argued elsewhere that the banners on the chariots were, indeed, representatives of the Indradhvaja (see IIJ. 11, p. 154ff., particularly p. 156, where this passage is quoted). Gonda, JAOS. 87 (1967) p. 416, quotes BhNŚ. 3.11ff. (pājā to the jarjara) for the ritual aspects of the festival, but that passage only describes the function of the jarjara, whereas here the cosmogonial character of the festival is stated expressis verbis.

84 Gonda, JAOS. 87 (1967), p. 414.

85 See J. J. Meyer, Trilogie altindischer Mächte und Feste der Vegetation, Ein Beitrag zur vergleichenden Religions- und Kulturgeschichte, Fest- und Volkskunde (Zürich-Leipzig, 1937), III, where all the relevant texts are quoted in German

mostly a bamboo pole (venumayi yaṣṭi VarBS. 43.8)86, as a representative of the axis mundi cannot, however, be questioned. Its association with Indra's cosmogonical act, for which evidence can be found in comparatively early texts87, then admits of no other interpretation than that the erection of the pole was a ritual re-enactment of Indra's "propping up" the sky by means of the cosmic pillar (see IIJ. 11, p. 156, 16, p. 248ff., and cf. p. 141). The banner festival was, consequently, a reiteration of the creation and must as such have inaugurated the new year, that is, a new life in a renovated world.

Two points of detail may here be stressed. First, several texts, both of earlier and more recent times, confirm that the banner festival was organized by the kings88. Secondly, many passages mention "spectacles" translation. The only two texts which Meyer had dismissed as worthless (p. 55 n.), viz. the Kauśikasūtra and Atharvavedapariśiṣṭa, have since been carefully translated and commented upon by J. Gonda, JAOS. 87 (1967), pp. 413–429. For the archaeological evidence see Odette Viennot, Le culte de l'arbre dans l'Inde ancienne (Paris 1954), pp. 96–98. While in Meyer's opinion the Indra-tree is a solar and vegetation tree (p. 134), which brings good luck and prosperity in this world and in the next (p. 116), Gonda, p. 416, formulates his conclusion in these words: "it is not only a seasonal rite and a form of tree cult, but also assures the king's victoriousness and invincibility". This is no doubt correct as far as it goes. It should only be observed that the hypothetical meaning of the world pillar, which may be inferred from its cosmogonical origin, viz. a symbol of life and guarantee of the cosmic order, may be formulated in very different ways according to the specific context in which it occurs. It is not advisable, therefore, to attach too much importance to a single passage. Thus Mbh. I.57.17 has iṣṭapradānam uddiśya śistānām paripālinīm, which I translate "for the fulfilling of desires [as a kalpataru]

and protective of the learned men" (otherwise Meyer, p. 4: "indem er sie als erwünschte Liebesgabe bezeichnete"). The Harivaṃśa 59.6, however, says (meghāh) sasyaṃ janayanti navāmbubhiḥ "(the clouds) cause the corn to grow by fresh rains", which reflects the main concern of an agricultural community. This has a striking parallel in the medhi which according to the (apparently late) Krṣiparāśara was erected in the field to ensure the growing of the corn (sasyavṛddhikara, see IIJ. 11. pp. 156, 214f.). This belief may have been as authentic as the kings' belief that the pole brought them victoriousness. Cf. Brhatsaṃhitā 43.11, where the pūjā to the Indradhvaja is performed narendrair balavd dhijayārthibhiḥ and, somewhat parallel to it, the pūjā to the jarjara, NŚ. 3.13: nrpatya vijayaṃ sāma r̥pūnāṃ ca parājayam, gobrahmaṇaśivaṃ caiva nātyasya ca vivardhanam, which is instructive for the variety of intentions. Gonda, p. 418, discusses the possibility that the "emphasis laid on victoriousness" may be of later date. The different character of the various literary sources may to a large extent have determined the different emphasis. In conclusion it may be observed that the character of the Indradhvaja, connected as it was with a very definite cosmogonical myth, was so specific that by subsuming it under the general heading of "tree worship" one runs the risk of missing the point. It was not the tree as such but the renewing of the world which it symbolized that stood central (see Odette Viennot, Le culte de l'arbre dans l'Inde ancienne, pp. 93–98). See also n. 123.

86 See Meyer, p. 18.

87 Viṣṇudharmottarap. II.154.17: tato labdhajayaḥ Sakrah pūjayām āsa tam dhvajam. For a translation of the whole passage see Meyer, p. 6ff.

88 Cf. Kauś. Sūtra 140.1 and, e.g., Mbh. I.57.19 tataḥ prahṛti cā 'dyā 'pi yaṣṭyāḥ kṣitipasattamaiḥ, praveśaḥ kriyate . . ., Harivaṃśa 59.18 tasmāt prāvṛṣi rājānah sarve Sakraṃ mudā yutāḥ, mahatīḥ suresaṃ arcanti . . .

130

(preksāh, preksanīyāni) performed during the festival 89 but, as Meyer observes, this refers to performances by dancers, wrestlers (malla) and jugglers. These festivities for the ordinary people clearly did not include dramatical performances and must, therefore, be kept distinct from the latter: they were limited to such popular amusement of the town's people as could take place in the streets 90. In all discussions of the origin of the Indian drama, as far as I can see, the essential fact has constantly been overlooked that the characteristic difference between the drama and all popular amusements is that the former had to be performed in a consecrated building.

About the time of the year when this festival was celebrated the texts have very divergent statements but Hopkins's conclusion still holds good: "it is at least clear that the festival occurred after the rains had ceased and when New Year's was celebrated, for in its installation it is especially said that the feast takes place at the end of the year (gate samvatsare)" 91. According to the Prakrit description quoted in n. 90 the festival apparently lasted eight days, the last one being the full moon, when the king himself worshipped the tree. The next day the tree was

89 Cf. Viṣnudharmottarapurāṇa II.155.17 sthāne sthāne deyā preksyā (read preksā, Meyer) . . . pūjayen nrtyagitena rātrau Śakram narādhipah (21) nityam nrtyena gītena tathā Śakram ca pūjayet, Bhavişyottarapurāṇa 139.25 preksanīyaih (Meyer, p. 38). Meyer (p. 10 n. 4) does not attach much importance to this detail because performances are mentioned with reference to every festival. As for the reading nrtya-, the distinction made by M. Krishnamachariar, History of Classical Sanskrit Literature (Delhi, etc. 1937/1970), viz. "The Nrtya is gesticulation without language, or pantomime; and the Nṛtta is simple dancing" is, it seems, of later date. It is based upon Daśar. 1.12-13. In Vedic texts the nouns corresponding to JB. II.69 yad viṇāyām gīyate, yan nrtyate "what is sung to the accompaniment of the lute, what is danced" gīta and nrtta, cf. VS. 30.6 (VSK. 34.6) gīdyā śailūṣám : TB. III.4.2.1 nrttāya śailūṣám and VS. ibid. nrttāya sūtám, ŚB. III.2.4.6 (yā . . .) nrttām gītám upāvavárta, KB. XXIX.5.16 nrttam gītam vādyam iti (v. 1. nrtyam only in the printed editions of Lindner and Poona). The Nātyaśāstra seems still to continue the Vedic usage, e.g. 5.163C (158B) nrttagītavidhim prati, v. 1. gītarnavidhim p.), 164 gite vādye ca nrṭte ca, 22.47 C. bahunrttagītā, and cf. 5.12, 93, 131, 148, 159C, but Mhbh. XIII.109.36 has (a)parasaram nrtyagitavinādite, only the South Indian manuscripts reading nrtta-. See n. 107.

90 Cf. the Prakrit story of Domuha (Jacobi, Ausgewählte Erzählungen in Māhārāshṭrī, p. 40,18): tao naccanti naṭṭiyāo, gijjanti sukaīraiā kappabandhāo, naccanti narasaṁ-ghāyāo, diśanti diṭṭhimohanāṁ indayāläṁ, indayāliṇo ya dijjanti tambolāim, khippanti kappūrakunikumajalacchāḍā, dijjanti mahāḍāim, vajjanti muingāiojjāim (drums and instruments), evam mahāmoena gayā satta vāsarā "Then the nautch girls danced, poetic compositions written by good poets were sung, the multitude of men danced, juggler's tricks that bewildered the eyes were seen, and betel and other things were given to the jugglers; a great deal of camphor, saffron, and water was thrown, great gifts were given, drums and other instruments were sounded. Thus seven days passed in great joy". (Translation by J. J. Meyer, Hindu Tales, p. 143). Cf. also the southern recension of the Mahābhārata I.57.507*, line 5ff. sabhājayitvā rājānam krtvā narmāśritāḥ kathāḥ, ramante nāgarāḥ sarve tathā jānapadaiḥ saha, sūtāś ca māgadhāś caiva naṭante naṭanartakaiḥ, etc. Var. Brhatsaṁhitā 43, 9–10. For similar festivals (stūpamahā) when a stūpa was erected see Avadānaśataka I, p. 385,7.

91 See Hopkins, Epic Mythology, p. 126. See below, p. 133.

pulled down and despoiled by the people. It can be inferred that the pole was not allowed to remain standing during the waning moon 92. The main reason, however, why the tree had to be removed will be discussed below. In spite of his correct statement that the tree was erected at the end of the year Hopkins finally explained the date of the festival from the fact that it coincided with the period of warfare 93. There are, however, no indications to show that this can have determined the date of the Indra-festival 94.

The oldest text, the Kauśikasūtra 140.2, gives as the exact date of the "entrance" of the tree into the town the eighth day of the light fortnight in the month Prauṣṭhapada (now: August-September) or Áśvayuja (now: September-October) 95. Although in later times the former two dates was considered to be characteristic of the Sāmavedins 96, it is here found in a work of Atharvavedins. The same date is given by Garga, the source of the relevant passages of the Bṛhatsaṃhitā and the Viṣṇudharmottarapurāṇa (Meyer III, p. 17). The passage which is of particular interest has been handed down in the commentary on VarBS. 42 (43).7: "Garga has it as follows: 'When the Asuras saw the dhvaja, they lost, dispelled by the power (tejas) of dhvaja, consciousness and, defeated in the battle, they ran away and disappeared (or, perished).

After the thousand-eyed god had slain the Asuras with his vajra (see below, p. 142) in the month Bhādrapada (māse Bhādrapade) ... the victorious one [read sajitvā? Cf. RS. III.12.4] went on his way to heaven, under Śravaṇa, accompanied by the twice-born' 97. The last words might be interpreted as referring to the end of the seasonal worship of Indra but according to most sources it is the erection of the pole, not its pulling down, that takes place "under Śravaṇa" (Kauś.S. 140.3, see Meyer III, p. 113f., Gonda pp. 418 and 420). Later texts give somewhat earlier or later dates, ranging from June-July to November-Deœmber 97.

92 Similarly Viṣṇudharmottarapurāṇa II.154. See J. J. Meyer, p. 113.

93 Hopkins, p. 125f.: "Indra's day comes when the rains are over and the roads are fit for war, and is the new moon's day of Saumya māsa (probably the end of September: amāvāsyā Śakradavatā 5,142,18)", but he thinks it is impossible to identify this "Indra's day" with the festival. Similarly Gonda, JAOS. 87 (1967), p. 420, who, however, rightly opposes the last statement; see p. 417 n. 25.

94 Although the end of the rains is traditionally known as the season for military expeditions, the months most suited for warfare are according to Mbh. XII.101.9f. Mārgaśīrṣa and Caitra, accordingly January-February (or December-January, see n. 112) and March-April (or February-March): caitryām vā mārgaśīrṣyām vā senāyogaḥ praśasyate, pakvasasyā hi prthivī bhavaty ambumati tathā. (10) naicā 'tiṣito na 'tyuṣṇaḥ kālo bhavati Bhārata.

95 See Gonda, p. 420.

96 See Auguste Barth, Oeuvres 4, p. 175 (quoting Jacobi, Rām. IV.27.10 Gorresio (21.10 N.W. recension)=IV.27.34 crit. ed.

97 See Meyer, p. 113ff., especially p. 117f., Gonda, p. 420, Bhat, The Vidūṣaka p. 33, M. M. Ghosh, Nāṭyaśāstra, Translation p. 9 n. 1 (ad 1.53). According to Keith, Sanskrit Drama, p. 41, the Indra-pole was originally the Maypole, erected at the end of the winter. In India, however, he thinks there has been a shift, the pole having lost its original character and symbolizing "a thanksgiving for Indra's victory

132

This is sufficient proof to show that the festival was not originally connected with either the culmination of the sun ("Johannisfest")98 or war-fare.

On the other hand, there is one passage in the Mahābhārata where (as Hopkins has pointed out) the actual meaning of the ceremony is stated in unambiguous terms. In the story of the Cedi King Vasu, to whom Indra had given the bamboo pole (yaṣṭi vainavī) and who was the first man to celebrate the pole festival, it is said: "In order to worship Śakra on earth the protector of the earth (the king) ordered the pole to be brought into the town when the year was gone"99. In explanation of this statement the commentator Nīlakanṭha adds the following note: "Nowadays one can still see the 'entrance' of the pole in Mahārāṣṭra and other countries at the end of the year"100. Since he here clearly refers to a custom that he knew from personal experience (that is, about 1675 A.D.) and since the words gate samvatsare of the Mahābhārata cannot possibly mean anything else but "at the end of the year" the doubts expressed by Meyer101 are unfounded.

It is a well-known fact that Indian sources mention different dates for New Year's day102. As has long been observed, a radical change took place

over the clouds, the Asuras". This mythological reconstruction will have to be restated: although it is indeed likely that the Indra-pole has always symbolized the beginning of a new year, the date of New Year's day must have shifted considerably after the oldest Vedic period.

98 Meyer, p. 118.

99 Mhhb. I.57.18 tasvāḥ Śakrasya pūjārtham bhūmau bhūmipatis tadā, praveśam kārayām āsa gate samvatsare tadā.

100 Nīlakanṭha: samvatsarānte yaṣṭipraveśo 'dyāpi Mahārāṣṭrādiṣu dṛśyate.

101 See Meyer, Trilogie III, p. 4 n. 2. Meyer refers to Bhavisyottarapurāṇa 139.7, where it is said that Vasu brought the pole (yaṣṭi) from heaven to earth during the rains, whereas verse 9 says that in later times people imitated him in worshipping the pole varṣānte. Here a translation "innerhalb der Regenzeit", which Meyer (pp. 36, 114) diffidently suggests, is plainly out of the question. Cf., e.g. Tamil varuṣāntam "close of the year". Similarly the commentator on Viṣṇupurāṇa V.10.24 prāvrṣi: "at the end of the rains" (see n. 115).

102 Cf., e.g., Weber, Jyotiṣam, pp. 27 (partly the winter solstice), 78, Oldenberg, Kleine Schriften pp. 28, 647 (spring as the first month), 663, Hillebrandt, Romanische Forschungen 5 (1889-90), p. 310ff., Rituallitteratur p. 115, Vedische Mythologie I2, p. 33, II2, pp. 182, 137, Macdonell-Keith, Vedic Index II, pp. 157, 412f., J. Hertel, Das indogermanische Neujahrsopfer im Veda (1938), passim, J. Filliozat, L'Inde Classique II, p. 724, C. d'Onofrio, Studi e Materiali di Storia delle Religioni 24/25 (1953-54), p. 141f. (with ample references). Claus Vogel, ZDMG, 121 (1970[1971]), pp. 284-326, and the references given in IIJ. 4, p. 219 n. 7. According to the Jaim.Up.Br. I.35.5ff. the year ended with hemanta and began with vasanta, which constituted the two "ends" of the year. With the first (hemantá) the three seasons of the Manes ended and with vasantá began the three seasons of the gods (ŚB. II.1.3.1).

[For similar speculations about the two Phalgunis and the two atirātras being the end and the beginning see A. Barth, Oeuvres 4, p. 174f. and, e.g., TS. VII.5.1.3, ŚB. VI.2.2.18]. The rainy season and the autumn are in the middle of the year (ŚB. VIII.3.2.7-8), which is not contradicted by RS. VII.103.9 (as Jacobi suggested): see, e.g., Oldenberg, Kl. Schr. p. 659f. According to Benveniste, Mnemos Charin I, p. 34, the word hāyaná "year" shows that the year began with

133

in the post-Vedic period, as a result of which the month Mārgaśīrṣa (now: mid-November to mid-December), in older times the tenth month of the calendar, became the first 103. It is not necessary for the purpose of this study to inquire further into the historical background of this change. Even though it may be true that certain Vedic sacrifices began after the rainy period 104, the fundamental change in the calendar of "L'Inde des moussons" seems to have taken place in post-Vedic times. In the Mahābhārata Mārgaśīrṣa is clearly the first of the months 105. In

the winter (*hayan : Av. zayan). This might be compared with Goth. twalibintrus "twelve-years-old" (and Latin bimus, trimus, Cl. Vogel, p. 284) but the Ṛgvedic poets use śatā́ṁ śarā́dah just as much as śatā́ṁ himā́h. Cf. bahvī́h sámāh and ŚB. XII.8.2.35 "All the seasons are first, all of them intermediate, and all of them last". Anyway, after a system of six seasons had been formed by splitting up hemantā́ into hemantā́ and śiśirá (Claus Vogel, p. 286), the winter ended with śiśirá, and New Year's day (ekāṣṭakā́, AS.III.10, TS. IV.3.11.3, etc.) fell on the eighth day of the second half of Māgha, that is, January-February or formerly December-January (māghakṛ̣ṣṇaṣṭamī́, Sāyaṇa ad AS. III.10.12 and introduction to that hymn). See Hillebrandt, Rituallitteratur, p. 94ff., Ved. Myth. I 2, p. 30, Caland, note on translation of PB. V.9.1, Keith, note on transl. of TS. VII.4.8.1 (p. 607 n. 3), Vedic Index I, p. 119, II, p. 157. This is, accordingly, round about the winter solstice, which fell on the new moon of Māgha (KB. XIX.1.28 māghasyā́ 'māvāsyā́yām). According to Caland New Year's day was ("as it seems") a week after Ekāṣṭakā́. But from MS. II.5.9 (59,3), IV.2.3 (25,4), KS. XIII.3 (182,15), AS. III.10.2 and 8, etc. it is apparent that it was New Year's night. In different calendars, again, the new year began in the month of Phālguna, i.e. February-March (cf., e.g., the commentary on PB. V.9.8 etad uktaṁ bhavati: yadā phālguni paurnamāsy uttara-phālguni yukta tāda phālgunamasaḥ samvatsarasya 'dir bhavatī 'ti) or in Caitra, i.e. March-April. Cf. ĀpŚS. XXI.15.4-6, Keith's note on his translation of TS. VII.4.8.1-2,

Heesterman, The Ancient Indian Royal Consecration p. 7f. and further Oldenberg, Kleine Schriften, pp. 647f., 663ff. (from 1894–95), C. D'Onofrio, l.c. (either the 8th or the 9th day of śuklapakṣa of Phālguna, or full moon of Caitra, or 8th or 9th day of śuklapakṣa of Jyaiṣṭha or Āṣāḍha; or generally in the period between spring and summer, cf. ŚB. XIII.4.1.2-3 (according to some one should begin the Aśvamedha in summer, for summer is the kṣatriya's season, but it is strongly advised to begin it in spring, which is the brāhmaṇa's season).

103 PW. V (1868), col. 745: "der zehnte (später der erste) Monat im Jahre". But now Mārgśīrṣ is the ninth month of the Hindu year, the first being Cait (Caitra), i.e. March–April. The Fasli year, however, commences with the month of Āśin (see Grierson, Bihar Peasant Life, p. 274), that is, in September–October. Similarly the Tamil-Malayalam year begins on the first day of the month of Cittirai (Mēdam) (10–14 April). Divination, however, here takes place at the time of the winter solstice during poṅkal, which marks the first day of the month of Tai (January) as the beginning of a new period. The festival on the preceding day (pōki(p)paṅtikai) is celebrated in honour of Indra ((L. Dupanis et L. Mousset1. Dictionnaire tamoul-français s.v. poṅkal), who is here, accordingly, associated with the winter solstice. The Cola kings, on the other hand, celebrated the Intiraviḷau festival in the month of Cittirai.

104 The cāturmāsya offerings started in autumn, in winter or at the beginning of the rainy season. See Jacobi, Festgruss f. Roth, p. 71, Hillebrandt, Rituallitteratur, p. 115f.

105 Mhbh. XIII.109.17–19 (the twelve months from Mārgaśīrṣa to Kārttika), Appendix I, 12 line 5 (in a list of upavāsas; it starts with dvādaśyā́ṁ Mārgaśīrṣe tu), 12A line 5 (an enumeration of the months starting with Mārgaśīrṣa). Cf. Bhagavadgītā 10.35 māsānā́ṁ mārgaśīrṣo 'haṁ rtūnā́ṁ kusumākarah.

134

this later calendar the main caesura in the annual cycle is the period of

the rains (about mid-June to mid-September), which coincides with

Viṣṇu's sleep. Viṣṇu's awakening on the eleventh day of the light half

of Kārttika (26 October) 106 inaugurates a new beginning and is celebrated

accordingly 107. The circumstance that Kārttika, the last month of the

year, became the specific time for dicing may also reflect this change in

the calendar. At least, Held's observations 108, if correct, would lead to

the supposition that the time for dicing must originally have been what

he calls "the period of the great tribal feasts". In terms of Vedic social

life this would mean: in the period of the contests round about

New Year's day. In any case, there can be no doubt that the rainy season,

was a most inauspicious time, surrounded by taboos 109.

As such it was, as far as its character is concerned, comparable to the

Vedic annual period of áṃhas. This must have been at the end of the

year and must have formed the transition to the new year, as may be

inferred from the fact that the thirteenth (intercalary) month was called

"lord of the áṃhas" 110. Just as the Vedic period of cosmic chaos was

followed by New Year's day, so the rainy period of classical India formed

• at least, in some of the later calendar systems - the transition from the

old to the new year. Diwali, which is celebrated on the day of the new

moon of Kārttika 111, is nowadays often regarded as a New Year's festival.

That the rainy season, as the main caesura in the annual cycle, has

come to determine the beginning of the new year is, therefore, probable.

The month Mārgaśīrṣa, it is true, does not coincide with the end of the

rainy season since nowadays it corresponds with mid-November to

mid-December. At the time of the Saurasiddhānta (± 350 A.D.), however,

which can in this respect be considered contemporaneous with the

Mahābhārata, the same month must have corresponded with October-

November 112.

When the dates of the various parts of India are taken into account,

the rainy season lasts from mid-June to mid-October and comprises,

106 Thus Hillebrandt, Ved. Myth. I 2, p. 361.

107 Cf. J. J. Meyer, Trilogie III, pp. 10, 34 (jāgarana), 20, 39 (jāgara), and, e.g.,

Skāndapurāṇa VII.4.24.35 nrtyaṁ gītam ca kartavyam saṁprāpte jāgare tava.

108 See G. J. Held, The Mahābhārata, p. 243ff., especially p. 277.

109 See, e.g., the references in Hillebrandt, Ved. Myth. II 2, p. 447, Held, The

Mahabharata, pp. 190, 185, 197 (not entirely correct in detail, for the rainy season

was certainly not the time of war). The taboo on travelling in the rainy season

was not restricted to Buddhists but also valid for Brahmins, cf., e.g.,

VaikhDhS. III.7.8.

110 See VS. VII.30, XXII,31, VSK. VII.12.1, XXIV.18.1 aṁhasáspatáye, MS.

III.12.13 (164,7), TS. I.4.14.1, VI.5.3.4, ĀpŚS. VIII.20.8 aṁhaspatyàya (=saṁsarpa,

Ved. Index II, p. 162).

111 On the day of the new moon (amāvāsyā) of Kārttika, that is, on the fifteenth

day of the dark half of that month. See, e.g., Platts, A Dictionary, Hindūstānī and

English, Meyer, Trilogie II, p. 90 n. 1 and passim, Hopkins, Encyclopaedia of

Religion and Ethics 5, pp. 868, 870.

112 According to J. Filliozat, L'Inde Classique II, p. 723.

135

accordingly, the months Āśādha, Śrāvaṇa, Bhādra (or Prauṣṭhapada) and Āśvina (or Aśvayuja) 113. The same months are according to J. J. Meyer (p. 117) mentioned in the texts as those in which the Indra-pole is erected. This would imply that the last and the first months of the calendar are not mentioned at all. In the beginning of his study (p. 5), however, Meyer had righty pointed out that two editions of the southern recension of the Mahābhārata state that the Indradhvaja was erected in Mārgaśīrṣa, the first month of the year. The Telugu and Grantha manuscripts collated for the critical edition read, indeed, instead of the pāda I.57.18d (praveśaṁ kārayām āsa) gate samvatsare tadā, the two pādas (502 *) sarvotsavavaramaṁ tadā Mārgaśīrṣe mahārāja. The two manuscripts in Malayalam script, on the other hand, add after v. 18 the line (503 *) Mārgaśīrṣe mahārāja paurnamāsyāṁ mahāmaham. The erecting of the Indra-pole in Mārgaśīrṣa was apparently a regional custom, which was restricted to South India. In modern Kerala (as Dr. A. Govindankutty informs me), every third year in Kumbham (= Phālguna, March-April) members of low castes bring a bamboo pole into the town, where it is erected in front of the temple of Bhagavatī and adorned with garlands, etc. After some weeks, but before New Year's day (14 april) the pole is removed. There are no doubt numerous parallels in Northern India. One of them is found in the Krṣiparāśara, a late text on agriculture, in which it is said that in the month Mārga (vv. 214, 219) a pole (medhi, vv. 214-220), adorned with flags (vaijayantisamāyukta, v. 217), is erected in the fields to protect the crops (see IIJ. 11, p. 214ff.).

It is clear, however, that older texts, such as the Kauśika Sūtra, give much earlier dates for the Indra festival and that it is not self-evident that our conclusion, based upon the words gate samvatsare of the Mahābhārata, also holds good for these texts. According to the Sūtra the festival was celebrated in one of the two months preceding the last month of the year of the later calendar. These months, Prauṣṭhapada and Aśvayuja, now correspond to August-September and September-October of our calendar, but must have corresponded to July-August and August-September about the beginning of our era. In this case it cannot be questioned that the festival was celebrated during the rainy season, whether or not this coincided with the new year. The term prāvṛṣi "during the rains" remained traditional in later texts, although there are indications to show that actually the date had changed 114. Thus the commentator of the Viṣṇupurāṇa interprets prāvṛṣi as meaning "at the end of the rains" 115. This curious explanation, which obviously reflects a shift of the date, is confirmed by the Harivaṁśa, which itself specifies the term prāvṛṣi by "in the autumn . . . at the end of the rainy season"

113 See Petersburger Wörterbuch and Monier-Williams s.v. prāvṛṣ.

114 Viṣṇupurāṇa V.10.24 and Harivaṁśa 59.18 crit. ed. (prāvṛṣi).

115 Similarly, e.g., Odette Viennot, Le culte de l'arbre dans l'Inde ancienne, p. 95. See also above n. 101 (varṣānte) and n. 100 (samvatsarānte).

136

and says that the pole is adorned with "autumnal flowers" 116. It would seem, accordingly, that prāvṛi, a remnant of older times, had come to be used as a rough indication for that period, at the end of, or after, the rainy season, which, marked by such festivals as Diwali and Viṣnu's jāgara and extending to Mārgaśīrṣa, had in later calendar systems become the beginning of the new year. It may be noted incidentally that in Vedic India as well as in Tibet, the New Year's festival was celebrated by contests and especially by chariot-races 117. See also above, p. 126. In modern India chariot-racing is also reported to occur on the occasion of the erection of a certain pole 118.

A few words must also be said about the conclusions which J. J. Meyer has drawn from his careful and important study of the evidence of the Sanskrit sources. His view that the banner festival celebrated the culmination of the sun and that the Indra-pole corresponds to the Johannisbaum in Germany has already been briefly mentioned above (p. 133). This theory was strongly influenced by Hillebrandt's interpretation of Indra as a solar deity, which Meyer had accepted as "one of the few certain results of Vedic research" 119. There is some contradiction in his interpretation since, on the other hand, he thinks that none of the Vedic poets was still aware of this character of Indra 120. That Meyer must assume considerable shifts in the date of the festival, from the summer solstice to the period of the rains 121, is clear. The main objection, however, that must be raised to his theory is that, while Hillebrandt disregarded the cosmogonical character of the Vṛtra-myth, a correct appreciation of that character necessarily implies that Indra was a seasonal god connected with the ceremonies that accompanied the transition from the old to the new year. In this light the importance of the classical testimonies to the effect that, whatever the calendar date of New Year's day may have been, Indra's festival was to be celebrated at the end of the year, should not be underestimated. Meyer ignored their vital importance or tried to explain them away. In this connection also Keith's explanation must be mentioned. In his opinion the drama was "once connected with

116 Harivaḿśa 59.118 tasmāt prāvṛi rājānāḥ etc. (see n. 88) but 31 śarat (cf. 47, 50, 55 referring to war), 57 śaradi . . . prāvr�ṣaḥ kṣaye, 58 puspaiḥ śāradikaiḥ. As for 62.46 varṣārdhe ca dhvajo nityam see Meyer III, p. 132.

117 See IIJ, 4, p. 217ff., 5, p. 169ff.

118 See Meyer, Trilogie III, p. 134 on "Wettrennen und Wettspiele" in connection with the erection of the pole of Zāhir-Pīr.

119 Trilogie III, p. 134: " . . . zu den wenigen ganz festen Erkenntnissen vedischer Forschung . . . gehört meines Erachtens die Entdeckung Hillebrandts, dass Indra ursprünglich ein Sonnengott sei", p. 135 "dass Indra ein Sonnen-, ein Frühlingsgott ist", p. 118 "was seit Hillebrandt für jedermann feststehen müsste". Hence his conclusion (p. 118): "Der Indrabaum wäre also genauer ein Johannisbaum" and further p. 191, where Indra is said to have been "ursprünglich ein Sonnen- und Frühlingsgott und ein Befruchtungsgenius". The word "ursprünglich" here seems to point to a certain reservation that is lacking in the other passages.

120 See p. 150 n. 1, which would seem to contradict p. 135.

121 See pp. 117f., 192.

137

the ceremonies of bringing in the Maypole from the woods at the close

of the winter, but in India the rite fell at the close of the rainy season,

and the ceremony was converted into a festival of thanksgiving for Indra's

victory over the clouds, the Asuras" 122. There is no reason to doubt

that in the old Aryan tradition Indra has always been the god who

conquered the Asuras, and that the festival has always been a symbolical

reiteration of that cosmogonical event 123. As such it must originally have

been celebrated during the winter solstice or the spring equinoctial,

before the beginning of the new year had been shifted and the tenth

month had become the first (see nn. 102 and 103).

In conclusion a few words must be said about the relation of the god

to his pole (yasti) or banner (dhvaja). This relation is more problematic

than might seem at first sight. The tree or bamboo pole is, first and

foremost, considered a symbol (cihna, laksman) of the god. During the

ceremony of the pürana (also called prapürana, āvahana, adhivāsana) 124,

however, the god was called upon to be personally present in his symbol.

Hence it is that the texts often refer to the tree as the god himself 125,

a belief that is still found as late as circa 1000 A.D. in Abhinavagupta's

commentary on the Nātyaśāstra 126. Although the identification of the

god and his symbol is, accordingly, an established fact 127, the question

of to what extent the Indra-pole was identical with the cosmic tree

(axis mundi) calls for some comment. The world pillar which keeps heaven

122 See Sanskrit Drama, p. 41.

123 Otherwise Gonda, JAOS. 87 (1967), p. 418: ". . . it is not beyond possibility that

Indra's relation to the pole-festival is not 'original', that Indra and--especially

among the higher classes -with him also the emphasis laid on victoriousness etc.

were associated with it at a comparatively recent date". As has been pointed out

above (n. 85) the position taken in the present study is different in that the

interpretation here given of the festival implies that quite different meanings could

be attached to the festival according to the social situation. The emphasis which

certain texts lay on victoriousness is easily explainable from the fact that the pole

was erected by the king and that his personal interest naturally centred around

victory in warfare. For the rest, the interpretation of the festival from an Aryan

myth does not exclude that other, apparently parallel, forms of "tree-worship",

such as that of the medhi referred to in n. 85 may have been rooted in historically

different religious traditions, even without an explicit interpretatio Arica. Even as

far as Indra's banner festival is concerned, it seems that the authors were no

longer aware of any connection with the Vedic Creation myth. At that date the

distinction between Aryan and non-Aryan, if it can then be made at all, is only

of historical interest.

124 See Meyer III, p. 23 n. 3 and Gonda, p. 421.

125 See Meyer III, pp. 13, 46, 52 n. 2, 101, 117, 131 n. 1, 134.

126 In his commentary (1. 54) on dhvajamahah . . . mahendrasya (see n. 83) he writes:

dhvajasye 'ndrasya yatra sa dhvajamahah. dhvajamahasya

viśeṣanadvarena sambhavam darśayati"nihatāsuradānava" ityādinā, "The Dhvajamaha

(occurs) where there is a mahana (worshipping) of the dhvaja, that is, Indra. In

the words 'after having killed the Asuras and Dānavas' etc. (the author) indicates

by means of an attribute the origin (occasion) of the Banner Festival". Here

Indrasya can only be meant as an explanation of dhvajasya.

127 See IIJ. 11, p. 155, Gonda, JAOS. 87, p. 422.

138

and earth apart has risen in the beginning of the creation as the result of Indra's demiurgic act but is nevertheless not identical with the god. It is sufficient to point out that the world tree, which symbolizes the totality of the universe, is conceived as standing in the centre of the cosmos, whereas Indra is in the cosmic classification localized in the East (or South, when Agni occupies the place in the East). While this was the god's normal place in the system, the situation may have been different during the few days when the pole was erected and Indra was particularly prominent. The central point is Indra's character of a seasonal god, which has been inferred above 128 from his mythic role. As early as the Rigveda, indeed, Indra is not only said to place his sign (ketu)—which may be interpreted as referring to his world pillar—but he is also said to be himself that ketu 129. Except during that short period round about New Year's day, however, the cosmic tree has little to do with Indra but all the more with the gods of totality, such as Viṣṇu 130, with whom in modern times the holy fig tree (pīpal) in the centre of the village is sometimes identified 131. This situation may be illustrated by a reference to the jarjara, which on the Indian stage represented the cosmic tree like the kekajon or gunungan in the Javanese wayang. Although according to the Nāṭyaśāstra it is Indra who gave the jarjara to the actors as a token of his contentment (see p. 144) - a trait which might be an innovation of the author of this chapter 132 - the gods who reside in the five nodes of the bamboo pole do not comprise Indra, as they are Brahmā, Śiva, Viṣṇu, Skanda, and "the great Nāgas Śeṣa, Vāsuki and Takṣaka" 133. It has already been pointed out elsewhere 134 that Viṣṇu's position between the high gods of the upper world and the deities of the nether world 135

128 See above, pp. 10, 30, 34, 42.

129 See IIJ.11, p. 155ff. on the use of ketu from the Rigveda onwards. It is still used for Indra's dhvaja in, e.g., Varāhamihira's Bṛhatsaṁhitā 43.2.

130 See Indological Studies—William Norman Brown, p. 150, IIJ. 13, p. 283, Gonda, JAOS. 87, p. 417, Viṣṇuism and Śivaism, p. 6 and passim.

131 For the identification of the Pipal and Viṣṇu see Dubois-Beauchamp, Hindu Manners, Customs and Ceremonies, 3rd ed., p. 652f.

132 Cf. BhNŚ. 1.59 prītas tu prathamam Śakro dattavān svadhvajam śubham. In passages which deal with the banner festival it is Viṣṇu who gives the dhvaja to the gods, cf. Viṣṇudharmottarapurāṇa II.154.13 (cf. v. 17 where Indra worships it !), Varāhamihira's Bṛhatsaṁhitā 43.5 and Devīpurāṇa (Meyer III, p. 8). The difference is, of course, that in the cosmogonical strife Indra needed the help of Viṣṇu who, as a representative of the unity of the universe, guaranteed that the dichotomy of upper and nether world which Indra was about to create would nevertheless be comprised in an all-encompassing totality (see p. 102). In the Nāṭyaśāstra, however, Indra is the central god of the festival. Moreover the jarjara, although said to be the dhvaja, seems also to be a form of the vajra, which is Indra's weapon. See following note and n. 141.

133 See BhNŚ. 1.92ff. jarjare tu vinikṣiptam vajram daityanībarhaṇam, tatparvasu vinikṣiptāḥ surendrā hy amitaujasaḥ (93) śiraḥpārśvāṁśu tato Brahmā dviṭīye Saṁkaras tathā, trtīye ca sthito Viṣṇus caturthe Skanda eva ca (94) pañcame ca mahānāgāḥ Seṣavāsukitakṣakāḥ, evam vighnavināśāya sthāpitā jarjare surāḥ.

134 See, e.g., IIJ. 13, p. 283.

135 As for Skanda cf., e.g., his place in the Vāstupuruṣa, discussed by V. S. Agrawala,

139

is quite in line with his character as the connecting link between the two worlds, which is exemplified by the old Vedic simile of the arrow, the head of which is Agni, the shaft Soma, while Viṣṇu is the part by which the head is attached to the shaft 136.

The reason why Indra does not occur among the gods in the jarjara is that he does not belong to the centre. The interesting point is, however, that here a conflict between the seasonal function of the god and his usual place in the classification comes to light 137.

The festival must have been a re-enactment of the god's primeval act of lifting up the sky and separating heaven from earth by means of the world pillar (skambhā). It need not be stressed that this is a modern explanation based upon what we think we know at this moment of Vedic religion as a whole: none of the texts that have been handed down speak of the festival in these terms.

The general view of their authors may be summarized as follows: it is traditionally performed for the good of the country and its king, it should, therefore, not be omitted and be done in the right way, or else disaster may threaten the king and the country.

There is only a difference in degree with the modern villagers who erect their Maypole every spring in the centre of a Dutch village, by the side of the Roman-Catholic church, for no other reason than that it has been done so from times immemorial.

The meaning which people attach to such a ceremony may, of course, have changed several times. On the other hand, whoever

Matsyapurāṇa (Ramnagar, Varanasi, 1963), p. 347, viz. in the south-eastern corner, together with the witch Pūtanā. Skanda's localization in the cosmic centre requires a closer examination. For the place of the nāgas at the bottom of the world tree see Mhbh. V.101.2, VII.69.48 and IIJ. 8, p. 108, India Maior (Congratulatory Volume - J. Gonda), p. 151.

136 See Indological Studies - William Norman Brown, p. 145 and cf. MS. III.8.1 (92,11) Agnim śṛṅgam Somaṃ śalyam Viṣṇum kūlmalam, KS. XXV.1 (102,12), KKS. XXXVIII.4 (208,5-243,1) Agnim śṛṅgaim, Somaṃ śalyam, Viṣṇum tejanam, TS. VI.2.3.1 tā iṣumim sām askurvata 'gṇim ānikam, Somaṃ śalyam, Viṣṇum téjanam, ŚB. III.4.4.14 Agnim āṇikam, Somaṃ śalyam, Viṣṇum kūlamlam (here transferred from the iṣu to the vajra), see Schlerath, Orbis 24 [1975], p. 508), Mhbh. VIII.24.84 iṣuś cāpy abhavad Viṣṇur jvalanāḥ Soma eva ca, Agṇiṣomau jagat kṛtsnam vaiṣṇavam co 'cyate jagat (which refers to the belief that all things in the Universe are characterized by either an Agni or a Soma nature, but that the world as a whole

is at the same time identical with Viṣṇu as the god of the all-embracing totality), VII.173.58 śaraṃ kālāgnisaṃyuktam Viṣṇusomasamāyutam, 173.1457* śalyam Agnim ca vai kṛtvā puṅkhe Somam apāṃ patim, sa kṛtvā dhanur omkāram sāvitrīṃ jyām Maheśvarah (cf. VII, App. 25.13 [= VII.202.77 Bombay ed.] Viṣṇum śarottamam kṛtvā rājabhṛṅguṃ Vaiśravaṇam kṛtvā tabhāya ca. Vāruṇam śalyam Agnin,jaṃ puṅkhe Somaṃ Hutāśam ca tasye 'ṣum samakalpayan, śṛṅgam Agnir babhūva 'sya bhallaḥ Somo viśāṃ pate, kuḍmalas cā 'bhavad Viṣṇus tasminn iṣvāre tadā (the commentator Nilakaṇṭha here quotes MS. as the śruti), XIII.145.27 Viṣṇum kṛtvā śarottamam, śalyam Agnim tathā kṛtvā puṅkham Vaivasvatam Yamam.

Against Whitney's translation of AS. IV.6.5 (and PW. VII, 109) see Zimmer, Altindisches Leben, p. 300: śṛṅga is the tip (= āṇika), apāṣṭhá the barb, śalyá the shaft, and kūlmala the "neck". Similarly Gonda, Early Viṣṇuism p. 35 n. 8.

137 For the difference that may exist between a god's function and his place in the system of classification see IIJ. 13, p. 283.

140

wants to understand the meaning of a festival - thereby presupposing that

it is possible, in that particular case, to find the original meaning - must

come to the conclusion that what the villagers have to say about the

meaning of their Maypole is of little help to him. The same is true of

the, mostly post-Vedic, texts on which our information about the banner

festival is based: either one restricts oneself to their explicit information

and gives up every attempt at a real understanding of the underlying

meaning of that festival, or one bases oneself on modern knowledge of the

older religion and tries to interpret the facts accordingly. In the light of

such an interpretation, then, it must be assumed that at the moment

when Indra, standing in the centre of the Universe, separated heaven

and earth, he was himself the skambhá. For the ambiguous relation between

Indra and the world axis, cf. on the one hand RS. X.111.5 "the great

sky he has propped with a pillar, the best supporter" and, on the other,

AS. X.7.29–30, ASP. XVII.9.10, 10.1 "Skambha, I (we) clearly know thee

to be contained entire in Indra" [or: I clearly know thee: all is set in

Indra"], "Indra, I (we) clearly know thee to be contained entire in

Skambha". The last words, taken from the well-known Skambha-hymn,

can only refer to the banner festival, during which the world axis, in

the shape of the Indra pole, could be considered an impersonation of

Indra himself. It is clear, however, that the Indradhvaja, connected as it

is with the re-enactment of a specific moment of Creation, is only a

momentary aspect of the world tree. The banner festival celebrates the

short moment when Indra coincided with the cosmic centre, but since he,

as the protagonist of the Devas, stood for only one of the two moieties,

the pole which was his symbol could not be allowed to stand any longer

and was pulled down and thrown into the water at the end of the

festival. There is a parallelism with the Christmas tree in Europe, and a

contrast with the Maypole, which, at least in the Netherlands, is left

standing the whole year in the centre of the village 138. Indra's pole is

associated with the inauspicious period of strife when the old year passed

into the new year. In the new harmony after the close of that period

there is no longer a place for that tree, which has to return to its origin,

the primeval waters. Cf. AS. X.7.38,41 and the stanza VI.80.3, Paipp.

XIX.16.13, which according to Atharvaveda Pariśiṣṭa XIX.1.10 should

be addressed to Indra's banner in case a vulture or a black bird alights

on it: "In the waters is thine origin, in heaven thy home, in the middle

of the sea and upon the earth thy greatness" (Paipp. "in the sea thy soul

(ātmā) on earth thy greatness") 139. The cosmic tree is, indeed, considered

to have risen from the primeval waters and to remain rooted in the

138 I refer particularly to the Maypole of Noorbeek (prov. Limburg, The Netherlands).

139 apsu te jánma diví te sadhástham samudré antár mahimá te prthivyám. See, e.g.,

Bloomfield, SBE. 42, p. 501 and Gonda, JAOS. 87, p. 425, who explains apsu as

the heavenly waters, the mythical prototype of the Indradhvaja having been in

heaven (?).

141

subterranean ocean, while its top reaches into heaven, illustrations of

which can be found in India and in Java 140.

The jarjara will be discussed in greater detail in a following section.

Here it may be stated that, although Bharata's Nāṭyaśāstra uses the term

dhvaja in connection with Indra's victory over the Asuras and Dānavas

and his festival (dhvajamaha), it is apparently necessary to distinguish

between the dhvaja of the Indra-festival and the jarjara on the stage.

Only the former could impersonate Indra, whereas the jarjara could not.

Although it is said to be the dhvaja, it actually has the function of his

vajra, as it is the weapon with which Indra destroys (jarjarikaroti,

BhNŚ. 1.70,72) the demons. It is this double function of the jarjara that

justifies its being denoted by a distinct term. This is particularly clear

when Bharata's account is compared with the fragment of Garga's work

quoted above (p. 132). In the latter the Asuras are depicted as fleeing,

frightened and losing consciousness at the sight of the dhvaja, whereupon

Indra kills them with his vajra 141. In the Nāṭyaśāstra, on the other hand,

it is not only said that the Vighnas, as soon as they see the jarjara

(1.74 drṣṭvaiva jarjaram) will flee, but also that Vighnas and Asuras are

torn to pieces (1.73 jarjarikrtāh) by the same weapon (cf. p. 151).

If the presence of Indra in the dhvaja at the banner festival is due to

his character of a seasonal god, his absence from the jarjara certainly

does not prove that Indra's importance was already on the wane. On the

contrary, from the fact that in the first chapter of the Nātyaśāstra he

is still the central god it may be inferred that the author of this chapter,

as far as his theological views are concerned, still stood in between the

Vedic and the post-Vedic world. This circumstance makes it possible for

modern research to reconstruct the religious background against which

the older parts of the Nāṭyaśāstra were written.

4. THE "FIRST" DRAMATIC PERFORMANCE (NŚ. 1.51-59)

From the preceding section it will be clear that the authors of the

oldest handbook on Indian dramaturgy considered the drama as having

originally been a ritual re-enactment of creation, consisting of an imitation

of Indra's victory over the Asuras 142. The tale of the first chapter is

essentially a cult legend, which explains the dramatical performance as

140 See India Major, p. 145 n. 2 (with references) and for the reconstruction of this

aspect of Indian cosmology also IIJ. 8, pp. 108, 116.

141 See the commentary on Var. Brhatsamhitā 43 (42). 7: tathā ca Gargaḥ : Asurās

tam dhvajam̀ dṛṣṭvā dhvajatejahaṁsahatah, visamjn̄ā̀s samare bhagnāḥ parābhūtāḥ

pradudruvuḥ. tān vajreṇa sahasrākṣo māse Bhādrapade 'surān, ghātayitvā sajyeṣṭhāyām

(read sa Jyeṣṭhāyām) . . . Similarly Viṣṇudharmottarāpurāṇa II.154.12 drṣṭamātreṇa

yeneha vidrāviṣyanti Dānavāḥ. See further Ch. I, notes 132 and 133 and below, p. 160.

142 This is not at variance with the fact that the theory of the Nāṭyaśāstra also

takes into account the nature of the later drama, whose goal and character are

defined by such terms and phrases as, e.g., 4.270 vinodakarāṇam or 36.11C lokasya

caritam nāṭyam. See n. 75.

a drōmenon, whose origin can be traced back to primordial time and which, for that reason, seems to have been considered a mystery 143.

The technical question as to whether the mythical first drama in Bharata's presentation was a full-fledged drama or an introductory part 144 is not relevant to our purpose. The Nāṭyaśāstra only says that it consisted of fight, tumult and mutual cutting off of members. The whole passage (NŚ. 1.51-59C), which is instructive when viewed in its context, is here given in full in translation (see p. 128): “Thus, after having well studied, together with all my sons as well as Svāti and Nārada, this drama (nāṭya) which was based on Vedas and Vedāngas, I approached with folded hands the Lord of the worlds (Brahmā) with a view to a performance: ‘We have completely learnt the drama. Tell me what I shall do’. When the Grandfather heard this word, he answered ‘An important (lit. great) time for the performance has just now (ayam) arrived: this is the time of the auspicious (śrimat) banner festival of the great Indra. On this occasion now perform this Veda which bears the name Nāṭya’. Thereupon at that banner festival, (celebrated because) the Asuras and Dānavas had been destroyed, at the celebration of great Indra's victory where the gods were assembled in joy, I first performed (kr-) the Benediction (nāndī) with blessings, which was full of variation as it consisted of words from the eight categories and was approved of by the gods. Thereafter I arranged an imitation of how the Daityas were defeated by the gods, which was full of altercation, tumult, mutual cutting off and piercing (of limbs). Then Brahmā and the (other) gods, who were pleased at the performance and glad, gave us all sorts of paraphernalia (upakarana)” 145.

143 BhNŚ. 36.10C tat sarvam nikhilene 'dam nāṭyam guhyam nidarśaya, 36.15C pratyuvāca punar vākyam guhyārthābhinayam (thus the MSS.) prati “He answered them again with respect to the dramatic presentation, whose meaning is secret” and the next lines, which read in KM3 (36.11) kathayāmi kathām guhyām yan mām prcchatha suvratah, pūrvarangavidhānāsya . . . “I will tell you the secret tale for which you are asking me, O pious men, namely about the performance of the Pūrvaranga”. [Similarly 36.15 B.]

144 Abhinavagupta ad 1.57B: tadanta iti, nāndyante parisamāptau, anukrtir iti nāṭyam, tatra ca baddhe 'ti gunanikā [“introduction to a drama”] yojitā, na tu prayoga ity etac casat, tatpūrvottaravyāghātāt [“contradiction”]. pūrvam hy uktam : evam nāṭyam idam ityādi nāndī krte 'tyantam [51-56], vaksyate ca [58b]: brahmādayah prayogaparitoṣitā iti. tasmād baddhe 'ti prasāvitā, na tu niṣpāditā. prasāvanā tāvat prayukte 'ty arthah. anye tu anukrtir iti nāṭyānukārārūpā prasāvitā 'ty āhuh, “krtā tadante 'nukrtīḥ” iti ca paṭhanti. etadupajivinena 'va cirantanā kavayo “nāndyante sūtradhāraḥ” iti pustake likhanti sma. kim prasāvitam ity āha daityā yathā iti, dīma-samavakāre-thāmṛgādinām anyatamaḥ prayogaḥ prāstāvī 'ty arthah. yady api Bharataputrais dasarūpakam abhyastam tathā 'pi na yugapat sarvaḥ prayoktum pāryata ity evam uktam . . . [For baddhā cf. 36.71-73 C.].

145 BhNŚ. 1.51C (KM2, 53B) evam nāṭyam idam buddhvā sarvaiḥ sutaiḥ saha (52) Svātināradasaṃyukto vedavedāṅakāraṇam,upasthito 'ham lokesām prayogārtham krṭāñjalīḥ (53) nāṭyasya grahaṇam prāptam brūhi kim karavāṇy aham, etat tu vacanam śrutvā pratyuvāca Pītāmahaḥ (54) mahān ayam prayogasya samayaḥ samupasthitaḥ, ayam dhvajamah śrīmān Mahendraśya pravartate (55) atre 'dānīm ayam vedo nāṭyasamjñāḥ prayujyatām, tatas tasmin dhvajamahane nihatāsuradānave (56) prahrṣṭā-marasamkirṇe mahendra-vijayotsave, nāndī krtā mayā pūrvam āśīrvacanasamyutā (57) marasamkirṇe mahendra-vijayotsave,

This, then, is Bharata’s account of the “first” performance of a drama. Some general conclusions that can be drawn from it have already been indicated above, p. 128. In a different context we shall meet with another account. For the present moment, however, the tale of the Nāṭyaśāstra has to be interrupted in order to study more closely the various presents of the gods.

5. THE PRESENTS OF THE GODS (NŚ. 1.59ff.)

As we have seen at the end of the preceding section, the gods, pleased with the performance by Bharata and his sons, gave all sorts of gifts to the actors. The Nāṭyaśāstra describes them as follows (1.59b-63): “First of all Indra, pleased, gave his glorious banner, and Brahmā the kutilaka. Likewise Varuṇa (gave) the golden pitcher (bhrṅgāra), Sūrya the parasol, Śiva success (siddhi) and Vāyu the fan. Viṣṇu (gave) the throne (siṁhāsana), Kubera the head diadem (mukuṭa). [And Sarasvatī gave the gift of being well audible and visible]146. And all the rest, Devas as well as Gandharvas, Yakṣas, Rākṣasas and Pannagas, very pleased in that gathering, the glad celestials, gave in accordance with their (?) different descent (jāti) and qualities147 to my sons, according to the latter’s various parts, the (appropriate) speech, the (various) psychic states, emotions, good looks, the right movements (?) and strength”148.

In general, the presents are clearly emblems characteristic of the various gods, e.g., the banner given by Indra and the fan given by the Wind. As for the throne which is presented to gods and kings149, it occurs in Buddhist iconography from the first century A.D. onwards and has been shown to be closely connected with the yūpa (sacrificial stake) and the cosmic centre150. It is not surprising, therefore, that Viṣṇu, who is a god of the centre, offers the throne as his gift. The only exception is, at first sight, the parasol, which is presented by the Sun god, although it gives protection against the heat of the sun (cf. the synonym ātapa-tra). Abhinavagupta was apparently also struck by this anomaly but his

aṣṭāṅgapadasaṁhyuktā vicitrā devasaṁmatā [v. 1. vedanirmitā], tadante 'nukṛtir baddhā yathā Daityāḥ surair jītāḥ (58) saṁpheṭa-vidrava-kṛtā chedyā-bhedyā-'havātmikā, tato Brahmadayo devāḥ prayogaparitoṣitāḥ (59) pradadur hṛṣṭamanasaḥ sarvopakaraṇāni naḥ . . .

146 This line is lacking in some manuscripts.

147 Read nānājātiguṇāśrayāṇ for -ān C?

148 1. 59b vṛītas tu varathamaṁ Śakro dattavān svadhyayami (v. 1. svam் dhvajam) subham (60) Brahmā kuṭilakam (v. 1. kamandalum) caiva bhrṅgāraṁ Varuṇas tathā (v. 1. naḥ subham), Sūryas chattram Śivaḥ siddhim Vāyur vyajanam eva ca (61) Viṣṇuḥ siṁhāsanaṁ caiva Kubero mukuṭaṁ tathā, [śrāvyatvaṁ prekṣanīyasya dadaṃ devi Sarasvatī], (62) śeṣe ye devagandharvā yakṣarākṣasapannagāḥ, tasmin sadasy atipṛītā (v. 1. abhipretān) nānājātiguṇāśrayān (63) aṁśāṁśair bhāṣitān (v. 1. bhāṣitam) bhāvān rasān rūpāṁ kriyābalam (v. 1. balam tathā), dattavantah prahrṣṭās te matsutebhyo divaukasaḥ.

149 Cf. 13.208 (12.167 KM) devānām nrpatīnām ca dadyāt siṁhāsanam dvijāḥ.

150 Jeannine Auboyer, Le trône et son symbolisme dans l'Inde Ancienne (Paris, 1949), pp. 34, 74ff.

144

ingenious explanation to the effect that the parasol is an imitation of

the clouds, which are a product of the sun 151 would seem astutius quam

verius. In a different context (Bhāg. Pur. XVIII.18.15) the chattra is a

present of the sky (dyauh).

As for the kutilaka, this is a bent stick. Abhinavagupta explains it as

"a curved stick used by the vidūṣaka, a staff which is the brahmins' (?)

weapon" 152. This must be based on NŚ. 13.143f., which describes the

natural gait of the vidūṣaka, who has the kutilaka in his left hand while

making the caturaka-gesture with his right one 153. The kutilaka is

apparently identical with the dandakāṣṭha, which is described, immediately

after the manufacture of the jarjara, in 23.179-182 (21.172ff. KM). Cf.

particularly verse 180: "It should be made curved in its third part [?]

and have (good) characteristics" 154. The passage does not mention how

this dandakāṣṭha was to be used but Ghosh rightly refers to Mālavikāgnimitra

IV.15.36 (29), where Nipunikā says: "I will, behind a pillar, frighten this

inferior brahmin, who is afraid of snakes, with this staff of his own, which

is crooked like a snake" 155. A similar, apparently stereotyped, expression

occurs in Ratnāvalī II.5.12, where the vidūṣaka says with reference to

151 chattram atra vītānāḥ : jaladānāṃ sūryodbhavatvāt tatpratīmaḥ.

152 vakradandaako vidūṣakopayogī, brahmāyudhatvā dandah.

153 svabhrāvajayāyāṃ vinyasyā kutilaṃ vāmake kare (144) tathā dakṣiṇahaste ca kuryāc

caturakam் tataḥ.

154 vakraṃ caiva hī tat kāryaṃ (kartavyaṃ 21.183B) tribhāge lakṣaṇānvitam. The v. 1.

tvaiva (21.172 KM) for caiva is apparently a mere misreading. As for tribhāga, it

denotes a third part of the eye, e.g. Kādambarī 165,16 (ed. Peterson) madanaśara-

śalyavedanākūnitastribhāgena nātimīlītena locanayugalena and pw.: "ein Drittel des

Auges bei Seitenblicken". It cannot mean, as far as I can see, "three bents", as

Ghosh translates it. Although it is tempting to read tribhaṅgalakṣaṇānvitam "having

three bents and (good) characteristics", none of the editions B C KM mention a

variant reading for tribhāge and it is not clear when tribhaṅga is attested for the

first time. The only occurrence that is known to me from literature is in the late

Gopālakelicandrikā (p. 46 line 1, cf. tribhaṅgin p. 124 line 32) but it is well-known

in archaeology. The word bhujaṅga(ma)kudila, which is used with reference to the

daṇḍakāṣṭha, may indeed indicate that it was curved like the Indonesian keris

(dagger), which has been explained as representing a serpent.

155 bhujaṅgabhīrūṇaṃ bamhabandhuṃ iminā bhujaṅga(ma)kudileṇa attaṇo daṇḍakaṣṭhena

tambhandaridā bhāiṣamī (bhāiṣamī). Shankar Pāṇḍurang Pandit omitted attaṇo,

see The Mālavikāgnimitra, A Sanskrit play by Kālidāsa, with the commentary of

Kāṭayavema, ed. with notes by Sh. P. P. (Bombay Sanskrit Series No. VI), 2nd ed.,

Bombay 1889. Of all his manuscripts, however, only one modern MS. omits the

word, whereas it occurs in his six manuscripts ABCDE and the Telugu manuscript G.

Whether or not Kāṭayavema read attaṇo in his text is not sufficiently clear from

his chāyā, which in the edition referred to does not contain ātmanah. In the recent

editions by R. D. Karmarkar (4th ed., Poona 1950) and Sane-Godbole-Ursekar

(2nd ed., Bombay 1959) attaṇo is also omitted. Since, however, it is the lectio

difficilior and has the support of the manuscripts, it must be accepted as the

authentic reading. If so, it must be taken, like Skt ātmanah, in the meaning "eius

ipsius" (see PW.). The alternative would be to assume that Nipunikā, while speaking

the words quoted, picks up a stick from the ground. This is very unlikely because

parallel passages show a stereotyped use of similar phrases with reference to the

staff of the vidūṣaka. See above.

145

the dandakāstha in his hand (6.12) "With this stick, which is crooked like

the heart of a deceiver I shall . . . throw you to the ground" 156 and in

Mṛcchakaṭikā I.42.15: (angrily raising his staff) . . . "Therefore I shall

smash him to pieces with this stick which is crooked like the lot of people

like us" 157. Also in the Śakuntalā there are references to the dandakāstha

as an attribute of the vidūṣaka 158. Outside the drama the dandakāstha

occurs, beside the danda, as a name for the ascetic's staff, e.g. in the story

of Uttañka in the Pausyaparvan 159. It is probable, therefore, that the

kuṭilaka or dandakāstha characterized the vidūṣaka as a brahmin. If so,

this may explain why it is Brahmā who gives the kuṭilaka.

Varuna's gift confronts us with one of the problems of Rigvedic

mythology, about the interpretation of which no general consensus has

been attained yet. Some of the technical aspects have been discussed at

greater length in a separate article, to which the reader may be referred 160.

According to Vedic cosmology there was a receptacle (pātra) of the cosmic

waters under the earth, in which the world tree was rooted. This receptacle

was identical with the pail (kóśa) or amṛta-jar(kalāśa) 161, to which the

texts often refer and which contained the elixir of the waters. Varuṇa

was imagined as residing in the day-time in this nether world, at the

bottom of the world tree. The jar is sometimes identified with the earth,

of which in a sense it forms part. Indra's demiurgic act of opening the

(primordial) hill in order to release the waters meant, therefore, the

breaking open of the hill, which was conceived as a jar. It will be shown

below that Indra's act was ritually imitated on the stage by breaking an

earthenware jar (kumbha) in which a small piece of gold had been laid

(NŚ. 3.72), apparently to indicate that this jar represented the golden

pitcher. Since gold is well-known from the brāhmaṇas as symbolic of life,

it seems a reasonable guess that the golden pitcher expresses that its

contents consisted of the waters of life. Perhaps there is a connection

with the golden pail (hiraṇyāya kóśa) in the "Unconquerable City"

156 iminā pisunajamahiakuḍilena danda(k)atṭhena . . . (bhūmīe) tumam் pāḍaissam்

(iti hantum udyatāh).

157 (sakrodham் dandakāstham udyamya) . . . tā edinā amhārisajanabhāadheakuḍilena

dandakatṭhena . . . kuṭṭaissaṁ. The use of kuḍila in the Ratnāvalī with reference

to the "crooked" stick and the deceitful heart of a cheat is more natural than

this passage, where the vidūṣaka likens the crookedness of the stick to that of his

own lot. This seems to be a later variation of a current expression. Since the words

do not occur in the Cārudatta, they may have been added by the author of our

version of the Mṛcchakaṭikā, which was probably composed after 600 A.D. Therefore,

the possibility that the author of this version here imitates the Ratnāvalī cannot

be ruled out a priori.

158 II.0.19 edd. Cappeller and Pischel iti dāṇḍakāstham avalambya tiṣṭhati (v. 1.

ethitah), VI.9.1-2 Pischel 1, 8.1-3 Kale bho vaassa, cittha dāva jāva dandaatṭhena

kandappabānamm் nāsemi (iti dandakāstham udyamya cūtānikuramm் pātayitum icchati).

159 Mahābhārata, interpolation 204* after I.3.138.

160 See India Maior, Congratulation Volume-J. Gonda (Leiden, 1973), pp. 144-155.

See also IIJ. 8 (1964), p. 107ff.

161 For these words see Hillebrandt, Vedische Mythologie I2, pp. 432-438, especially

p. 435.

146

(pûr ayodhyâ) of the gods162 but if so, the character of this divine city163

and its relation to Varuna's nether world164 remain to be elucidated. In

later times the use of the golden pitcher was widely spread, not only in

ritual. In the Tamil poem Cilappatikâram the talaikkōl, which was

obviously identical with the jarjara and which was identified with Indra's

son Jayanta (Cil. 3.119 intiraciruvan cayantan âk' ena), was bathed in

the holy water from a golden pitcher before it was carried into the theatre

(3.121-122 puṇniya naṇṇīr porkuḷatt' ēnti, maṇniya piṇṇar). In iconography

the connection between Varuṇa and the amṛta-cask or jar is but seldom

expressed. However, in the reliefs of Cave IV of Badami, dating from

the sixth century A.D., Varuṇa is depicted as "a male seated on a makara

with a mace in his left hand and a cup in his right hand"165 and elsewhere

as "seated beside the amṛta-bowl"166. In this connection the words

vāruṇaṁ kumbham of Agnipurāṇa 64.4 should be noted. The Badami

reliefs may well be based on the Vedic mythological conception but

iconography seems to offer no parallels to the reliefs167.

6. DESCRIPTION OF THE PLAYHOUSE

After Viśvakarmā had built the playhouse he went to Brahmā's sabhā

(1.80) to announce the completion of his work and to invite the World-

Father for a visit. Brahmā, along with Indra and with "all the gods

accompanied by others"168 then came to look at it. Brahmā thereupon

ordered the gods to contribute to the protection of the theatre with "parts"

(aṁśabhāga) of their essence169.

The following list of gods (1.83–95) shows that the theatre was considered

a replica of the cosmos. The italicized names also occur in chapter III,

where the consecration of a new playhouse is described (3.1ff.):

1. Candramas 'Moon': rakṣaṇe mandapasya, 'for the guarding of the

(whole) theatre'. Whereas the Sun is not mentioned at all, the most

162 Cf. Ath. Sam̉h. X.2.31 and Hillebrandt, Ved. Myth. I2, p. 326.

163 In the third heaven? Cf. AS. V.4.3 aśvatihó devasádanas tríyas(y)ām itó diví, tátrā

'mŕtasy a cákṣaṇam devấḥ kúgham avanvata "the holy fig tree which is a seat for

the Devas is in the third heaven from here: there the Devas won the kústha-(herb)

as a manifestation of the amŕta".

164 On this fundamental problem see India Maior, p. 156 and Hist. of Rel. 15,

p. 118 n. 22. The third heaven apparently represented what Brede W. Kristensen

called "the absolute life". As a world which transcended the dualism of the

phenomenal world it had some traits in common with the primordial undifferentiated

world which preceded it.

165 See R. D. Banerji, Basreliefs of Badami, Memoirs of the Archaeological Survey

of India, No. 25 (Calcutta 1928), p. 56 and Plate XXVIIb.

166 Plate XXIIIa. See also P. K. Agrawala, The Pūrṇa Kalaśa, fig. 48.

167 See A. Gopinath Rao, Elements of Hindu Iconography, vol. II-2 (1916).

168 NŚ. 1.81 tataḥ saha Mahendreṇa suraiś sarvaiś ca setaraiḥ. Abhinavagupta explains

itare as vidyādharagandharvādyāḥ.

169 1.82 aṁśabhāgair bhavadbhiḥ tu rakṣyo 'yaṁ nāṭyamaṇḍapaḥ. Abhinavagupta

explains aṁśair by yānī bhajanāny adhiṣṭhānāni taiḥ "with the sharing (possession?),

that is, the power of control".

147

important function, viz. the protection of the whole theatre, is here assigned to the Moon. For the moon as the guardian deity of the playhouse see also Raghavan, note ad Sāgaranandin 1110 (= NŚ. 5.109 jitaṁ Somena vai rājñā) and comm. on Mālatīmādhava (Lévi II, p. 24). The moon is satified by the nāndī (NŚ. 5.51, Sāgaranandin 1120). Any connection between the theatre building and the night or nether world cannot, however, be proved.

2. Lokapālas “Regents of the points of the sky” : dikṣu “in the quarters”. It should be noted that Agni, Yama and Varuṇa, all of them lokapālas, are also mentioned separately.

3. Mārutas (Maruts 3.5) “the winds”: vidikṣu “in the intermediate points of the compass”.

4. Mitra and Varuṇa: nepathyabhūmau “in the floor of the dressing room”, respectively ambare (= ?). The line 1.84 BC nepathyabhūmau Mitras tu nikṣipto Varuṇo 'mbare poses several problems. In the parallel list of Chapter III Varuṇa is omitted (3.6 Mitram Agniṁ surān Rudrān) and from the prayer to Varuṇa in 3.63-64, which characteristically differs from those to the other gods, it can be inferred that he was still dreaded as an ominous god. The dressing room was the place where the introductory part of the pūrvarañga, the nirgīta, was performed to appease the demons (5.11, 36f.). The reason was no doubt that this part of the pūrvarañga was too inauspicious for the stage, which was a consecrated place. There was, accordingly, a contrast between the stage and the nepathya, a contrast which probably must be defined in terms of sacred versus profane, rather than of the celestial world versus the nether world. In any case, hell and demons could be represented on the stage (14.6), and the performance of the Amṛtamaṁthana (see ch. I, § 18) shows that demons had been common on the stage from earliest times—but only to enact their defeat. As for the term nepathyabhūmau, the dressing room is always denoted by the words nepathya, nepathyagrha (e.g., 2.35, 62; 14.2, 33. 221 line 2) or nepathyagṛhaka (2.69, 95). So bhūmi must here have the specific meaning of “floor” (cf. 2.25). In view of the old contrast between Mitra and Varuṇa, this might contain a clue to ambare “in the sky (?)”, if this should be understood as the ceiling of the dressing room, but both the occurrence of Varuṇa in this context and the place where he is located remain puzzling 170. Abhinavagupta does not explain ambare.

5. Vahni (3.6: Agni), the Fire God: vedikārakṣaṇe “appointed to protect the raised floor of the stage”. Abhinavagupta's comment (vol. I, p. 31, line 10) does not help, except that it states that the vedikā was part of the stage: raṅgavedikā, tatra tīkṣṇo 'dhiṣṭhāte 'ty arthah. The vedikā is also mentioned in the description of the wooden ornamentation of the theatre, the so-called dārukarma or kāṣṭhavidhi (2.77 nānāgrathitavedikā) and at

170 Since no variant reading is given in B, the reading Varumeśvarah in KM seems to be corrupt. Ghosh translates “the space [within the building]”, but Raghuvaṁśa takes it as khāli ākāś bhāg mē Varuṇa. Similarly M. Chr. Byrski (see n. 1), p. 34f.

148

2.98, 99, where Ghosh renders it by "plinth". Mankad does not mention it. The translation "raised floor" is only tentative 171.

6. Sarve divaukasas "all celestial gods": bhānde "in the musical instruments (specifically the big drum)". For the worship of the instruments, and especially of the mṛdaṅga (drum) see Clifford R. Jones, JAOS. 93 (1973), p. 295 n. 29 and cf. IIJ. 16, p. 258. There can be little doubt that divaukasah, here just as elsewhere (e.g., 1.63, 71; 5.58), means the gods. Cf. also the parallel passage 3.6 Mitram Agniṁ surān Rudrān 172. It was a general belief that the sound of thunder of the big drum was caused by the celestials. The nātyamaṇḍapa should be constructed in such a way that the deep sound of the drums comes out clearly. Cf. 2.82 gambhīra-svaratā yena kutāpasya bhavisyati.

7. The four varṇas: stambheṣu "in the pillars". The four pillars represent the four social classes, cf. 2.46-49 brāhmaṇastambha, kṣatriyastambha, vaiśyastambha and śūdrastambha. The classes are here deified, just as in 3.6 Mitram Agniṁsurān Rudrān varṇān Kālam Kalim tathā 173.

8. Ādityas and Rudras: stambhāntareṣu "in the interstices between the pillars". It deserves notice that the Vasus are omitted and that Mitra-Varuṇa and Indra also occur elsewhere separately.

9. Bhūtas: dhāraṇiṣu "in the rows", cf. 2.78 supīṭhadhāraṇī "row of good seats" 174. For the Bhūtas see below sub 20.

10. Apsarasas: śālāsū "in the rooms" (?). It is not clear which rooms are meant. See below for dvāraśālās.

11. Yakṣiṇīs: sarvaveśmasu "in all apartments". It is not clear what exactly the term denotes, nor in what respect the veśman differed from the śālā. It cannot mean "in the entire house" (Ghosh) because of the plural form and because it would then be synonymous with maṇḍapa (see above sub 1).

12. Mahodadhi "Ocean" (: 3.7 Samudra): mahīprṣṭhe "on the (bare) ground)" (lit. on the surface of the earth). The ground (bhūmi) is mentioned with regard to the very first preliminaries of the building of a theatre (2.24, 25), cf. 2.27 vasumatī. "Ocean" (samudra) was from Vedic times

171 The translation "the stage" (Ghosh) is inexact; in 2.77 he does not translate the word nānāgrahitavedikam. Raghavaṁśa explains it by yah sundar vyavasthit vedikāo par ādhārit. For vedikā "(small) platform" see Mānasāra 15.54 and 58, 19.31 and 32. Less likely is the meaning "railing" (see, e.g., Pādatāditaka of Śyāmilaka, The Hague 1966, p. 272. Implausible Byrski (n. 1), p. 35.

172 Curiously, Abhinavagupta takes divaukasas in the sense "clouds": meghā mandragambhīraśabdasiddhaye (etc.), which was accepted by Ghosh, vol. I, Introd. p. liii, vol. II, p. 12, although sarva speaks against it. The celestials could give the big drum (mṛdaṅga) the sound of thunder. Was the correct reading perhaps meghamandragambhīraśabda "having the deep and low sound of clouds"?

173 Abhinavagupta explains this as follows: varṇā iti: tadadhisṭhātaro devatāviśeṣāh. Ghosh's objection that the four classes are nowhere deified (vol. II, p. 38 n. 1) is refuted by the fact that they are so in the Nāṭyaśāstra (see 3.6 and cf. n. 182).

174 dhāraṇiṣu is not found in two manuscripts. KM reads dhāraṇeṣu, whereas Ghosh translates "the railing (of seats)".

149

on the specific term for the primordial waters under the earth. From

the description of the building it appears that only the surface of the

ground was ritually purified by ploughing and removing all bones and

potsherds (2.26). Since the guardian deities do not go beyond the boundaries

of the sacred area, the surface of the earth is here assigned to the Ocean

as one of the cosmic entities.

13. Kṛtānta and Kāla: dvāraśālāniyuktau "in charge of the lodge at

the entrance" 175. There is a text-critical and interpretative problem.

Ghosh (1.87C) reads with some manuscripts dvāraśālaniyuktas tu kṛtāntaḥ

Kāla eva ca, where the last words may be taken to imply that kṛtāntaḥ

is here an epithet of Kāla. In 3.68C, however, all editions (B C KM) read

tathā Kṛtāntaḥ Kālaś ca sarvaprāṇivadheśvarau. In spite of the dual Ghosh

here translates "O Yama, the Fate . . ., the end of all actions" but since

Yama is specifically mentioned in 3.26 and 61, the author apparently refers

to Kṛtānta and Kāla as two different deities. As the variant reading for

3.68 sarvaprāṇadhaneśvaraḥ (in two MSS.) shows, there has long been

some divergence in the interpretation of this passage.

14. Nāgamukhyau (v. 1. Nāgarājau) "the two chief Nāgas" : dvārapārśve

tu "beside the door(s)", (that is, the entrance of the theatre?). Variant

readings dvārapatre tu, dvārapatreṣu "in the door-blades" (B KM Ragh.),

the latter variant being the reading which Abhinavagupta explained by

the gloss dvārabahutvāc ca bahuvacanāṁ 176. Since the Nāṭyaśāstra elsewhere

mentions three chief Nāgas (Śeṣa, Vāsuki and Takṣaka, see 1.92f., 3.78f.),

it remains doubtful which two can have been meant here. Abhinavagupta

takes them to be Ananta and Gulika (?), whereas Ghosh and Raghuvaṁśa

identify them as Ananta and Vāsuki.

15. Yamadaṁḍa (: 3.6 Kāladanḍa): dehalyāṁ "on the threshold".

16. Śūla "(Śiva's) pike": upari ("the top of the door" Ghosh).

17. Niyati and Mṛtyu: dvārapālau "door-keepers".

18. Mahendra (: 3.4 Indra): pārśve raṅgapiṭhasya "by the side of

the stage".

19. Vidyut (daityaniṣūdanī): mattavāraṇyāṁ "in the Mattavāraṇī". See

Clifford R. Jones, op. c., p. 295: "the covered veranda or balcony-like

structure with seats and backrests . . ."? In 3.69 it is the vāstudevatāḥ

who are residing in the mattavāraṇī.

20. Bhūtas, Yakṣas, Piśācas and Guhyakas: stambheṣu mattavāraṇyāḥ

"in the pillars of the mattavāraṇī". In 3.65f., however, the Yakṣas and

Guhyakas are connected with Kubera (as usual) and in 1.95 they are

located adhastād raṅgapiṭhasya rakṣaṇe. The Bhūtas were already

mentioned sub 9. For the Yakṣas and Guhyakas see also sub 24.

175 dvāraśālā B C Ragh., cf. Mrech. III.2.3, 21.2 bāhiladvālaśālā, Mudr. I.10.5

dvāraprakosṭhaśālā. KM dvāraśākhā, which is corrupt.

176 The translation "the two blades of the door" (Ghosh) is hardly correct. In

Nāgārjunakoṇḍa the entrance has two doors, one behind the other. See Tarlekar,

Studies (n. 1), plate XI.

150

21. Vajra (daityanibarhanam): jarjare “in the jarjara”. It is one of the characteristics of the jarjara that, although it stands for the Indra-tree (indradhvaja), it has at the same time the function of Indra’s weapon (praharana), the vajra. Cf. 1.69–70 grhitvā dhvajam uttamam . . . jarjarī-krtadehāṁs tān akaroj jarjareṇa saḥ : 72 aho praharaṇam divyam idam āsāditaṁ tvayā and 3.12 (jarjarapūjā) tvaṁ mahendra-praharaṇam sarva-vighnanibarhanam, while according to our passage (1.91) the vajra is one of the divine powers that is located in the jarjara (jarjare caiva nikṣiptam vajram daityanibarhanam). Cf. 3.79 nirmitas tvam mahāvīryo vajrasāro mahātanuḥ.

22. Brahmā, Śaṅkara, Viṣṇu, Skanda and the three Mahānāgas Śeṣa, Vāsuki and Takṣaka: parvasu “in the (five) joints (of the jarjara). Similarly 3.78f., where the gods mentioned are Brahmā, Hara, Janārdana, Kumāra, but only one Pannagottama is mentioned (see sub 24). The bamboo staff which is used as jarjara must be pañcāparvā caturgranthiḥ (23.174), that is, it must have five joints and four nodes. It has been pointed out above177 that in the jarjara, as the cosmic axis, all the gods of the totality are located and that Viṣṇu’s place in the middle of that axis corresponds to his function of connecting link (p. 140) between upper world and nether world. Only Skanda’s (Kumāra’s) place below Viṣṇu calls for a closer examination. If the interpretation of Viṣṇu’s middle position is correct, Skanda is on the side of the nether world but the Nāṭyaśāstra is rather contradictory in this respect. While in 3.24 he is located in the East together with Indra, Sūrya, etc. (Skandārkau, v.l. Skandah Sūryo), in 3.31 his place is in the western pillar. Such a detailed investigation cannot be undertaken here.

23. Brahmā: raṅgapīṭhasya madhye “in the centre of the stage”. From the description of the consecration of the stage in Chapter III it is apparent that the stage as a whole is considered a replica of the cosmos and that its centre, indicated by lines, is the brāhma maṇḍala. Since in the ideal picture of the cosmos the indradhvaja should stand in this centre, there is no contradiction between Brahmā being located both here and at the top of the jarjara. See also IIJ. 16, p. 253.

24. The Pātālavāsins, viz. the Yakṣas, Guhyakas and Pannagas : adhastād raṅgapīṭhasya “under the stage”. The Pātāla was situated at the lower end of the world axis, see p. 89. Since the centre of the stage, the brāhma maṇḍala, was identical with the cosmic centre, it is natural that the Pātāla was situated under the stage. In 3.79 the Pannagottama “chief of the serpents” is said to be in the lowest joint of the jarjara, which place is cosmologically identical with the one here assigned to the Pannagas and in 1.93 to the three Mahānāgas Śeṣa, Vāsuki and Takṣaka. On account of 3.7 nāgarājāmaṁ ca Vāsukim it may be guessed that the “highest Pannaga” of 3.79 is also Vāsuki. It is hard, therefore, to identify the two Nāgamukhyas,

177 See p. 139 and IIJ. 13 (1972), p. 283, 16 (1974), p. 252, India Maior, Congrat. Vol. J. Gonda (1973), p. 151.

discussed above sub 14, although there are several repetitions in this list. The Yakṣas and Guhyakas have already been mentioned above, sub 19. They are commonly associated with Kubera (see, e.g., Sørensen's Index to the Names in the Mahābhārata).

All the italicized names of the preceding list recur in Chapter III, where instead of a legendary tale an exact description of the ceremony of consecration of a newly built playhouse is given. Therefore, it is not correct to say with regard to the names just mentioned "Such deities are nowhere to be met with" (Ghosh, Transl., p. 12 n. 1). The authenticity of the list cannot be doubted. It is, however, far from complete, as is apparent from a comparison with Chapter III. Some details may strike us as curious, such as the fact that the Moon is mentioned first of all but that the Sun is omitted, that the Ādityas and Rudras are recorded but not the Vasus, Mitra and Varuṇa but not Viṣṇu (except, of course, in the jarjara). Is there a predominance of elements of the nether world? Why do the Ocean, Kṛtānta and Kāla, the Nāgas, Yama's staff, the inhabitants of the Pātāla and the Bhūtas 178 occur? There are many details which would deserve a closer examination than is possible here. If Fate (Niyati) and Death (Mrtyu) were "doorkeepers of the stage" (Ghosh), this could throw an interesting light on the character of the nepathya, but the "threshold" mentioned in close connection with the door suggests that the entrance of the theatre is meant. In that case, however, the question recurs: Why had one to pass by Death when entering the theatre? And why should the playhouse be built in the shape of a cave and was the stage on the west side so that the audience was looking in the direction that was associated with Varuṇa's nether world? 179 But then, why was it forbidden to give a performance at midnight and at noon? 180 On the other hand, during the consecration the gods got their respective places on the stage assigned in the evening

178 They occur twice, in 1.86 dhāraṇiṣu sthitā bhūtāḥ (v. 1. dhāraneṣv atha bhūtāni KM) and in 1.90 stambheṣu mattavaranyāḥ sthāpitāḥ paripālane Bhūtā Yakṣāḥ Piśācāś ca Guhyakāś ca mahābalāḥ. This raises the question as to whether two different recensions have got mixed up, or if the term bhūta denotes two different categories. The same problem recurs in 3.23 as against 3.27. The first line reads ādau niveśyo bhagavān sārdham Bhūtagaṇair Bhavah, where the last word must be the correct reading (as against the variant readings navaih, Śivah, see KM) on account of 3.9 sarvalokeśvaram Bhavam. Here, as often, the Bhūtas form the retinue of Śiva, cf. 4.11 tato Bhūtagaṇā hṛṣṭāḥ karmabhāvānukīrtanāt, Mahādevaś ca suprītaḥ Pitāmaham athā 'bravīt. In 3.27, however, nairṛtyām Rākṣasāṃś caiva sarvabhūtān niveśayet inauspicious beings are meant, just as in 3.17 dinānte dāruṇe ghore muhūrte bhūtadaivate "in the evening, in the hard and horrible hour over which the Bhūtas preside". On the other hand, in 3.30 the brahmaṛṣi-bhūta-saṅghas are located in the most divine region, the North-East. For a non-Sanskrit tradition concerning the Bhūtas cf. Dikhitar's note ad Cilappatikāram III.95–114 (Translation p. 102 n. 1, p. 115 n.): "The names of the deities of būtās worshipped are Vacciradēhan, Vacciradandan, Varuṇan and Prattakēśvaran (Śila, p. 115 n. Cf. Jivakacintāmaṇi st. 672)".

179 2.80 kāryah sailaguhākaro dvibhūmir nātyamaṇḍapah. For the stage on the west side of the theatre see IIJ.16, p. 266ff.

180 27.95 ardharātre na yuñjīta na madhyāhne tathaiva ca.

152

(3.17 niśāyām), that is, at a dangerous moment of the day, and at this same time (2.38) offerings to the lokapālas are brought during the laying of the foundations. At night, too, the gods were asked for protection and aid in the performance.

7. THE CONSECRATION OF THE THEATRE I

It has already been noted above that Chapter III also contains two lists of tutelary deities of the theatre. Although these lists on the whole agree fairly well with that of Chapter One, they show some interesting differences. In the first chapter the two central gods are Brahmā and Indra and there is an entire lack of Śiva-worship. References to a worship of Brahmā are generally and rightly taken as evidence of an early date. See, e.g., Ghosh, Text I, Introduction p. lxxvii, who, however, failed to notice the characteristic differences in this respect between the first books of the Nātyaśāstra, which discrepancies inevitably point to the conclusion of a different authorship. Just as Chapter Two, the third chapter is apparently the work of a devoted Śivaite. His list contains the names of the gods to whom the nātyācārya, after fasting for three days, is to pay homage (namas). It begins with

A) 1. Mahādevam sarvalokodbhavam Bhavam “the great god Bhava, the origin of the whole world” 181.

2. Padmayoni (=Brahmā). From the fact that this god here occupies the second place it may be inferred that the list has been reworked and originally began with the World-Father.

3. Suraguru (=Bṛhaspati).

4. Viṣṇu.

5. Indra.

6. Guha (=Kārttikeya).

He then continues with: 1. Sarasvatī 2. Lakṣmī 3. Siddhi 4. Medhā 5. Smṛti 6. Mati

B) 1. Sarasvatī 2. Lakṣmī 3. Siddhi 4. Medhā 5. Smṛti 6. Mati and ends up with the following gods:

C) 1. Soma 2. Sūrya 3. Maruts 4. Lokapālas 5. Aśvins 6. Mitra 7. Agni 8. Suras 9. Rudras 182 10. Varṇas 11. Kāla 12. Kali 13. Mṛtyu 14. Nīyati Vāsuki 18. Vajra 19. Vidyut 20. Samudra 183 21. Gandharvas and Apsarasas 22. Munis 23. Nātyakumārīs 24. Mahāgrāmanī (also 33.272!).

The names have here been enumerated in the same order in which they occur in the text. It seems clear that the list has been composed of three different lists: A) is the group of great gods who reside in the Jarjara (1.92-93), with the sole difference that Śiva here occupies the first place

181 Thus B KM Ragh. C reads sarvalokeśvaram “Lord of the whole world”.

182 3.6 Mitram Agnim surān Rudrān varṇān Kālam Kalim tathā C, surān varṇān Rudrān B, svarām varṇān Rudrān KM Ragh. This corresponds to 1.84 Mitra . . . Vahni, divaukasah, varṇāḥ . . . Ādityāś caiva Rudrāś ca. Therefore Ghosh’s translation “other gods such as Rudra” cannot be correct.

183 vajravidyutsamudrāṃś ca C. Probably one ocean is meant.

153

and that his original place has been assigned to Bṛhaspati, who is very rarely mentioned in the Nātyaśāstra 184. Similarly the Serpents have been left out, but Indra (Mahendra in the preceding list) has been inserted between Viṣṇu and Kārttikeya. Apart from traces of sectarianism and a few alterations, this group is basically not very much different from that of Chapter One.

Of group C) almost all of the first twenty items are found in the same order in Chapter One. The last ones (21-24) can easily be accounted for as later additions.

Most interesting, however, is group B). There can be no doubt that this list contains the nātyamātarah. Ghosh, in a note ad 3.44, remarks that they do not occur in the Purāṇas and (ad 3.86) that their number is seven. I doubt the correctness of the latter statement as it seems impossible to fix their number, the divergence between the various lists obviously being due to metrical considerations 185. As for the nātyakumāris in 3.8 (variant reading grāmādhidevatāḥ), they seem to be different from the Mothers but, as far as I can see, no further information about them is available.

8. THE CONSECRATION OF THE THEATRE II

A second list of gods enumerated with reference to the system of classification also occurs in Chapter Three. The gods are here mentioned in connection with the drawing of lines on the floor of the stage, which characterize it as a replica of the cosmos, and with the placing of images of the gods in their appropriate places. The whole cosmological configuration might be called a maṇḍala, although the text reserves this term for the centre, which is denoted as “Brahma’s maṇḍala” 186.

The symbolical representation of the cosmos was performed by drawing two lines, which crossed in the centre 187. In this way four main divisions

184 Cf. Ghosh, Text, Vol. I, Introduction p. lxxv, who, in addition to this passage, refers to 34.98, 35.56.

185 In 3.86b–87a siddhi is not personified, as she is in 5a (cf. also 3.63 diśantu siddhim) but other names are added: Sarasvatī Dhṛtir Medhā Hṛih Śrīr Lakṣmīr Matih Smṛtih, pāntu vo mātarah sarvāḥ siddhidaś ca bhavant u vaḥ. While here both Dhṛti and Smṛti occur, 3.5 has only Dhṛti (B KM Ragh., v. 1. Smṛti C). Smṛti is also lacking in 3.51f. devadevi mahābhāge Sarasvati haripriye, pragṛhyatām balir mātṛ mayā bhaktyā samarpitaḥ (52) Lakṣmīḥ Siddhir Matir Medhā sarvalokānamaskṛtāḥ, mantra-pūtam imam devyāḥ pratigṛhṇantu me balim. In 3.66 the name Brāhmī occurs for Sarasvatī: namo ’stu nāṭyamātrbhyo brāhmyāddhyo namo namaḥ. In 3.24, where Śraddhā is added, the same group must be meant: Sarasvatī ca Lakṣmīś ca Śraddhā Medhā ca pūrvataḥ. Without specification the group is mentioned in 3.44 (consecration of the mattavāraṇī) nāṭyasya ca tathā mātṛ Dhanadam ca sahānugaiḥ and in 3.29 uttarasyām diśi tathā Dhanadam samniveśayet, nāṭyasya mātṛś ca tathā Yakṣān atha saguhyakān (KM). The last word (v. 1. sahānugān) is at variance with 3.26, where the Guhyakas, along with the Pitaras, Piśācas and Uragas, are located in the South.

186 5.74 brāhme ’tha maṇḍale.

187 3.22 madhye ca ’tra kartavye dve rekhe tiryagūrdhvage, tayoḥ kakṣyāvibhāgena daivatāni niveśayet. Mankad’s discussion of the kakṣyāvibhāga (Ancient Indian Theatre, Vallabh, Vidyanagar 1950, p. 21f.) does not lead to a convincing result.

(kaksyā) were formed. Since the theatre was facing West, the stage facing East188, the lines were no doubt drawn in the direction West-East (along the longitudinal axis of the theatre) and North-South and thus pointed to the four quarters. The four corners of the stage accordingly represented the intermediate points of the compass (pradiśah or vidiśah 1.83). In these divisions of the stage the gods are placed as follows (3.20ff. devatānāṃ niveśanam):

Centre: Brahmā seated on a lotus (23a).

East: Bhagavān Bhavah (Śiva) with his bhūtaganas (23b), Nārāyaṇa, Mahendra, Skanda, Arka (or Sūrya, variant reading in B KM Raghuv.), the Aśvins, Śaśin; further Sarasvatī, Lakṣmī, Śraddhā, Medhā.

Of these gods Śiva, Viṣṇu and Skanda occupy the second, third and fourth joints of the bamboo jarjara, as gods of the centre. The first two are also mentioned apart, together with Brahmā, in the pūrvaraṅga, where the sūtradhāra, after turning round and saluting the gods of the four cardinal points (5.95–97), greets them: “Thereupon the sūtradhāra, with his face towards the audience, should perform the salutation of Rudra, Brahmā and Upendra” (5.98f.). On the other hand, Indra has from early times been located in the East (cf. 5.95). For the addition of Sun, Moon and the Aśvins I know no parallels. Although the female deities are no doubt the nātyamātarah (see n. 185), they are also separately mentioned as a group in the North.

South-East: Vahni, Svāhā, Viśvedevāḥ, Gandharvas, Rudras and Serpents (?)189.

South: Yama, Mitra sānuga, Pitaras, Piśācas, Serpents (uragāḥ), Guhyakas. On the location of Mitra in the South see below. It is doubtful if uragāḥ can be taken to prove that the reading sarpaganāḥ (see n. 189) is corrupt. The Guhyakas recur in 3.29, where they are located in the North.

South-West: Rākṣasas, Bhūtas (27a).

West: the samudras and Varuṇa “Yādasāṃ patih” (27b).

North-West: saptavāyus, Garuḍa with the birds (28ab).

North: Dhanada, nātyasya mātarah, Yakṣas and (v. l.) Guhyakas (29). As we have seen, the last group is also mentioned in 26b (South), and four of the nātyamātarah also occur separately in 24b (East).

North-East: Nandin, etc., Gaṇeśvaras, Brahmarṣis, Bhūtasamghas. The last group occurs as bhūtaganāḥ in 23b (East). For the problem of the bhūtas in general see n. 178.

This elaborate list differs from the short traditional group of lokapālas (1.83) as found in 5.95–97 (E.: Śakra, S.: Yama, W.: Varuṇa and N.: Dhanada). The most remarkable detail, however, is that Mitra, along with Yama, is located in the South. This combination is characteristic of the

188 See IIJ. 16, p. 266ff.

189 3.25 sarpaganāḥ B Raghuv., sarvaganāḥ KM, ca rgayaḥ C.

155

third chapter, where it occurs thrice 190, with the only exception of 3.6a,

where Mitra is mentioned conjointly with Agni (cf. Mitra, Varuṇa and

Agni in 1.84a) 191. They are always referred to in the fixed order Yamo

Mitraś ca (-as tu), Yamamitrau. For the author of this chapter they seem

to have been a kind of dual deity but none of the handbooks of epic and

classical mythology which I consulted nor Gonda's recent books on Mitra

and the dual deities 192 make mention of this group. Since in the Vedic

system of classification Mitra, as an Āditya, belongs to the West, the

association with Yama is hard to explain. The Vedic dual deity Mitrā-

Várunā had in any case long since ceased to exist. As late as the epic,

however, Mitra had retained some traces of his ancient function of

deliverer 193 – a function which, as far as I can see, has not been refuted

by Gonda 194. In the Rigvedic period Varuṇa and Yama ruled over the

realm of the dead but in later Vedic literature Varuṇa, although still

associated with the notion of death, lost the specific function of King of

the dead, which then became the function of Yama alone. The question

may be considered whether in Yama-Mitráu the former of the gods stands

for Varuṇa in his Rigvedic function, the dvandva in a way continuing

Vedic Mitrā-Várunā. In my opinion the basic contrast between these

Vedic gods was that of a binding and a delivering god. Yama, who is

armed with the noose (pāśa) that originally characterized Varuṇa, is

clearly the binding god. As for Mitra, his connection with defecation

(e.g. Mhbh. XII.301.2, XIV.42.34) cannot be explained as a later

development of a god who in classical literature does not seem to have

been much more than a reminiscence of the Vedic period. In the Veda

an idea more or less akin to this can perhaps be found in MS. III.10.6

(138,8), where it is said that by turning round the intestines (? gudá) of

the sacrificial animal one kills people by constipation (udāvartá), for gudá

is equal to the “vital airs” (prāṇáh) (III.10.4: 135,3 etc. prāṇá vai gudáḥ)

and prāṇā-'pānā́ are equated to Mitra and Varuṇa. The turning round

(cf. TS. VI.4.1.1) must imply the idea of producing the Vāruṇic aspect

of these dual entities. However that may be, the three passages in

Chapter III, isolated as they are, provide no basis for further speculations.

As for Varuṇa, it may be added that although his role in religion, as

reflected by the Nāṭyaśāstra, was seemingly reduced to that of lokapāla

of the West (3.27, 5.96) and god of the oceans 195, yet he had not, no

190 See 3.26 dakṣiṇena niveśyas tu Yamo Mitras tu sānu-gaḥ, 38 Yamamitrau

samabhyarcyau apūpair modakais tathā, pit-n piśitcā́n uragān sarpīḥkṛireṇa tarpayet,

61 Yamo Mitrás ca bhagavān lokapāljitatām (82) imáṃ me pratigṛhṇītām balim

mantrapuraskṛtam, rasātalacarebhyas tu panṇagebhyo namo na-maḥ.

191 nepathyabhu māus Mitras tu nikṣipto Varuṇo 'mbare (cf. 85, which mentions

Ādityāś caiva Rudrāś ca).

192 See The Vedic God Mitra (Leiden 1972) and my remarks in IIJ. 15, p. 232,

and cf. the same author, The dual deities in the religion of the Veda (1974).

193 See IIJ, 5, p. 51.

194 See J. Gonda (see n. 192) and my remarks in IIJ. 15, p. 227.

195 3.63 sarvāmbhasāṃ patir devo Varuṇo haṃsavāhanah. The last epithet corresponds

to the classical iconography. Cf. also his title Yādasāṃ patih (3.27), which is quite

common in the epic.

more than in the epics, lost his inauspicious character. (See below).

It can hardly be merely accidental that when during the consecration of

the play-house the bali is offered to the various gods, the formula for

offering it to Varuna is completely different from the other formulas.

Whereas in almost all other cases the formula is “accept this offering”

(pratigrhnatu me balim, pragrhyatām eṣa baliḥ, pragrhyatām baliḥ deva) a

special formula is used in addressing the serpents and Varuna (3.62-64):

“Homage to the serpents who live in the nether world (rasātala)! May

they who live on the air, now that they have been worshipped, bestow

success on the play! May the lord of all waters, God Varuna, who has

a goose for his vāhana, be propitious (pritimant), along with the oceans,

streams and rivers, now that he has been worshipped” 196.

9. THE ROLE OF THE JARJARA IN THE CONSECRATION

The details of the ceremony of the consecration leave no doubt that

the stage (rañgapīṭha) was considered a sacred space, which symbolically

represented the cosmos. It is impossible to ascertain how much of these

ceremonies and the underlying religious notions had survived up to the

time of Kālidāsa and the classical drama, but it is clear that the

consecration was based on old traditions. This is apparent from the fact

that the god who is located in the centre of the stage is Brahmā 197. Several

scholars have already surmised that in an early phase of Indian culture

the worship of Brahmā must have been more important than may be

inferred from the classical evidence 198. The data of the Nāṭyaśāstra are

the more interesting because most chapters of this text must have been

composed by one or more Sivaites, and even in the third chapter Siva is

once mentioned first, even before Brahmā 199. It must have been an old

and very strong tradition which in spite of Sivaité sectarianism of the

author(s) maintained itself in Chapters One and Three of our text.

In this connection the curious contradiction between NS. 1.19-20 and

4.5 is particularly interesting. In the former passage Brahmā is the supreme

196 rasātalacarebhyas tu pannagebhyo namo namaḥ (63) diśantu siddhim nāṭyasya

pūjitāḥ pavanāśanāḥ, sarvāmbhasaṃ patir devo Varuṇo hamsavāhanānaḥ (64) pūjitah

pritimān astu sasamudranadinadaḥ.

197 3.23 padmopaviṣṭaṃ Brahmāṇaṃ tasya madhye niveśayet, 5.74 brāhme ’tha maṇḍale,

which is followed by a spurious line raṅgapīṭhasya madhye tu svayaṃ Brahmā

pratiṣṭhitah. This is lacking in some manuscripts. It may be a later addition (based

upon 3.23 ?), dating from a time when the term brāhma maṇḍala was no longer

understood (but see n. 227 !). See also Feistel, Das Vorspiel auf dem Theater, p. 55.

198 Marie-Thérèse de Mallmann, Les enseignements iconographiques de l'Agni Purāṇa,

p. 119 and D. Seyfort Ruegg, IIJ. 8, p. 226, who refers to recent studies by Hacker.

T. Bhattacharya's book The Cult of Brahma (Patna, 1957) was inaccessible to me.

199 3.4 namaskṛtya mahādevam sarvalokeśvaram Bhavaṃ, padmanyoniṃ suraguruṃ

Viṣṇuṃ Indraṃ Guhaṃ tathā, see above p. 153. Cf. 1.1 praṇamya śirasā devaṃ

pitāmahamāheśvarau and 3.23 padmopaviṣṭaṃ Brahmāṇaṃ tasy a madhye niveśayet,

ādau niveśyo bhagavān sārdaṃ bhūtagaṇair Bhavaḥ, 3.46-47 (mantras addressed

to Brahmā, Siva, Viṣṇu), 5.99 Rudrabrahmopendra bhivandanam, 100-101 vandeta

pauruṣeṇa 'saṃ tripadena Janārdanam, napuṃsakapadena 'pi tathaivā 'mbuja-

saṃbhavam.

157

god, whose authority is not challenged and who, after creating the drama as a special genre, asks Indra to perform it with the other gods (which Indra declines). In 4.5, on the other hand, Brahmā composes the samavakāra Amṛtamanthana200 and after having respectfully paid obeisance to Śiva, orders it to be performed for that god on the Himālaya. Thereupon Brahmā asks Śiva graciously to act as an instructor in the aṅgahāras ("limb-movements, major dance figures"), cf. 4.16ff. That the ṇatarāja Śiva201 should teach the dance figures is not surprising, although their optional character (4.16) suggests the conclusion that they do not belong to the older type of performance but are a later addition to it. The main point, however, which has been overlooked by Ghosh as well as by Kane202, is the fundamentally different character of Chapter One, where (apart from the maṅgalaśloka 1.1 see n. 202) Śiva is only one among the devas, a dancer (1.45), who confers success (1.60 siddhi) and is the protector of all the characters save the three chief ones (1.96). The only "supreme deities" in this chapter are Brahmā and Indra. The circumstance that the dramatic performance is here connected with the latter's banner festival (dhvajamaha) is no satisfactory explanation for Śiva's subordinate position in this chapter. In comparison with the archaic character of the first chapter, the Śivaite tradition of the fourth chapter seems clearly to reflect a later cultural stage (cf. Śāradātanaya p. 285, 6-121). In its first four lines, it is true, no conclusive indications for this assumption can be found but nothing compels us to assign them to a different author.

As for the consecration of a new playhouse as described in ch. III, the principal objects of worship are here Brahmā and the jarjara. This conclusion, which follows from the preceding analysis, is confirmed by the ritual of the pūrvaraṅga, which is a kind of worship of the stage (raṅgapūjana)203. The identity of the jarjara and Indra's "banner" (dhvaja) is clearly stated in 1.69 and 73. It must be observed in this connection that from a purely text-critical point of view it might be argued that in the text as we have it 1.58-63 and 67-76a are interpolations. Even if this is true, there are no indications to show that the interpolations are of a much later date than the rest of the text.

It is one of the indications of the archaic character of the first chapter that here the Indra festival is still represented as a celebration of the

200 See below, p. 195.

201 Cf. 1.45 Nāṭakāpyaṁ nṛtyatah.

202 See Ghosh's remarks in his text edition I2, pp. LXXV, LXXVII, Translation I2, p. XLIX, and P. V. Kane, History of Sanskrit Poetics (11971, 31973), pp. 20 and 22, according to whom the first chapter, probably even the first five chapters are of later date, added some centuries before the fifth century A.D. That the Nāṭyaśāstra is the work of different authors is also Ghosh's opinion, cf. Translation I2, p. XXVII. As for the maṅgalaśloka (NŚ. 1.1), this is in all likelihood a later addition. Christopher Byrski's theory that Maheśvara here denotes Indra (Festschrift Sluszkiewicz, p. 57ff., and Concept of Ancient Indian Theatre, pp. 24-26) has not convinced me.

203 See "The worship of the jarjara on the stage", IIJ. 16 (1974), p. 241ff.

158

god's victory over the Asuras204, although the cosmogonial nature of this victory seems no longer to have been recognized. This is particularly instructive when it is compared with such late Vedic texts as Kauśika Sūtra 140 and Atharvaveda Pariśiṣṭa 19, which characterize the festival in a purely ritual way as "a royal rite for the benefit of the king's welfare and superiority in war"205.

It is obvious that the older meaning of a ritual can easily be obliterated or re-interpreted in later times. In Europe the tree of life that was erected about the winter solstice to promote (as we may assume) the resurrection of life and the return of sunlight has been accepted by means of an interpretatio christiana as Christmas tree and can be seen inside Christian churches. The Maypole, however, continues to belong to folklore and in a village in Limburg (The Netherlands), for instance, it is every year carried into the village and erected outside, but by the side of the Roman Catholic church, where it is left standing for the rest of the year. Folklore and religion thus remain separate domains in the life of the community.

In India poles were erected on various occasions. Although they do not bear Indra's name, their character is so much similar to that of the Indradhvaja that they may be mentioned in this connection. Temporary posts were annually erected in the month of Mārga (c. November), when agriculturists put a post adorned with flags on their fields to protect the crops206 and the cattle-breeders made a Mārgapāli for the lustration of the animals207. Non-temporary posts were, e.g., the banner on top of the

204 See 1.55 tasmin dhvajamahe nihatāsuradānave.

205 Cf. J. Gonda, "The Indra festival according to the Atharvavedins", JAOS. 87 (1967), pp. 413–429. The words quoted occur according on p. 414. See also p. 423. While there is admittedly no point in speculating about the prehistory, I cannot see any reason for doubting that the Indian banner festival has always been associated with Indra (cf. Gonda, pp. 416, 418). For the description in later texts see J. J. Meyer, Trilogie altindischer Mächte und Feste der Vegetation (Leipzig-Zürich 1937), pt III (to which references to the classical Tamil literature, such as Cilappatikāram 3.119ff. [cf. 28.98ff.], 144, Manimēkalai 1.7.8 should be added). For festivals in Western Bengal and Nepal see the references given by Meyer, III, p. 196 and Gonda, p. 413 n. 4. An eye-witness's account of the ceremony in modern Nepal has been given by A. A. Bake in a paper "Het feest van de rode en de witte Matsyendranath in Patan en Kathmandu", read in the Royal Dutch Academy in 1959 (see Jaarboek der Kon. Nederlandse Akademie van Wetenschappen 1959–1960, p. 148). As far as I know, it has never been published. Mr. John Irwin informed me about celebrations of the festival in other parts of North India and according to Dr. Michael Witzel the Indrajātra in Nepal (in September) is the usual time for dramatic performances (letter 21.1.1977).

206 See the data in the Kṛṣiparāśara about the medhi vaijayantīsamāyukta, discussed in IIJ. 11, p. 214f. The word vaijayantī may be a direct reference to Indra, as J. J. Meyer, op. c., III p. 4 n. 1 suggests. The word for erecting is here āropayati, which is also, but rarely, used with reference to the Indra pole, cf. Agnipurāṇa 102.8. In Cilappatikāram 3.119 the staff is said to represent Indra's son Jayanta: intiraciruvan cayantan āk ena. Cf. Skt. jayanta "flag".

207 Meyer, op. c. II, pp. 160–174.

159

chariot of a warrior, which was to guarantee his life and victoriousness 208,

and the “pinnacle” erected on stūpas 209 and theatre buildings 210.

As for the Nāṭyaśāstra the dhvaja itself, although associated with

Indra, had lost its character of world tree and had been attributed, as

Indra's weapon, the function of the Vedic vajra. Indra, indeed, uses the

dhvaja to dispel his enemies and on the other hand the vajra is said to

reside in the jarjara 211. Therefore the main function of the jarjara is

said to be the dispelling and destroying of the “obstacles” which might

disturb the performance (cf. 5.83 jarjaro vighnajarjarah). It is addressed

as “weapon of the great Indra” 212, which is a reminiscence of the mythical

tale of the first chapter, where Indra is said to have driven away the

“obstacles” (1.69) and the Asuras and Dānavas, and to have given his

dhvaja as a present to the actors to protect them (1.59). But in the same

verse 3.12 it is also said to have been created by all the gods 213 and to

be worshipped as a sacred object in its own right 214.

It follows that the pole could be erected with various intentions. The

aim of the rite was different according as to whether it was performed

on the stage, in the fields or in the town, when the main concern was

the king's welfare and victoriousness. The relative chronology of our

sources, therefore, is of minor importance in this respect and it would

be wrong to think that the “original” meaning of the rite is best preserved

in the earliest texts. When it was performed with a view to the welfare

and prosperity of the whole country, it was natural that the king, in

whose person they were concentrated, was mentioned first of all, and

success of the king almost automatically meant victoriousness. This is

true, not only of the two Atharvavedic texts, the importance of whose

testimonies can easily be overrated unless they are carefully interpreted

in their particular contexts, but also of the Nāṭyaśāstra. The prayer in

NŚ. 3.11–13 “for the success of the performance” 215 ends with the words:

“Do thou announce victory to the king and defeat to his enemies, welfare

of cows and brāhmins and success (vivardhana) of the drama” 216. The

208 jayāspada. See IIJ. 11, p. 154ff. This banner was called dhvajayaṣṭi, ketu or

ketana in the epic.

209 The stūpayaṣṭi, see Avadānaśataka, p. 387, 10-11.

210 For this stūpi (stūpi), Tamil tāliṅka-k-kuṭam see Clifford R. Jones, JAOS. 93

(1973), p. 294. For the stūpikila see Mānasāra 18.169ff.

211 1.91 jarjare caiva nikṣiptaṁ vajraṁ daityanibarhaṇam and 3.78 vajrasāraḥ

(see n. 221).

212 Cf. 3.12 mahendra praharaṇam. For Mahendra see Hillebrandt, Vedische

Mythologie II 2, p. 190f.

213 See 3.12 nirmitaṁ sarvadevaiś ca sarvavighnanibarhaṇam and cf. 3.77 vighna-

jarjaraṇārtham tu jarjaraṁ cā 'bhimantrayet, vighnānāṁ tathāṃnārtham hi devair

brahmapurogamaiḥ (78) nirmitas tvaṁ mahāviryo vajrasāro mahātanuḥ and 3.80

nityaṁ sarve hi pāntu tvāṁ surās tvāṁ ca śivo bhava.

214 pūjā, see 3.11 and 73–77 and 5.102 pūjanaṁ jarjarasya.

215 3.11 nāṭyaprasiddhaye.

216 nrpasya vijayaṁ saṃsa ripuṇāṁ ca parājayam, gobrāhmaṇaśivaṁ caiva nāṭyasya

ca vivardhanam.

160

military aspect, which is expressed in the word jaya "victory", and which also accounts for the function of banners on war-chariots, is only a special facet of the more comprehensive notion of "good fortune, success". It occurs in conjunction with "victory" in, e.g., 3.80f. Here the jarjara is addressed as follows: "do thou give victory and success to our king" 217. Protection of the crops 218 (for which J. J. Meyer's description in terms of a fertility cult is not fully adequate, unless it is seen in the light of the cosmogonical prototype) is yet another aspect of the same function, which is ultimately founded on the role of the world pillar in the cosmogony.

At the same time, however, we are here faced with the problem which has already been touched upon above (see p. 42, n. 135). Indra's erecting of the cosmic pillar has its particular place in the cosmogony, and in the few days at the beginning of the new year (Mbh. I.57.18 gate samvatsare) Indra, as a seasonal god, repeats his act and is temporarily considered identical with his pole. During the rest of the year, however, the gods of the totality are considered to be connected with it and residing in it. The same is true of the jarjara. Although it is Indra's weapon, the gods who reside in its five joints are, from top to bottom, Brahmā, Śiva, Viṣṇu, Skanda and the three Mahānāgas, viz. Śeṣa, Vāsuki and Takṣaka 219.

It has already been pointed out above 220 that Indra's absence from the jarjara is due to his character of a seasonal god. His relation to, and identity with, the cosmic tree was restricted to the few days when the renovation of the world was celebrated by the erection of the Indradhvaja. After that period Indra reassumed his normal function of lokapāla of the East. The statement of the Nātyaśāstra that Indra's vajra is present in the jarjara 221 concerns the general character of the jarjara as the protector of the stage against disturbing powers (vighnas). In contrast with the gods who reside in the various joints the vajra has no special place because it is functionally identical with the jarjara as a whole. As far as I can see, this is not a characteristic of the world axis or the indradhvaja in general but only an aspect of the jarjara on the stage.

217 3.80 naksatre 'bhijite śreṣṭhe jātās tvam ripusūdanah (81) jayam cā 'bhyudayaṁ caiva pārthivāsya prayaccha nah. For the importance of Abhijit see Hillebrandt, Romanische Forschungen 5 (1889–90), p. 301.

218 For the medhi see above n. 206.

219 See 1.92f. śirahparvā 'sthito Brahmā dṛitiṣu Śankaras tathā, trāye ca asthito Viṣṇuḥ caturthe Skanda eva ca (93) pañcame ca mahānāgāḥ Seṣavāsukītakṣakāḥ, 3.78f. śiras te rakṣatu Brahmā sarvadevagaṇaiḥ saha (79) dvitīyaṁ ca Harah parva trtīyaṁ ca Janārdanaḥ, caturthaṁ ca Kumāraś ca pañcamaṁ pannagottamāḥ (-ottamāḥ C.). Cf. the similar cosmic symbolism with regard to the sacrificial stake in TS. VI.3.4.7, ŚB. IV.4.13 (see IIJ, 16, p. 248 n. 43 and p. 259 n. 80).

220 See p. 140 and cf. IIJ, 16 (1974), p. 252.

221 See 1.91 jarjare caiva nikṣiptam vajram daityanibarhaṇam, tatparvasu vinikṣiptāḥ surendrā hy amitaujasaḥ. Cf. 3.78 nirmitas tvam mahāvīryo vajrasāro mahātanuḥ, which is, however, ambiguous: either there is an intentional double entendre or it may simply mean "hard as adamant" (Ghosh).

10. THE CONCLUSION OF THE CONSECRATION: COSMOGONICAL TRAITS

The archaic character of the role of the jarjara on the stage is particularly apparent in connection with the ceremonies which conclude the consecration of the playhouse as described in the third chapter.

After the consecration of the jarjara (NŚ. 3.73-81) ghee is poured into the sacrificial fire (homa) and then the king, the female dancers and the musical instruments (ātodya) are purified (mārj-) or illuminated (abhi-dyotayati) with torches, lit (in the sacrificial fire) in order to enhance their brilliance (dipti). They are then sprinkled with water that has been purified with mantras "for the happiness and well-being (bhūti) of the king" and a benediction is uttered for the success of the performances (3.81-87).

From this passage it may be inferred that the consecration of a new theatre took place immediately before the first performance, at the moment when the king (probably as the preksāpati) 222 and the audience were already present. The technical term adhivāsanam veśmanah 223 describes its function correctly, for adhivāsana denoted "the preliminary consecration of an image, its invocation and worship by suitable mantras, etc. before the commencement of a sacrifice" 224. This implies, however, that it was immediately followed by the pūrvarañga, which in some respects was a doubling of the consecration (it also comprised a pūjā to the jarjara, NŚ. 5.102f.), although its main purpose was different. It seems doubtful that in such cases the ritual of the pūrvarañga was performed in toto.

The end of the consecration is particularly instructive. It consists of three consecutive acts:

a) the first act: the breaking of the jar (kumbha).

The jar used in this ritual is different from the golden pitcher (bhrṅgāra), which is used in the pūrvarañga for ritual purification (5.81) and elsewhere in connection with the pole festival 225. For the purpose of this ritual an earthen jar is used, which is filled with water and adorned with a garland of leaves. After having placed it in the middle of the stage the nātyācārya puts a piece of gold in it 226, which may indicate that the earthen jar stands for a gold one. The place where the jar is put down (the raṅgamadhya) is the point of intersection of the two lines (East-West and North-South), which have been drawn on the floor of the stage. It is the centre of

222 See above, p. 123.

223 Cf. 3.2 tatra 'dhivāsayed veśma raṅgapīṭhaṁ tathaira ca, mantrapūtena tūena proksitāṅgo niśāgame. Abhinavagupta has the following comment: adhivāsayed iti : devatā mandapam adhivaśanty, āgacchanti. nātyācāryo dharmaḥ (ryakṛtam upari)tana-devatopanimantraṇam adhivāsanam veśmanah . . . rakteti sambandhaḥ. The reading nātyācāryakṛtam is apparently an emendation of the editors of the Baroda edition.

224 See Apte, The practical Sanskrit-English Dictionary, s.v.

225 See Cilappatikāram 3.121 puṇṇiya naṅṇir porkuṭatt' enti.

226 3.71 kumbhaṁ salilapūrṇaṁ ca parṇamālāpuraskṛtam (72) sthāpayed raṅgamadhye tu evarṇaṁ cātram dāpayet. The parṇamālā is also referred to in 2.54. For the (usual) meaning of puraskṛta ("accompanied by, with") see, e.g., 3.87. The translation "with a garland of leaves in its front" (Ghosh) is no doubt incorrect.

162

the stage, which is known as the brāhma mandala. According to Abhinavagupta a lotus, symbolizing Brahmā, used to be drawn at this point227. Under it another piece of gold had been buried in the ground when the theatre was built228. Ever since the brāhmaṇas, gold was considered a symbol of life229 and the notion of a golden pitcher in the nether world as the receptacle of amṛta, the life-giving substance, was well-known in the Veda230.

The ritual of breaking the jar must be performed with great care (prayatnatah), for much depends on it. Should the nātyācārya fail to break it, then the king would be menaced by danger from his enemies. On the other hand, if the director succeeds, the king can be sure of destroying his rivals231. It is clear that the author of this chapter considered the ritual a beneficial act. That it was in origin a mere analogical magic, the smashing of the jar symbolizing the crushing of the enemies, is improbable for several reasons. The jar, filled with water and adorned with leaves, is clearly a pūrṇakumbha or pūrṇaghaṭa232, also called a maṅgalakalaśa233.

227 3.22 (see n. 187) and 23 padmopaviṣṭam Brahmāṇaṃ tasya madhye nivesayet. From Abhinavagupta's commentary on the last line (ed. Baroda I3, p. 74 line 5), viz. padmopaviṣṭam iti : madhye padmaṃ kāryam iti it might be inferred that at his time the symbolism of the centre of the stage and ritual acts connected with it were still known (against the suggestion made above in n. 197). Other data, however, point to the conclusion that about 1000 A.D. the whole pūrvarāṅga was skipped or performed, as far as the nāndī was concerned, behind the scenes. Cf. also Raṅganātha (1665 A.D.) ad Kālid. Vikram. I.1.1 alam ativistāreṇa : pratyāhārādi-nikhilamūrvarāṅgāṅgakarane hi kathārambhoktathitāṃ janan jādvatā syāt, atas tiṣṭhatv ayam etatprapañca 'iti bhāvah. It is possible, however, that the consecration of a new playhouse was still performed in the absence of the audience.

228 Cf. 2.73 ratnāni cā 'tra deyāni pūrvam vajram (diamond) vicakṣanairh (74) vaidūryam dakṣiṇe caiva sphatikam paścime tathā, pravālam uttare caiva madhye tu kanakamb bhavet.

229 Cf. ŚB. V.2.1.20, 3.5.15, 4.1.14, VI.7.1.2; X.4.1.6, XII.8.1.22.

230 See Hillebrandt, Vedische Mythologie, I2, pp. 326, 437, Lüders, Varuṇa, p. 379, Kuiper, India Maior (Fs. Gonda), p. 148f.

231 Cf. 3.88 bhī(n)dyāt kumbhaṃ tataś caiva nātyācāryaḥ prayatnatah, abhinne tu bhavet kumbhe svāmino satruto bhayam (89) bhinne caiva tu vijñeyaḥ svāminah satrusamkṣayah. In this connection it is interesting to note that in the Rigveda Indra is said to have broken open the hill like a new (=unfired? J. J. Meyer, Trilogie III, p. 139) jar. Cf. X.89.7 bibhēda girim nāvaṃ in nā kumbhám, á gā Índro akṛṇuta svayúgbhih).

232 For the pūrṇakumbha or pūrṇaghaṭa, pūrṇakalasa see, e.g., A. K. Coomaraswamy, Yakṣas II (1931), p. 61f. (cf. ch. I, n. 327), Jeannine Auboyer, Le trône et son symbolisme dans l'Inde Ancienne (1957), pp. 84, 100f., Marie-Thérèse de Mallmann, Les enseignements iconographiques de l'Agnipurāṇa, p. 242f., S. Al-George et A. Roṣu, "Pūrṇaghaṭa et le symbolisme du vase dans l'Inde", Arts Asiatiques 4 (1957), pp. 243-254, F. D. K. Bosch, The Golden Germ (1960), pp. 110-113, P. K. Agrawala, Pūrṇa Kalaśa or the Vase of Plenty (Varanasi, no date, 62 pp.). It is sometimes counted among the aṣṭamaṅgalānī, see von Hinüber, ZDMG. 1974, Supplement II, p. 361. Two of them were placed on the sides of the door, Kāś. 68,4 pūrṇakalaśādhṛṣṭadvārapakṣake, 96,14 dvārāvasthitasitapūrṇakalasaṃ, (Utt.) 281,4 ubhayapārśvasthāpitotpallavamukhapūrṇahemakalaśam.

233 Mṛch. IV.27.20 cūda-pallava-lalāma-phaṭiha-maṅgalakalasābhirām-ohaa-pāssassa

163

Therefore it is placed in the very centre of the stage. Mythologically this was the place where the golden amrta-jar was hidden in the primordial hill, at the root of the world axis 234. The piece of gold which, as we have seen, had been buried beneath this centre, must have symbolized this world of potential life. According to Rigvedic mythology there is under the earth (represented by the primordial hill) and ocean (samudrá) of cosmic waters 235. Indra's demiurgic act consisted in making a hole (khá) in this hill so that the subterranean waters could stream down from the top of the hill in four directions 236.

That the earthen jar of the director is rightly equated to the cosmic jar or bucket under the earth is apparent from the fact that the director must crush the jar and that the prosperity of the king (and implicitly of his kingdom) is believed to depend on the water flowing from the jar. The nātyācārya thus repeats Indra's cosmogonic act. After the stage has been ritually converted into an image of the sacred world, its religious meaning is emphasized by a re-enactment of the "second creation" 237.

A golden pitcher (bhrṅgāra) is also one of the main paraphernalia (upakaraṇa) used by the actors in the ritual ceremonies which precede the dramatic performance. The fact that this pitcher is said to have been the present given by Varuṇa to Bharata's sons after their legendary first performance of a play supports our thesis that it symbolizes the subterranean jar 238. In the pūrvaraṅga it is used to pour out water on the director immediately before he is to erect the bamboo staff. Although the Nātyaśāstra describes this rite in terms of ablution and sipping of water 239, it is possible that originally here, too, the act of pouring out was the most important aspect (see p. 168).

Mythologically the opening of the hill was immediately followed by the rising of the world axis. It is not surprising, when viewed in this mythological context, that the two main paraphernalia of the actors were the golden pitcher and the bamboo staff. The more remarkable, therefore, is the fact that the director, after breaking the jar, does not erect the bamboo staff in the centre of the stage. The jarjara is only mentioned twice in the third chapter, first when, as part of the "worship of the deities on the stage", a pūjā is offered to the staff (3.11-13) and secondly when it is consecrated for its use on the stage (3.73-81). It has not,

... vasantaseñā-bhavana-dvārassa, Kāś. 68.7 sirobhāgavinyastadhavalanidrāmaṅgala-kalaśam, 178.13 saptalokanidrāmaṅgalakalaśe, Gītagoṿ. 12.18 (II.1) with comm.: maṅgalarthakālaśo hi payaḥpurṇo bhavati santatamrddhiphala-dair upacitaś ca.

234 For the amrta-jar at the root of the world tree see, e.g., E. B. Havell, A Handbook of Indian Art (1920-27), p. 43, F. D. K. Bosch, The Golden Germ (1960), pp. 110–113, 156f. (and the plates), and cf. above, ch. I, n. 327.

235 samudrá = āpah, see e.g., Lüders, Varuṇa p. 284ff., B. H. Kapadia, Purāṇam 4 (1962), pp. 146-153. Cf. India Maior p. 148 (with further references).

237 For the "second creation" see History of Religions 10 (1970), p. 104ff., 117, Renou, EVP. 16, p. 169.

238 For Varuṇa and the kumbha see above, p. 147.

239 śauca and ācamana (5.81).

164

however, a fixed place assigned to it in the “installation of the gods”

(3.20–32). As will be shown in the next section, the staff was only temporarily erected in the pūrvarañga and for that purpose it had to be consecrated as a sacred object during the consecration of the theatre.

b) the second act: the illumination of the stage (3.89–91).

The director now takes fearlessly 240 a burning lamp to illuminate the stage 241. With a roaring sound and snapping of the fingers he places the lamp with much noise in the centre of the stage 242. In this description three points are significant. First, the illumination takes place immediately after the breaking of the jar; second, the lamp is placed in the centre, and third, the illumination is accompanied by much ostentatious tumult (the technical term for which is uddhava). In the Ṛgveda the winning of Agni or, more specially, of the sun (svàrṣāti) 243 is a cosmogonical event which is entirely parallel to the winning of water. When viewed in the light of the preceding act, this rite must obviously be interpreted as an imitation of the first rising of the sun from the world mountain, which is the cosmic centre. In this respect, too, we are accordingly dealing with the enactment of an aspect of the “second creation”. As for the roaring sound, however, no unambiguous explanation can be proposed. On the one hand it is reminiscent of the shouts that seem to have accompanied the “return” of the sun after the winter solstice 244. On the other hand it reminds us of the legendary imitation on the stage of the primordial fight of gods and Daityas, which is said to have consisted of “altercation and tumult” 245.

c) the third act: the fight.

“Then one performs on the stage fights which are accompanied by the sound of conch-shells and drums, big drums (mṛdaṅga) and small ones (paṇava) and all kinds of instruments of percussion (ātodya). In these fights wounds which are struck, incised and gaping, with blood streaming from them . . . will porténd success” 246. The words bhinna and chinnā in

240 3.89 nāṭyācāryo 'py apetabhīḥ C (with the manuscripts pha ma ta). The variant reading prayatnataḥ (B KM Raghuv.) is a lectio facilior, which may have crept in from 88a nāṭyācāryaḥ prayatnataḥ.

241 raṅga is not “auditorium” (Ghosh), since only the stage as a sacred space can be meant.

242 3.89b bhinné kumhé tataś caiva nāṭyācāryo 'py apetabhīḥ (90) pragṛhya dīpikāṁ dīptām sarvam raṅgam pradīpayet, kṣveditaiḥ sphoṭitaiś caiva valgitaiś ca pradhāvitaiḥ (v. 1, vipradhāvitaiḥ) (91) rañgamadhyé tu tāṁ dīptāṁ saṃkṣdbdāṁ saṁprapojayet.

243 For svàrṣāti see the comments in IIJ. 4, p. 220, 5, p. 173.

244 Cf. Chānd. Up. III.19.3 ghoṣā ulūlavo and see IIJ. 4, pp. 235, 272.

245 1.57 tadante 'nṛktir baddhā yathā Daityāḥ surair jitāḥ (58) saṁpheṭavidravakṛtā chedyabhedyāhavātmikā, cf. 14.56C (13.54 KM), and 35.63 (26.28B, 35.20+, p. 654 line 1 KM) yat tv āviddhāṅgahārāṁ tu (-antam) chedyabhedyāhavāmakam (sattvā-viddha-C) and 27.13B C yac chedyabhedyāhavāmakam (n. 328).

246 3.91 saṅkhadundubhinirghoṣair mṛdaṅgapaṇavais tathā (92) sarvātodyaiḥ prana-ditaiḥ raṅge yuddhaṁ kārayet, tatra bhinnāṁ ca chinnāṁ ca dāritāṁ ca saṅsitam (93) kṣatam pradīptam āyastam nimittaṁ siddhilakṣaṇam. B (3.92f.) has the same text but KM reads raṅgayuddhaṁni and has kṛttam for kṣatam.

combination with kṣvedita “roaring sound” of verse 90 exactly echo the

word chedyabhedyāhava quoted in n. 245, where āhava means “challenging”.

This terminological parallelism is an indication that the fight on the stage,

introduced by the illumination, is an “imitation”247 of the cosmogonial

fight between Devas and Asuras. Thus the consecration of the stage

ended with three main aspects of the cosmogony. Only the jarjara,

although ritually consecrated, was not actually erected. As we shall see

below, this formed part of the ceremonies of the pūrvarañga.

Summing up it can be said that the end of the consecration was a drama

in a nut-shell, for the legendary first dramatic performance is said to have

also been a representation of the cosmogony (see p. 143). That there is

also some kind of parallelism with the pūrvarañga is not surprising. The

two ceremonies, however, are only partly parallel for, whereas the former

aims at transforming the stage into a sacred place, the central meaning

of the pūrvarañga lies in the ritual of the jarjara. There is a clear contrast,

accordingly, between the consecration, which comprises a representation

of the release of the waters, the birth of Agni and the fight between Devas

and Asuras, and the pūrvarañga, which, although not entirely lacking the

element of strife, concentrates upon the unity of the ordered world as

personified by Brahmā. It is he, not Indra, who erects the jarjara and

it is under his patronage that the verbal contest between the protagonists

of the cosmic moieties takes place. This shows that the two rituals were

basically different in character.

11. THE JARJARAPRAYOGA IN THE PŪRVARAN̄GA

The “performance with the jarjara” is discussed in detail in ch. V,

which deals with the pūrvarañga. The interpretation of the text, howerer,

presents some difficulties and consequently opinions differ. Since these

problems have been treated in detail elsewhere 248, this section will only

briefly summarize those conclusions that are of interest in the context

of this study.

The pūrvarañga is a religious act of high importance, which according

to the Nātyaśāstra is equal to a sacrifice: 1.126= 3.96. Cf. 3.93, 5.112 and

5.57 “The performance of the pūrvarañga is a religious (meritorious) act,

which confers fame and long life and serves to please Daityas and Dānavas

and all the gods”. After playing music behind the curtain to appease the

Daityas and Dānavas, the musicians enter the stage and make some music

in praise of the gods.

The director (sūtradhāra) then enters accompanied by his two assistants,

who carry in their hands the two paraphernalia (upakaraṇa), viz. the

jarjara and the golden pitcher. As far as I can see, the following ritual

cannot be understood unless the two assistants are taken to stand for

247 1.57 anukṛti (see n. 245). In this way anukṛti and anukarana (1.111, 121)

became terms for the drama in general. Cf. Abhinavagupta (B I2, p. 26 line 7)

anukṛtir iti nāṭyam and Ghosh, Translation I2, Introduction p. XXIX.

248 See n. 203.

166

the divine world. If so, however, the two objects which they are carrying characterize them as Indra and Varuṇa, for we have seen above that the upakaranas are the legendary presents of these two gods. In that case the fact that these gods were chosen to represent the world of the gods reflects Vedic mythological thought since only in the Veda Indra and Varuṇa stood for the two cosmic moieties and thus represented the whole ordered world. Apart from the internal evidence of the ritual itself, an argument in favour of this interpretation is the fact that at the end of the pūrvaraṅga one of the assistants becomes the vidūṣaka, whose traditional make-up, as described in the Nāṭyaśāstra is remarkably similar to the outward appearance of the Vedic jumbaká, the human representative of Varuṇa in the Vedic ritual. As we shall see below, the vidūṣaka and the nāyaka were the only male characters who had a special divine protector, the nāyaka being Indra's protégé.

The sūtradhāra, who has flowers in his hands, goes between the two assistants. Since this fact is expressly stated in our text (5.70), it must have had a special meaning. This can only be understood if the two assistants actually represent (as was assumed above) the upper world and the nether world. If so, the sūtradhāra's position between them characterizes him as a god of totality (see p. 176f.). It must be stressed in this connection that the pūrvaraṅga was a ritual and for that reason considered equal to a sacrifice. Hence the special meaning referred to above must have been a ritual meaning, and the sūtradhāra must consequently have impersonated god Brahmā. It then follows from the well-known symbolical meanings of "right" and "left" for "higher" and "lower" that (although the text does not say so) the assistant with the jarjara must have gone on his right and the bhrṅgāradhāra on his left.

Although the sūtradhāra had entered the stage "with the wish to worship Brahmā" (5.71 Brahmaṇo yajanecchayā), it is characteristic that the word pūjā is nowhere used in this context: he throws his handful of flowers as a donation in the centre, which is dedicated to Brahmā.249 and then performs three salutations250. The normal "salutation" (abhivādana) consists of rising from one's seat, clasping the feet of the superior and mentioning the name of the person addressed and one's own.

249 5.74 puṣpāñjalyāpapargas ca kāryo brāhme 'tha maṇḍale. According to Sāgaranandin (c. 1200-1250) this ceremony is meant to avert evil: raṅgeṣu puṣpaprakarah kriyah svastyayanam bhavet (Nāṭyakaalakṣaṇaratnakośa 2185-263). A similar definition is sometimes given of the pūrvaraṅga in general, e.g. vighnapraśāntaye (see n. 254), Mālatīm. I.5.2 suvr̥hitāni raṅgamaṅgalāni. Such definitions, which refer to the aim of the pūrvaraṅga, are not contradicted by those which describe its formal aspect as a devārcanavidhi (Sāgaranandin 2159-250). Cf. NŚ. 5.57 sarvadaivatapūjanam and below, n. 255, where the benediction of the world is described as devastutisamāśrayā "consisting in a praise of the gods". In the account given above the various dance steps are left out of consideration because their meaning is still, as far as I know, obscure. Each of them, however, had no doubt its own specific ritual meaning.

250 5.75 abhivandyah Pitāmahah, abhivādanāni kāryāṇi triṇi hastena bhūtaie (76) . . . vandanābhinayāntakaih.

text only specifies that, instead of clasping the feet (which is impossible),

the sūtradhāra should, while kneeling in the centre, touch the ground

thrice. The number three symbolizes the idea of totality associated

with the centre.

The sūtradhāra then rises, leaves the brāhma maṇḍala in which he had

made his salutations, makes a pradakṣiṇa around it and calls his assistant.

Actually, as the situation indicates, both assistants must have been meant.

The one who carries the golden pitcher (and who has been called according

to the text) pours water from the pitcher on him. The text interprets this

as an ablution and rinsing of the mouth. There is, however, some reason

to doubt if ritual purification was actually the original meaning of this

act, the more so as the sūtradhāra was already consecrated (dikṣita) at

the moment when he entered the stage. As has been pointed out above

(p. 164), in the consecration of the playhouse the water streaming from

the broken jar and portending happiness and prosperity for the king and

his kingdom can be interpreted as a reiteration of the creation of the

ordered world. Likewise the water poured out from the pitcher, symbol

of Varuṇa's nether world, can, in view of the unmistakable cosmogonical

meaning of the following ceremony, originally have had a mythological

meaning different from the ritual interpretation given by the Nāṭyaśāstra.

Immediately after his purification the sūtradhāra seizes the jarjara,

apparently from the hand of the other assistant. Muttering mantras the

director performs particular steps. This part of the pūrvaraṅga ends with

his taking five steps in the direction of the orchestra. Although the text

does not specify how this "seizing of the jarjara" (jarjaragrahaṇa) was

to be executed, it must have comprised the erecting of it. This appears

from the fact that this first part of the pūrvaraṅga to be performed on

the stage is called Utthāpana, which is the traditional term for erecting

the Indra-pole, and further from the fact that the director later lowers

the jarjara. At this moment he apparently erects it in his (right?) hand,

which explains why preferably a bamboo staff was used for this purpose

(23.172ff.)251.

When entering the stage between his two assistants "with the wish to

worship Brahmā", the sūtradhāra could be considered to be himself the

god of the totality. Now that, after "greeting" god Brahmā, he stands

in or near the maṇḍala in the centre of the stage, he must have impersonated

Brahmā and the totality of the cosmos. He then turns round and "salutes"

the four quarters with their lokapālas (1.83), viz. the east and Indra,

the south and Yama, the west and Varuṇa, and the north with Dhanada

(Kubera). An example of these salutations from a much later time is

given by Sāgarandin (c. 1200–1250), NLRK. 11:34/118: "Accept my

salutation, you gods, lords of the regions; come nearer, O illustrious ones,

and protect my stage". Thereupon he performs with three steps, described

as masculine, neuter and feminine, his "salutation" to the three gods of

251 However, also for the pole of Indra's banner festival a venuyaṣṭi was often used.

See Meyer, Trilogie III, passim.

the centre Rudra, Brahmā and Viṣṇu. Since these gods also reside in the three upper joints of the jarjara, it remains uncertain just how they are "saluted": whether in the mandala of the centre or in the bamboo staff. It has already been remarked above that nowhere in the description of the ritual the word pūjā is used. This term is reserved for the homage which is to be paid by the "fourth man" (caturthakāra). The conclusion seems warranted, therefore, that the sūtradhāra actually represents the divine world, and that he and his two attendants impersonate the gods Brahmā, Indra and Varuṇa.

In addition to the divine honours which the "fourth man" pays to the sūtradhāra, which will be discussed below, there are other indications which seem to confirm this conclusion. One of them is the curious way in which, according to the Handbook, the names of the author and the drama should be announced to the audience. Since this was a profane task, the sūtradhāra and his attendants leave the stage at the end of the pūrvarañga (5.166) and "The sthāpaka, who has the qualities and the outward appearance of the sūtradhāra, then enters" (5.167b). In the Diamond Jubilee Number of the Annals of the Bhandarkar Or. Res. Inst. (1978), p. 173ff. it has been pointed out that sthāpaka or kāvyaprastāvaka was the name of the sūtradhāra in his profane function (thus also Abhinavagupta), which explains why the sthāpaka has the same appearance as the sūtradhāra. In the classical drama, however, this form of prologue is entirely unknown and the interpretations of Dhanamjaya and other theorists, though usually taken as an accurate description of the earlier practice, must necessarily have been mere guess-work. As the pūrvarañga fell into disuse, the need was felt for a more elaborate introduction than the Prologue, which was only an interlude. Hence the conversation (saṁlāpa) was substituted for the Prologue. This talk of the sūtradhāra with his wife (naṭī) or an attendant was technically called āmukha (NŚ. 22.28) but in most dramas the name for the older prologue, sthāpanā (in the prologues of the "Bhāṣa"-plays) or prastāvanā, was maintained, and the use of prastāvanā for āmukha was sanctioned by NŚ. 22.29 (possibly an early interpolation). It will be clear that the German term "Vorspiel auf dem Theater" is neither correct for the pūrvarañga as a whole, nor for the Prarocanā and the Prastāvanā, but only for the Āmukha.

After this digression, necessary to illustrate the sūtradhāra's sacred character in the pūrvarañga, we will now return to the ritual itself. What happens next is that a "fourth man" (caturthakāra), who is not ritually consecrated, enters the stage. His only task is to offer a pūjā of flowers to the jarjara, the orchestra and the sūtradhāra himself. It should be noted that this pūjā is entirely different in character from the preceding salutations. The sūtradhāra is treated as an equal of the bamboo staff and the orchestra (or: musical instruments), that is, as a divine being. It thus confirms the view taken above, that the sūtradhāra, when entering the stage with his two attendants, represented the world of the gods, and indirectly it may support the inference that he stands for Brahmā. The pūjā accentuates the character of the salutations given to the gods of

169

the four quarters and those of the centre. It is true that they formed

part of a kind of consecration of the stage as an image of the divine world

• a consecration which was prerequisite for the following ritual act. It

must also be admitted that at the moment the sūtradhāra paid obeisance

to the gods, including those of the centre, he at best represented a lower

deity but he was not yet the highest god. However, after circumambulating

the centre and seizing the jarjara he must have become god Brahmā,

and the pūjā of the "fourth man" (apparently a representative of the

human world) confirms his divine function. The text does not say just

where he is standing at this moment but if it is not in the very middle,

the brāhma maṇḍala (which would be the only legitimate place), it must

have been quite near it.

It is now that the sūtradhāra/Brahmā pronounces his benediction

(nāndī)252. In the oldest practice, accordingly, as it is described in the

Nātyaśāstra253, the nāndī was not the benedictory verse recited at the

beginning of a dramatical performance but the most solemn part of the

pūrvarañga. Its function is to appease all obstacles254. On hearing its

sound, which is equal to that of the explanation of a Vedic mantra (36.26),

evil spirits flee (36.25). It is a benediction255 of the gods, the king, the

cows and brahmins (5.108-112)256. Since the performance by the sūtradhāra

culminates in this act, it becomes clear why such high demands are made

upon his person (36.66-74): a man who was to bless "the whole world"257

had to fulfil the highest requirements.

After the nāndī the sūtradhāra pronounces the jarjarasloka (5.117f.)

which is immediately followed by another śloka which accompanies the

inclining of the jarjara. The two assistants then retreat to the back of

the stage. Apparently the director has acted a banner festival with the

bamboo staff in his hand, which he first raises (to his forehead? 9.78)

and then inclines. As a result the following dramatic performance can

be considered to take place just as in the legendary first performance

(1.54), on the occasion of Indra's banner festival. After lowering the

jarjara, the director holds (the lower part of) his left arm in a horizontal

252 For its importance see Bādarāyaṇa apud Sāgara 1092f.

253 For the later practice see, e.g., S. Lévi, Le théâtre indien I, pp. 136, 379, II,

pp. 27, 64, Kuiper (p. 169), pp. 179–182.

254 Cf. Bādarāyaṇa vighnapraśāntaye, Sāgaranandin 1090-115 sarvavighnapraśāntaye.

See n. 249.

255 5.24 āśirvacanasamyuktā (cf. Sāgaranandin, NLRK. 1123-116. Sāh. D. 5.24,

35.96 āśirvāda(na?)yuktair madhurair vākyaś ca samanggalācāraib,

S. Lévi II, p. 24), 35.96 āśirvāda(na?)yuktair madhurair vākyaś ca samanggalācāraib,

sarvam stauti hi lokam yasmāt tasmād bhaven nāndī, Sāgaranandin, NLRK. 1095

pradhānam aṅgam pūrvarañge sūtradhāraprayojyadevastutisamāśrayā āśirvacanātmikā

dvādaśapadā aṣṭapadā vā.

256 5.108 namo 'stu sarvadevebyo dvijātilbhyah śubham tathā (109) jitam Somena vai

rājñā ārogyam gobhya eva ca, brahmo 'itaram tathaicā 'stu hatā brahmadviṣas tathā

(110) praśāstv imām mahārājah prthivīm ca sasāgarām, rājyam pravardhatām caiva

raṅgaś cā 'yam samṛdhyatām, etc. For the connection between Moon (Soma) and

nāndī see 5.51, for the Moon and the rasas see Bhāvaprakāśana, p. 197 (G.O.S.),

lines 6–9.

257 35.96 sarvam stauti hi lokam.

170

position near his navel, while the palm of his left hand, in which he seems to have held the horizontal jarjara, is turned downwards. All this can be explained as symbolic of the laying down of Indra's banner, which is now desecrated. After having “turned round” (5.101 parivartana) and recited two ślokas, the sūtradhāra joins his two assistants and returns the jarjara to one of them. Whatever the way in which the latter may have got rid of it (which is not clear from the text), it cannot be doubtful that this act meant the visarjana of the desecrated object, the technical term for which was jarjaramokṣa (27.40). At this moment, it seems, the acting was taken to have begun, and the jury of experts (prāśnika) started evaluating the acting (27.40). This moment, after the jarjaramokṣa (27.40) or after the first jarjaraśloka (5.118), was called raṅgadvāra “beginning of the acting” (5.27, 118)258 or, it seems, nāṭyadvāra. It cannot have implied, however, that now the pūrvaraṅga lost its ritual character. Passages in which this view is taken, such as 32.460

prayujya ca bahirgitaṃ pūrvaraṅgaṃ prayojayet pūrvaraṅge pravṛtte tu nāṭyadvāraṃ samaśrayet “After the bahirgīta he performs the pūrvaraṅga, and after the pūrvaraṅga the nāṭyadvāra” cannot represent a common opinion of older times. Usually the pūrvaraṅga is said to commence with the bahirgīta, the first member, in which music was played behind the scenes to appease the demons, and to end with the prarocana. If, therefore, nāṭyadvāra is actually a synonym of raṅgadvāra, the verse must represent a later practice, which would not be surprising in this appendix. (See p. 190). The alternative is that nāṭyadvāra (not attested elsewhere, as far as I know) denoted the beginning of the play proper, after the completion of the pūrvaraṅga.

However, the more of the pūrvaraṅga was skipped, the stronger became the tendency to consider the remainder as part of the play. According to Śāradātanaya's Bhāvaprakāśana, p. 288, 7–10 the sūtradhāra must announce the play, the hero, the story and the rasa in the nāndī. The first traces of this trend can be found in the definitions of the last two “members” of the pūrvaraṅga in Chapter V.

12. NĀYAKA, VIDŪṢAKA AND TRIGATA

In the preceding sections 5, 9 and 11, which dealt with the presents of the gods, the jarjara in the consecration of a new playhouse and the worship of the jarjara in the pūrvaraṅga respectively, it has been pointed out that the presents of Indra and Varuṇa, the jarjara and the golden pitcher, were particularly important. On the other hand, the ritual of the consecration and the jarjaraprayoga will, it is hoped, have made clear that the stage represented the sacred world and that the central deity in this world was Brahmā.

This long discussion of various religious aspects was necessary as an

258 5.26 yasmād abhinayas tatra prathamam hy avatāryate (27) raṅgadvāram ato jñeyam vāgaṅgābhinayātmakam. Cf. IIJ. 16, pp. 260 and 264f.

introduction to one of the basic problems of the oldest form of the Sanskrit drama, viz. its chief characters. The key to this problem can be found in chapter I, where a separate passage deals with the divine protection of the theatre (I.82–95). After a lengthy enumeration of all the gods who protect the various parts of the building this passage concludes with briefly mentioning the protectors of the leading characters: “As for the nāyaka, he is protected by Indra, and the nāyikā by Sarasvatī. The Oṅkāra protects the vidūṣaka, and Hara [Śiva] the remaining characters” 259.

There is no reason to doubt that the Nāṭyaśāstra has here preserved a very old tradition, which is in accordance with the general archaic character of the first chapter. This tradition recognizes three protagonists, who are considered distinct from the rest of the troupe. In the following discussion the nāyikā will be left out of consideration. The final section is exclusively devoted to her.

The first conclusion that can be drawn from the oldest native tradition is that it does not lend the slightest support to modern theories which explain the vidūṣaka as a buffoon who at a comparatively late date intruded from the popular into the higher drama. This raises the question as to whether the modern interpretation of the character of the vidūṣaka, which is mostly regarded as self-evident and on which these theories are based, is correct. This point will be discussed below (§ 16).

That the tradition preserved in the Nāṭyaśāstra is actually very old follows from the fact that Indra is still regarded as the protagonist of the Devas, be it only on the occasion of the banner festival. The close relation between the nāyaka and his protector has already been referred to in the preceding section, where it has been argued that of the two assistants of the sūtradhāra in the pūrvaraṅga, the one who carries the jarjara, represents Indra, the giver of that religious symbol. There is, indeed, a double close relation, between Indra and the nāyaka on the one hand, and between the god and the jarjara on the other. It would seem a reasonable guess, therefore, that it is the nāyaka who in the ritual part of the pūrvaraṅga functions as the jarjaradhara, that is, the assistant who carries the jarjara. Apart from other arguments to be discussed below, it is significant that at the end of the pūrvaraṅga one assistant comes again to the front of the stage with the make-up of the vidūṣaka. At this moment his opponent may well have been recognizable as the nāyaka, but the text does not give any particulars.

A similar close relation must have existed between Varuṇa, the golden pitcher and the assistant who carried it. In view of the unmistakable

259 1.96 nāyakam் rakṣati 'ndras tu nāyikām tu Sarasvatī, vidūṣakam atho 'ṅkārah seṣe tu prakṛtīr Harah. An interesting parallel for the importance of the protection of the actors during the performance is provided by the dalang, who performs the wayang (shadow play) in Bali. As Hooykaas remarks: “A dalaṅ is prone to all sorts of dangers during a show” (Cosmogony and Creation in Balinese Tradition, p. 117). On the parallelism between dalang and sūtradhāra see J. J. Ras, BTLV 129 (1973), p. 456 (with references).

172

similarity between the make-up of the vidūṣaka and the outward appearance of the Vedic jumbakā, who represented Varuṇa (see § 18), the assistant who at the end of the pūrvarañga comes to the front as the vidūṣaka can only have been the bhrṅgāradhāra. It might be expected, then, that the personal protector of the vidūṣaka was Varuṇa. In fact, however, the text gives a different and enigmatic name for the god who protects the vidūṣaka. It cannot be accidental that references to a god Oṅkāra are only seldom met with 260.

As a name for Viṣṇu it seems to occur only once in the Mahābhārata 261, which rules out the possibility that this god could have been meant. The same is true of Śiva 262, since he is mentioned in the same verse as the protector of all the other actors and thus is contrasted with Oṅkāra. The circumstance that Oṅkāra as the name of an individual god is only found in a few late Vedic texts may, therefore, be considered significant in the light of the archaic character of the first chapter of the Nāṭyaśāstra and its many Vedic reminiscences.

In one part of the Gopathabrāhmaṇa, which has gained a certain amount of independence under the name Praṇava-Upaniṣad 263, oṅkāra is both a term to denote the sacred syllable Oṁ 264 and, in the same passage, the name of a son of Brahmā. In I.1.23 the text relates how the gods, frightened by an attack of the Asuras on Indra’s city Vasor dhārā 265, wondered who could repel the Asuras 266. The text then runs

260 M. M. Ghosh only remarks that “Oṅkāra as a deity is very rarely to be met with” (Translation I2, p. 13 n.).

261 In the Mahāpurusastaya, Mbbh. XII.325.83. Different is Hariv. App. I. 31.3648 tvam oṅkaro vaṣṭkāras tvam yajñas tvam Pitāmahah.

262 The Petersburg Dictionary refers to Wilson, who in the Works by the late Horace Hayman Wilson vol. I (London 1862), p. 223 n. I quotes Aufrecht's Catalogue of the Bodleian, and Weber's of the Berlin manuscripts and writes, with regard to the legendary twelve liṅgas mentioned in the Nandi-Upapurāṇa (Śivapurāṇa ch. 44-61), that “Oṅkāra is said to have been in Ujjain, but it is probably the shrine of MAHĀDEO at Oṅkāra Mandatta [Māndhāttā] on the Narmadá”. For an ample description of this shrine see Central Provinces Gazetteer, 2nd ed., Captain James Forsyth, The Highlands of Central India, 2nd ed. (London 1872), p. 171f.: “Far to the west of Puchmurree, in the district of Nimár, is a rocky island in the Narbadá river called Māndhāttá, on which is situated the shrine of Śivá called Omkár—one of the oldest and most famous of all India”. Forsyth quotes from the “Narmadá khandá” (said to be part of the Skānda Purāṇa) that this may have been an old centre of worship of Kālī and Kālá-Bhairava by the Bhils before the first Rājput introduced Oṅkāra. Whether the name Oṅkāra-tīrtha (Verzeichniss der Oxforder Handschriften 68, b, 22) refers to the same place I am unable to verify, but the same name is mentioned several times in the Matsyapurāṇa, cf. 22.27 Oṅkāraṁ pitṛtīrthaṁ ca Kāverī-kapilodakam (cf. 189,6), 186.1–2 idānīṁ Narmadāyās tu māhātmyam vada sattama, yatro ’ṅkārasya māhātmyam kapilā-saṅgamasya ca, 195.1 ity ākarṇya sa rājendra Oṅkārasya ’bhivarṇanam.

263 GB I.1.16–30, cf. Deussen, Sechzig Upanishad's des Veda3, p. 858ff.

264 GB I.1.22 purastād oṅkāram prayuṅkte. Cf. NS. 34b. 178B ādāv oṅkāra ucyate.

265 For Vasordhārāṁ aindram nagaram (ed. Gaastra) read Vasor dhārā nāmai 'ndraṁ nagaram. Cf. MS. III.4.1 (45,1), KŚ. XXI.11 (51,5 etc.), TS. V.4.3.1, ŚB. IX.3.2.1 (etc.) vásor dhárām, PB. XXI.3.7.

266 Read: ka imān asurān apahaniṣyati.

173

as follows: "They saw Oṁkāra, the eldest son of Brahmā and said to him: 'May we, with thee as leader (mukha), be victorious over these Asuras'. He then said . . ." 267. This passage stresses the fact that the oṁkāra precedes all other things (pūrvam, āditalah) 268, a notion which runs parallel with that of oṁkāra as the essence of everything 269. At the end of the passage, indeed, the oṁkāra is said to be the ṛc in the ṛc (etc.), the sūtra in the sūtra, the pranava in the pranava. It is described, accordingly, as the core and essence of every solemn utterance and as such also as different from, and even transcending, the praṇava Oṁ. In this speculation the oṁkāra seems to come very close to the all-embracing notion of Brahman 270 and this is, indeed, stated expressis verbis in the Śaunaka-Upaniṣad 271. Although in its first paragraph the syllable Oṁ (praṇava)is compared and then fully identified with Indra 272, in the third paragraph this merges in an identification with Brahman: "Der Praṇava ist Indra und darum gross . . . Da erkannte der Praṇava: . . . und ich bin die Erscheinung des Brahman und nicht von ihm verschieden". In the two Upaniṣads the Praṇava/Oṁkāra occurs in the characteristic Vedic frame-work of a fight between Devas and Asuras. Hence its being equated in the Śaunaka-Upaniṣad, as the king of syllables, with the king of the gods Indra. On the other hand this work, whose central motif is the glorification of the praṇava, thus introduces the notion of power. This is at variance with the basic idea of immanence, which comes out clearly

267 ta Oṁkāraṁ Brahmaṇaḥ putraṁ jyeṣṭhaṁ dadrśus, te tam abruvan : bhavatā mukhena asurān jayeme 'ti sa ho 'vāca . . . Note the modern expression bharatā mukhena "with thee as leader" for MS. III.4.1 (45,1) etc. trayā mukhena.

268 GB. I.1.23 tasmād oṁkārah pūrvam ucyate, 25 tasmād vai tad bhadram oṁkāraṁ pūrvam ālebhe, 24 oṁkāram āditaḥ kurvanti, 26 ādita oṁkāro vrikiyate. Cf. Śaunaka Upaniṣad (Deussen, Sechzig Upaniṣad's, p. 876) "Da machte Indra den Praṇava zum beständigen Anfang der Gāyatrī". The meaning of mukha is, accordingly, different from Chānd. Up. III.10.1 tat (scil. amṛtam) Śādhya upajīvanti Brahmaṇā mukhena ("with Brahmā as their mouth") and Brh. Ār. Up. V.14.8, where Agni is the devouring mouth (mukha) of the gāyatrī-metre.

269 GB. I.1.30 ātmabhaiṣajyam ātmakaiyalyam oṁkārah.

270 In general oṁkāra and praṇava were, of course, identical. In the Ṣaḍviṁśa-brāhmaṇa (IV.5.3) Prajāpati is said to have seen as a "remedy" against the attacks of the Asuras: ṛta, satya, brahman, oṁkāra (commentary : praṇava), "and he (also) saw the gāyatrī, consisting of three pādas as the beginning (?mukha) of the brahman". Bollée translates "He saw the face (i.e. the most important part) of Brahman", taking brahman in two different meanings. In the current equation brahma gāyatri and Chānd. Up. III.12 (to which Eelsingh in his edition of the ṢB, p. 218, refers). Bollée refers to PB. XI.11.8-9, where however brahman denotes the priesthood. In the passage of the ṢB quoted above the words tripadām gāyatrīm apparently also denote the metre, but then the meaning of mukhaṇo brahmaṇo mukham by the side of brahman remains a crux. I doubt whether the gloss pradhānam of the commentator is correct. See also above, nn. 267 and 268.

271 Deussen, Sechzig Upanishad's des Veda3, p. 875ff.

272 "Dieser Praṇava ist Indra, ist alles was ist" . . . "Indra ist das Bewegte und das Unbewegte, so sagen sie, Indra aber ist der Praṇava".

174

in the recognition that the praṇava is identical with the Brahman and

in the words "denn der Gipfel der Grösse ist der Pranava. Alle Wesen

sind in ihm enthalten, und seine Stätte ist in dem Nachhall, denn in diesem

hatte er sich verborgen". The same idea is stressed in the Gopatha

Brāhmaṇa, which not only depicts the Devas as invoking the assistance

of Oṅkāra, the eldest son of Brahmā, but also describes the non-personified

oṅkāra as comprising the totality of the divine world 273.

The only later relevant passage which has been quoted as referring to

the personified Oṅkāra occurs in the Kāśikhaṇḍa. It has, however, an

incidental character.274 Two passages in the Mahābhārata, VII.173.1457 * 275

and V.106.14 276 can also be left out of consideration.

After this digression we can return to the protector of the vidūṣaka.

The Nāṭyaśāstra has fortunately preserved a precious indication as

regards the light in which the Oṅkāra should be considered. This is the

word ādau in 33.224 caturnām api vedānām ādāv oṅkāra ucyate "The

oṅkāra is said to be at the beginning of the four Vedas, too". This reminds

us of the speculations about the oṅkāra being āditah etc. in GB. I.1.26

and 28 (see n. 268). This parallelism may not be accidental and may show

that in this respect, too, the Nāṭyaśāstra is still rooted in the Vedic

religious tradition. However, whether or not the vague figure of the divine

Oṅkāra who protects the vidūṣaka is a reminiscence of the "eldest son

of Brahmā" in the Gopatha Brahmaṇa I.1.23, the problem remains that

the three other protectors, viz. Indra, Sarasvati and Śiva, are well-known

deities, whereas Oṅkāra can at best be associated with a very vague

and incidental personification, somehow related to Brahmā but not

273 In GB. I.1.25 the four morae of the oṅkāra are equated to Brahmā, Viṣṇu, Śiva

and all the gods (sarvadevatya).

274 H. H. Wilson, op. c., p. 4 n. 1 (see PW. V, col. 1225) points to Kāśíkhaṇḍa 31.

In this passage the Pranava is introduced as speaking after the four Vedas, for

which purpose he must assume a human figure (31.31) : tad udīritam ākarya

praṇavātmā saṅitanah, amūrto mūrtimān bhūtvā hasamāna uvāca tau. Praṇava

uvāca : . . . Much later than this passage from a purported part of the Skandapurāṇa

is Māyā's account of the first being, which Wilson (p. 81) cites from the Bijak

of Bhagodās, one of Kabir's disciples: "the Omkāra did not witness his beginning,

how then can I explain it . . ."

275 In Mhbh. VII.173.1457* crit. ed. there is no reason to assume a personification:

śatyam Agnim ca vai kṛtvā puṅke Somam apām patim, sa kṛtvā dhanur oṅkāram

savitṛīṁ jyām Mahesvarah, hayāṁś ca caturo vedān sarvavedamayam ratham,

Prajāpatim rathasṭhe viniyujya sa sārathim. In the Bombay edition a longer

passage is interpolated (VII.202.71-80). The oṅkāra here occurs in v. 76 kṛtvo

'mkāram pratodaṁ ca Brahmāṇaṁ caiva sārathim. Both interpolations (and

Mhbh. VIII.24.257*) are elaborations of an ancient Vedic equation, see n. 136.

In contrast with this older version, which contains the names of three gods, the

later one contains the four Vedas and the Sāvitrī-verse, which is never personified

(PW. VII, col. 964f.). There is no reason, therefore, to consider oṅkāra as a

personification.

276 Mhbh. V.106.14 oṅkārasya 'tra jāyante sūtayo daśatir daśa, with a variant reading

sṛtayo which Nilakaṇṭha explains as follows: sṛtayo : mārgāh. Oṅkāro vai sarva vāg

iti śruter, vedasya sākhāpraśākhādibheda iti arthah. Neither reading is

clear to me.

actually a god of the pantheon. Abhinavagupta refrains from giving any explanation. Since our difficulties in identifying this "god" can hardly be due to a deficient knowledge of the Hindu pantheon on our part, it is reasonable to conclude that this name has been chosen on purpose, just because it does not evoke the image of an individual god.

The reason for this remarkable choice must have been taboo. It is the object of this book to demonstrate that the vidūṣaka impersonated Varuṇa, who was too inauspicious to act (or, to be named) as the protector of a dramatic character. Was, then, Oṅkāra a different god, substituted for Varuṇa, or merely a euphemistic name, used because the real name was tabooed? As a mere guess it may be suggested that, since the oṅkāra was identified with the brahman (n. 270), and since bráhman could denote in Vedic the whole social class of brahmins, the vidūṣaka, who had to be a brahmin, was protected by an impersonation of the brahmins as a whole (see p. 222). According to Byrski (n. 1), p. 27 Oṅkāra stands for Brahmā.

The basic difficulty in studying Varuṇa is this element of taboo. An intentional reticence can seldom be proved. In his case, it is true, there are sufficient indications pointing to feelings of fear and awe with regard to him. We have some reason, therefore, to expect a restraint, dictated by taboo, in what Vedic and post-Vedic authors tell us about Varuṇa. Although it is argued again and again that the only sound and scholarly method of studying a god is to confine oneself to what the texts explicitly say, this is only fully true for the study of gods of a less complex character, such as Indra. It is seldom realized that the same principle, when applied to the study of Varuṇa, is tantamount to condemning oneself to an imperfect understanding, if not a complete misunderstanding, of the god. In this case it would be an unsound method to ignore the aspect of taboo and a possible reticence, and to proceed as if the direct data of the evidence are the only reliable basis for an interpretation of his character.

As we have seen above, in the pūrvaraṅga the sūtradhāra and his two assistants enter and leave the stage as a group. Since this is described by similar formulaic lines 277, the specification given for their entrance, viz. that the sūtradhāra goes between his assistants 278, must also be true of their exit. From the fact that the pūrvaraṅga is a ritual, it has been inferred that this position had a ritual meaning. The symbolism of the middle is too well known to need an ample discussion. In the Yajurveda Prajāpati is said to have created the Devas with his right hand (or, on his right) and the Asuras with (or, on) his left 279. In modern illustrations of the Mahābhārata Yudhiṣṭhira is represented as seated in the middle, with Bhīma and Arjuna sitting to the right of him, and Nakula and Sahadeva on his left 280. Right is always associated with the higher group

277 See 5.69 praviseyuh samam trayah and 5.142 niṣkrāmeyuh samam trayah.

278 See 5.70 madhye tu sūtradhṛk tābhyām vṛtah.

279 See MS. I.9.3 (132,16) dákṣiṇena hástena devân ásṛjata savyéna 'surān, té devā vīryàvanto 'bhavan mṛddhâ ásurāḥ, KS. IX.11 (112, 18) dakṣiṇaim hastam anu asurā̀ms te vīryàvanto 'bhavanto savyàm hastam anu devā̀n asṛjata te vīryāvanto 'bhavan.

280 See Numen 8, p. 41f.

and the divine, left with the lower order. In the ordered world the contrast between Devas and Asuras in the cosmogony has been replaced by that of upper world and nether world 281. For that reason the jarjaradhara, as he represented the world of the Devas, must have gone on the right of the sūtradhāra. The ambiguous position of Varuna has been amply discussed in Ch. I. Although he had become a Deva (be it a devá ásura, who remained associated with the nether world), the Vedic theological system leads to the supposition that at the end of every year he became again an Asura. But apart from this conclusion, which cannot be supported by direct evidence of our texts, the fact that he stood for the nether world was sufficient reason to make him, in this respect, inferior to the representative of the upper world. It may not be superfluous to warn against the misconception that Varuṇa was therefore considered morally inferior to Indra. Moral appreciations according to human standards are out of place here. In the last hundred years they have contributed much to misinterpreting and distorting the image of the Vedic pantheon. That the Vedic Varuṇa, as the guardian of the cosmic order (ṛtásya gopáh) had a majestic character does not contradict the fact that as a god of the nether world he was of a lower order than the gods of the other moiety. The assistant who carried his present and symbol, the golden pitcher, must therefore have gone on the left of the sūtradhāra.

This conclusion is of direct importance for the interpretation of the next part of the pūrvaraṅga, which follows the desecration of the jarjara, viz. the so-called Trigata (5.136-140) or Trika (?5.16). After a detailed description of the dance steps of the Mahācārī (5.128-135) the Nāṭyaśāstra describes the Trigata as follows: “When this is finished he addresses 282 his assistants while taking three steps 283. When they have come to him, those who know (the tradition well) sing a narkutaka song 284. [While this is being sung he performs (the Sūcī [Vedha] Cārī) by putting forward the left foot first and the right foot afterwards] 285. He must also perform a Trigata in the Bhāratī style 286. The vidūṣaka holds a conversation which is ekapadā(?) 287, mainly consists of incoherent phrases and evokes a smile of the sūtradhāra. [He must make critical comments 288, with abrupt

281 See History of Religions 15, p. 113.

282 The use of vyā-hṛ- seems uncommon but it can hardly have a different meaning from the one assumed here. Influence of ā-hṛ-“herbeiholen” (Feistel) is improbable.

283 tripadyā: cf. 5.99a, where the reading tripadair C is against the manuscripts. The meaning is unknown but is obviously different from tripad, name of a metre. Feistel, p. 80, translates “mit einem Dreischritt”.

284 See for the narkuṭaka or natkuṭaka (dhruvā) 32.322-347 (and vv. 395, 440, 476-478), 34.204 + KM (p. 643 line 19, p. 644 line 3) = 33.237 +, 238 + C.

285 This verse, only found in B and C, is the same as 5.135a C.

286 See n. 307.

287 The reading ekapade C seems to be a mere conjecture for ekapadām of the other editions. B does not mention any variant reading. For this word see Feistel, pp. 83, 120.

288 vitandā: Ghosh translates it by “controversial topic”, Feistel by “frivole Diskussion’ and refers to Nyāyakośa (Poona 1928), s.v. See below, p. 190ff.

177

remarks 289 and enigmatical utterances 290, and questions such as 'Who is standing (there)?' (or 'Who has won?', etc., which having a bearing (?) 291

on the (following) play. The trigata is performed as a conversation 292 by one assistant. It is distorted 293 by the vidūṣaka but established by the

sūtradhāra]' 294.

This curious intermezzo is, apart from the following Prarocanā, the final "member" of the pūrvarañga. Hardly any attempt has been made

to understand its meaning. As far as I can see, scholars have without

289 For gaṇḍa see 20.129ff. (18.125B), Lévi, p. 115 and Konow, p. 22 "Wenn ein von einem Aussenstehenden gesprochenes Wort sich in ein Gespräch hineinfügt,

und z.B. wie eine Antwort auf eine Frage aufgefasst wird". This is based on Daśarūpaka 3.18 gaṇḍam prastutasambandhibhinnārtham sahaso 'ditam, where the commentary refers to Uttararāmacarita 1.38-†. Haas renders it by "abrupt remark"

(cf. Feistel, p. 83). Similarly Sāhitya Darpana 527 (VI.260 gaṇḍam prastutasambandhi bhinnārtham satvaram vacah) and Nāṭyadarpaṇa 2.97, which K. H. Trivedi, The

Nāṭyadarpaṇa p.110 paraphrases 'When a relevant speech suggests, all of a sudden,

some other sense it is called Gaṇḍa". But Sāgaranandin, Nāṭakalakṣaṇaratnakośa 2990 (366) takes it in a different sense.

290 For nālikā (nālikā KM) see 20.123C hāsyenā 'vagālarthā prahelikā nālike 'ti vijñeyā, Daśarūpaka 3.19 sopahāsā nigūḍhārthā nālikāiva prahelikā, Sāhitya Darpana 529 (6.261) prahelikāiva hāsyenā yuktā bhavati nālikā, S. Lévi, p. 116 "paroles énigmatiques où se mêle la plaisanterie et dont le sens est caché", similarly Konow,

p. 21. Sāgaranandin 2935 (360–61) adds an alternative definition. B (5.135a) reads tālikām without indicating on which manuscripts this is based. The reading has been adopted by Raghuvaṃśa and Feistel. The latter's guess (p. 83) " 'Schnapsideen',

Rede die vom Teddy (tālī) beeinflusst ist" is rather far-fetched. The only word tālī which frequently occurs in the NŚ. denotes musical time, e.g. 5.132, 144–46,

148. (For Hindi tīr, Engl. toďy see Turner, CDIAL. no. 5750 [the Kādamabari has both tāďiputa p. 126,23 and (tālapatra 20,8), tālīpaṭṭa 188,3, tālīpuṭa 98,16], and Hobson-Jobson, p. 927f.). KM. only records the v. 1. nāmikām (in KSS). The verse occurs exclusively in one (or two) manuscript(s) of the Madras Gov. Or. Mss. Library

and the library of Trivandrum.

291 prarūpinīm : a (probably late) hapax, from *prarūpayati, used instead of ni-rūp-, cf. 5.141 kāvyavasthānirūpaṇam. Perhaps corrupt for -prarūpanīm, but see Debrunner,

Altind. Grammatik II-2, p. 342f.

292 samjalpa, cf. 5.28–29 vidūṣakah sūtradhāras tathā vai pāripārśvikah, yatra kurvanti samjalpam tac cāpitam smr̥tam. Konow, p. 24, gives the correct translation

"Gespräch", cf. p. 191. Not "(amüsantes) Geschwätz" (Feistel, pp. 34, 83, 120), "gemeinsames Geschwätz" (idem, p. 151).

293 vidūṣitah C KM, virūpitah B Raghu. No edition records a variant reading.

The verse is not commented upon by Abhinavagupta. The former word occurs elsewhere in the NŚ., cf. 34.217 KM. (carmalakaṣaṇam) na bhūmāgnividūṣitam (but 33.251C na ca dhūmagniďūṣitam). The word may be meant as an etymological

explanation of vidūṣaka (n. 320) and virūpita may be the lectio facilior, due to a copyist but this cannot be decided without a better knowledge of the evidence of

the manuscripts. See nn. 320, 355 and 391.

294 5.136 tasyā 'nte tu tripadyā 'tha vyāhareta pāripārśvikau, tāyor āgamane kāryam gānam narkutakam budhaiḥ (137) tatā 'pi vāmaṛedhas tu vikṣepo dakṣiṇāsya ca, tathā ca bhāratībhede trigataṃ samprayojayet (138) vidūṣakas tu ekapadāim sūtradhāra-smitāvahām, asambaddhakathāprāyām kuryāt kathanikām tataḥ [(139) vilandāṃ gaṇdasainyuktāṃ nālikāṃ ca prayojayet, kas tiṣṭhati jitaṃ kene 'tyādi kāvya-prarūpinīm (140) pāripārśvikasamjalpo vidūṣakavidūṣitah, sthāpitah sūtradhāreṇa trigataṃ samprayujyate].

exception taken it for granted that it was a meaningless farce. The later Indian theorists did not understand it any more, and from the fact that they even omit it in their lists of the “members” of the pūrvarañga295 it may be inferred that not only its performance on the stage had become obsolete (which was probably true for almost the whole pūrvarañga), but no memory of it had even survived.

In the light of our conclusion, however, that the whole pūrvarañga was a ritual which aimed at a restoration of the primeval sacred world on the stage and a re-enactment of at least a part of the cosmogony by celebrating the erection of the Indra pole, the question as to the meaning of the Trigata cannot be lightly dismissed. The ritual character of the context in which it appears leads to the conclusion that this quasi-comical intermezzo must originally have been completely different from the clownish act which it is generally taken to be. The author of ch. V of the Nāṭyaśāstra himself, to whom we owe the last echo of what at one time must have been a meaningful performance, does not seem to have had any idea as to why this Trigata had to be performed. From his description we can only infer that he simply transmitted a tradition which had been observed from times immemorial. At this point the modern student of the Sanskrit drama is faced with a basic methodological dilemma. Trying to understand and interpret the scanty data of the Indian handbook means, in this case more than in others, going farther back into the past, beyond the Nāṭyaśāstra, and reconstructing, on the basis of what we think we can understand of the ancient pūrvarañga, a meaning for which even our earliest dramaturgical source can provide no corroboration.

The alternative is to confine oneself to what the text says expressis verbis and to the conclusions that can be drawn from it, which would imply that one would have to abstain from any guess at a possible meaning of the Trigata and leave it for what it is in our text: an incomprehensible corpus alienum in a ritual of high religious importance. No scholar has shown this self-restraint, for interpreting the Trigata as a farce means saying more than the text itself permits. The following sections are based on the conviction that the bounds set to research are not always necessarily dictated by the nature of our sources and that an effort to reconstruct,

295 It is interesting to see which of the twenty-two members (against twenty-one in NŚ. 33.226+, see IIJ. 16, p. 243 n. 10) Sāgaranandin, NLRK. 1126f. considers worth mentioning: pratyāhāra (placing of the musicians), mārjanā (playing of the drum), gītavidhi (song of praise), ōrmādyogacārī (graceful song and dance), mahācārī (forceful song and dance), prarocanā (inducement), nāndī (benediction), jarjarastuti (praise of the jarjara), digvandanā (salutation of the cardinal points), etc. (here quoted according to the translation by Dillon-Fowler-Raghavan). The order is strange but from the words (v. 1088) nāndy api tatraiva kartavyā we may infer that the nāndī at this time (1200–1250?) was recited after the prarocanā, immediately before the sūtradhāra entered (1094 yad ucyate nāndyante sūtradhāra iti). While Sāgaranandin quotes illustrative texts of the prarocanā, nāndī, jarjarastuti and digvandanā, his reticence about the Trigata was not, perhaps, entirely due to its character of an improvisation in prose. It may have been a mere name known from the Nāṭyaśāstra.

179

if possible, a pre-literary stage can at times be as legitimate as the reconstruction of a prehistoric linguistic stage.

A serious difficulty, however, is the circumstance that the word trigata, apart from denoting one of the last members of the pūrvarañga, also occurs in a seemingly different meaning in the list of the thirteen constituents of the vithī, the so-called vithyañgas 296. It is therefore necessary, in order to prevent wrong conclusions, first to examine the use of the term trigata.

13. THE TRIGATA I.

A definition of the vithyañga Trigata has been handed down in two different versions, both in the āryā metre and both corrupt. They will be distinguished as A and B. The version A is found in 20.128 C

yad udāttavacanam iha ca tridhā vibhaktam bhavet prayogeṣu hāsyarasasampprayuktam tat trigatam nāma vijñeyam

and in 18.179–180 KM (where it is given as a variant reading in the critical apparatus)

yathā 'nudāttavacanamm tridhā vibhaktam bhavet prayoge tu hāsyarasasampprayuktam tat trigatam nāma vijñeyam

In the second quotation the first word yathā must be corrupt for metrical reasons. One may conjecture yatrā or, in view of yad in C, yady. For metrical reasons, however, the latter reading would necessitate a further alteration of the text by reading yady anudāttam vacanam, which is improbable 297.

Version B is found in 18.179b–180a KM

śrutisārūpyārthatmani bahavo 'rthā yuktibhir niyujyante yad dhāsyam ahāsyam vā trigatam nāmā 'pi vijñeyam 298

296 See Konow, p. 21f.: "Bestandteile der vithī"; Haas, Daśarūpa, p. 84: "division", Ghosh, Translation I, p. 371 n. 1: "type"; Feistel, pp. 103, 105: "Glieder der Vithi".

Cf. 22.26 bhedāh . . . aṅgatvam āgatāḥ "varieties . . . have become its component parts" (Ghosh), "sind zu Gliedern geworden" (Feistel).

297 For udātta "exalted" ("erhaben", Konow, p. 34) with reference to persons cf., e.g., 20.10 and 64 prakhyātodāttanāyaka, 34.20 senāpatir āmatyaś ca dhirodāttau prakīrtitau, 34.26 divyā ca hṛpātmā ca kulastri yaḥ tān tu nāyikā jāneyā nānāprakṛtilakṣaṇāḥ (27) dhīrā(ś) ca lalitā(ś) caiva udāttā nibhṛtā caiva tathā, divyā rājāṅganā hy etā gunair yuktā bhavanti hi (28) udāttā nibhṛtā caiva bhavet tu kula-jāṅganā, lalitā cā 'py udāttā ca gaṇikā śilpakārikā and passim. On the other hand, it is true, udātta is met with in such expressions as NŚ. 20.32 udāttavacana(krta) "exalted speech", 20.45 udāttā bhāvāḥ "exalted situations", 26.11 gambhīrodātta-samyuktān "deep and exalted (feelings)", Dhanamjaya, Daśar. 3.60 anudāttokti, Sāradātanaya, BhPr. p. 197, 16 prakhyātodāttavastunah, Sāgaranandin, NLRK. 321–22 tayor udāttam api vacanam (with Raghavan's note, p. 61). See further below, p. 188. Ghosh's translation "exalted words" is, therefore, also possible.

298 For yad the edition has the conjecture etad, which is against the metre.

180

and in 18.264 B (here quoted from a foot-note in C)

śrutisārūpyād asmin bahavo 'rthā yuktibhir niyujyante

yad dhāsyam ahāsyam vā trigataṁ nāmā 'pi vijñeyam

Version A can be translated as follows: “When in a performance a talk of (non?-exalted persons is divided over three (characters) and it has the comical sentiment, it is to be distinguished as a Trigata”. As for version B, whatever may have been meant exactly by its composer, it clearly refers to something quite different from A: “In it many meanings are artfully attached to (a sentence, etc.) owing to a resemblance of sound. This, which may have a comic or non-comic character, is to be distinguished by the name Trigata”. It is interesting to see that one of the earliest among the later theorists, Dhanamjaya (between 974 and 996 A.D.), did not try to conciliate the two contrasting definitions but simply juxtaposed them (Daśarūpaka 3.16):

śrutisāmyād anekārthayojanam trigataṁ tv iha

naṭāditritayāl lāpaḥ pūrvarañge tad iṣyate

which can only mean that he still recognized, beside the modern vīthyañga, a different Trigata that was limited to the pūrvarañga. His commentator (and brother?) Dhanika only explains the first meaning (like the Nāṭyadarpana, which takes it in the sense samdeha “doubt”, see K. H. Trivedi, The Nāṭyadarpana, pp. 114f., 124) and quotes Vikramorvaśī I. 3 (N.S. Press): here the sūtradhāra is in doubt whether he bears bees, a cuckoo or kinnaris (v.l. divine women). Similarly Śāradātanaya, Bhāvaprakāśana, p. 231, lines 10-15 (between 1100 and 1300 A.D.), who quotes the Daśarūpaka but then adds the line etat prastāvanātme 'ti kathyate nāṭyavedibhiḥ “experts call it a kind of prologue”. This shows that he mixed up the Trigata and the āmukha (prologue). Significant for this confusion is the fact that in his definition of the older Trigata (Bh. Pr. p. 197 line 19) he substitutes sallāpa (that is, samlāpa) from the definition of the āmukha (NŚ. 22.28) for samjalpa. In the first passage quoted here there can be no doubt that Śāradātanaya, like Dhanika, is only referring to the later Trigata. It is, however, difficult to decide whether he regarded the two Trigatas (the older one of which was a purely traditional name to him) as identical. Unfortunately, Raghavan's book on Bhoja, who was Śāradātanaya's main source, is inaccessible to me so that it is impossible to determine to what extent Śāradātanaya is here dependent on his predecessor. He illustrates his definition with a quotation from a vīthī, which is similar in character to Dhanika's example. Viśvanātha (between 1300 and 1350?), as usual, follows Dhanamjaya in paraphrasing the Daśarūpaka: trigataṁ syād anekārthayojanam śrutisāmyataḥ “The Trigata is the application of more than one meaning to a sentence owing to a similarity of sound” (Sāhityadarpana 523/VI.257). The commentary refers to it as a different opinion and quotes as an example Vikramorvaśī IV. 21 (NS Press), where the echo of a question is taken as the answer to it. Only Sāgaranandin (between 1200 and 1250?) in his Nāṭakalakṣaṇarat-

181

nakośa 2981/364 presents an entirely different definition, viz. sphuṭa-bhāvyarthakathanam "telling clearly what is going to happen".

On the strength of these later texts and their commentaries Konow (p. 22) interprets this definition as follows: "wenn dieselben Worte ver-schiedenen Sinn haben, z.B. wenn ein Echo als Antwort auf eine Frage aufgefasst wird; verschieden hiervon ist der ununzehnte Bestandteil des pūrvarañga". Cf. Lévi, p. 114, Śāradātanaya p. 231 l. 13ff.

Konow's last words confront us with the fundamental problem: was there any connection between the Trigata of the pūrvarañga and the vīthyañga of that name, or were they indeed, as Konow assumes, entirely different? Feistel, p. 81f., arrives after an ample discussion at the conclusion that the definition of version B is a later interpolation, inserted by one who did no longer know the difference between the two Trigatas. Like Konow, accordingly, he assumes that the two Trigatas were originally different. His arguments will be discussed below.

The first difficulty that faces us when we try to disentangle this complex of problems is the circumstance that the manuscripts give two entirely different definitions. That version B is also found in the Daśarūpaka and the Sāhityadarpana is no proof for its being the older form. Its words yad dhāsyam ahāsyam vā "which can have a comical character or not" rather seem to contain an element of polemics directed against hāsyarasa-samprayuktam "with the comical sentiment" of A. If that is correct, A is the older definition and has been replaced by a later one, which covered a wider use of the rhetorical figure.

Another and stronger argument in favour of the older age of A is the fact that it apparently still describes the Three Men's Talk of the pūrvarañga, whereas B refers to something different. Some more conclusions can be drawn from A. According to the reading anudātta- in KM., the samjalpa between the three men was considered a conversation between lower characters. In this view the sūtradhāra who, as the word tridhā implies, was placed on the same level with the two participants in the conversation, must have lost his lofty character when coming back to the front of the stage. This raises a number of questions, because there are no indications to show that at this point the pūrvarañga had already ceased to be a religious ceremony. Unfortunately we have no means of answering such questions as: on what manuscript evidence is this reading based, and, is it probable that the one who for the first time wrote anudātta- "non-exalted" had still seen the whole pūrvarañga (including the Trigata) performed on the stage, at a time when the Trigata had already been degraded to a mere farce? Or was he writing about a literary tradition to which nothing corresponded any longer in practice? These questions are important because the same definition has been handed down in a form which says exactly the opposite, viz. that the Three Men's Talk was performed by exalted characters. The variant form of the verse in which udātta-vacanam "consisting of words of exalted characters" (C) occurs is certainly a bit clumsy, with its stop-gap iha ca, but then, this is not uncommon in such technical definitions which had to be given in

182

a metrical form. Judging from what it says, the latter reading is no doubt the lectio difficilior. No attempt will be made, however, with the scanty data that are at our disposal, to decide this text-critical problem. It is sufficient to state that there is a tradition according to which the three actors of the jorjaraprayoga (including the vidūṣaka), when coming back to the front, had remained exalted characters or spoke an “exalted” text.

For determining the relation of the Trigata as described in definition A to the vīthyaṅga of that name it is important to note that the words tridhā vibhaktam “divided over three (characters)” are irreconcilable with the general rule that all vīthyaṅgas had only one or two actors: “The vīthī consists of one act and has one or two actors” 299. This unambiguous definition (quoted by Konow, p. 32) is, however, curiously contradicted by a line which precedes it in the Kāvyamālā-edition, but follows it in C: “It has three characters who can be of the high, the middle or the low category”. (adhamottama-madhyābhir yuktā syāt prakrtibhis tiṣrbhir) 300. Konow refers to the Sāhitya Darpaṇa 301. That this line, which directly contradicts the general rule has been inserted at a later time into the text of the Nāṭyaśāstra is indicated by its place in the manuscripts, before or after the general rule. It must have become necessary at a time when to the list of the vīthyaṅgas a new one was added with three characters. As a result the older definition was no longer generally valid and had to be widened so as to comprise the Trigata with its three characters. Sāgaranandin, as a matter of fact, describes the vīthī in general as performed by three actors 302 and quotes as an example for the Trigata a passage from a classical drama in which a female character predicts (or rather hints at) future events 302a.

299 Cf. 20.112 vīthī syād ekāṅkā dvipātrahāryā tathai ’kahāryā ca. Cf. 18.163 KM vīthī syād ekāṅkā tathai ’kahāryā dvihāryā vā.

300 This is, it seems to me, the only possible translation. Ghosh renders it as follows: “And it is to include characters of the superior, the middling or the inferior type” but ignores the crucial word tiṣrbhir. In the introduction to his translation, vol. I2, p. XLVI, he paraphrases these words as follows: “It may contain any of the three kinds of characters superior, middling and inferior (XX.112-113)” . The reason for this remarkable interpretation can be found in the words which precede it: “Vīthī should be acted by one or two persons”.

301 Konow, p. 32, who refers to SD. 6.253f.: “SD. bemerkt aber, dass nach der Ansicht eines Unbenannten auch drei auftreten können, ein Hochstehender (uttama), einer mittleren Ranges (madhyama) und ein Niedrigstehender (adhamā). Der Kommentar zum Mālatīmādhava, ed. Bhandarkar3, S. 81, schreibt diese dem Bharata zu. Das dort gegebene Zitat lässt sich aber in der Ausgabe nicht nachweisen, und nach SD. handelt es sich wohl um eine andere Autoritāt”. The text reads: vīthyam eko bhaved aṅkaḥ caid eko ’tra kalpyate, ākāśabhāṣitair uktaiś citrām pratyuktim āśritaḥ, etc. and the commentary explains kaiścid by uttamo madhyamo ’dhamo vā.

302 Sāgaranandin, NLRK. 2906 (357 +) atha vīthī. sā ca tribhīḥ pātraiḥ pratraiḥ prayoktavyā. yathā Bakulavīthikā. uttamadhamamadhyamānāyikābhūṣitā triprakṛtiyutā bijabindukāryair arthaprakṛtibhir yuktā. ekāṅkā saṃdhiadvayayuktā mukhanirvahaṇayutā nānārasābhinidhītā ca. asyā aṅgāni trayodaśa, (etc.). Cf. 1855 vīthikāyāḥ, udghātyakādi trayodaśa.

302a Ibidem (v. 2981/364 +) sphuṭābhāvyarthakathanam trigatam. yathā Kadalīgrhe 183

Our tentative conclusion is that definition A, which corresponds to the Trigata of the pūrvarañga and is more concrete, is older than version B, which is vaguer, more comprehensive and which seems to contain a correction of A. Although none of these arguments is in itself decisive, the correspondence of B with the Daśarūpaka and the Sāhitya Darpana shows that this definition was accepted in later times. Whether this reconstruction is correct or not, line 20.112b is, as far as I can see, only explicable as a later interpolation. Since the general rule in 20.112a prescribes for all vithyañgas a maximum of two actors, one is bound to conclude that the words tridhā vibhaktam in definition A of the Trigata (20.128) are irreconcilable with the older definition of the vithyañgas. Thus the Trigata as defined in version A cannot originally have been a vithyañga. Konow and other scholars have arrived at the same conclusion but in their opinion this discrepancy proved that the vithyañga Trigata as a rhetorical figure had from the outset been different from the Trigata of the pūrvarañga. The only text, however, which Konow had at his disposal in 1920 was the first edition of the Nātyaśāstra in the Kāvyamālā Series. It seems now possible to reconstruct the historical development in greater detail.

All “members of the vṛthi” are as such also kinds of the verbal style (bhāratī vṛtti) 303. As bhāratībhedas all vithyañgas can be used in, e.g., the prologue (āmukha or prastāvanā), which is itself also a bhāratībheda 304. The “laudation” (praśocanā), which follows the Trigata and is the end of the pūrvarañga, is expressly mentioned as a form of bhāratībheda 305. In the same manner the verbal character of the Trigata of the pūrvarañga involved that, even at a time when it was not yet inserted in the list of vithyañgas, it was performed according to the verbal style (bhārati vṛtti). When the theorists began to list the many cases of the verbal style as special “species” of the vṛthī, the Trigata was also marked as a special vithyañga 306. It should be noted that the wording of 5.137 tathā ca bhāratībhede trigataṃ prayojayet “and further he should perform the Trigata in a kind of verbal style” 307 does not necessarily imply that the

[Ratnāvalī II.0.61] Susamgatā: esā una etha sāriā saṅkidavā. kadā vi gahidaatthā kassa vi purado paḍeedi “Trigata is clearly to tell what is going to happen. For example, in Kadambari: Susamgatā: But this sārikā is to be mistrusted. If she hears what we say, she may at any time make it public before anyone” (Transl. M. Dillon etc.). Cf. Śāradātanaya p. 231 ll. 14-15.

303 Cf. 20.107 vithyañga and 22.26 bhedās tasyās tu vijñeyāś catvāro 'gatvam āgataḥ. 304 22.29 vithyañgair anyathā 'pi vā “they adopt any type of the vṛthī or talk in any other way” (transl. Ghosh). 305 22.27 jayābhyudayini caiva maṅgalyā vijayādvahā, sarvapāpapraśamanī pūrvarange praśocanā. 306 Similarly A. Barth, Oeuvres IV, p. 149. Cf. Feistel, pp. 103 and 104. 307 Similarly Feistel, p. 30. Ghosh translates “Then, in case of a play in the Verbal Style (bhārati vṛtti) the Three Men's talk should take place”. This is impossible because the Trigata is part of the pūrvarañga, which is performed irrespective of the nature of the following play.

184

author was already acquainted with the theory of the vīthyaṅgas. The fact that he explicitly mentions the bhāratibheda may even indicate that at his time the theory of a "verbal style" had not yet been developed to such an extent as to make this characterization of the Trigata superfluous.

It is, indeed, attractive to accept Feistel's assumption308 that the elaboration of a systematic theory of the bhārati vrtti in ch. 22 (20 KM) dates from a later period (circa 250 A.D. according to Feistel). At that time four bhāratibhedas "kinds of the verbal style" were distinguished, viz. the parocanā, the āmukha, the vithī and the prahasana309. The first belonged to the pūrvaraṅga and the second, immediately following it, was the introductory part of the drama. It would not be surprising, therefore, if the Trigata had also been inserted as a "member" of the vithī, which would explain why it is not expressly listed here. This brings us with due caution to the conclusion that the "two Trigatas" were historically one and the same.

While, however, the older Trigata of the pūrvaraṅga fell into disuse at an early date, theorists started looking for instances of the vīthyaṅga Trigata and quoted scenes of an entirely different character (cf. the quotation from Sāgaranandin in n. 302). It may be true that according to the later theory all bhāratībhedas belonged to the literary art of the dramatist310.

This would only prove that the improvised Trigata of the pūrvaraṅga, which did not form part of the drama proper and which had to be performed bhāratībhede, dates back to a time when such theoretical distinctions were not yet made.

Definition A (see p. 180) still seems to refer to the old Trigata but its characterization of this scene as "performed with the comic sentiment" (hāsyarasasamprayukta) indicates that the Trigata had already come to be regarded as a farce. It is further difficult to decide which of the two variant readings udāttavacanāṁ (C) and anudāttavacanāṁ (KM) is correct.

Arguments have been brought forward above for the correctness of the former on the assumption that the definition actually refers to the old Trigata. As for definition B (18.264 B, 174 KM), it apparently refers to Trigatas of a more general character, which occurred in the plays and had not necessarily a comic character. It has probably taken the place of A at a time when there was no memory any more of the older Trigata.

Even in the theory of chapter XXII there are some details which seem to be inconsistent. Among the characteristics of the verbal style are mentioned the absence of female characters and the use of Sanskrit. One of the four categories of this style, however, is the āmukha, the introduction of the prologue of a drama, which can consist of a talk between the naṭī and the sūtradhāra311.

As for the Trigata, not even the restriction that 308 See Feistel, pp. 136, 125.

309 Similarly Daśarūpaka 3.5, Sāhitya Darpana 6.30.

310 See Feistel, p. 81, who considers the possibility that the 'Trigata of the pūrvaraṅga might have got its name from the rhetorical figure. The evolution has rather been the other way round.

311 22.28 naṭī vidūṣako vā 'pi pāripārśvika eva vā, sūtradhāreṇa sahitāḥ samlāpam

185

only Sanskrit should be used (22.25) was later considered valid, for Sāgaranandin quotes a scene from the Ratnāvalī in which a female servant is speaking Prakrit.

It must be admitted that the evidence is too scanty for an irrefutable reconstruction of the historical development between the Nāṭyaśāstra and the dramaturgists after 1000 A. D. In such a situation it is in general advisable to abstain from even tentative conclusions. In this particular case, however, no attempt at an explanation of the ancient Trigata can be made without determining to what extent later definitions can have had a bearing on this Trigata. Whatever the merits or demerits of this reconstruction which will here be briefly summarized, one conclusion seems sufficiently well founded, viz. that the Trigata of the pūrvarañga stands quite apart. It has to be taken as an episode of the pūrvarañga in its own right and must be interpreted in the light of the only authentic evidence available, viz. the passage at NŚ. 5.137–140 and the definition given at 5.28–29.

To the authors of these two passages its meaning seems no longer to have been fully clear but there is nothing in their wording to justify its being characterized as “amusing small talk”312. There is, however, one word in one of the interpolations which betrays that a later conception had already begun to gain ground: the “controversial topic” (vitandā) is said to have some bearing on the following play (kāvyaprarūpinī). First, this is against the general character of the pūrvarañga, which is a worship of the gods (devārcanavidhi NLRK. 2159), in which the audience is not addressed directly and which had no relation to the play (kāvya). Secondly, it was seen above (p. 181) that in later times the disputation (samjāpa) of the Trigata got mixed up with the samlāpa of the prologue (āmukha), whose proper function was to introduce the first actor (NŚ. 22.28–36). For these reasons it may be suggested that alluding to the following play was not originally a function of the Trigata and that the word kāvyaprarūpinī, occurring in the interpolation which is only found in two South Indian manuscripts, dates from a time when the proper character of the Trigata was no longer understood.

A similar term, viz. kāvyavastunirūpanam “mentioning the plot of the play” is also found in the definition of the next (and last) member of the pūrvarañga. This is the prarocanā, and here, again, the tradition has got confused. If we first consider the two definitions of Chapter Five, we find:

5.141

prarocanā ‘tha kartavyā siddheno ‘panimantranā (-ṇam)

raṅgasiddhau punah kāryam kāvyavastunirūpaṇam

yatna kurvate. Feistel, p. 125 is probably right in explaining this as the result of a later development as against 5.167ff., where only the sthāpaka is mentioned. See the Diamond Jubilee Number of the ABORI (1978) p. 175.

312 See above, n. 292. Cf., e.g., DR. 3.16.

186

"The the prarocanā is performed. (Then), however, an "invitation" with siddham is performed

• for success of the performance - which mentions the subject of the play".

From kartavyā and kāryam we may infer that the invitation is different from the preceding prarocanā. If, however, the audience was addressed at this moment, this would imply that the "invitation", which "defined" the plot, had a similar function as the subsequent prastāvanā, the prologue of the play. upanimantrana+instr. is common in the sense "invitation with, offer of". That siddhena (Ghosh: "expert") must actually be construed with upanimantrana is apparent from 5.29-30:

upakṣepena kāryasya hetuyuktisamāśrayā

siddhenā 'mantranā yā tu vijñeyā sā prarocanā

"An invitation with siddham, based on reasoning and arguments (?) by alluding to the (following) action, is known as prarocanā".

For kāryasya C (against kāvyasya Abhinavagupta, Śāradātanaya) cf. 21.98 and n. 329. Abhinava I, p. 219, 12 here explains siddhena by siddhyo 'palakṣitam. However, later theorists confuse this pūrvarañga-prarocanā with both the prastāvanā and the vimarsa-saṃdhi-prarocanā (which during the "crisis" points ahead to the denouement, NŚ. 21.96). See also T. Venkatacharya, Rasārnavasudhākara (forthcoming), p. 342f. Thus the definition of the latter prarocanā in DR. 1.47 is siddhāmantranato bhāvidarśikā syāt prarocanā, which is made up of NŚ. 5.30 and 21.96. Śāradatanaya's interpretation as āmantranām yat sadhyasya siddhāvat "announcement of (future events) as though they had already taken place", however, would seem excluded for siddhena at NŚ. 5.30. In other words, āmantrana has here a new meaning, different from NŚ. 5.30. On the other hand, the two definitions of Ch. V are strikingly different from the third in 22.27 (cf. Sāgaranandin 1070/114):

jayany udayini caiva maṅgalyā vijayāvahā

sarvapāpapraśamanī pūrvarañge prarocanā

"The prarocanā which occurs in the pūrvarañga serves to attain success, prosperity, good luck, victory and extinction of all sins".

According to this definition the "Recommendation" (prarocanā) was a religious act, not a prologue. If it did contain an "invitation", who exactly were addressed? Not before the end of the 10th century do we find the interpretation "appeal to the audience". Cf. Abhinavagupta (I2, p. 219 line 11f.) tena kāvyopakṣepena hetubhūtena tadviṣaye sāmäjikānām yā āmantranā, nimantranam. About the same time Dhanamjaya (between 974 and 996 A.D.) wrote (Daśarūpaka 3.6) unmukhīkaraṇam yatra praśaṁsātaḥ prarocanā, which Dhanika explains as prastutārthapraśaṁsanena śrotṛṇām pravṛttiyunmukhīkaraṇam. Somewhat later Śāradātanaya, BhPr. p. 197, 16-17 paraphrased it by prarocanā sā yatraiva prakhyāto-dāttavastunah, praśaṁsayā prekṣakānām unmukhīkaraṇam tu yat "where

the attention of the audience is drawn by the praise of a well-known and elated subject" and p. 228, 20-21 prekṣakādy-unmukhīkārah prastulārtha-praśaṁsayā, prarocana sā 'Śriharṣo Nipuṇe 'tyādino 'cyate (the last words of which refer to Dhanika's quotation from the Ratnāvalī).

It is necessary to stress the fact that the current view, which holds that the Prarocanā aimed at "exciting interest by praising an author in the prologue of a drama" (Monier-Williams) is, first of all, incorrect in that the Prarocanā had originally no connection with the Prologue. And secondly, this view is exclusively based on the word unmukhīkaranam in a definition written at a time (between 974 and 996 A.D.), when there was not even a faint memory of the Prarocanā. This can be seen from the fact that Dhanamjaya's brother (?) and commentator illustrated Daśar.

3.6 by a quotation (Ratnāvalī 1.5) which is, in fact, a stanza from the Prastāvanā! This is, of course, fully intelligible because at that time next to nothing had survived of the old pūrvarañga and no information could be found in old manuscripts, since the ritual had never formed part of the play. As Sylvain Lévi (Le théâtre indien I, p. 379) saw long ago, all difficulties are due to the fact that the theorists continued to pay lip-service to Bharata and to copy his directions, even though nothing corresponded any longer to them in the practice of their time.

If, then, we put aside the evidence of Dhanamjaya and the later theorists and confine ourselves to the only and last testimony about the Prarocanā, which is found in the Nāṭyaśāstra, we are faced with the contradiction between 22.27, which describes it as a ritual, and the two passages in Chapter V - one of the oldest chapters, in my opinion, of the whole work - , which include in the definition an allusion to the following play. In 5.141 this is formulated as a kind of "afterthought", which does not form part of the definition of the preceding line. In 5.29-30, however, the two definitions seem to have been amalgamated into one, or, to put it in a more neutral way, the aspects that are discernible in 5.141, are here inseparable. Clearly, these facts admit of different explanations. I am fully aware of the fact that in the following account an element of circular argumentation cannot be avoided. The pūrvarañga is a religious ceremony in which only the gods are addressed, but nowhere the audience. This makes it very unlikely that āmantranā and upani-mantranam can mean an appeal to the audience; nor can āmantranā, for that reason and in view of the apparently synonymous word upani-mantrana, have meant a "farewell" of the sūtradhāra. An "invitation" addressed to the sthāpaka, who was to introduce the play after the completion of the pūrvarañga is not likely either, if it is correct that the sthāpaka was the same person as the sūtradhāra. On the other hand although āmantrāyati is sometimes used in a context where a god is addressed (e.g., ŚB. IX.1.2.16), our text would no doubt have been more specific if this had been meant in these two passages. No convincing interpretation, therefore, can in my opinion be proposed for the two words. There remains, then, the second question: Is it likely that the prarocana had originally the function of "alluding to" (5.29) or "explaining"

188

(5.139) the plot? Not only the character of the pūrvarañga, as described in the preceding pages, but also the fact that it was the task of the sthāpaka to introduce the play lead to the conclusion that this definition of the Praocañā must have arisen at a time, when the distinction between pūrvarañga and the play was no longer fully maintained. For the same reason, then, the allusion to the play in the Trigata must be considered alien to the original character of this episode. It will be clear that this conclusion cannot be drawn from the material but rests entirely on a general theory of the pūrvarañga.

14. THE TRIGATA II

The following attempt to interpret the character of the old Trigata will be based exclusively on the data of ch. V of the Nāṭyaśāstra. All theories, therefore, which derive the Trigata and the vidūṣaka from the mime313, or which hold that the Trigata and the Praocañā had a profane character314 and that the vidūṣaka had only secondarily intruded from the pūrvarañga into the play, there to become the friend and companion of the nāyaka315, will be left out of consideration. They are all based on the premise that the vidūṣaka was, and had always been, a buffoon ("lustige Person, Hanswurst") and hence had necessarily originated in the "popular theatre" (about which nothing is known) and could not have had a religious origin316. They all disregard the fact that the only Sanskrit drama that we know of belonged to court art and was performed in a theatre whose religious character is beyond dispute, and that the vidūṣaka according to the Nāṭyaśāstra had always been one of the three leading parts in this drama (cf. Pischel, Puppenspiel, p. 17).

Let us first summarize a few conclusions of the preceding sections. The pūrvarañga has been shown to be, up to this Trigata, a religious ceremony (5.142 vidhi). The stage represented the world in its sacred aspect and the religious importance of the various acts, including the raising and worship of the jarjara and culminating in the nāndī, cannot be questioned. There is no indication to show that after the desecration of the Indra staff the character of the pūrvarañga had changed. As has been pointed out above, statements in the Nāṭyaśāstra to the effect that after the dismissal of the jarjara the acting with words and gestures was considered to begin (5.27 rañgadvāram) and that at that moment the judges started evaluating the acting according to aesthetic standards (27.40) do not imply that this was the end of the ritual part - quite apart from the question as to when this aesthetic evaluation may have been introduced. As for the late passage 32.460, which says that the nāṭyadvāram

313 Thieme, "Indisches Theater", p. 40, Feistel, pp. 113, 120, 131. 314 E.g., Thieme, p. 40, Feistel, p. 116. 315 Feistel, pp. 105, 125. 316 Feistel, although accepting Thieme's theory (p. 120), admits that the argument of the participation of the vidūṣaka is "nur bedingt beweiskräftig" (p. 119).

(beginning of the acting?) was at the end of the pūrvaranga317, this may date from a time when the seemingly clownish and meaningless Trigata and the Prarocanā, which had become a useless double of the Prastāvanā, were skipped. There can be no doubt, howerer, about their having been the last parts of the pūrvaranga318.

As we have seen, in the first part of the pūrvaranga the two attendants had been characterized by the sacred objects they were carrying as impersonations of Indra and Varuṇa, and thus as standing for the two cosmic moieties. During the jarjaraprayoga and the nāndī they were clearly subordinate to the sūtradhāra. In the Trigata they are still so but now the active role is theirs.

In interpreting the relevant passage (5.137-140) we are faced with the text-critical problem that it contains several lines which occur in a few manuscripts only and must therefore be considered interpolations. As a rule such interpolations are either amplifications or corrections, dating from a later time and reflecting changes in the practice of the dramatic performance. Since, however, there is no appreciable contradiction between the lines of this passage, use has been made of the whole passage as it appears in C., even though the text-critical basis on which 5.140 rests seems to be slender. It has been adopted in B KM and C but was rejected by Grosset. As we shall see, something more special can be said about it.

The characteristic point of the interpolated verses 5.139 and 140 is that they seem to describe the intermezzo of the pāripārśvika and the vidūṣaka in terms of a formal disputation according to the terminology of the oldest stage of the vāda-doctrine that is known to us. As G. Oberhammer, WZKSO. 7 (1963), p. 63ff. has shown, the oldest passage preserved, viz. Caraka-Saṃhitā III.8.28, is merely “ein Organon für die Disputation”, the disputation being divided into two parts, the jalpa and the vitaṇḍā. In the jalpa each of the opponents defends his thesis, e.g., one defends the thesis that there is transmigration, while the other argues that there is not. In contrast with the jalpa, in which each party uses arguments, the vitaṇḍā “only consists in criticizing the opponent” (jalpaviparyayo vitaṇḍā. vitaṇḍā nāma parapakṣa doṣavacanamātram eva). The commentary of Cakrapāṇidatta (13th century) is here particularly interesting: tena vitaṇḍāyām apy uttaravādinah parapakṣadūṣaṇalakṣaṇah pakṣo 'sty eva, parāṃ svamatiṃ nā sādhayati 'ti bhedaḥ.

In view of the term sthāpita in 5.140 it must be added that Caraka attaches a high importance to sthāpanā “establishment, dialectical proof of a proposition” and pratiṣṭhāpanā “counter-assertion, statement of a counter-thesis”. They belong to the earliest stage of the vāda-doctrine and got out of use before the Nyāyasūtras (Oberhammer, pp. 73, 77f.), whereas jalpa and vitaṇḍā disappear after them (ibidem, p. 81 n. 40).

317 See above, p. 171.

318 5.15 trikaṁ prarocanā cā 'pi pūrvararnge bhavanti hi (16) etāny aṅgāni kāryāṇi pūrvarañgavidhau tu ca [but see Corrigenda p. 236: trigataṁ prarocanā ca].

190

Although no definite date can be fixed for this stage, it can hardly have been later than the first centuries of our era, accordingly the same period in which also the core of the Nāṭyaśāstra was presumably composed. The verses 5.139-140 thus define (redefine?) the “talk” of the Trigata both in terms of rhetoric (ganḍa, nālikā), the date of which remains to be settled, and as a formal disputation with the terminology of the oldest “vāda-tradition” (vitaṇḍā, vidūṣita and sthāpita). As for the word samjalpa, although it is a common term for “conversation, talk” in classical Sanskrit, Bharata (5.29, 140) distinguishes it from samlāpa (22.28). Is it used here in the sense of the old technical term jalpa?

A few words must be added on the term sthāpita, which is here used in a sense that is clearly different from sthāpanā in Caraka. Unlike the Vedic ritual brahmodyas, in which one of the two opponents automatically gained the victory when the other could no longer answer and remained silent (see J. C. Heesterman, WZKSO. 12–13 [1968], p. 171ff.), there is in this disputation a highest authority, who stands above the parties and gives his decision by “establishing” the words of the assistant (5.140, lacking in most manuscripts). Heesterman has already suggested that the “unsophisticated abuse and the equally artless praise” of the Vedic ritual may perhaps be considered “an early predecessor of the distinction between jalpa and vitaṇḍā” (p. 178). In view of the apparently religious origin of the Trigata I am inclined to think that this disputation in the presence of a “judge” was a direct reflex of the Vedic vīvāc as a social phenomenon, which could be described in terms of the post-Vedic (vi)vāda because the latter was a direct continuation of the older verbal contest. Oberhammer (p. 81) has already concluded that the sthāpanā and pratiṣṭhāpanā must have belonged to a pre-Caraka stage, and that they were notions that were known from the practice of disputations. In Caraka the five categories of demonstration (Beweisglieder) are still elements of sthāpanā and pratiṣṭhāpanā. It seems to me that the Trigata preserves the older Vedic stage, in which sthāpayati had nothing to do with a logical proof but only with one’s winning acceptance for one’s contentions in a verbal contest. The decisive sthāpanā is here pronounced by the highest authority.

The preceding verse 5.138 quite suddenly introduces the vidūṣaka and refers to his incoherent speech. He had apparently the role of the uttaravādin in the later formalized vitaṇḍā. According to Cakrapāṇidatta the uttaravādin confined himself to criticizing (dūṣaṇa) the proposition of the opponent without proving his own (see notes 320, 391). Since the two assistants had remained all the time at the back of the stage and had now come again to the front (apparently without their paraphernalia), one of them must in the meantime have made himself recognizable as the vidūṣaka, probably by means of a make-up to which the text refers elsewhere (36.25 KM)319. It has been argued that the only assistant that

319 Ghosh’s translation (p. 94) “During it the Jester should suddenly come in” is of course not based on what the text says.

191

can have functioned as such is the one who had carried the golden pitcher. If it is correct that the two assistants have all the time impersonated Indra and Varuna, this must still be the case, since nothing indicates that they have changed their former characters. Unfortunately it cannot be decided whether the Nāṭyaśāstra considers them "exalted" (udātta) or "non-exalted" (anudātta) characters. As far as I can see, howerer, the remarkable fact that the vidūṣaka speaks Sanskrit in this Trigata has never been given due attention. (See Abhinavagupta p. 219 l. 8 ad NŚ. 5.28-29!). For a correct insight into the character of the disputation between the assistant and the vidūṣaka it may be well to call to mind that the consecration of the stage in chapter III, which shows some striking parallels to the pūrvarañga, ended with a fight which has been interpreted above as a re-enactment of the cosmogonical strife between Devas and Asuras. See p. 165 and cf. NŚ. 3.91-93. The disputation of the Trigata, too, has a strong element of contest. The vidūṣaka "criticizes" the words of his opponent. Although the authenticity of the reading vidūṣakavidūṣita, a variant of -virūpita "deformed"320, is not certain (but see n. 391), it is obviously intended to give an etymological explanation of the word vidūṣaka (see n. 320) and in view of the archaic terminology preserved in these verses, there is no reason to question its correctness. See above on the dūṣaṇa of the uttaravādin. The sūtradhāra, who must have been standing between the two opponents, decides the contest by confirming ("establishing") the words of the assistant. The fact that the latter must be victorious is significant. It shows that this dispute was not simply an improvised comical intermezzo but a ritual act, such as, e.g., the Mahāvrata-ceremony, in which the Ārya (or Arya) had to defeat the Śūdra. Such a verbal contest (5.137 bhāratibhede!), as the Trigata obviously was, is strongly reminiscent of the verbal contest between Indra and Varuṇa in Rigveda IV.42. This was a cosmogonical vivāc321 in which Varuṇa probably still (or again)322 had his Asuric character. In other words, the contest of the Trigata between the leader of the Devas and the highest Asura is only a mythological variant of the cosmogonical strife between all Devas and Asuras, represented at the end of the consecration of the stage. In that consecration as well as in the pūrvarañga the ceremonies ended with a renewal of the cosmos by a reiteration of the primordial mythical fight.

320 See n. 293. virūpita is, it seems, first attested in the epics. It may originally have been a gloss for vidūṣita, which has later ousted the correct reading. It is important to note that vidūṣayati has been in use ever since the Rigveda so that it is quite possible that the word vidūṣaka was created in the Vedic period. For this type of formation cf. VS gaṇaka "astrologer": ep. gaṇayati (Debrunner, Altind. Grammatik II-2, p. 146f.). See further n. 355. 321 See above, p. 22f. and cf. IIJ. 4, p. 270 n. 80b, Numen 8, p. 38. For a different interpretation of the hymn as referring to the royal consecration see H. Lommel, Festschrift Schubring p. 32ff., Schlerath, Das Königtum im Rig- und Atharvaveda (Wiesbaden, 1960), p. 160, J. Gonda, The dual deities in the rel. of the Veda (1974), pp. 234-248. 322 See History of Religions 15, p. 116 on Varuṇa as a seasonal god.

192

As for the incoherent talk of the vidūṣaka, which evokes a smile of the sūtradhāra and which has greatly contributed to his being considered a buffoon in modern times, it may have its origin in the fact that the vidūṣaka must be defeated in a verbal contest. In the vivāc, the Vedic prototype of the saṁjalpa, the victorious speaker "outtalks" (áti-vad-, cf. bṛhád vad-) or "defeats" (ji-) his opponent by his successful speech 323. Just as Indra must win the cosmogonical contest, it is the task of the sūtradhāra, who still impersonates Brahmā as the god of totality and who as such stands above the two parties, formally to confirm his victory.

The vidūṣaka's negative role as criticizer of the assistant's speech should be noted because we find here the prototype, as it were, of the pattern of the relation between the nāyaka and the vidūṣaka in the plays (see p. 208).

The preceding interpretation of the Trigata was mainly based on one verse, which seems to have been added as an amplification to the older description and which, if I am not mistaken, is only found in one of the few manuscripts used for the edition of Baroda, the one written in Old Malayalam script. It does not, however, contradict the other verses and even though it is apparently an interpolation, it may contain an old and authentic tradition.

15. THE CHURNING OF THE OCEAN

According to the Nātyaśāstra the origin of the Sanskrit drama dates back to the mythical primordial time when Brahmā created the dramatical art and "composed" the first drama. The tale of the legendary first performance occurs in two versions, which are both instructive as they show us what, according to the Indian tradition, was the oldest form of drama.

The first version describes the performance by Bharata and his sons, on the occasion of Indra's banner festival. At that time no theatre had been built yet by the divine architect Viśvakarmán. The sage Bharata tells the munis about this performance in the following words (1.57-58): "Thereupon [that is, after the nāndī, 1.56] I organized 324 an imitation 325

323 Cf., e.g., II.1.16 bṛhád vadema vidáthe suvírāḥ "May we with our valiant sons outtalk (our rivals)", VII.18.13 jéṣma Pūruṁ vidáthe mṛdhravācam "May we during the potlatch surpass Pūru who speaks injuriously" (see Indologica Taurinensia 2, p. 131), AB. VI.33.20 ativādena vai devā́ asurā́n atyudya "thainān atyāyan "by means of the Ativāda [AS. XX.135.4] the gods outspoke the Asuras and overcame them" (Keith, ŚB. XI.6.2.5 té hocuh, dvi vāi no 'yáṁ rājamṛdbandhur āvídd dhántai 'naṁ brahmódyam āhvā́yāmahā iti "They said, 'Surely, this fellow of a Rājanya has outtalked us: come, let us challenge him to a theological disputation!'" (transl. Eggeling).

324 baddhā : the PW. gives the meanings "Verse abfassen, componiren", "bewirken, hervorrufen" and "machen". Cf. 1.18 sambaddha =1.4 grathitā. Since "to compose" would seem excluded by the context, one might consider translating "I devised" (Ghosh). Abhinavagupta thinks that no prayoga can be meant here and takes it to mean that a prastāvanā was performed: tatra ca baddhe 'ti guṇanikā yojitā, na tu prayoga ity, etac cā 'sat tatpūrvottaravyāghātāt . . . (see n. 144), tasmād baddhe 'ti prastāvitā ; na tu niṣpāditā ("performed"). prastāvanā tāvat prayukte 'ty arthah.

of how the Daityas were defeated by the gods, (which imitation) consisted

of altercation 326 and tumult 327 and of cutting and piercing (of limbs)

and roaring sounds” 328. The second line has a more or less stereotyped

character. See the parallels quoted in n. 245 and cf. 27.13

diptapradesam kāvyam yac chedyabhedyāhavātmakam

savidravam athau 'tpātam tathā yuddhaniyuddha kam

which Ghosh translates "If the play has [a plot containing] burning

hostility, cutting and piercing [of limbs], fight, portentous calamity,

terrific happening or minor personal combat . . . 329.

According to this legend the first dramatical performance, after Brahmā

had created the art as a fifth Veda, took place immediately after Indra's

victory over the Daityas and Asuras and consisted of an imitation of his

fight and victory. It is quite understandable that, while the gods were

elated, the Asuras were disconcerted because (as we may add) this

"imitation" was in fact a new defeat by magical-religious means. They

therefore disturbed the performance by means of a host of "Obstacles"

(vighna). This again induced Brahmā to order the building of a theatre

as a sacred space, well defended all around against demoniacal attacks.

This first legend confronts us with two important facts: the first

performance not only took place immediately after Indra's cosmogonical

Brahmā's composing plays our text uses the terms 1.56, 111 krta, 1.19 utpādayati,

1.19, 4.12 srsta, 4.3 pragrathita. Here a performance improvised by Bharata is

referred to. Hence "staged, arranged the performance" rather than "devised"

must be meant.

325 See n. 247.

326 sampheta "conflict". Cf. 22.61 "Conflict is known to include excitement, many

fights, personal combats, deception, betrayal and much striking of weapons" (Ghosh)

and 20.79 "angry conflict" (Ghosh). More succinct is the definition of Daśarūpaka

2.58 sampheṭas tu samāghātaḥ kruddhāsamrabdhayor dwayoh, and similarly Sāh. D.

420 = 6.135. Sāgaranandin's definition (NLRK. 1380ff.-137 +)

this view: sampheṭaś ca vīraudrādbhutāprayāyir yuktas sasambhramamayo yuddhani-

yuddhābahulah kapāṭamayah satpratāpaṭavişayah, which he illustrates by a reference

to the fight of the gods and Asuras in the Samudramathana and the hand-to-hand

fighting and fighting with weapons (bāhuyuddhair astrayuddhaih) of Indra and Vṛtra.

Hence the translations "ungestümer Zusammenstoss" (Konow, p. 21), "fighting"

(Keith, Sanskrit Drama, p. 328). This is supported by sampheṭa, samspheṭa "war,

battle" (comm. on Hemacandra) and sampheṭaka "drummer" (not listed in the

lexica but occuring in Abhinsvagupta I2, p. 64 line 5 kutapah : sampheṭaka-gāyana-

vādaka-samūhah). However, the translation "leidenschaftlicher Wortwechsel"

(PW,

pw) has the support of Sāgaranandin 806 roṣāgrathitām vakyaṁ sampheṭaḥ "discussion

conducted in anger". Cf. Drav. Et. Dict. 3601.

327 vidrava "tumult" (Konow, p. 29, Ghosh, who also translates "excitement",

cf. 20.79). Three kinds are traditionally distinguished, cf. 20.70, DR. 3.67, Sāh. D.

6.240 and Sāgaranandin, NLRK. 766 saṁkābhayatrāsakṛto vidravaḥ "panic is caused

by apprehension and fear and trembling" (M. Dillon, etc.); cf. 2811ff.

328 See pp. 143 and 165.

329 This is based on Ghosh's doubtful emendation diptapradvesam. Cf. 20.85 diptarasa

"with exciting sentiments", Mhbh. V.182.7 krodhādipta "excited with anger".

Different is 3.93 kṣatam pradiptam āyastam. Note kāryam B KM for kāvyam, and

athotphullam B KM, athātpātam K for athautpātam.

194

fight (here apparently an annual re-iteration of the primordial fight, connected with the celebration of his banner festival) but it was itself a performance of this fight and therefore a creative act, a renovation of the world owing to the efficaciousness of imitation. A second conclusion that follows from this tale is that every dramatic performance has potentially a dangerous “magical” power which arouses the resistance of and obstruction by fiendish spirits. It will be remembered that the aim of the nāndī, for instance, was said to be “the appeasement of all obstacles” (see n. 254).

When the theatre has been built (1.81) and the consecration of a new playhouse has been described in detail (ch. III), the legendary tale is taken up again in 4.1: “After having thus carried out the pūjana, I said to the Grandfather (Brahmā): ‘Give me now quickly your orders, O Lord, as to which play must be performed!’ Thereupon the Holy One said to me: ‘Perform the Churning of the Ocean. It stimulates efforts and is a powerful means of giving pleasure to the gods’ ” 330.

Again it is a cosmogonical fight—although of a special character, in which cooperation and competition go together—which is the theme of this performance. In contrast with the previous one in ch. I, which can only have been improvised by the actors, this one is said to have been a samavakāra (4.3), composed by Brahmā himself (4.4), which was performed to the delight of both gods and demons (4.4). It is somewhat surprising to find the latter here assembled together with the gods but the playhouse had only been built to safeguard the actors and the performance against evil influence of fiendish powers (1.76-79). On the other hand, Brahmā had consoled the enraged demons (or rather, their “Obstacles”) by saying that the drama would represent acts and feelings of the demons no less than of the gods (1.105) and that it would be instructive for all (1.113).

The genre samavakāra must have become obsolete at an early date. Our knowledge, as well as that of later dramaturgists, is exclusively based on the definition of the Nāṭyaśāstra (20.64-65), which seems to describe the characteristics of a single play still known to the author, rather than those of a whole genre 331. In Ghosh’s translation this definition is as

330 4.1 evaṁ tu pūjanaṁ kṛtvā mayā proktah Pītāmaheḥ, ājñāpaya prabho kṣipraṁ kaḥ prayogaḥ prayujyatām (2) tato 'smy ukto Bhagavatā yojayā 'mṛtamanthanam, etad utsāhajanananāṁ surapritikaraṁ mahat.

331 See S. Lévi, pp. 30, 143, 153 (“éteint avant la période classique”), Konow, p. 28f. (“Es ist nach alledem klar, dass Bharata ein bestimmtes Werk, und nicht eine ganze Gattung, vor Augen hatte, und seine Nachfolger haben einfach seine Definition wiedergegeben und zum Teil weiter ausgesponnen. Bh. 4.2f. nennt den samavakāra Amṛtamanthana, und dasselbe Werk ist gewiss mit iem Ambhodhimanthana des Daśarūpa (3.58) und dem Samudramanthanā des Sāhityadarpaṇa (zu 6.240) gemeint.), Keith, p. 97 “that dubious kind of play”, p. 291f. “the Samavakāra, for instance, is described in terms which, with the precise definition of the time to be occupied by the acts, can be interpreted only as based on a single drama”, p. 353 etc. Only Sāgaranandin diverges in referring to a drama Śakrānanda (NLRK. 2812/355f.), but from his ample description (2811ff.) it is clear that he tried to identify some

195

follows: “It should have the [exploits of] gods and Asuras as its subject matter and one of them as its well-known and exalted Hero, and it is to consist of three Acts [presenting] the three kinds of deception, the three kinds of excitement or the three kinds of love. [Besides this] it should have as many as twelve dramatis personae and require a duration (lit. length) of eighteen Nāḍikās [for its performance]” 332. In this translation the words “one of them as its well-known and exalted Hero”, for which the first edition had “and an Asura as a well-known and exalted character as its Hero”, can hardly be correct. The text clearly says that there are twelve nāyakas, who are all traditionally well-known and exalted characters. The implication is, as Abhinavagupta explicitly states, that there was no viḍūṣaka in this play 333.

The samavakāra was a sub-class of the āviddha type of play, which is described as follows (14.56-59, cf. 35.53-56): “The play which requires energetic (āviddha) gestures and dance movements (āṅgahāra) to represent cutting, piercing and challenging, and contains the use of magic and occult powers as well as artificial objects and make-up, (57) and has more men and less women [among its dramatis personae] and applies [in its production] mostly the Grand [Sāttvatī] and the Violent [Ārabhaṭī] Styles, is of the energetic type. (58) According to the [expert] producers, [plays of] the Dima, the Samavakāra, the Vyāyoga and the Īhāmṛga [classes] are known to be of the energetic type. (59) Production of a play of this type should be made by [an impersonation of] gods, Dānavas and Rākṣasas who are majestic and haughty, and have heroism, energy and strength” (Ghosh) 334.

In this definition of the more comprehensive category, too, the presence of demons is stressed, even though Ghosh's translation of the last two lines is not very exact 335.

passages in the dramas known to him as instances of the samavakāra. On the question as to whether Bhāsa's Pañcarātra is a samavakāra see Konow, p. 29, Keith, pp. 96f., 346, Ghosh, Transl. I, p. 363 n. For an explanation of the name see, e.g., Dhanika ad Daśarūpaka 3.62 samavakiryan­te 'sminn arthā iti samavakāraḥ, accordingly an overwhelming by fate.

332 devāsurabījākṛt pra­khyātodāttanāyakaś caiva, trya­ṇkas tadā (tathā KM.) trikapātas trividravah syāt triśṛṅgā­raḥ (65) dvādaśanāyakabahulo hy aśā­da­śanā­dikāpramāṇaś ca . . . Cf. DR. 3.62-68 and SD. 6.234–239, which will be discussed below.

333 Abhinavagupta I2, p. 32 line 11 (ad NŚ. 1.97) has the following comment: hāsyaśṛṅ­gārāṅgata­vād “viḍūṣakam” ity uktam, ata eva daśarūpakaprayogasūcanam etat, samavakā­re viḍūṣakā­bhāvāt.

334 sattvā­viddhāṅ­gahā­rāṃ tu cchedyā­bhedyā­havā­tmakaṃ mā­vendrajālā­bahulaṃ tus­tanepathyasamyutam (57) puruṣair bahubhir yuktam alpastri­kaṃ tathaiva ca, sāttvaty-­ārabhaṭī­prayam nā­ṭyam ā­viddhasaṃjñitam (58) ḍimah samavakā­raś ca vyāyoge 'hāṃrgau tathā, etā­ny ā­viddhasaṃjñā­ni vijñeyā­ni prayok­tṛ­bhiḥ (59) eṣāṃ prayogaḥ kartavyo dai­tya­dānava­rākṣasaiḥ, uddhatā ye ca puruṣāḥ śau­rya­vī­rya­balān­vitāḥ.

The same text recurs in 35.53C1, 26.28B, where C has yat tv ā­viddhāṅ­gahā­ntam as the initial words, yat tv (v. 1. tatra) ā­viddhāṅ­gahā­raṃ tac 13.54 KM., na cā' vidyāṅ­gabhā­vas tu (!) p. 654 line 1 KM.

335 A more literal translation would be: “These genres must be performed by Daityas, Dānavas and Rākṣasas and by men who are dominating personalities and have heroism, courage and strength”. For cchedyā­bhedyā­havā­tmakaṃ see above n. 245.

196

The subject-matter of the Samavakāra is, accordingly, the devāsura, which the later dramaturgists took as a dvandva "gods and Asuras"336 but which the author may have meant in the Vedic sense "fight between Devas and Asuras, well-known (from tradition)"337.

If the general opinion that the composer of these lines has based his definition on a single play is correct, it can be concluded that he still knew a Samavakāra which represented the fight of Devas and Asuras and had twelve "famous and exalted heroes", who, then, must have been partly gods and partly Asuras. This conclusion is indeed drawn by the author of the Daśarūpaka338. This definition is perfectly in line with the situation found in the Churning of the Ocean, where the chief gods and Asuras, all "famous and exalted", took turns pulling the rope of the churn-staff, thereby symbolizing the pravṛtti and nivṛtti of the cosmic creative process (see n. 209). This situation excluded, however, the presence of other characters and it remains obscure why the commentator of the Daśarūpaka introduced a new element by adding "etcetera" after "Devas and Asuras"339.

The number of twelve heroes is not exceptional for one of the old and extinct genres. For two other genres of the āviddha type, viz. the Dima and the Īhāmṛga, a late theorist mentions the numbers of sixteen and ten (or six) heroes respectively340. They differed, however, from the Samavakāra in that, e.g., in the Īhāmṛga the three leading parts were the hero (nāyaka), his antagonist (pratināyaka) and a woman, whereas in the Samavakāra there seem to have been no protagonists. All heroes, probably six Devas and six Asuras, were on an equal footing and they all pursued their own purposes341.

336 Cf. DR. 3.63 khyātam devāsuram vastu, SD. 6.234 vrttam samavakāre tu khyātam devāsurāśrayam, NLRK. 2811 (355+) sa ca devāsuravīryakṛtab.

337 Cf., e.g., JB. I. 247 lines 6 and 9.

338 Dhanamjaya 3.63 netāro devadānavah (64) dvādaśo 'dātavikhyātāḥ phalam teṣām prthak prthak, bahucirā rasāḥ sarve yadvaṭ ambhodhimathanāthe. Similarly SD. 6.240 yathā samudramathanam. These late theorists may simply repeat an old tradition and combine NŚ. 20.63ff. with 4.2 (see above n. 331) but the detailed description of SD. is anyway remarkable.

339 Cf. Dhanika ad DR. 3.63 devāsurādayo dvādaśa nāyakāḥ. It is not clear what he had in mind. The two editions of the Sāhityadarpaṇa accessible to me, viz. the one by Pandit Durgāprasād Dvivedi and Kāśīnāth Pandurang Parab (Bombay 1902) and by Kṛṣṇamohan Śāstrī (Benares, Chowkhamba Sanskrit Series, 2nd ed. 1955) both read in 6.235 dvādaśo 'dātṛ akhyātā devamānāvāḥ. The commentator Rāmacaraṇa Tarkavāgīśa Bhaṭṭācharya (1544 A.D.) glosses this reading by divyādivyāḥ but quotes a variant reading devadānavāḥ, which does not seem to occur any longer in modern manuscripts. The curious reading -mānavāḥ is possibly due to the definition of the general category of the āviddha, of which the samavakāra was a sub-group. In the āviddha-category puruṣas did occur. This may have led to the assumption of human heroes in the samavakāra, too.

340 See SD. 6.243 śodāśa 'tyantam uddhatāḥ and for the Īhāmṛga 6.248 patākānāyakā divyā martyā vāpi daso 'ddhatāḥ and 250.

341 DR. 3.64 phalam teṣām prthak prthak, which the commentator Dhanika illustrates with the words "just as in the Churning of the Ocean Vāsudeva etc. won Lakṣmī etc." (yathā Samudramathane Vāsudevādinām Lakṣmyādilābhah).

197

In this connection there is certainly, in the light of the origin of the Greek drama, some reason to raise a point to which, as far as I can see, little attention has been given by students of the Sanskrit drama. Greek tradition has preserved the memory of the exact time when the drama was born, as the first protagonist stepped forward as a separate actor, to speak and act as an individual against the anonymous collectivity of the chorus. The question must be raised if something similar may have happened in India and if in the Samavakāra with its twelve heroes the prototype of later forms of drama, with a nāyaka and a nāyikā, has been preserved. Perhaps it will never be possible that question in a satisfactory way. I do not think that we are at this moment, with the available evidence, in a position to discuss it exhaustively. I will, therefore, confine myself to the following brief comment.

Although in Greece as well as in India the drama has developed from a religious drōmenon, there is a clear difference. To begin with, the Samavakāra was different from the chorus because the actors were not an anonymous collectivity. The twelve heroes are described as well-known personalities, each trying to fulfill his own desires. Just as in modern illustrations of the Churning of the Ocean each god and Asura is marked with his own name, so they were in the ancient play khyāta, "renowned" and "well-known", that is, recognizable for the audience as individuals. Secondly, the cosmogonical strife which they are said to have represented was ever since Vedic times described in a twofold way. While in the Rigveda Devas and Asuras as groups remain in the background, the fight being primarily depicted as one between two protagonists (particularly Indra and Vṛtra), the later brāhmaṇas mostly ignore Indra's individual exploit and use the fixed formula (with slight differences from one śākhā to another) "The Devas and Asuras were at strife" (devás ca 'surāś ca pasprdhire [aspardhanta, ayatanta, samyattā āsan, etc.]) 342. There are no indications to show that one way of presenting the sacred history was anterior to the other. In the light of this fact, the absence of the vidūṣaka in the Samavakāra is no more surprising than that of the nāyaka. If we are right in identifying them with Varuṇa and Indra respectively, both were present among the (six) Devas and (six) Asuras. There was no need, therefore, for them to act separately, disguised as special characters of the drama. It is true that in some forms of the classical drama the vidūṣaka does not occur either. This, however, is a different problem which will be discussed separately in § 17, after the character of the vidūṣaka has been examined more closely.

To conclude this section may be raised which, it is hoped, will show that reasonable guesses are sometimes possible in these matters, even though they cannot be definitely proved with the help of our sources. This point is the place on the stage occupied by the two parties during the performance of the Churning of the Ocean. It has been suggested above that at the moment when, at the beginning of the pūrvaraṅga, the

342 For references see, e.g., S. Lévi, La doctrine du sacrifice p. 44 n. 1, 45ff.

198

sūtradhāra entered with his two assistants, the jarjaradhara must have gone on his right and the bhrṅgāradhāra must have been on his left. In the same way the relative posicions of Devas and Asuras as the higher and the lower group must have necessarily have involved that the first stood right of the centre (the brāhma maṇḍala) and the latter on its left. What was "right" and what "left" was determined by the nātyācārya, the stage-manager, while he was standing on the stage and looking towards the audience. It may be assumed that the situation in India was in this respect not much different from that in Indonesia and Malaysia, where in the wayang purwa (shadow play) all the characters of the "noble type" are on the right of the dalang (puppet handler) and those of "violent type" on his left 343.

16. THE CHARACTER OF THE VIDŪṢAKA IN THE SANSKRIT DRAMA

Although much has been written on the part of the vidūṣaka344, it is remarkable that most scholars have started from a premise which to them was apparently so self-evident as to be in no need of being stated explicitly, viz. that the vidūṣaka was a clown ("Spassmacher, Hanswurst"), who therefore could not originally have belonged to the drama. The further conclusion was that in that case he must have originated in a "popular

343 See W. H. Rassers, "Over den oorsprong van het Javaansche tooneel", Bijdragen Taal-, Land- en Volkenkunde 88 (1931), pp. 338, 360, 363, who mentions the terms wayang tengen (on the right of the dalang) and wayang kiwa (on his left).

344 Sylvain Lévi, Le théatre indien (Paris 1890), p. 122f., [Francesco Cimmini, Il tipo comico del Vidūṣaka (Napoli 1893), not accessible to me], J. Huizinga, De vidūṣaka in het indisch tooneel (1897), Montgomery Schuyler, "The origin of the Vidūṣaka and the Employment of this character in the plays of Harṣadeva", JAOS. 20 (1899), p. 333ff., R. Pischel, Die Heimat des Puppenspiels (1900), p. 17ff., A. Hillebrandt, "Über die Anfänge des indischen Dramas", Sitzungsber. d. kön. bay. Akad. d. Wiss. Philos.-philol. und histor. Klasse, Jahrg. 1914, 4. Abhandlung pp. 24–26, M. Winternitz, Geschichte der indischen Litteratur III (1922), pp. 171, 172 n. 2 (references), 176, 178 n. 1, J. Hertel, "Der Ursprung des indischen Dramas und Epos", WZKM. 18, pp. 59f., 137f., 24, p. 118, S. Konow, Das indische Drama (Berlin-Leipzig 1920), p. 14, A. B. Keith, The Sanskrit Drama (1924), pp. 39f., 50f. (references p. 51 n. 1) and passim (see index), D. R. Mankad, Types of Sanskrit Drama (Karachi 1936), p. 211ff., J. Gonda, "Zur Frage nach dem Ursprung und Wesen des indischen Dramas", Acta Orientalis 19 (1943), pp. 329ff.(especially pp. 395–417), [R. V. Jagirdar, Drama in Sanskrit Literature, Bombay 1947], [J. T. Parikh, The Vidūṣaka: Theory and Practice, Surat 1953], G. K. Bhat, The Vidūṣaka (Ahmedabad, 1959), passim, Henry W. Wells, The Classical Drama of India (Bombay, 1963), p. 14ff., Paul Thieme, "Das indische Theater" in Fernöstliches Theater (ed. H. Kindermann, 1966), pp. 38, 50, 53f., etc., Paul Horsch, Die vedische Gāthā- und Śloka-Literatur (Bern 1966), p. 343, H.-O. Feistel, Das Vorspiel auf dem Theater (no place, 1969), pp. 119f., 125, The same, "The pūrvarāṅga and the chronology of pre-classical Sanskrit theatre", Samskrita Ranga Annual VI (Special felicitation volume in honour of Dr. V. Raghavan, Madras 1972), pp. 7f., 14, S. Weeratunge, "Some Remarks on the Vidūṣaka and Imagery in Sanskrit Drama", Añjali (O. H. de A. Wijesekera Felicitation Volume, Peradeniya 1970), pp. 182–184. Cf. also Louis Renou, La recherche sur le théâtre indien depuis 1890 (Paris 1964), p. 29 n. 4, J. C. Heesterman, Spel der Tegenstellingen (Leiden 1964), p. 23 n. 39 and J. Clifford Wright, Non-Classical Sanskrit Literature (London 1966), p. 21.

drama", about which, however, nothing is known. It is, of course, possible to assume by way of hypothesis that such a popular dramatic art once existed but it is clear that further conjectures about its character necessarily lack any foundation. Such theories are particularly hazardous if they are construed on the analogy of the 18th century "Possentheater" in Germany. Some authors, indeed, went so far as to define the character of the hypothetical popular theatre as a mime (e.g. M. Schuyler 1899, p. 338, Hillebrandt 1914, p. 26, Konow 1920, p. 14f.). The theory of a popular drama, again, created the further dilemma as to whether or not this drama had had a religious origin, a question that Hillebrandt 1914, p. 27, left open.

It may be useful to state clearly that this whole scholarly discussion, which in German-speaking countries was dominated by the word "volkstümlich" as the vitium originis, was primarily based upon a questionable conception of the character of the vidūṣaka. The problems with which several generations have struggled, such as the question "popular or not?", "religious or non-religious?", "mime or something else?" are modern constructions, which can be ignored when an attempt is made to interpret the data of the Nāṭyaśāstra. It is characteristic that some modern scholars, while accepting some of the theories, felt compelled to make an important reservation as to the foundation upon which they were built. Thus Winternitz 1922, p. 172, remarks about the vidūṣaka "Im Nāṭyaśāstra wird er viel grotesker beschrieben, als er in unsern Dramen erscheint. In manchen der klassischen Dramen tritt das Groteske fast ganz zurück und die Freundestreue viel mehr in den Vordergrund".

Hillebrandt 1914, p. 27, had already observed that the description of the vidūṣaka in NŚ. 24.106 differs from his character in later texts such as Aśvaghoṣa's drama and rather resembles a kind of devil.

Thieme, p. 53345, has elaborated the theory of an original mime, in which the vidūṣaka did not take part in the action. The latter was not, accordingly, a dramatis persona but a member of the troupe, whose natural place can only have been in the "Vorspiel" (pūrvaranga). In the course of time, however, he became for unknown reasons mixed up in the play, as an actor who had to play his part. Since the approach of the present study is entirely different, there is no use in amply discussing here Thieme's ingenious theory. It has already been observed that the assumption of the mime has no foundation in historical data. It may be added as a general remark that in many modern studies on the Sanskrit drama a somewhat confusing use is made of the term "Preliminaries" or "Vorspiel", in so far as the introduction to the play (the āmukha or pratāvanā) is insufficiently distinguished from the pūrvaranga, which stood apart as a religious ceremony.

Feistel, p. 125, has elaborated Thieme's theory in a different way. He suggests that the introduction (āmukha) has arisen in imitation of the Trigata of the pūrvaranga. This is very well possible. According to

345 See also pp. 38, 40, 43, 50.

200

NŚ. 22.28 the āmukha consists of a conversation of the sūtradhāra with either an actress or the vidūṣaka or the assistant (pāripārśvika). The fact that the two characters of the Trigata here return with the same names, but now in a profane function, may be considered a strong indication that the āmukha was somehow a continuation of the preceding Trigata. This āmukha was not, however, a Three Men's Talk and the fact that one of the two characters of the Trigata could be used as a partner in the conversation of the Introduction may have been due to simple practical considerations. After the exit of the three as impersonations of the divine world and the sūtradhāra's renewed entrance as sthāpaka or prastāvaka, it was a convenient device to take one of the assistants as a partner in the dialogue (as is sometimes done in the prologues of Bhāsa's plays). There was, however, nothing to prevent an actress taking his place. It is hard to find in the facts an argument for Feistel's assumption that in the āmukha, once it existed as such, the actress was introduced as a new element. Still harder is it to conceive how the actress, after she had been admitted to the Prologue, could have ousted the vidūṣaka from it. The further hypothetical development, in which the vidūṣaka, after being ousted from the āmukha, became one of the three leading characters of the Indian drama, cannot well be visualized.

There is only one point in these theories which is rightly stressed by Thieme and others and which will have to be seriously considered in the context of this study. This is the absence of the vidūṣaka from the dramas with an epic subject-matter, the “genuine” nātakas ascribed to Bhāsa, and from one of Aśvaghoṣa's plays.

Against the current theory that the vidūṣaka originated in the popular mime the following arguments can be brought forward:

1. In the oldest Indian tradition, as found in ch. I of the Nāṭyaśāstra, the vidūṣaka, the nāyaka and the nāyikā are the only leading characters of the drama (1.96). Abhinavagupta's comment may here be quoted (I, p. 32 line 10) : pradhānapatrāṇi prthag rakṣaṇīyāni 'ty aha "nāyakam" iti "Because the leading parts are in need of a personal protection, (the author) says nāyakam etc." (= 1.96).

2. Ever since the oldest dramas extant the vidūṣaka is one of the dramatis personae. In two of Aśvaghoṣa's plays he acts already his traditional part. In the Śāriputraprakaraṇa he is the friend and companion of a monk, who is here the nāyaka (see n. 374).

3. The theory of a (non-religious) mime-of which no trace exists—346

346 See Keith (n. 344), p. 50f.: "The popular origin of the Vidūṣaka is obvious but the point is whether this origin is religious or secular . . . . It is manifestly unnecessary and illegitimate, when the descent of this figure from the Vedic literature is clear, to insist that it was borrowed directly from popular usage, for which there is no proof, but only conjecture". Similarly Gonda (n. 344), p. 403f.: "Warum, schliesslich, soll man . . . diese Typen als 'volkstümlich' bezeichnen, falls man mit diesem Adjektiv meinen möchte, dass sie anfangs zu profanen Bühnaufführungen und Volksspielen gehörten (von denen wir aber nichts wissen) und aus diesen in das

201

is based on the interpretation of the vidūṣaka as a clownish character. As we have seen, several scholars have already expressed their doubts about the correctness of this interpretation of his part in the classical drama. In the preceding sections it has been argued that in the pūrraraṇga the vidūṣaka is identical with the assistant who carries Varuṇa's golden pitcher in the Utthāpana (the second part of the pūrvarañga) and represents the nether world in the Trigata. In the latter his relation to the other assistant is structurally similar to that of the bhrṅgāradhāra to the jarjaradhara. The fact that as vidūṣaka he speaks Sanskrit (overlooked by Abhinavagupta ad 5.28 B vidūṣakareṣabhāṣācaro!) would seem significant.

4. The so-called grotesque character ascribed to the vidūṣaka in the Nāṭyaśāstra (Winternitz, p. 172, Thieme, p. 50, etc.) is mainly based on the description of his make-up in the Nāṭyaśāstra 347.

It is true that this text in some passages (to be discussed below) does refer to his comical (hāsya) appearance, but the reason for his strange appearance (described in the Nāṭyaśāstra in such detail as can hardly have been actually represented on the stage) will become clear from § 18. There must, indeed, have been a wide gap between the theory of the prescriptions of the Handbook and the practice of the dramatic performances. One indication is that according to the Nāṭyalocana 348 the vidūṣaka must wear a turban, like the ministers, chamberlains, big merchants and the chaplain, whereas the Nāṭyaśāstra prescribes that he must be bald.

Although it will be shown that the curious appearance of the vidūṣaka, as described in the Handbook, had not originally any connection with the stage nor aimed at a comic effect, it cannot be denied that in the classical drama his behaviour was intended to evoke laughter. Cf. Śāradātanaya, p. 289,2 vidūṣako 'pi sarvatra vinodeṣu 'payujyate. Terms, however, in which his clownish character is expressed, do not appear before the ninth century A.D. Two different lines of tradition can be distinguished, one which uses hāsyakārin, hāsyakṛt and similar expressions, and another which uses the word vaihāsika.

Of the material that has come to my notice the following instances may be quoted.

Opferrituell aufgenommen worden sind !" and Horsch (n. 344), p. 343 n. 1: "Das Fehlen älterer Volksstücke macht seine Verbindung mit dem Hanswurst der Volksbühne (s. Konow) rein hypothetisch".

347 See, e.g., Winternitz (n. 344), p. 172, Thieme, p. 50 and Bhat (n. 344), p. 21 : "The Vidūṣaka appears in the preliminaries of a dramatic representation, known as Pūrvaraṇga, and provides laughter by his appearance. talk and gestures. This role that the Vidūṣaka plays is irrespective of the drama proper, which may or may not have contained comic elements. Its purpose is pure fun; and it must have been all the more necessary, we may presume, if the drama were of a serious or exalted type. The role of the Vidūṣaka in the preliminaries of the drama is determined solely by the psychological and social expectation of amusement by way of laughter . . . ". 348 Quoted by S. Lévi, II, p. 70 from Rāghavabhaṭṭa's Arthadyotanikā.

202

hāsyakārin, etc.: circa 850 Rudraṭa, Śṛṅgāratilaka I.31 (41) kriḍāprāyo

vidūṣakah, svavapurveṣabhäṣādhīhāsyakārī svakarmavit (vv.11. ca karmavit,

ca narmavit), Kāvyālamkāra 12.15 vidūṣakah kriḍanīyakaprāyah, nijaguṇa-

yukto mūrkho hāsakarākāraveṣavacāh, circa 975–1000: Dhanamjaya,

Daśarūpaka 2.8 hāsyakṛ ca vidūṣakah and Dhanika's commentary:

hāsyakārī vidūṣakah, asya ca vikṛtākāraveṣāditvaṃ hāsyakāritvenāiva

labhyate (Schmidt: “Weil er der Spassmacher ist, so ergibt sich daraus,

dass er sein Äusseres, seine Kleidung usw. entstellt''), c. 1175–1250:

Śāradātanaya, Bhāvaprakāśa (GOS. p. 289) hāsyānūkavibhūṣitah [Bhat,

p. 48 "possessed of a funny sort of back-bone (that is, 'hunch-back')”],

c. 1300 (?)-1350: Viśvanātha, Sāhityadarpana 3.79 karmavapurveṣabhā-

ṣadyaih, hāsyakarah kalaharatir vidūṣakah syāt svakarmajñaḥ (cf. above

Rudraṭa), Pratāparudrīya I.38 hāsyaprāyo vidūṣakah, c. 1330 A.D.:

Śiṅgabhūpāla, Rasārṇavasudhākara I.92 vikṛtāngavacovair hāsyakārī

vidūṣakaḥ 349. As an anonymous quotation the last definition is found in

Kavīndracārya Sarasvatī's commentary Pācandrikā on the Daśa-

kumāracarita (c. 1650), p. 42 line 8 (ed. Agashe, B. S. S.), where the

Bhūṣaṇa reads aṅgādiveṣavaikṛtyahāsyakārī vidūṣakah and the Laghudīpikā

samdhānanipuṇaś caiva hāsyakārī vidūṣakah.

vaihāsika ("jester, buffoon, actor in general" Apte, Dict.; Böhtlingk

still quoted the lexicographers Halāyudha and Hemacandra as his only

authorities). As we now know, it was in rather common use in classical

Sanskrit. As a gloss for vidūṣaka it seems to have been used for the first

time in the Kāmasūtra I.4.46 ekadeśavidyas tu kriḍanako viśvāsyaś ca

vidūṣakah, vaihāsiko vā. In VI.1.22 the vaihāsika is mentioned side by side

with the pīthamarda "parasite": bhavajjñasārtham paricārakamukhān

samvāhakagāyana-vaihāsikān gamye tadbhaktān vā praṇidadhyāt, tadabhāve

pīthamardādīn. The gloss of the commentary vaihāsiko vidūṣakah raises

questions that will be discussed in § 19. Murāri, Anargharāghava, p. 146,

quotes the line vaihāsikah kelikarah prahāsī ca vidūṣakah, which he

attributes to Bharata (Lévi II, p. 19). On the great variety of verses

which are assigned to the Nāṭyaśāstra see in general S. Lévi, p. 27. Murāri

is generally supposed to have lived circa 900 A.D. The use of the word

vaihāsika would seem to exclude that this line was actually taken from

the Nāṭyaśāstra. On the other hand, a variant form vaihasika, not

registered in any dictionary, occurs in Agnipurāṇa 339.40 vidūṣako

349 See K. H. Trivedi, The Nāṭyadarpaṇa of Rāmacandra and Guṇacandra, A critical

Study (Ahmedabad 1966), p. 186 n. 3, who quotes this passage while summarizing

the data of the ND in the words: "In short the Vidūṣaka is ludicrous in physical

appearance, dress and speech". The Nāṭyadarpaṇa dates from about 1100–1150 A.D.

Sylvain Lévi, II, p. 19 quotes the same text from Rāghavabhaṭṭa's commentary

on the Śakuntalā, as taken from the Sudhākara. As for Śiṅgabhūpāla's date see

S. K. De, Sanskrit Poetics I2, p. 239, and for that of the Pācandrikā see

Wilhelm Printz, KZ. 44 (1911), pp. 100, 104. Of unknown date is Nityanātha's

Rasaratnāhāra 54, which has been quoted as saying "Der vidūṣaka erregt Gelächter

wegen der Entstehung seines Körpers, seiner Sprache und seiner Kleidung"

(Richard Schmidt, Beiträge zur indischen Erotik, p. 146).

203

vaihasikah. This is the reading in the text of Poona 1900 as well as in that of the Kashi Skt. Series 174, Varanasi 1966.

In contrast with these later theorists the Nāṭyaśāstra does not stress, as far as I can see, the clownish character of the vidūṣaka except in one passage. No importance, it seems, can be attached to the circumstance that this passage (13.137–142 C) is omitted in KM 350. It deals with gait (gati) and distinguishes three kinds of comic effect (hāsya), which are specified according to a distinction well-known in the Handbook as aṅga-vākya-krtam and nepathyajam “produced by corporeal defects, by words and by the costume and make-up”. This kind of distinction has become stereotyped since Rudraṭa and is also found in the discussion of success 351.

Apart from this passage I can only quote 27.8 C vidūṣakocchekakṛtam (?) bhavec chilpakṛtam tu yat, alihāsyena tad grāhyam prekṣakair nityam eva hi 352 and 27.61 bālā mūrkhā(s) striyaś caiva hāsyanepathyayoh sadā (viz. tuṣyanti) 353. The last passage can refer in general to dāsas and other characters, who have to be characterized by some deformity.

It is not necessary to discuss in this context the theory that the deformity of the vidūṣaka is connected with the superstitious belief that disfigured persons bring good luck 354.

After these preliminary remarks an attempt will be made to explain the status and the function of the vidūṣaka in classical drama. Only then his name may become clear. There is, it is true, a general agreement that vidūṣaka means “corrupter” (“Schlechtmacher, Schimpfer”) 355 but this etymology does not explain very much, as long as it is not clear what exactly he “corrupts”.

350 The passage occurs in B (12.137ff.) and KM1 (12.121ff.). In KM2 thirty lines have been omitted after 12.114, but neither the critical apparatus nor Kane, History of Sanskrit Poetics4, p. 13 mentions this gap in the text. It is, therefore, probably due to mere inadvertency.

351 While 27.3C mentions (siddhiḥ) dārirī vāñmayā tathā, 27.3B has vāñnaipathya-śarīrajā and 27.2 KM vāñmanahkāyaoambhūtā. See further below, p. 217.

352 The correct reading of the first word is very doubtful, see below, p. 216. For śilpa see Ghosh, Translation, foot-note: “For example, the art of comic make-up” and NŚ. 27.58, where śilpa occurs beside nepathya.

353 For this stereotyped group see ILJ. 16, p. 243 n. 10.

354 Gonda, pp. 410–416.

355 See in general Gonda (n. 344), p. 402f., K. H. Trivedi, The Nāṭyadarpaṇa, p. 187, and cf. PW. (“Verunglimpfer”), Konow, p. 14 (“Schimpfer”), Keith, p. 39 (“given to abuse”) and n. 2 on “ritual cathartic cursing”), Gonda, p. 416 (“Tadler, Schimpfer, derjenige, der jemand. oder etwas in den Augen anderer schlecht macht”), the vidūṣaka being “gewissermassen eine Verschmelzung des Hofnarstypus und des apotropäischen Phthonoswehrers”), A. N. Upadye, Candralekhā (Bombay 1945), Introduction, p. 26f. (“The name Vidūṣakaḥ is just a hyper-Sanskritic back-formation of Prakrit viuso or viusao [with k-suffix] which is to be connected with vidvas”), similarly J. C. Wright (n. 344), p. 21, whose remark on “the highly stereotyped figure of the pedant (Sanskritized vidūṣaka, i.e. vidūṣ, Pkt. vidū)” is based on a misconception of the character of the vidūṣaka, G. K. Bhat (n. 344), p. 88 (“one who has a characteristic mode of fault-finding, or spoiling, with a view to evoking laughter”). Thieme, p. 38, conjectured “Verpatzer” (gambling away).

204

As for his social status in general, several theories have been proposed,

all necessarily without any foundation in facts. It may be sufficient to

mention the not uncommon views that he was used by the dramatists

as a persiflage of the brahmins356, or that the court-jester had stood model

for him357. All such theories disregard the undeniable fact that actually

we know the vidūṣaka exclusively as a stage character. This cannot be

due to a deficient knowledge of urban life in ancient India. His purported

role in Indian society will be examined in § 19. Here we will first try

to answer the question as to what can be learnt about the status of

the vidūṣaka in the only surroundings where we meet him, that is, in Sanskrit drama.

Status in the drama. It is curious that seldom attention seems to have

been given to the fact that according to the Nāṭyaśāstra the vidūṣaka

is the only character on the stage who meets the king on an equal footing.

This follows as an inevitable conclusion from the circumstance that the

king and the vidūṣaka address each other as "friend" (vayasya), which

is a form of address that is exclusively permitted between people of

equal standing358. This throws a new and unexpected light on Abhinava's

statement that the nāyaka and the vidūṣaka are the two male leading

characters. Bhat, p. 183, points out that only in late dramas the strict

rule is disregarded and that the vidūṣaka here ranks with the servants,

who address their master as deva. The uniqueness of vayasya as a form

of address comes out best in the context of the whole passage359. It is

356 Cf., e.g., R. Pischel, Die Heimat des Puppenspiels (1900), p. 18 ("die Priester

aller indischen Religionen mit Vorliebe verspotten"); A. Hillebrandt (1914), pp. 24

and 26: "eine Verspottung der höchsten Kaste", H. Güntert, Der arische Weltkönig

und Heiland (Halle-Saale, 1923), p. 235: "Ist es doch bezeichnend, dass in den

späteren indischen Dramen die komische Figur fast immer ein Priester sein muss,

dessen Appetit meistens unvergleichlich viel mehr ausgebildet ist als sein Verstand",

W. Ruben, Kālidāsa, The human meaning of his work (Berlin, 1957), p. 19, etc.

357 Hertel (n. 344); but Winternitz, p. 172, n. 1, rightly objected that in no Old Indian

description of court life a court-jester is mentioned and that the vidūṣaka, like

the viṭa, rather belonged to Indian town-life (which point will be taken up below

in § 19). It should be noted here that rājya-vidūṣaka "court-jester" in Sāgaranandin's

Nāṭakalakṣaṇaratnakośa 2014 is a corrupt reading, see Raghavan's note in

Translation, p. 66 (Bābūlāl Shukla reads accordingly rājā vidūṣakam avadat, p. 194,

line 1 from the bottom). At the court of the Mughuls there seem to have been

professional jesters, as Bīrbāl was famous as the court-jester of Bābar in Delhi.

In Old Tamil literature, however, the professional jester seems to have been as

unknown as in Sanskrit literature. How the Telugu poet Tenāli Rāmakrṣṇa, who

lived under the kings Kṛṣṇadevarāya and Veṅkaṭāpatirāya (first half of the

16th century A.D.) has become in Tamil tradition the court-jester of Kṛṣṇadevarāya

I have been unable to find out (S. M. N. Sastri, Tales of Tenali-rama was unfortunately

not accessible to me). Dr. Kamil Zvelebil, to whom I owe the data on the Telugu

poet, referred to A. S. Panchapakesa Aiyar, Tenāli Rāma (Madras, 1947). See p. 226.

358 19.10 (17.72 KM) samāno 'tha vayasye 'ti.

359 19.16 deve 'ti nrpatir vācyo bhṛtyaiḥ prakṛtibhis tathā, bhoṭṭe 'ti sārvabhaumas tu

nityam pari janena hi (17) rājann ity ṛṣibhir vācyo hy apatyapratyayena vā, vayasya

rājann iti vā bhaved vācyo mahīpatibhiḥ (21) vidūṣakeṇa. rājnī ca ceṭī ca bhavatī 'ty api,

205

also significant for the ancient character of this stage convention that later theorists, such as Sāgaranandin (between 1200 and 1250?), did not understand it any longer and misinterpreted the theory: "A companion is called vayasyaka (friend), and this term is used also for the jester" 360. The Sāhityadarpana gives also a different version 361. See further Bhat, p. 106f.

Function in the drama. This leads us to consider the central problem: What exactly was the function of the vidūṣaka in the play?

Those who have studied his part, as it is known to us from the classical plays, without preconceived ideas about buffoons, have more than once stated that the vidūṣaka interferes with the plot, but often in a specific way. J. Huizinga, who wrote his doctor's thesis on the vidūṣaka, gave evidence of his perspicacity even in his early years, when he characterized the role of the vidūṣaka as follows: 362 "The vidūṣaka is a counter-stroke of the spirit of the play itself; the loftiness of the sentiments comes out the more strikingly by the contrast of his vulgarity". He points out that the vidūṣaka, while trying to help his friend, on the contrary gets him into trouble (p. 25): "His clumsiness manifests itself in his committing silly blunders in the very circumstances in which all depends on him - blunders with which he gets his friend into trouble, especially because he always lets his tongue run away with him". (See also pp. 23 and 128). In numerous situations he is scolded as mürkha or vaidheya "fool" 363.

In connection with an observation by Huizinga on the Mālavikāgnimitra Gonda, p. 406f., remarks: "... aber ich frage mich, ob es nicht zum Charakter eines wahren Vidūṣakas gehört, dass er die Sachen stört und verdirbt, und auch, ob gerade die Tatsache, dass die Lösung des Knotens ohne sein Zutun zustande kommt, nicht vollkommen im Einklang ist mit der Aufgabe des Vidūṣakas in diesen Stücken, die u.a. darin bestand,

nāmnā vayasye 'ty api vā rājñā vācyo vidūṣakaḥ (19) sarvastribhih patir vācya āryaputre 'ti yauvane, anyadā punar ārye 'ti vācyo rājñiyā 'pi bhūpatiḥ

360 Sāgaranandin, Nāṭakalakṣaṇaratnakośa, translated by Myles Dillon-Murray Fowler-V. Raghavan (Philadelphia 1960), p. 44. The text (NLRK. 2199) reads vayasyakaḥ sahacarah sa eva ca vidūṣakaḥ. Rañganātha in his commentary (dated 1655 A.D.) on Vikramorvaśī II.0.1 and the edition by Bābūlāl Śukla (Varanasi 1972, śl. 270) have the variant reading vayasyakaḥ cātupatuḥ sa eva ca vidūṣakaḥ; cātupatu or cāṭubāṭu lexicographers gloss with bhandaḥ "jester". According to Sāgaranandin the vidūṣaka belonged to the court servants (see 2206=270), which must be due to a late attempt of theorists to fix his position (see § 16). This author does not mention him as a leading character together with the nāyaka and nāyikā in 2170ff.

361 See SD. 6.145 rājarṣibhir vayasye 'ti tathā vidūṣakena ca, rājann ity ribhir vācyaḥ so 'patyapatyayena ca.

362 J. H. Huizinga (n. 344), p. 41: "De vidūṣaka is een tegen slag van den geest van het stuk zelf; de verhevenheid van de sentimenten komt door het contrast van zijn platheid des te sterker uit", p. 25: "Zijn onhandigheid bestaat daarin, dat hij juist in de omstandigheden, waar het op hem aankomt, onnozele blunders maakt, waarmee hij zijn vriend in ongelgenheid brengt, vooral doordat hij nooit zijn mond weet te houden". Cf. also pp. 23 and 138.

363 See Huizinga, p. 32, Indu Shekar (n. 344), p. 79.

206

dass er die von anderen in Gang gebrachten und geführten Vorgänge zwa,

förderte, aber zugleich einen allzu glücklichen Ablauf ein wenig hemmte

oder doch die grosse Freude des Helden über den schönen Verlauf der

Ereignisse einigermassen verleidete. Diesen letztgenannten, schon oben

erwähnten Charakterzug, die Neigung zum Stören von freudvollen und

begeisterten Stimmungen durch nüchterne und kleinliche Bemerkungen,

seine Tadelsucht, seine Platheit der Erhabenheit des Helden oder der

Situation gegenüber, finden wir auch wiederholt in anderen Stücken".

Bhat, p. 171, gives a similar characterization: "He, however, miscarries

them [viz. the tasks of the hero], or makes a mess of them, and often

creates complications for the hero". Elsewhere in his book, it is true, he

proposes an unacceptable explanation for this behaviour but the facts

referred to support Huizinga's description and deserve to be quoted

here (p. 106): "The Vidūṣakas enjoy the confidence of the heroes, who

turn to them with their secrets of love. And if Gautama blabbers the

secret of the king in his sleep, as does Vasantaka in Priyadarśikā; if

Māṇavaka loses the love-letter, as Maitreya allows the pot of ornaments

to be stolen : these are blunders affected for humorous turn, or for story-

development; and cannot be interpreted as a betrayal of the trust placed

by the heroes in the Vidūṣakas".

It is not surprising, in view of such characterizations of the vidūṣaka

as a leading part of the drama, that some connection has been sought

between his generical name and his function in the plot of the play. Mostly,

however, his name was taken to mean "scoffer, jeerer", which has long

been associated with the ritual scoffer who is known from the Vedic

ritual of the Mahāvrata. Those scholars, however, who have drawn

attention to this ritual as a possible forerunner of the classical drama,

have in general concentrated too one-sidedly on the abusive language used

in the ritual talk between a brahmacārin and a courtesan 364.

More instructive is the ritual contrast between the "praiser" (abhigara)

and the "reviler" (apagara). In this case the function of the reviler is

clearly formulated in the words (PB. V.5.13) "He who finds fault with

them, drives away their evil lot" 365 or (JB. III.356) "The ninditṛ takes

364 Thus Konow, p. 14, Keith, pp. 24f., 39. Gonda, p. 402f. criticizes this one-sided

approach. On p. 359 he discusses the abhigara.

365 The whole passage runs as follows: abhigarāpagarau bhavato, nindaty enān anyaḥ,

prā 'nyaḥ śaṁsati. ya enān nindati pāpmānaṁ eṣāṁ so 'pahanati, yaḥ praśaṁsati

vad evaśāṁ sustutāṁ tuṣkastamīṁ tat so 'bhigṛṇāti. See J. J. Meyer, Trilogie altindischer

Mächte und Feste der Vegetation (Zürich-Leipzig 1937), III, p. 247, Gonda, AO. 19

(1943), p. 352, J. C. Heesterman, WZKSO. XII-XIII (1968), p. 177, Pierre Roland,

"Le Mahāvrata" (Nachr. Akad. Wiss. Göttingen 1973-3), p. 67, and cf. TS. VII.5.9.3

ā 'nyāḥ krośati parā 'nyāḥ śaṁsati, yā ākróśati punar evāṛ 'nānt sā, yāḥ praśāṁsati

pūtésv evā 'nnādyāṁ dadhāti, JB. II.405 tayor ā 'nyāḥ krośati, prā 'nyaś śaṁsati.

ya ākrośati punar evaiṣā sa, atha yaḥ praśaṁsati pūtésv evaiteṣu sa indriyāṁ

viryaṁ dadhāti, KS. XXXIV.5 (39,1) abhigarāpagarau bhavataḥ, pra vā anyas

sattrināś śaṁsati, nindaty anyo. yaḥ praśaṁsati yad evaśāṁ suṣṭutāṁ suśastāṁ tat

sa praśaṁsaty, atha yo nindati yad evaṣāṁ sustutāṁ su(śa)śastāṁ [r: duṣṭ. duśś.]

tat so 'pahanti, TB. I.2.6.7 (ĀpŚS. XXI.19.10-11) imé 'rātsur imé subhūtám akramn

207

upon him the impurity of (the one he reviles)". In this connection it is

interesting to note that what is badly offered in the sacrifice is seized

by Varuna366. These words express more than a mere association of

Varuna and evil. They obviously imply that Varuna, like Nirti, is not

only dreaded but also functions as a divine scapegoat who can take away

evil and sin. This may also throw some light on a hidden aspect of

the brahmin-vidūsaka (see p. 222).

The importance of the Vedic "reviler" is primarily that it reveals in

a general way in what context the role of the "criticizer" in the drama

should be considered. It becomes more and more clear to what extent

the figure of the vidūsaka is rooted in Vedic religion. For a more direct

study of the part of the vidūsaka in the drama, however, we need not

rely on speculations on possible Vedic prototypes or "forerunners", since

the Nātyaśāstra contains more concrete data than any theory about

the prehistory of the drama can give us. Here we have to consider once

more the role of the vidūsaka in the Trigata.

In the definition of the Nātyaśāstra he is described as the "spoiler"

of the conversation of the assistant367. The assistant and the vidūsaka

are identical with the two assistants of the first part of the pūrvarañga,

ity anyataró brūyāt. imá udvāsikārīna ime durbhūtám akrann ity anyatarāh. yád

evaíṣām sukr̥tā́m yā́ rād dhíḥ tád anyataró 'bhíśr̥ṇāti, yád evaíṣā́m duṣkr̥tā́m yā́ 'rād dhíḥ,

tád anyataró 'pahanti (accentuation according to the Ānandāśrama ed.), Baudh.

ŚS. XVI.22 (p. 268,7) brāhmaṇas cā 'śūdraś cā 'dre carmakarte vyāyacchete. ime

'rātsur ime subhūtam akrann iti brāhmaṇa, ima udvāsikārīṇa ime durbhūtam akrann

iti vrsaló, brāhmaṇaś caṃhaiyati, naṣvati vrsaláh. Lāty. ŚS. IV. 3.1 brāhmano 'bhigarāḥ

pūrvasyā́ṃ sadasó 'vāri pratyaṅmukha upaviśet (2) vrsaló 'pagaro 'parasyā́m

prāṅmukhaḥ (3) sa brūyān nā 'rātsur ime sattrinā iti (4) arātsur ity abhigarāḥ . . .

(13) sarveṣā́m karmāṇi niṣṭhite tad evā 'bhigaras trír brūyā́t [commentary : arātsur iti]

(14) avasrjya śūdraḥ pradrāvet, Kāty. ŚS. XIII.3.6 antraved abhigarāpagarau (7)

ākrośaty ekāḥ, praśaṅsaty aparāḥ. For the possible historical relations between the

abhigarāpagarau and the opponents in the disputation of the later vāda-tradition,

see p. 191 and J. C. Heesterman, WZKSO. 12-13 (1968), p. 177.

366 According to JB III. 356 line 24 the apagara takes upon himself the impurity

of the one whom he reviles (śamalám asyá ninditā́ pratimunciate). In this connection

it may be suggested that the Sūtras which connect the reviling with the fight for

the symbol of the sun (see the preceding note) may reflect a later conception of

this ritual. It will be argued below (p. 222) that the vidūsaka had to be a brahmin

because as such he could take upon him the impurity of the sponsor of the dramatic

performance. The same is true of the vendor of the Soma plant and of the apagara.

I do not think, therefore, that the śūdra or vrsalá of the fight was

originally identical with the person of the apagara. The parallelism in this respect

with Varuna as the divine scape-goat is clear. According to ŚB. IV.5.1.6 Mitra

seizes what is well offered (svistam) of the sacrifice, and Varuna what is badly offered

(dúristam see Ch. II n. 260). The same role of Varuna can, it seems, be seen in the

formula which the priest mutters while accepting the dakṣinā, viz. varuṇas tvā

nayatu devi dakṣine (KS. IX.9, PB. I.8.2ff.) "Let Varuna lead thee, o divine

Dakṣinā". Heesterman, IIJ. 3 (1959), p. 243, rightly stresses the danger inherent

in accepting a dakṣinā: the priest, while accepting it, should turn away from it

(vyāvŕtya) and assign the gifts to various gods. This is expressly done in the two

pages referred to above, where each assignment is preceded by the formula quoted.

367 See p. 191 and the foot-notes 293 and 320.

who have been interpreted as standing for Indra and Varuṇa. As we have seen above, the highest authority in this vivāc368 declares the "assistant" victorious, which implies that he stands for Indra, who has to win the contest. The vidūṣaka must then be the impersonation of Varuṇa. There is, accordingly, a mythical identity of Indra = jarjaradhara = pāripārśvika on the one hand, and of Varuṇa = bhrṅgāradhāra = vidūṣaka on the other. It has already been pointed out in passing that the relation between these two mythical "complexes" contains the basic pattern for that between the nāyaka and the vidūṣaka in the play (see p. 193). These two are the leading parts and the nāyaka is protected by Indra. That their relation was a curious one has long been noted by modern scholars. If we disregard the protector of the vidūṣaka for the reasons stated above (p. 176), the two groups of equations are Indra = jarjara-dhara = pāripārśvika = nāyaka and Varuṇa = bhrṅgāradhāra = vidūṣaka. In other words the relation between pāripārśvika and vidūṣaka in the Trigata is basically the same as that between nāyaka and vidūṣaka in the play. The structural approach thus throws a new light on the so-called "Freundestreue" of the vidūṣaka and shows that in the relation that must originally have existed between the hero and his "friend" there was in fact an element of contest.

The Vedic verbal contest between Indra and Varuṇa must have taken place at a time when Varuṇa was still the representative of the primordial world of the Asuras. Their vivāc was an aspect of the strife between the two cosmic powers, out of which the ordered world was to be born. The contest was in general considered a creative act, as is beautifully symbolized in the myth of the Churning of the Ocean. Held has pointed out that the creation of the cosmos is an act in which both Devas and Asuras participate. By taking turns in pulling the rope that is wound around the churn-staff they perform the two cosmic movements of pravṛtti and nivṛtti, as a result of which the primordial waters (later called the Milk Ocean) produce the goods of life369. In Ch. I, we have already met with this combination of cooperation and contest between gods and Asuras, who need each other and yet must necessarily fight each other in order to promote the process of life370. The analysis of the Trigata leads to the conclusion that the same ambiguous relation must originally have existed between the nāyaka and the vidūṣaka. It is Huizinga's merit to have detected traces of this relation even in the behaviour of the vidūṣaka in the classical drama and to have divined that here the most essential and characteristic trait of his character has been preserved. He was a "tegenspeler", a fellow-player and "counter-actor".

It is inevitable, in view of the current opinions on Varuṇa and the vidūṣaka, that the idea that the latter is an impersonation of the former

368 See n. 321.

369 See G. J. Held, The Mahabharata, pp. 171, 306.

370 Cf., e.g., Dhanika ad DR. 3.64 teṣāṃ ca phalāni prthak prthag bhavanti, yathā samudramanthanena Vāsudevādīnāṃ Lakṣmyādilābhah.

must appear wildly grotesque and absurd. This is due, on the one hand,

to a misinterpretation of the vidūṣaka as a mere buffoon and, on the

other hand, to a confusion of a superhuman cosmic order (Ṛtá), of which

Varuṇa was the guardian, with human ethics. In the Vedic conception

one who unwittingly transgresses the cosmic order is punished for a sin

which can only be understood metaphysically, not ethically. Varuṇa is a

majestic god but in his own way, not as many great Vedists of the 19th

and 20th centuries used to depict him, because he impersonates the

ambiguity of cosmic life, in which death is comprised. Just as no pravṛtti

can be conceived without its negative counterpart of nivṛtti, so the nāyaka

can only be understood in relation to his counterpart, the vidūṣaka. The

basic flaw in many studies on the vidūṣaka has been that he was considered

separately, not in the interrelation which constitutes his real character.

This also provides an answer to the fundamental question, which, however,

seems never to have been asked: Why was it that this purportedly

grotesque and comical character is stated in our oldest source to have

been the leading part of the drama, on a par with the hero? This fact,

which none of the existing theories can explain, becomes clear when it

is realized that in the interplay of action and counteraction his relation

to the nāyaka was that of Varuṇa to Indra. For basically the action of

the oldest Sanskrit drama must have been a reiteration of the cosmogony

and a ritual act aiming at a renewal of life.

It is instructive to compare the ambiguous character of the vidūṣaka

with that of the “opponent” of the nāyaka, the “counter-nāyaka”. From

the handbooks on dramaturgy we know that in the Īhāmṛga a pratināyaka

stood over against the nāyaka. According to the commentaries this term

referred to the “hero of the opposite party” (patipakṣanāyaka) in plays

where, e.g., Rāma and Rāvaṇa, or Yudhiṣṭhira and Duryodhana were

opposed to each other371. This theoretical distinction does not seem

have been very important in practice.

17. DRAMAS WITHOUT A VIDŪṢAKA

From an analysis of chapters I and V of the Nāṭyaśāstra it has been

inferred that in an older form of the Sanskrit drama the nāyaka and the

vidūṣaka were the leading parts372. There is a more or less general

agreement on the older parts of the Nāṭyaśāstra dating back to a time

371 See DR. 2.9 (atha pratināyakah) lubdho dhīroddhatah stbdhah pāpakṛd vyasanī

ripuh, and Dhanika’s commentary: tasya nāyakasye 'ttham bhūtaḥ pratipakṣanāyako

bhavati, yathā Rāmayudhiṣṭhira yo Rāvanaduryodhanau. Cf. also the definition of the

Īhāmṛga, DR. 3.73 naradivyād aniyamān nāyakapratināyakau, khyātau dhīroddhatāv

antyo viparyāsād ayuktakṛt. The idea that the pratināyaka's wrong deeds are due

to “error” (viparyāsād, Dhanika: viparyayajanānād) introduces a novel element.

SD. 6.246 changes this into anyo gūḍhabhāvād ayuktakṛt.

372 Thieme, who states that “das Nāṭyaśāstra einerseits von einer 'Rolle' des vidūṣaka

im Schauspiel nichts sagt” must have overlooked such data as the term of address

vayasya (see n. 359), which presupposes scenes in which the vidūṣaka and the king

were both on the stage.

210

anterior to the oldest extant dramas, including the fragments of Aśvaghoṣa's 373. In view of this fact and of the theory that the combined action of vidūṣaka and nāyaka reflects a certain structure of Vedic mythology one is led to assume that all old dramas must needs have had a vidūṣaka. The tradition which required that he should be one of the leading characters is, indeed, faithfully maintained in one of the oldest dramas, fragments of which have come down to us, viz. Aśvaghoṣa's Śāriputraprakaraṇa, which Lüders has in a masterly way reconstructed from the central Asian fragments. The vidūṣaka is here the friend and companion of a mendicant monk, and he speaks words which according to the Buddhist tradition have been spoken by Maudgalyāyana. His character is not without dignity but on the other hand his action results in checking and retarding the action of the nāyaka 374. In Bhāsa's and Kālidāsa's dramas the vidūṣaka was "ein kluger Berater des Königs" 375.

From the dramatic literature that has been handed down to us, however, it has rightly been concluded that the vidūṣaka only occurs in the Prakaraṇa, a genre often characterized as "bürgerliches Drama". He does not occur in the nātaka, which takes its subject-matter from the epics. Since far-going theories have been founded on this fact, it may be well to repeat that, as far as we can see, the Sanskrit literary drama has always been a form of court art. This is as true of the Prakaraṇa as it is of the Nāṭaka. The difference between the two genres was according to the Nāṭyaśāstra (20.50-51) one of plot and characters. The theory that on account of these characteristics the Prakaraṇa must have originated in a hypothetical popular theatre is unfounded 376.

In the following discussion data of the theorists about other genres, such as the Dima, the Īhāmṛga, etc. will be left out of consideration as they have only theoretical importance.

The vidūṣaka does not occur in the following dramas:

a) the Samavakāra (see Abhinavagupta ad 1.97).

b) the Rāma- and Kṛṣṇa-dramas of Bhāsa.

373 See, e.g., Thieme, p. 35, Feistel, p. 137f.

374 See H. Lüders, Philologica Indica, p. 208. To Lüders the introduction of the vidūṣaka in such a drama was absurd: "An und für sich ist natürlich die Idee, einem nur der Hoffnung auf Erlösung lebenden Bettelmönnich–denn das war Śāriputra schon, ehe er in den buddhistischen Orden trat–die lustige Person zum Gefährten zu geben, absurd. Wenn es doch geschieht, so beweist das, dass die Verbindung des Vidūṣaka mit dem Helden, er mag sein wer er will, zu Aśvaghoṣas Zeit schon ein so festes Gesetz der Bühne war, dass der Dichter sich ihm nicht entziehen konnte".

375 See Konow, p. 62.

376 For an ample and excellent discussion of the origin of the prakaraṇa see Thieme, p. 38, where it is derived from the mime, and, e.g., p. 81, where it is characterized as "die Gattung des auf die literarische Ebene gehobenen Volksstücks", which is older than Aśvaghoṣa. Konow, p. 43, sums up the arguments for the existence of a mime: "Es kann nach alledem nicht zweifelhaft sein, dass sie wirklich Mimen waren, obgleich wir uns von ihrer Kunst keine vollständige Vorstellung machen können" (cf. pp. 14, 15, 47).

c) the Mudrārākṣasa.

d) Bhavabhūti's dramas.

e) the shadow-plays (chāyānāṭakas) Mahānāṭaka and Dūtāṅgada.

Of these five categories the last three can be ignored for various reasons377. What remains is the Samavakāra and the epic dramas. The Samavakāra has been discussed above, where it was pointed out that here Indra and Varuṇa must have appeared without disguise. In its structure the Samavakāra with its twelve nāyakas was nearer to the Dima with sixteen nāyakas378.

The Nāṭaka was the typically heroic drama. Kālidāsa has been the first to introduce a vidūṣaka in this genre, who had up to that time been excluded from it. The subject-matter of these plays is characterized as mythical, in contrast with the Prakaraṇa. To a certain extent this is correct but it may give rise to false impressions and misunderstandings.

There is a wide gap between the Vedic myth and the epics. Even though the war between the Pāṇḍavas and Kauravas, as told in the Mahābhārata, is a replica of the fight between Devas and Asuras379, Yudhiṣṭhira and Duryodhana380 are not, from a mythological point of view, of prime interest. The arguments brought forward in favour of the theory that Arjuna stands for Indra, and Dhṛtarāṣṭra for Varuṇa, seem to me convincing381. Besides, the so-called epic dramas do not seem to date back to a remote past. In any case, those of “Bhāsa's dramas” which are based on the Mahābhārata, viz. Madhyamavyāyoga, Dūtaghaṭotkaca, Pañcarātra, Ūrubhaṅga, Karnabhāra and Dūtavākya deal in fact with legendary tales, not with the old mythology, traces of which are even in the epic hardly perceptible. Hence the absence of the vidūṣaka is not surprising382.

377 Viśākhadatta's Mudrārākṣasa is in more than one respect quite unique in dramatic literature: although its author introduces it as a nāṭaka, it has some characteristics of a prakaraṇa and its plot can be said to be based on two nāyakas (not a pratināyaka and a nāyaka !) without a nāyikā. The severe virarasa prevented the introduction of a vidūṣaka, which is not striking in the most uncommon of classical dramas. In the case of Bhavabhūti there was apparently “eine persönliche Abneigung” against the vidūṣaka (Lüders, Philologica Indica, p. 428). The shadow-plays Mahānāṭaka (11th cent.) and Dūtāṅgada (13th cent.) are too late to prove anything in this connection.

378 See for the first 20.65, DR. 3.64, SD. 6.235, and for the second 20.88, DR. 3.58, SD. 6.243.

379 See the references in Numen 8 (1961), p. 40.

380 Quoted by Dharmika, see n. 371.

381 For Arjuna see W. Caland, “De incarnaties van den god Vishnu”, Mededeelingen van het Provinciaalsh Genootschap voor Letterkunde (Utrecht 1927), p. 27. Cf. Mhbh. V.58.11, where Arjuna and Kṛṣṇa are called indrāviṣṇusamau. In a more general sense, however, all five Pāṇḍavas are impersonations of Indra (: 32, 28 etc.), III.254.21 Pārthāḥ pañca pañcendrakalpāḥ, IV.19.18 yūyam indrasamāḥ sadā, V.24.8 Pāṇdoḥ sutāḥ sarve 'ndrakalpāḥ, V.33.103 Pāṇdoḥ putrāḥ pañca pañcendrakalpāḥ. For the equation of Varuṇa and Dhṛtarāṣṭra see Gösta Johnsen, IIJ. 9 (1966), pp. 245–265.

382 For some of these dramas it is moreover doubtful whether they are nāṭakas or vyāyogas. See Konow, p. 52f., M. M. Ghosh, Translation I2, p. 367 n. 2, Thieme, p. 74.

212

The same is true of 'Bhāsa plays' based on the Rāma and Krṣṇa legends. In the Pratimānāṭaka, and Abhiṣekanāṭaka, which belong to the Rāmaic tradition, and in the Bālacarita, taken from the Krṣṇa-tradition there is nothing to suggest that we have an old type of dramas before us. It is even possible that in these dramas the development of the Nāṭaka from the recitation of the epic (according to Lüders's theory)383 can still be seen. Renou stressed the fact that the classical drama is principally based on Rāma worship384. In dramas based upon Vālmīki's Rāmāyaṇa, however, there was very little room for a vidūṣaka. His absence, far from proving anything about the origin and prehistory of the Sanskrit drama385, would rather seem an indication of the comparatively late rise of the Nāṭaka as a genre.

On the other hand, the circumstance that traditionally the vidūṣaka belonged to the Prakaraṇa and the kathā-dramas leads to the unexpected conclusion that as a genre the Prakaraṇa was probably older386 than the Nāṭaka. This leaves several questions open, which cannot be answered for lack of evidence.

18. vidūṣaka AND JUMBAKA

Since the Nāṭyaśāstra only seldom refers to the comic effect of the costume and make-up of the vidūṣaka (see p. 202), it is remarkable that there existed a detailed and fixed tradition about the corporeal particularities which characterize the vidūṣaka387. References to this tradition are to be found in various parts of the Nāṭyaśāstra, which leads

383 Lüders, Philologica Indica, p. 427: 'Dadurch, dass man das Wort dem Rezitator nahm und den bis dahin nur durch die Geste wirkenden Figuren selbst in den Mund legte, ist das Nāṭaka im engeren Sinn entstanden'.

384 See Renou, La recherche sur le théâtre indien depuis 1890, p. 39. He, however, also leaves the possibility of a Krṣṇaite origin open (pp. 30 n. 2, 39).

385 Cf. Lüders, Philologica Indica, p. 428: 'Glücklicherweise sind wir jetzt in der Lage, mit Bestimmtheit behaupten zu können, dass das episch-mythologische Drama in vorklassischer Zeit den Vidūṣaka noch nicht aufgenommen hatte'.

386 For the theory of its originating in the popular mime see above, n. 376.

387 G. K. Bhat, The Vidūṣaka (Ahmedabad 1959), pp. 48-52 gives a survey of the particulars of the appearance of the vidūṣaka as described in the Nāṭyaśāstra and found in the classical dramas: 'Later theorists do not mention anything about the appearance of the Vidūṣaka except Śāradātanaya who repeats Bharata's adjectives, that the Vidūṣaka is bald-headed, red-eyed, possessed of a funny sort of back-bone (that is, 'hunch-back') and adds that his hair is tawny and he has a beard that is either yellow or green in colour'. This passage in the Bhāvaprakāśana (p. 289, lines 5–7, dating from between 1175 and 1250 A.D.) runs as follows: vedavin narmavedī yo netuḥ sa syād vidūṣakaḥ, khalatiḥ piṅgalākṣaś ca hāsyānukā vibhāṣitaḥ, piṅgakeśo hariśmaśrur nartakaś ca vidūṣakaḥ. It should be added that Dhanika in explanation of DR. 2.9 hāsyakṛc ca vidūṣakaḥ remarks asya ca vikṛtākṛtiveṣāditvam hāsyakāritvenai 'va labhyate. SD. 3.42 mentions his appearance in a very general way: 'causing laughter by his actions, his figure, his make-up and his speeches' (karmavapurveṣādyair hāsyakaraḥ). The term vikṛtākṛta used by Dhanika occurs in NŚ. 31.17 with reference to the servants (dāsabhūmi).

to the conclusion that it must have been part of an old collection of prescriptions which have been compiled in Bharata's Handbook.

As we have seen above (p. 204), these descriptions have generally been interpreted as primitive means to achieve a comic effect, although the Nātyaśāstra only twice refers to this effect. Seldom has due consideration been given to the practical consequences of the fact that the Handbook distinguishes between three kinds of comic effects, viz. those due to corporeal defects, to words and to the costume and make-up of the vidūṣaka (p. 205). The first of these three categories, however, requires an entirely deformed person for the part of the vidūṣaka. There can be no doubt that in the classical drama this aspect has been ignored, but if so, the basic question that calls for an answer is : Why does the Handbook consistently stress such strict requirements which are only understandable in the case of servants (p. 221)?

Since the Nāṭyaśāstra constantly deals with this practice, and goes into the most minute details of how movements, feelings, etc. should be represented on the stage, it is inconceivable that it should have given such impracticable prescriptions unless there were absolutely compelling reasons to lay down these strict rules, in the face of the obvious impossibility of fulfilling them in practice. It is the object of this section to inquire into these reasons. To this end it will be necessary to quote the exact wording of the Sanskrit passages.

The first passage occurs in chapter XIII, which deals with the various gaits. Immediately after a passage which describes how the gait of lame men, cripples and dwarfs is to be represented the text has the following verses :

vidūṣakasyā'pi gatir hāsyatrayasamanvitā

(138) aṅgavākyakṛtam hāsyam hāsyam nepathyajasam tathā danturāḥ khalatīḥ kubjaḥ khañjaś ca vikṛtānanāḥ

(139) ya idṛśaḥ praveśaḥ syād aṅgahāsyam tu tad bhavet yadā tu khagavad gacched ullokitavilokitaiḥ

(140) atyāyatapadatvāc ca aṅgahāsyo bhavet sa tu vākyahāsyam tu vijñeyam asaṃbaddhaprabhāṣanāt388.

388 13.137–139 C., 12.137–138 B., 12.121–122 KM1 (omitted in KM2, see above, p. 204. Variant readings: yadīdrśaḥ KM1 for ya idṛśaḥ ; smṛtam and -aṅgakāvyakṛtam for tathā and aṅgavākyakṛtam C. Cf. also 12.142 B (125 KM1) kāvyahāsyam for vākyahāsyam, which must be the correct reading: "vākya is the text or song" (Raghavan, note on NLRK., p. 65), whereas kāvya is "drama, play", cf. 27.24, 42, 43, 44. Cf. 27.58 yad yasya śilpam nepathyam karma vāk ceṣṭitam tathā, tasya tenaiva tat sādhyaṁ svakarmaviṣayādṛśrayam, 13.191 strī pumāṁsam tu abhinayed veṣavākyaviceṣṭitaḥ and Siṅgabhūpāla vikṛtāṅgavacoveṣair hāsyakāri (see p. 203).

214

however, he walks like a bird, looking up and down, he is (also), owing

to his taking excessively wide strides, laughable due to his limbs.

Laughter due to words is the effect of incoherent talk . . ." etc.

The last words are reminiscent of the asaminbaddhakatathāprāyāṁ . . .

kathanikāṁ (5.138) of the vidūṣaka in the Trigata (p. 178 n. 294).

In a brief characterization of the various parts of the sūtradhāra,

pāripārśvika, viṭa and śakāra, the vidūṣaka is described in a similar way:

vāmano danturaḥ kubjo dvijihvo vikrtānanah

khalatiḥ piṅgalākṣaś ca sa vidheyo vidūṣakaḥ 389

"The vidūṣaka must be represented as a dwarf with protruding teeth,

hunch-backed, double-tongued, with a distorted face, bald-headed and

yellow-eyed".

In Suśruta the double-tonguedness is a disease. It may be conjectured

that here it refers to a stuttering speech, similarly to Dutch zijn tong

slaat dubbel "he speaks thickly, in a thick voice", but this must remain

a guess. See above, p. 193.

Two further passages must be mentioned. The first runs as follows:

vidūṣakasyā khalatịḥ syāt kākapadam eva ca 390

"The vidūṣaka is bald-headed and has lines in the skin (? of his face?)"

and the second:

pratyutpannapratibho narmakṛtair narmagarbhanirbhedair

chedavibhrāntavadano vidūṣako nāma vijñeyah 391

"A vidūṣaka is ready-witted by disclosures made as a pleasantry

(narma 22.50) or a covert pleasantry (narmagarbha 22.53) and has

stripes of colour (?) put on the face".

The last verse is for several reasons of dubious value. It appears as a

second definition (35.92-93) after the traditional definition, which is

given a few lines before (35.79), it has further been handed down in many

variant readings and, finally, the meaning of cheda- (a variant reading

of KM), which has here been translated in accordance with Bhat

(The Vidūṣaka, pp. 49, 102 n. 16), is quite uncertain.

389 35.79 C1. Similarly 24.103 KM2 (dvijanmā) and ibid., p. 655, interpolation after

35.20 (dvijihvo), 35.57 K. (v. l. dvijihvo according to Bhat, The Vidūṣaka, p. 48 n. 1).

390 23.151. Similarly 22.126 KM1 (khalatīḥ), whereas 21.155 B., 21.148 KM2, 23.148 K.

read vidūṣakasyā kartavyam khallikākāpadam (khallī kākapadam) tathā. Cf. Abhina-

vagupta kākapadam: kākapakṣavad yatra keśavicchedah and see Bhat, p. 49.

391 35.93 C1. While 35.25 KM (both editions) reads narmakṛtā, C. has narmakṛtair,

-nirbhedai(ś ca), (B. narmakṛto -nirbhedah) and cheko vidūṣitavacano, which might

be connected with (saṁjalpa) vidūṣakavidūṣitah in 5.140 (138 KM2), see the foot-notes

293 and 320. The Kāśi edition (35.71 K) has nirbhedyah yas tu vibhūṣitavacano . . .,

see Bhat, p. 102 n. 16. His assumption of a word narmakṛti with the same meaning

as narmakṛti is unacceptable.

The same variant reading also occurs in 27.8

vidūṣakocchekakrtaṁ bhavec chilpakrtaṁ ca yat

atihāsyena tad grāhyaṁ prekṣakair nityam eva hi

Ghosh's translation "Laughter created by the bragging of the Jester

and by some artifice (śilpa) should be received by the spectators with

an Excessive Laughter (atihāsya)" is based upon his conjecture utseka

"haughtiness, pride". Although the v.l. uccheda (27.8 KM1,2, 27.8 B1,

which reads tu for hi, 27.20 K [?]), seems to occur in some MSS., it cannot

be correct, since uccheda means "cutting off". Abhinavagupta reads cheda,

which according to him means "irony, wit, repartee" (vidūṣakacchedakṛtam iti cchedo 'tra vacanabhaṅgī', B. III1, p. 310 line 5). This interpretation is

not, as far as I can see, supported by any other evidence. If it is correct,

it would be possible to conjecture for 35.25 KM. chedavidhūṣitavadano

vidūṣako nāma vijñeyaḥ the emendation chedavidūṣitavacano "who criticizes

the other's words by his witty interruptions" (see n. 391). This would

come near to Ghosh's reading (conjecture?) cheko vidūṣitavacano vidūṣako

etc. (35.93) but the first word is probably not correct, since cheda vs. śilpa

seems a contrast like veṣa-vākya, vaco-veṣa (see n. 388).

As we have seen above (p. 202), there must have been a wide gap

between theory and practice. In the classical dramas there are no

indications to show that the stage-managers cared much for these rules

of the Handbook392. In practice the prescriptions about the aṅgakṛtaṁ

hāsyam, in so far as they referred to deformity, seem to have remained

a dead letter, and the actor apparently confined himself to choosing a

costume and make-up that were appropriate to his part. Along with the

traditional theory also this practice has found expression in our text: the

locus classicus for the outward appearance of the vidūṣaka in ch. XIII,

the beginning of which has been quoted above, then goes on to describe

the nepathyajaṁ hāsyam in the following words: "One who is dressed in

a tattered cloth or in a skin, or smeared with soot, ashes or red chalk,

such a man, O brahmins, will cause laughter owing to his costume and

make-up"393. The text states expressly that it is left to the actor, after

carefully considering the character (prakṛti) he has to represent, to adapt

his bhāva ("state", in Ghosh's translation) to it. Nowhere else is the gap

between tradition and reality so manifest as here. The author begins

with two different definitions of the comic effect achieved by the "limbs".

The first is obviously an old mnemotechnical verse, which is quoted with

some variation in several places of the Handbook but has, in fact, hardly

any bearing upon the gait. It is purely theoretical and impracticable and,

therefore, complemented by a second definition, which can, indeed, be

392 Cf., e.g., Bhat, The Vidūṣaka, p. 50: "The Sanskrit dramatists tacitly assume

the peculiar appearance of the Vidūṣaka and do not bother to mention it in their

texts. And when they do so, they refer not always to the attributes given by the

theorists, but to the general appearance of the Vidūṣaka only".

393 13.141 (12.141 B., 140 K., Grosset) ciracarmamaṣibhasmagairikādyais tu maṇḍitaḥ,

yas tādrśo bhaved vīprā hāsyo nepathyajas tu saḥ. Bhat, p. 50, points to this passage.

216

realized by gait, but only incidentally. In reality the vidūṣaka must have achieved the comic effect that was expected from him mainly by means of his talk (which was part of the vākya, the literary text) and his make-up.

A further argument for the purely theoretical nature of the traditional prescriptions is that none of the later theorists mention them except Śāradātanaya (c. 1175–1250 A.D.) 394.

In view of some modern studies in which the appearance of the vidūṣaka has been connected with deformity 395 it may be noted that the Nātyaśāstra never classes him with deformed people. It distinguishes the gait of cripples and dwarfs (13.131–136) from that of the vidūṣaka (13.137–140) and in the description of the accompaniment of the orchestra (vādyavidhāna) 396 it makes a distinction between the way in which the walking movements of the lames, the cripples, dwarfs, hunch-backs, etc. are accompanied 397, or the walking of yatis, munis, Pāśupatas and Buddhists 398, and the accompaniment of the walking of “the vidūṣakas, the nirmundas, servants and eunuchs” 399, accordingly a group of court-servants. As far as his corporeal characteristics are concerned, the vidūṣaka has, indeed, more in common with the servants 400 than with the cripples.

There is, however, another figure with whom the vidūṣaka has particularly striking features in common and who shows that the curious description of the vidūṣaka, inexplicable in a handbook for dramatic performances, is also rooted in an old Vedic tradition. It is the Vedic jumbaka 401, on whose head, at the end of the horse sacrifice and in conclusion of the purificatory bath, an offering is made. The similarities between the

394 Cf. Bhāvaprakāśana (GOS. No 45, p. 289), quoted in n. 387.

395 E.g., Gonda, Acta Or. 19 (1943), pp. 410–416.

396 33.232+C. (p. 176 line 14). Very corrupt in KM 2, p. 642 line 3. [Cf. 34, p. 486,23B].

397 khañjavikalavāmanakubjādīnāṁ pravṛttagamanēṣu.

398 yatimunipāśupataśākyādīnāṁ liṅgināṁ ca.

399 vidūṣakanirmundaposthāyakavarsadharādīnāṁ. For nirmuṇḍa cf. 34.77 aupasthāyika-nirmundān strīdṛnāṁ preṣanakarmani, raksane kumāriṇāṁ ca bālikānāṁ ca yojayet (the second line is omitted in 24.52 KM.) “And the Aupasthāyika-nirmundas are to be employed in escorting women, and in guarding maidens and girls” (Ghosh). They live in the harem, just as the snātakas, kañcukīyas and varṣadharas (34.72) “have no sexual knowledge” and are hermaphrodites: 34.80–81 napuṁsakā ye puruṣāḥ stribhāvena varjitāḥ, nirmundā nāmato jñeyāḥ kāmavijñānavarjitāḥ definition “one who is completely impotent and idle” (K. H. Trivedi, The Nātyadarpana of Ramacandra and Guṇacandra, Ahmedabad 1966, p. 188) seems only approximately correct.

400 See above n. 387 and cf. 35.16 mantharami vāmanami kubjami vikṛtami vikṛtānanam, viṣṭabdhanetrami kānākṣami sthūlami cipitanāsikam (17) durveṣami duḥsvabhāvami ca vikṛtākāram eva ca, dāsabhūmau prayuñjīta budho [k]asamiyutam. The reading dāsāṅka- C. ex conjectura: KM. has dāsāṅga- and further miṣṭabdhanetrami, durjanami (for durveṣam), whereas K (according to C and KM.) reads viṣṛtanetrakāryāṇāṁ sthūlami vihanurāsikam, durveṣi etc. It is not surprising that Huizinga (n. 342), p. 62, who had only the corrupt first edition of the Kāvyamīlā Series at his disposal, thought that these lines referred to the vidūṣaka and that one or more lines had been lost.

401 A variant form Jumbuka, Cumbuka is only found in Hiraṇyakeśin.

217

descriptions of the vidūṣaka and the jumbaká are the more interesting because the latter impersonates Varuṇa 402.

In the Śatapatha Brāhmaṇa the relevant episode of the purificatory bath (avabhr̥thá) is described in the following words: 403 “With ‘to Jumbaka hail!' he offers, at the purificatory bath, the last oblation; for Jumbaka is Varuṇa: by a sacrifice he thus manifestly redeems himself from Varuṇa. He offers it on the head of a white-spotted 404 bald-headed (man) with protruding teeth 405 and reddish brown eyes 406 ; for that is Varuṇa's form : by (that) form (of his) he thus redeems himself from Varuṇa” (translation by Eggeling).

Kātyāyana in his Śrautasūtra adds that the jumbaká is standing in the water, and Baudhāyana, that he must dive naked into the water. Most detailed is Śāṅkhāyana's description, according to which the jumbaká is taken to a river, where people make him descend into the water and make an offering on his head at the moment the water has reached his mouth. Some authorities add that he must be a brahmin of the gotra of the Ātreyas (Baudhāyana, Hiranyakeśin, Śāṅkhāyana), who according to Śāṅkhāyana must have been bought for a thousand (pieces of gold) 407.

402 See for the Jumbaka especially K. F. Johansson, Über die altindische Göttin Dhiṣaṇā und Verwandtes, Beiträge zum Fruchtbarkeitskultus in Indien (Skrifter utgifna af K. Humanistiska Vetenskaps-Samfundet i Uppsala 20: 1), Uppsala 1917, p. 126f., Lüders, Acta Orientalia 16 (1938), pp. 142-145. Cf. also Geldner, Vedische Studien 2, p. 292 n. 2, Eggeling, SBE. 44, pp. XXII, XXXIXf., Hillebrandt, Ved. Myth. II 3, p. 227 n. 3.

403 ŚB. XIII.3.6.5 jumbakāya svāhé 'ty avabhr̥thá uttamám áhutim juhoti, Váruṇo vái jumbakáh. sākṣād evá Váruṇam áva yajate. śukláya khalātéḥ viklidháya piṅgākṣásya mūrdháni juhoty, etád vái Váruṇasya rūpám. rūpénaí 'vá Váruṇam áva yajate.

404 śuklá 'hellfarbieg' (Lüders, Acta Or. 16, p. 134). The word seems to refer to a fair complexion, cf. TB. III.4.1.17 (Puruṣamedha) sāryāya haryakṣám ... ahné śuklám piṅgalám (comm. śuklavarnadeham) as opposed to rātriyai kṛṣṇám piṅgākṣám, Kāty. ŚS. XX.8.17 śuklo 'tigaurah, KB. 25.10.25.9.15 śukra etasyā 'hnaḥ piṅgākṣo hotā syād iti haika āhuḥ “The Hotṛ for this day should be white, with red eyes” (Keith).

405 viklidha is explained as dantura by Harisvāmin in his commentary on ŚB., in the commentary on Kātyāyana by Mahīdhara in his commentary on VS. XXV.9. See Lüders, op. c., pp. 142–145.

406 piṅgākṣa “with yellow eyes”, Lüders, op. c., pp. 132–134, contrasting with piṅgalá (see n. 317).

407 The relevant passages are Kāty. ŚS. XX.8.17 avabhrtheṣtyante 'psu magnasya piṅgalakhalativiklidhaśuklásya mūrdháni juhoti jumbakāya svāheti, on which Mahīdhara has the following comment (ad VS. XXV.9): avabhrthayāgānte, evám-vidhásya puṁso mūrdháni jumbakāye 'ti mantrenaḥ 'jyam ákr̥d grhītám juhuyāt, kidṛśasya puṁsaḥ? jale magnásya piṅgalākṣásya khalateḥ (khalātásya) viklidhaśya (danturásya) śuklásya ('tigaurásya) 'ti sútrārthaḥ. Varuṇadevatyā dvipadā, yajur gāyatry, Udanyaputramunḍíbhadṛṣṭā, whereas Uvaṭa gives some interesting details: avabhrthe, apsu magmása piṅgalakhalativiklidhaśuklásya mūrdháni juhoti “jumbakāya svāhā” iti. varuṇī dvipadā Śundibhaudanyadṛṣṭā.

eṣā cāntar jale japtā pāpanāśini. tad uktam Hāritena : “jumbakā nāma gāyatry vede vājasaneyake, antar jale sakṛj japtā brahmahatyaín vyapohate 'ti, (which is based on TB. III.9.15.3 etád ha vái Muṇḍibhá Audanyaváh/bhrūṇahatyádyai práyaścittím vidā́m cakāra/yó hā 'syā́d 'pi prajá́dyām brāhmaṇā́m hánti/sarvásmai tá́smai bheṣajáṃ

The commentator on the Taittirīya Brāhmaṇa states that this sacrifice makes Varuṇa pāparūpa disappear408. Somewhat apart from the stereotyped descriptions stands TĀ. I.2.3, where he is described in the words:

paṭáro viklidhaḥ píṅgáḥ | etád varuṇalakṣaṇam | yátraitád upadṛśyate | sahásram் tatra nīyate 409.

In view of the many details that have since become known it can no longer be reasonably doubted that Albrecht Weber was right in explaining this ritual as a last trace of human sacrifice 410. By drowning a brahmin as the impersonation of the god the latter was induced to return to his

karoti, iti, TB. III.9.15.3 Fáruṇo vai jumbakā́h. anītalā evá Váruṇam ávayajate, iti. khalatér viklidhasya śuklásya píṅgākṣásya múrdhāṅ juhoṭi. etád vai Varuṇasya rūpám, khalatyādiyuktaḥ śarīrami Fruṇasya váraksasya pāpasyā rūpam, atas taducitenaiva rūpena Varuṇam pāparūpam avayajate, vināśayati, BaudhŚS. XV. 37 athaiṣa Ātreyo vīḥṛtaḥ śuklo viklidhas tilakavān píṅgākṣaḥ khalatiḥ vikaṭaḥ kunakhī kubjaḥ śipiviṣṭo

nagma upamajjati. tasya mūrdhni juhoṭi “jumbakāyā svāhé” ‘ty, atrā 'smā etac chataṁ vipathamáṁ dadā́ty, athainam iṣujālāt pradhamaní (? reading uncertain) “mā me rā́ṣṭre vā́tsīr” iti, ŚāṅkhŚS. XVI.18.18 athā 'treyam sahasrenā 'vakrīya yaḥ śuklaḥ píṅgākṣo

valīnaḥ (“shrivelled, wrinkled”) tilakāvalo viklidhaḥ khando bandaḥ khalatiḥ, tam ādāya nadíṁ yanti. athai 'nam udake 'bhipragā́hya yudā́ 'syo 'dakáṁ mukham ā́syandatā 'thā 'smā adhvaryur mūrdhni jumbhakāyā svāhé” 'ti, ĀpŚS. XX.22.6 (based upon TB. and ŚāṅkhŚS.) avabḥṛthena pracaryā 'treyam

śipiviṣṭam khalatim viklidhaṁ śuklam píṅgākṣam tilakāvalam avabḥṛtham abhyavaníya ttasya mūrdhaṁ juhoṭi “mrtyave svāhé”, bṛhṇahatyā́yai svāhé”, jumbakā́ya svāhé” 'ti tisraḥ, HirŚS. XIV.4 Ātreyam khalatim viklidhaṁ śuklam píṅgākṣam śipiviṣṭam tilakāvalam avabḥṛtham avanīya . . . tasya mūrdhaṁ juhoṭi, Vādhūlasūtra (Acta Or. 4, p. 203) tasmā anoyuktā́ṁ ca śatā́ṁ ca dattvā 'ha : kṣattar, etad etasmā upakuru, athai 'nam் nirvaha, mā me vijite vā́tsīr iti.

408 Caland's conclusion in his note on Āp. XX.22.6 (“Die meisten [viz. Epitheta] beziehen sich offenbar auf das lichte Wesen des Varuṇa”) was rightly rejected by Lüders, op. c. p. 143 n. 4.

409 Quoted with the accentuation of the edition of the Ānandāśrama Series 36 pt I3 (1967), p. 9 (except yátraitád). The explanation which the commentary suggests for paṭára is plainly mere guess-work and devoid of sense. Lüders, p. 143 n. 2, rightly pointed to vaṭara, which the Kāśikā adds to the gaṇa kadārádi and for which Lüders proposes to read vaṭhara. It may be doubted, however, whether in such a native word for a defect a text emendation is required. The meaning may have been the same as of banda “maimed, defective, crippled” in Śāṅkhāyana. See also Turner, Comparative Dictionary of the Indo-Aryan Languages 9124 and 11236.

410 See A. Weber, ZDMG. 18, p. 268 (=Indische Streifen I, p. 63), Hillebrandt, Altindien, p. 33f. (quoted by K. F. Johansson, Über die altindische Göttin Dhiṣáṇā (see n. 402), p. 131f.), Vedische Mythologie II2, p. 27 n. 3, K. F. Johansson, op. c., pp. 125–132, A. K. Coomaraswamy, Yakṣas II (1931), p. 28, and further S. Lévi, Doctrine du sacrifice, p. 132f., Keith ad TS. VI.1.11.6 puruṣaniṣkrāyaṇa iva hi (p. 500 n. 3), J. C. Heesterman, “The case of the severed head”, WZKSO. 11 (1967), pp. 22–43, D. Schlingloff, IIJ. 11 (1968), p. 175 with references for the puruṣamedha. On Keith's interpretation of the Jumbaka as “a mystical view of the nature of Varuṇa” (Religion and Philosophy of the Veda, p. 96, on the basis of ŚB. XIII.3.6.5) see Hillebrandt, Ved. Myth. II2, p. 28 n. 3. Keith, p. 304, only accepted the scapegoat character of the Jumbaka. For the scapegoat in general see also Hertha Krick, WZKS. 16 (1972), p. 31 n. 27.

219

own element, the water. In this connection the words yadā 'syo 'dakam

mukham āsyandeta "when the water reaches his mouth" are particularly

significant. Equally eloquent is the prescription, handed down by

Baudhāyana and Śāñkhāyana, that at the end of the ritual the Ātreya

must be banished. After the ritual drowning of the victim (who had been

bought for this purpose) had become only a symbolical act, other means

were needed to get rid of the scapegoat, who was burdened with evil

(Varuṇa pāparūpa!).

An exact comparison of the terms which occur in the various descriptions

of the vidūṣaka with those of the jumbaka shows that four features of

the former have direct correspondences with the latter:

dantura "with protruding teeth" NŚ. : viklidha ŚB. TB. TĀ. Hir. Āp.

Śāñkh. Baudh.

khalati "bald-headed" NŚ.

: khalati ŚB. TB. Hir. Āp. Śāñkh.

Baudh.

kubja "hunch-backed" NŚ.

: kubja Baudh.

piṅgalākṣa "red-eyed" NŚ.

: piṅgākṣá ŚB. TB. Hir. Āp.

Śāñkh. Baudh. piṅgala Kāty.,

piṅgá TĀ.

Four other features can be paralleled with similar characteristics,

although the correspondence is at best partial:

vikṛtānana "with a distorted face" NŚ.: vikata Baudh.

khañja "limping" NŚ.

: khaṇḍa "crippled, defective"

Śāñkh.

vāmana "dwarfish" NŚ.

: baṇḍa "maimed, defective,

crippled" Śāñkh., cf. paṭāra TĀ.

chedavibhūṣitavadana (? p. 216) NŚ.

: tilakāvala Śāñkh. Hir. Āp.,

tilakavant "furnished with

marks" Baudh.

The precise meaning of khaṇḍa and baṇḍa/vaṇḍa is, it is true, doubtful:

the first means "mutilated" but bhagnadanta according to the commentary

ad ĀpŚS. VII.12.1, the latter also "mutilated" (yo na vivardhate) but

bhinnapuccha "tailless" according to the Nighaṇṭu and some commentators,

"impotent" (nirvīrya) according to Sāyaṇa ad AthS. VII.67(65).3. Both

characteristics were, in any case, inauspicious and therefore not allowed

in the case of sacrificial animals (such as the he-goat, ĀpŚS. VII. 12.1, cf.

MS. IV.3.8: 47,15) or of the Soma-cow. On the other hand, the latter had

to be piṅgākṣá "red-eyed", because this was the appearance (rūpá) of

Soma (TS. VI.1.6.7).

Apart from two old words whose meaning is unknown (śipiviṣṭa Baudh.

Hir. Āp. and paṭāra TĀ), there are only two adjectives in the description

of the jumbaka to which no parallel can be found in the Nāṭyaśāstra,

viz. śukla "bright-coloured" (in all texts) and kunakhin "having diseased

nails"Baudh.411.

411 Since for a comparison with the Vedic texts only the oldest description of the

vidūṣaka is relevant, later sources (n. 387) have been ignored. Cf. Śāradātanaya

(c. 1175-1260), Bhāvaprakāśana (GOS. No. 45, p. 289, lines 5-7) vedavīn narmavedī

220

It is curious that even Bhat, who explained the vidūṣaka as originating in the Vedic Asura, has not noticed these similarities and dismissed the outward appearance of the vidūṣaka as a "stage trick" 412. It is the more remarkable that this misconception did not blind him to one of the fundamental questions, which had almost universally been ignored in previous studies, viz., Why is it that only for the vidūṣaka such a detailed and rigid prescription is given?

It is true that the vidūṣaka shares some features with the servants (dāsa) but this only brings out more clearly the basic problem. It is natural that the low social order of the dāsa was represented on the stage by visual means, by making him a dwarf (vāmana), a hunch-back (kubja) or one with a distorted face (vikṛtānana). This, however, makes it the less explicable why a brahmin should be characterized by the same deformities. That Bhat's reference to a comic function cannot be correct, is apparent from the stereotyped and rigid character of Bharata's prescription and its impracticability. They lead to the conclusion that this is rather the last echo of an old tradition.

At this point the line of argument of this book has reached its completion. The comparison with the Vedic jumbaka has provided the final proof, it is hoped, that both the character and the traditional (but theoretical) outward appearance of the vidūṣaka date back to the Vedic ritual. It shows that there is, indeed, some reason to question the hāsya-character which even the Nāṭyaśāstra in some places ascribes to his strange appearance - at least, when its origin is considered. It now appears that it was still prescribed, in spite of its purely theoretical character, because it was dictated by a religious tradition. In its origin the deformity of the vidūṣaka had no connection whatsoever with dramatic performances but was simply the deformity of the Vedic scapegoat.

yo netuḥ sa syād vidūṣakaḥ, khalatik piṅgalākṣaś ca hāsyānūkavibhūṣitaḥ, piṅgakeṣo hariśmaśrur nartakaś ca vidūṣakaḥ, where "bald-headed" is contradicted by "tawny-haired" (see Bhat, p. 48), which in the Nāṭyaśāstra is said to be a characteristic of Daityas and Rākṣasas. Cf. 13.52f. raktākṣaḥ piṅgakeṣaś ca asito vikṛtasvarah, rūkṣo nirbhartsanaparo raudraḥ so 'tha svabhāvajaḥ (so 'yam 12.49 KM). For further references see R. Schmidt, Beiträge zur indischen Erotik, p. 145.

412 Bhat, The Vidūṣaka, p. 44f.: "The physical deformity of the Vidūṣaka, on the contrary, is artificial; it is a stage trick to provoke laughter; or better perhaps, it can be traced to the hideousness of the Asura type, or the unusual tuft of hair of Nārada. Many of these traits were convenient and handy for evoking fun. A repeated use of them must have settled the outline of this character and turned him into a conventional figure of the stage". P. 53: "The appearance of the Vidūṣaka could have been a legacy from the earliest Asura type, as previously suggested. But it is possible that Bharata provided several attributes, to which a little addition was made afterwards, mainly with the intention of guiding the actor in preparing his make-up. It cannot be imagined that each and every detail in the prescription was intended to be strictly followed. The underlying idea in the appearance of the Vidūṣaka seems to be of a physical deformity only, which Bharata suggested by means of several attributes. It must have been naturally left to the actor (or dramatist) to emphasise one or more of these attributes in the actual dramatic performance. After all, some incongruity in physical appearance would certainly be a source of laughter on the stage".

221

Needless to say, the Vedic jumbaká was not ridiculed for his corporeal defects, which were considered a varunalakṣaṇam (TĀ), a mark of Varuṇa. His deformity was due to his being a human image of Varuṇa and was the outward manifestation of the evil with which he was burdened. This is apparent from such a verse as AS. VII.65 (67).3

śyāvā́datā kunakhinā bandéna yát sahá 'simá ápāmārga tváyā vayám sárvam tád ápa mr̥jmahe "If we have been together with one dark-toothed, ill-nailed, mutilated, by thee, O off-wiper, we wipe off all that" (Whitney),

where the ominous character of the defects is manifest. Cf. also, e.g., TS. II.5.1.7, where śyāvádat and kunakhin occur side by side with khalati, kāṇá "one-eyed" and other defects.

There is, however, more to it than the problem of a traditional appearance. The jumbaká was a scapegoat but he was also a brahmin, who moreover impersonated Varuṇa. On p. 208 it has already been pointed out that Varuṇa, by taking upon himself that which is "badly sacrificed" (dūriṣṭa) acts as a divine scapegoat. Here we find finally an answer to a question which has been ignored in most studies on the vidūṣaka, viz. Why had the purported buffoon by definition to be a brahmin? The only exception is NŚ. 34.21, which also allows avarajāḥ. Heesterman has in various papers (e.g. WZKSO.8 (1964), p.3) stressed the character of the brahmin in general as the purificator, who takes impurity upon himself. The equation vidūṣaka=brahmin=jumbaká= Varuṇa seems finally to disclose the original full function of the vidūṣaka in the dramatic performance. It may be suggested that this not only consisted in cooperating-in-con

test with the nāyaka, but also in purifying the king, as the sponsor of the performance, by taking upon himself the latter's impurity. The statement of the Nāṭyaśāstra that the pūrvaranga was equal to a sacrifice may originally also have been true of the play itself. In the introduction it has been pointed out that the sponsor of a performance had the status of a yajamāna and must usually have been the king. Only a brahmin was able to redeem the king from evil by accepting the role of scapegoat. It was, it seems, because of his function of scapegoat, that the vidūṣaka had to be a brahmin. This lends some support to the explanation suggested on p. 176, that the Omkāra was named instead of Varuṇa as the protector of the vidūṣaka because the latter was a brahmin. In classical iconography no trace has been preserved of Varuṇa's deformity as the jumbaká. Varuṇa is "of white colour . . . draped in yellow garments, possesses a pacific look . . . karaṇḍamakuṭa, yajñopavīta" (Rao, l.c.), he has two or four arms and is seated on a haṁsa, makara or jhaṣa413.

413 The white colour (śveta) is also prescribed in the rules for the make-up of the Nāṭyaśāstra (23.91). For the classical iconography see Gopinath Rao, Elements of Hindu Iconography (Madras 1914-1916), II (1916), pp. 529-531 and 260 (texts); A. K. Coomaraswamy, Yakṣas (Washington D.C., 1928-1931), II, pp. 36-37, J. J. Meyer, Trilogie altindischer Mächte und Feste der Vegetation III (1937), p. 228,

222

19. VIDŪṢAKA AND VITA

In 1897 J. Huizinga concluded a critical discussion of the question as to whether the vidūṣaka was a real character in social life with the words: "Outside the drama, accordingly, the figure of the vidūṣaka proper does not seem to have developed" 414. Since that time there has been a long but fruitless discussion on whether the vidūṣaka was a court jester or a character of town life - a question that seems at best of minor importance. Besides, the conclusions at which some authors arrived were far from clear. Thus Winternitz, Geschichte der indischen Litteratur III (1972), p. 172, wrote about the vita: “Er gehört jedenfalls dem städtischen Leben an und wird wohl auch aus volkstümlichen Spielen, in denen das Treiben der Hetären und ihrer Gefolgschaft dargestellt wurde, herübergenommen sein”. One cannot but conclude that the vita, whatever his “origin” (which is Winternitz’s main concern) was a stage character and nothing else.

As for the vidūṣaka, it will be sufficient to quote here two authors to show that a reappraisal of the problem is not out of place. Dasgupta writes: “It is thus not true that the vidūṣakas are Fools who attended the courts of kings only. They are mere comedians, who made their livelihood by their witticisms and also by friendly advice . . . were real characters in social life in the 2nd century B.C. and were not merely

Marie-Thérèse de Mallmann, Les enseignements iconographiques de l'Agni-Purāṇa (Annales du Musée Guimet, Bibliothèque d'études, Tome 67, Paris 1963) pp. 131f. As for the haṁsa (goose) cf. Varāhamihira's Brhatsaṁhitā 58 (57).57 haṁsārūdhas ca pāśabhrd Varuṇah, Agnipur. 64.3 haṁsaprasthaṁ (but makare, makarārūdhas: 51.15, 56.23, 96.29). According to M.-Th. de Mallmann, p. 131 the haṁsa belongs to Varuṇa as sovereign of the Waters, whereas the makara characterizes him in his function of lokapāla. Cf. p. 132: “A cette tradition primitive [viz. of the Brhatsaṁhitā] se rattache la notion du roi des Eaux telle que l'atteste le Viṣṇudh., notion encore préservée par l'AgP., mais uniquement lorsqu'il s'agit des rites de consécration des puits, mares et réservoirs. Partout ailleurs, dans les textes comme sur les images l'ansér est supplanté par le makara, tandis que Varuṇa se réduit progressivement à n'être plus que le Gardien de l'Occident”. Cf. also p. 230. For the makara in general see A. K. Coomaraswamy, Yaksas II, p. 13, who stresses his aquatic character (p. 34, etc.), similarly M.-Th. de Mallmann, p. 233). On the other hand, attention has been drawn (with reference to F. D. K. Bosch, De Gouden Kiem, p. 170=The Golden Germ, p. 146) to the fact that the makara represents the nether world, his open mouth symbolizing the gate which is opened at sunrise and sunset. See Kuiper, Bijdragen Taal-, Land- en Volkenkunde 107 (1951), p. 71 n. 5. This explains why it was forbidden to look at the rising and setting sun, cf. VaikhDS. III.1.114 udaye 'stamaye oa sūryaiṁ ne 'kṣeta Manu 4.37 ne 'kṣeto 'dyantam ādityam nā 'sthaṇvantam. At those moments, indeed, the gate of the nether world was opened. A third vāhana of Varuṇa is the jhaṣa, cf. Matsyapurāṇa 261.18 jhaṣāsanagatam, which is the correct reading and need not be emended (see de Mallmann, p. 131 n. 6).

414 See J. Huizinga, De vidūṣaka in het indisch tooneel (Groningen, 1897), p. 57: "Buiten het drama schijnt zich dus de eigenlijke vidūṣaka-figuur niet ontwikkeld te hebben" and p. 63 "Voor een type uit de samenleving schijnt het woord vidūṣaka niet te zijn gebruikt. Een plaats uit het Daśakumāracarita kan kan een oogenblik doen denken, dat het woord een omgangstype aanduidt, maar zoowel de commentaren als de tekst zelf bewijzen, dat er van een acteur wordt gesproken". Gonda (n. 344), p. 408 contested this, but see below, p. 229.

223

dramatic invention”. Bhat seems to hold a somewhat different view, as may be inferred from the following words: “It was customary for kings to maintain a court-jester. It is also possible to say that professional fools were fashionable in ancient India”415. Elsewhere (p. 134f.) he discusses the Vidūṣaka's ‘Function of a Court-Jester”.

Bhat's bold statement naturally raises the question as to how it is possible to say anything definite about professional fools in ancient India and on what evidence the assumption of court-jesters is based. As far as I can see, the answer must be that what we actually know at this moment about Old Indian court-jesters is next to nothing. See n. 357. The reading rājyavidūṣaka at Sāgaranandin, NLRK 2014 is certainly corrupt and does not mean “court-jester” (Translation, p. 66). It is also clear that if the drama imitated contemporary court life, it would be even more difficult to explain why, then, a brahmin had to act as a jester. On the other hand, it will have been noticed that the picture of the “real characters in social life” that Dasgupta confidently presents is (and has to be) entirely based on the conventional part of the stage character of the drama. As the literature on the subject shows, it is, indeed, difficult to avoid circular argumentations when writing on this aspect of life in Ancient India, for which evidence is entirely lacking.

Although the distinction does not seem to be essential, it may be convenient for a critical discussion to follow those scholars who think the main problem is whether the vidūṣaka was a court-jester or a character in social life.

Gonda, l.c., points in a very general way to the existence of court-jesters in countries other than India. For the rest (pp. 410-415) he classifies the vidūṣaka on account of his physical deformity with the dwarfs, hunch-backs and cripples, who lived at royal courts, in India as well as elsewhere. See, however, above, p. 217. Bhat does not give the slightest argument for his statement quoted above. A. B. Keith416 had already pointed out that in the epics with their profuse descriptions of court-life there is not a single reference to professional jesters. The same is true of Kauṭalya's Arthaśāstra (first century A.D.?)417, where the word vidūṣaka does not occur either.

The situation is different in the Daśakumāracarita by Daṇdin (c. 550 A.D.? see n. 432) where, apart from a single occurrence of the word vidūṣaka, which will be discussed below, the following passage must be mentioned. It is the description of a man who had access to the harem and who was sitting there near the king: “a skilful mind-reader, a royal favourite, an adept in song, dance, instrumental music and related arts, a connoisseur

415 See Dasgupta in: S. N. Dasgupta-S. K. De, A History of Sanskrit Literature, Classical Period, Vol. I (Calcutta 1962), p. 654f., and G. K. Bhat, The Vidūṣaka, p. 39f. Cf. also Bhat, p. 25: “The professional fool and the court-jester were known in ancient India; the tradition continues”.

416 The Sanskrit Drama, p. 28.

417 See Hartmut Scharfe, Untersuchungen zur Staatsrechtslehre des Kauṭalya (Wiesbaden 1968), p. 334. Cf. also IIJ. 13, p. 286.

224

of unconventional women, shrewd, talkative, clever in periphrastical and enigmatical speeches (bhañgi), critical, a buffoon (parihāsayitṛ), a scandal-monger, an adept in calumny, ready to take bribes even from ministers of state, an instructor in all naughtiness, a pilot in the science of love (kāmatantra) and the king's servant from the time the former was a prince" 418. Only in one point does this person resemble the vidūṣaka: while he is said to be a "pilot in the kāmatantra", the vidūṣaka is once called a kāmatantraśaciva "counsellor in matters of the kāmatantra, or in amorous adventures" (Mālavikāgnimitra IV.17.10). The passage has never to my knowledge been quoted as proof for the existence of a professional court-jester, and rightly so. Ryder's translation of parihāsayitṛ as "buffoon", which has been adopted in the preceding rendering, is hardly correct. The literal meaning to be expected would rather be "who elicits mockery, who induces others to mock or ridicule (persons)". Therefore, Agashe's note in the commentary to his edition (p. 342) "causing merriment or jovial" is less correct than "Spötter" in J. J. Meyer's translation. In other words: the parihāsayitṛ is not identical with the vaihāsika (p. 203). But even if he were a jester, professional or not, this passage shows eloquently what differentiates the vidūṣaka from any jester, viz. the latter's intelligence and quick wit. The vidūṣaka is never characterized as paṭu "shrewd" (see n. 360). It has generally been overlooked that he lacks the characteristic traits of the court-jester.

No more than in the works quoted can any reference to the vidūṣaka as a court-jester be found in the very detailed descriptions of court life in Bāṇa's Kādambarī and Harṣacarita (shortly after 600 A.D.). On the other hand, a comparison of the Kādambarī with the earlier Arthaśāstra shows that enumerations of persons living inside and outside the royal court are so much similar that one is bound to conclude that at an early date lists of this kind must have been conventional in literary use. Cf. on the one hand Kādambarī (ed. Peterson), p. 70,11 -kubja-vāmana-kirātāh, p. 74,4 praṇṛtta-kala-mūka-kubja-kirāta-vāmana-badhira-jada-jana-p. 89,6 aneka-kubja-kirāta-varṣadhara-badhira-vāmanaka-mūka, p. 89,10, kubja-vāmana-kirāta, p. 90,23 purāṇapuruṣa-vāmanā-, and, on the other hand,

Kauṭalya 1.12.9 kubja-vāmana-kirāta-mūka-badhira-jadā--ndha-cchad-mano naṭa-nartaka-gāyana-vādaka-vāgjivana-kuśīlavāh, 2.27.25 etena naṭa-nartaka-gāyaka-vādaka-vāgjivana-kuśilava-plavaka-śaubhika-cāraṇānām, 7.17.34 naṭa-nartaka-gāyaka-vādaka-vāgjīvana-kuśilava-plavaka-śau(b)hika-. Cf. further also Kauṭalya 1.12.9 sūdā-r'ālikā-snāpaka-samvāhakā-'staraka-kalpakā-prasādhako-'dakaparicārakāḥ, 1.21.13 snāpaka-saṃvāhakā-'staraka-kalpakā-

418 See Daśakumāracarita p. 130,21 ed. Agashe (p. 190, 3 ed. M. R. Kale, 1966). For bahubhaṅgivisārada cf. p. 137,1-198,13 bhaṅgibhāṣanaratāḥ. The kumārasevaka must have been different from the tutor (kumārādhikṛta NŚ. 34.83, 97, kumārādhyakṣa Kauṭalya, cf. Ghosh, Text I2, Introduction pp. liii, lv, lxxi. Translation II1 p. 210 n. 3). For kāmatantra cf. madanatantra Daśak. p. 53.11 and NŚ. 16.32; 24.150, 191, 201, 205 (not 220); 25.38, 53, 62, 65. Cf. further Mālavikāgnimitra IV.17 (10) and the Caturbhāṇi, passim (Schokker, The Pādatāḍitaka, pp. 53, 246, 341).

225

'starakā-rajaka-mālākāra-, 7.17.40 sūdā-'rālikā-snāpaka-saṁvāhakā-'stara-ka-kalpakā-prasādhako-'dakaparicārikair.

These lists are quoted here because a work written about 970 A.D. shows in this respect a curious divergence from the Arthasāstra. Somadeva Sūri gives in his Nītivākyāmṛta 14.8 a long list of persons who may act as spies for the king. After some other persons, he enumerates the following: kirāta, yamapaṭṭika, ahitundika, saundika, saubhika, pāṭaccara, // viṭa, vidūṣaka, piṭhamardaka, // naṭa, nartaka, gāyaka, vāḍaka, vāgjīvaka, . . ., sūta, ārāḷika, saṁvāhaka . . ., jaḍa, mūka, badhira, andha 419.

It seems obvious that here Somadeva Sūri has simply copied a traditional list, such as found in Kauṭalya, of persons living inside and outside the court, irrespective of whether or not they could actually be used for the purpose of spying. Into this list, however, he has inserted the triad indicated above which does not occur in the Arthasāstra. Nothing indicates that they were living at the court, nor that one or more of them were court-jesters. As far as I can see, nothing is known about the time when the jesters mentioned in late Indian tales made their entrance at the courts. Hertel refers to the Sanskrit "Erzählungsliteratur" 420 but Jīnakīrti, the author of the Pāla-gopāla-kathānaka, probably lived about 1450 A.D. The famous North-Indian jester Bīrbal, said to be of Āndhra origin, is also of late date. As for the Tamil collection of anecdotes about Teṅṅālu Irāmaṉ, it is still entirely obscure how one of the great Telugu epic poets, Teṉāli Rāmakṛṣṇa (alias Rāmuḍu) became a court jester in Tamil legend, the less so as court jesters seem to have been completely unknown in classical Tamil literature. See above, p. 205 n. 356.

The triad, although quoted in an unusual order by Somadeva Sūri, had been well known long before him. It is, indeed, as traditional as the other lists, which he has compiled here. Its first occurrence is, it seems, in Vātsyāyana's Kāmasūtra, which is tentatively assigned to the first centuries of our era.421 In this work the group always occurs in the compound pīthamarda-viṭa-vidūṣaka, except in 1.4.44-46, where a definition of each of the three terms is given and the last sūtra runs as follows: ekadeśavidyas tu kriḍanako viśvāsyaś ca vidūṣakaḥ, vṛihāsiko vā 422. In all other passages, with a single exception, the three words always occur in this compound, often in combination with terms denoting garland makers, etc. Cf. I.4.21 bhojanānantaram . . . pīthamarda-viṭa-vidūṣakāyattā vyāpārā

419 Quoted from Oscar Botto's translation Il Nītivākyāmṛta di Somadeva Sūri (Torino 1962), p. 95.

420 J. Hertel, Literarisches Zentralblatt 1917, 1198ff., "Jinakīrtis Geschichte von Pāla und Gopāla", Ber. sächs. Gesellsch. d. Wiss. LXIX, 1917, p. 121 n. (both referred to by Winternitz, Gesch. d. indischen Litteratur III, p. 172 n. 1).

421 "Er wird wohl an den Anfang unserer Zeitrechnung gehören", R. Schmidt, Beiträge zur indischen Erotik2 (1911), p. 12; fourth century A.D. (?) according to Winternitz, op. c., III, p. 540.

422 Cf. Daśarūpaka 2.9 (13) ekavidyo viṭaś cā 'nyo hāsyakṛc ca vidūṣakaḥ.

226

divāśayyā ca, I.5.37 rajaka-nāpita-mālākara-gāndhika-saurika-bhikṣuka-gopālaka-tāmbūlika-sauvarnika-pīthamarda-viṭa-vidūṣakādayo mitrāṇi, II. 10.48 tatra pīṭhamarda-viṭa-vidūṣakā nāyakaprayuktair upaśamitaroṣā tair evā 'nunītā 423, VI.1.8-9 yair nāyakam āvarjayed anyābhyaś cā 'vacchindyāt

... te tv ārakṣapuruşā dharmādhikaraṇasthā daivajñā ... pīṭhamarda-viṭa-vidūṣaka-mālākara-gāndhika- 424 śauṇḍika-rajaka-māpita-bhikṣukāḥ. The only exception occurs in VI.1.22-23, where the vaihāsika is mentioned along with the pīṭhamarda 425.

Before discussing later occurrences of the triad it may be useful first to consider the evidence of the Kāmasūtra more closely. It has long been observed that Vātsyāyana in his Handbook of Erotics apparently imitates Kauṭalya 426. Winternitz remarks: “Das Kāmasūtra spricht da z.B. von den Freunden des Liebhabers und den Liebesbotinnen nicht viel anders als das Arthasāstra von den Freunden des Königs, den Spionen und Gesandten”. (l.c.).

Thus, after giving separate definitions of the pīṭhamarda, the viṭa and the vidūṣaka, the Kāmasūtra characterizes them together in the words (I.4.34) ete veśyānāṁ nāgarakānāṁ ca mantriṇaḥ sandhivigrahaniyuktāḥ “These are the ministers of the courtezans and men about town, to whom treaties and wars are entrusted” 427. It is instructive to compare this definition with the one given in a literary treatise (Agnipurāṇa 339.39)

pīṭhamardo viṭaś caiva vidūṣaka iti trayaḥ, śṛṅgāre narmamacivā nāyakasyā 'nunāyakāḥ

“The pīṭhamarda, viṭa and vidūṣaka, these three are the companions of the nāyaka in his erotic amusements and his attendants”.

On the other hand, there is a clear interrelation between the Kāmasūtra and the dramaturgic handbooks. The central figure of the Kāmasūtra is

423 Cf., e.g., SD. 3.40 kupītavadhūmānaban̄janāb and further Rasamañjarī, Kāvyānuśāsana and Rasaratnahāra (Schmidt, Beiträge, p. 143f.).

424 Cf. also NŚ. 34.55 (24.36 KM) chattraśayyāsane yuktā tathā vyajanakarmani (56), saṃvāhane ca gandhe ca tathā caiva prasādhane, tathā 'bharaṇasamyoge mālya-saṃgrahaṇeṣu ca (57) vijñeyā nāmatas sā tu nrpateḥ paricārikā=24.53 B1 śayyāpālī chatradhārī tathā vyajanadhāriṇī (54) saṃvāhikā gandhayoktrī tathā caiva prasādhikā, tathā 'bharaṇayoktrī ca mālyasamyojikā tathā (55) evaṃvidhā bhaveyur yās tā jñeyā paricārikāḥ. Lüders, Philologica Indica, p. 416 n. 1 quotes Vidhurapaṇḍitajātaka (545), Gatha 59 mālākāre ca rajake ganthike atha dussike, suvannakāre maṇikāre maṇimhi passa nimittam and gives in the Addenda the correct explanation of ganthike = gāndhikaḥ.

425 bhāvajijñāsārtham paricārakamukhan saṃvāhaka-gāyana-vaihāsikān gamye tad-ganthike = gāndhikaḥ.

bhaktān vā praṇidadhyāt, tadabhäve pīṭhamardādīn, where the commentary paraphrases vaihāsiko : vidūṣakah and pīṭhamardādī 'ty : ādiśabdād viṭa-mālākār-gāndhika-śauṇḍikādī sahāyāḥ.

426 For stylistic reminiscences see, e.g., R. Shamasastry, Kauṭilya's Arthaśāstra translated 5 (Mysore, 1956), p. xi and further Winternitz, op. c., III, p. 537 n. 1 with references. Cf. also Friedrich Wilhelm “Nāṭyaśāstra and Nītiśāstra” in Festschr. Sluszkiewicz, pp. 271–276.

427 The commentary says ete nāgarakānāṁ pārśvavartitvād upanāgarakāḥ. Gonda (n. 344), p. 409 n. 8, had already pointed to similarities between the nāgaraka and the svāmin and referred to Kāmasūtra 308–310 and Arthaśāstra 96.

227

the nāgaraka “man about town, elegant”, a member of a refined leisure class, of whose daily life this work gives an idealized picture. Rather than describing, however, the Kāmasūtra prescribes. It lays down the rules which one has to comply with in order to become a perfect gentleman, a vidagdhajana 428 as the commentary paraphrases the word nāgaraka. Several times, however, the commentary refers to him as the nāyaka, the technical term for the hero in the drama, and this term is also used in the Kāmasūtra itself, e.g., I.5.28ff., II.10.46, 47, etc. In other words, there is a tendency in the Kāmasūtra to describe the ideal “reality” in terms of stage characters. Inversely, it is a well-known fact that a small Kāmatantra has been inserted in the Nāṭyaśāstra 429.

All this must be borne in mind when one tries to evaluate the importance of the fact that the Kāmasūtra mentions the vidūṣaka as a synonym of the vaihāsika, the friend and confident of the nāgaraka, for whom he is a plaything (krīḍanaka) 430. Two things must here obviously be distinguished, viz. the formalization of the enumeration of groups of occupations and, on the other hand, the institutionalization of functions attributed to certain characters who may or may not have existed in Old Indian society.

To the first category belong, e.g., sūdārālika-, snāpaka-samvāhaka-(nāpita-Kāmas.), rajaka-mālākāra-gāndhika- (Kaut, Kāmas.), naṭa-nartakagāyaka-vādaka-, kubja-kirāta-vāmana-jada-mūka-badhira- (Kauṭ., Kāḍ.). The actual existence of these figures in Indian social life cannot, of course, be questioned but it is characteristic that every author, be it of theoretical treatises such as the Kāmasūtra and the Arthaśāstra, or of literary compositions, such as the Kādambarī and a Pāli Jātaka, when referring to these figures, automatically fell back upon traditional lists. However authentic and “real” every member of the groups enumerated may have been in social life, it cannot be denied that every author here follows a fixed literary pattern. His description of reality, accordingly, is not based on his own direct observation but is indirect and derivational. Nothing proves more impressively the artificiality of these lists than the fact that in a description of the rejoicing in a palace at the birth of a prince a prose kāvya sums up almost the same groups, in more or less the same conventional order, as are found in the Arthaśāstra as lists of persons who can be used as spies. An instructive instance is also Somadeva Sūri’s list of prospective spies, which is a mere concoction of the various existing lists of his predecessors, without much concern for the possibility of using them as such in actual life.

428 “ein gebildeter Mensch” (R. Schmidt). The same is no doubt meant by the term nāgarikapuraṣa in Daśakumāracarita p. 42,8 (ed. Agashe), where all the three commentaries explain nāgarikā nipunāḥ.

429 24.94–319C.=22.95–323.B., 22.89–312 KM. For the use of the term kāmatantra see above n. 418 and cf. P. V. Kane, History of Sanskrit Poetics 4, p. 43. As for nāyaka in the sense of “husband” (Daśak. 49,15 Agashe), this is clearly a “literary” word.

430 Bhat (n. 344) p. 135 rightly observes that the vidūṣaka is an “object of jest, rather than being a jester”.

As for the institutionalization, it is sufficient to point to the "amusement-companions" (narmasacivas) of the "elegant". In normal conditions of life it is natural that a young man of this class should have had some friends who may have been living at his expense since he, as a nāgaraka, had by definition some wealth. Occasionally one or more of these friends may have been a bit of a comic. When, however, it is prescribed that in the company of a nāgaraka there must be a vaihāsika, whose characteristics are specified, it would seem that we have left the solid ground of reality and are moving in a fictitious world. When, further, this comic is generally called a vidūṣaka, while the nāgaraka is sometimes denoted by the term nāyaka, the influence of the drama and dramaturgic theory can hardly be denied. It has been argued above that in the drama the vidūṣaka must originally have had a definite function, which was to help and, at the same time, unintentionally to counteract the hero. Even in the most abstract theory of a man about town, however, it would be difficult to conceive of a particular function of a comic, let alone of a vidūṣaka.

And when, finally, this vidūṣaka is in the Kāmasūtra almost exclusively mentioned as a member of the group pithamarda-vita-vidūṣaka-, there can be little doubt, as far as this text is concerned, about the purely theoretical character of this group. On the other hand, we have seen (p. 226) that the Kāmasūtra was apparently the first text in which this triad was added to the list taken over from Kauṭalya and probably other theoretical handbooks. The conclusion, then, must be that, since Vātsyāyana cannot have coined this compound himself but must have borrowed it from some dramaturgic treatise, and since it is still unknown to Bharata's Nāṭyaśāstra, the group must have been created by the author of a handbook on dramatic art which was in current use in the first centuries of our era.

A final remark may not be out of order here. Later dramaturgists often refer to this triad as the narmasacivas "boon companions" or "friends in sport" 431. The same word, however, is also used in stories for the companions of a young prince, such as Udayana. Gonda has rightly remarked that the use of the same term does not imply that such a narmasaciva was identical with the vidūṣaka (op. c., p. 407).

As we have seen, the compound pithamarda-vita-vidūṣaka- must have been in current use from the first centuries of our era onwards. One of the oldest texts in which it occurs, besides the Kāmasūtra, is Daṇḍin's Daśakumāracarita (c. 550 A.D.?) 432. Among the various ways in which Monier-Williams renders narmasaciva by "'amusement-companion', promoter of the amusement of a prince". Cf. Sāgaranandin, NLRK. 2200 antaḥpuracaro rājñām narmāmātyah prakīrtitah. A similar expression is Pkt. kāmatantasaciva used with reference to the vidūṣaka Mālavikāgnimitra IV.17 (10) etc. (see Schokker, Pādatāḍitaka, p. 246).

431 Thus A. B. Keith, Sanskrit Drama, pp. 188 and 311, as a translation of narmasuhṛd.

432 Daśakumāracarita, p. 42.8 ed. Agashe (p. 67, 3 ed. Kale). For the date of the Daśakumāracarita see Mirashi, ABORI. 26, pp. 20-31, Studies in Indology, pp. 176-177 ("not long after 550 A.D."). See also S. K. De, History of Sanskrit Literature, p. 209 (contemporaneous with, inter alia, the Caturbhāṇī, the oldest of which is dated between 455 and 510 A.D. by G. H. Schokker, The Pādatāḍitaka

229

a young courtesan is "advertised" in the town the author mentions pīthamarda-viṭa-vidūṣakair bhikṣukyādibhiś ca nāgarikapurusaśamavāyeṣu rūpaśilamādhuryaprastāvanā "having her beauty, character and sweetness praised in the gatherings of men about town by pīthamardas, viṭas, vidūṣakas and female mendicants, etc." 433. It is again the same stereotyped

of Śyāmilaka, p. 25), Kuiper, Journal Madras University 28-2, pp. 121-125 (anterior to the Mṛcchakaṭikā). Otherwise J. Nobel, ZII. 5, p. 149 and cf. JORMadras 19 (1952), p. 161 (end of the seventh century), M. Krishmachariar, History of Classical Skt. Literature (1937-1970), p. 457 (between 635 and 700 A.D.). These dates would seem too late.

433 See p. 42,7 ed. Agashe, v. 1. rūpaśilaśilpasaundaryamādhuryaprastāvanā, p. 69,10 ed. Godbole. p. 67.2 ed. Kāle. nāgarikapuruṣaḥ (commentary: nipunāḥ) is "men about town" (Ryder), not "fatsoenlijke lieden" (decent people, Huizinga, op. c., p. 63), nor "Männer der Stadt" (J. J. Meyer).In the Daśakumāracarita nāgara and nāgarika are used in three different meanings:

1. "townsman, citizen", cf. 54,13 nāgaramukhyeṣu (: 54.21 pauramukhyaiḥ), 58,18 nāgarikebhyaḥ (Padacandrikā: nāgavāśilokebhyaḥ, read nagara-), 63,9 nāgārikavara (Padac.: nāgarikaśreṣṭha) and Pūrvapiṭhikā 23,20 nāgarikajāna (: 22,23 paurajāna) "townspeople". Cf. also, e.g., Bṛhatkathāślokasaṅgraha, ed. F. Lacôte, 1.2, 17.102 nāgara, 17.93 nāgaraka and Pali nāgāra "citizen". Pkt. nāyara and see Turner, CDIAL. 7043. More common is paura, e.g., Daśak. 47,2 pauravrddhānām (Padac.: nāgarikāśreṣṭhānām), 64, 19 pauravrddhaiḥ, 54,21 pauramukhyaiḥ, etc.

2. "City Superintendent" (Jolly-Schmidt, Arthasāstra II, p. 17 ad Kauṭalya 6.2), "police officer" (Beng. nāyer, nāyar "agent, steward", Turner, l.c.). Cf. Śivarāma Tripāṭhin in his Bhūṣaṇā ad Daśak. 49,11 nāgarikabalam : nagararakṣākartuḥ, bhaṣayā "kotwāl" ityakhyāyate : balam : sainyam. In connection with the last words attention may be drawn to Mrcch. VI.16.1 saṅcāri bhadraṃdadādhārāṃ and IX.23.3 nagararakṣādhikṛta (the last word also occurs in a commentary on Śakuntalā VI.0.1 as a gloss on nāgarika), which are used in the same sense. In spite of the incorrect gloss kārapatiḥ : nāgarikaḥ, which the commentaries ad Daśak. 58,16 quote from the Vaijayantī, the word nāgarika is used in the sense "Chief police officer, superintendent of the city police" in 58,16 nāgarikaḥ, 49,11 nāgarikabalam [=51,6 rakṣikabalam], 57,10 nāgarikapuruṣān, 59,11 nāgarikapuruşaiḥ (Padacandrikā kārapatipuruṣaiḥ ; = 63,12 rakṣikapuruşaiḥ). Whether the nāgarika also functioned as a judge, as is suggested by the gloss nāgarikasamādi on 42,14 adhikarane in the commentary Bhūṣaṇā, must here be left open. As for 51.3, where Agashe's text has samuccalantau nāgarikasampātena and the third edition by N. B. Godabole and K. P. Parab (Bombay 1898) samuccalantau nāgarikasampāte it is clear that the Padacandrikā (by Kavīndrācārya Sarasvatī, c. 1650 A.D.) misinterprets the passage by paraphrasing nagarathalokasaṃmarde(na). sampāta no doubt means "appearance" (see PW.) and what Daṇḍin's adventurers are afraid of is not the appearance of townsfolk but of policemen, cf. 139,17 rājapuruṣasaṃpātabhītah. Therefore, only the variant reading nāgarikapuruṣasampāte (see Agashe's critical apparatus) makes sense: the Bhūṣaṇā rightly paraphrases nagararakṣakapuruşasampāte. As for sam-uc-cal-, its meaning can (in spite of PW. II, col. 970) hardly have much differed from that of uc-cal- "to start" (46,5 nagarāyo 'dacalam, 49,10 tricaturāṇi padāny udacalam, 57,10 asidvitīyo ramhasā pareṇo 'dacalam). The words samuccalantau nāgarikapuruṣasampāte accordingly mean "being off at the appearance of policemen" (Meyer: "uns davon machend, da die Stadtwächter herbeistürzten" ; incorrect Ryder: "we fell in with policemen"). Cf. also Kālid. Śak. VI.0.1, Vikram. V.4.1.

3. nāgara, nāgaraka, nāgarika "refined; fine gentleman, elegant, man about town" (PW. "gewandt, geschickt, gerieben, fein"; cf. A. K. Coomaraswamy,

230

triad, which cannot, therefore, prove very much for social life in Indian

towns of that period.

In later times the group frequently recurs in the writings of theorists,

such as Dhanamjaya (between 974 and 996 A.D.) in his Dasarupaka

II.8–9434, probably about the same time in the Agnipurana 339.39435,

later (c. 1175–1250) in Saradatanaya's Bhavaprakasana, p. 93, 21f.436 and

(in the 14th century?) in Vagbhatta's Kavyanuśasana, p. 62437.

Not before the fourteenth century an innovation was introduced by

the addition of the ceta “servant” as a fourth member to the traditional

group. As early as NŚ. 35.77–80, in a passage which must reflect an old

tradition because it contains the description of the vidusaka as a jumbaka,

the characteristics of respectively the vita, the sakara, the vidusaka and

the ceta are summed up. From such passages the ceta may, in course of time,

have come to be incorporated in the triad. It is significant, however, that

this is only found in works on literary aesthetics. Cf. (between 1300 and

1384. Visvanatha, Sahityadarpaṇa 3.39–40 and 3.46438, (between 1300

and 1350) Vidyanatha, Prataparudriya I.38439, (between 1340 and 1360)

Singabhupala, Rasarnavasudhākara I.89440, (between 1450 and 1500)

JAOS. 48, p. 265). Pali has nagarika “citizen-like, urbane, polite”. Cf. Latin urbanus,

Greek asteïos “polite, refined, elegant, witty”. In this sense it is used in Daśak.

42,8 nagarikapuruṣasamavāyeṣu (Ryder: “in gatherings of men about town”)

above, and 52,23 sandrādarah samāgaman nagarajanah “the connoisseurs gathered

with tense anticipation”) (Ryder: “gay society”; not “die Leute” Meyer. Different

Cf. Bhattkathaslokasamgraha 24.44 nagarikah, v. 1. nagarakah “amateur”, 9.19

(cf. vv. 32, 34, 35) nagarakah “un roué, un coureur de bonnes fortunes”, 7.45

nagarakaṃmanyah “qui se croit bel esprit”, 32.123 nagarikamman ya “femme qui

se juge fine”, 9.34, 102 nagarakatā “politesse”, 9.81 nagarakatvam, idem. Cf. also

Kalidasa, Śak. V.1.9 nagarikavrttyā samjñāpayai ‘nam “tell her politely”, Vikram.

II.11.20 naario si “you are a refined person”, III.13.3 ṇarī “gallants”. According

to Sagaranandin, NLRK. 2226 the names of nagarah in dramas should end in -ila :

Ramita-Kamilahvana nagarah prakirtitah and they always address a courtesan with

the term vasu (2228). The nagaraka is the central person of the Kamasutra

(see p. 228).

434 patakanayakas tu anyah pithamardo vikṣaṇaḥ, tasyāivā 'nucaro bhaktaḥ kiṃ-

cidunaś ca tadgunaih (9) ekavidyo vitaś cānyo hāsyakṛc ca vidūṣakaḥ.

435 pithamardo vitaś caiva vidūṣaka iti trayah, śṛṅgāre narmasacivā nāyakasyā

'nunāyakāḥ. For the date of the Agnipurana see P. V. Kane, History of Dharma

Sastra I, p. 170ff. (about 900 A.D.), History of Sanskrit Poetics4, p. 9 (after 1050 ?),

S. K. De, Sanskrit Poetics I, p. 99 (“later than the middle of the 9th century”).

436 According to Bhat (n. 344), p. 177 n. 2.

437 See R. Schmidt, Beiträge zur indischen Erotik, p. 144.

438 3.39 dārunvartini syāt tasya prasaṅgiketiṛte tu, kiṃcit tadgunahīnāḥ sahāya

evā 'sya pithamardākhyaḥ (40) śṛṅgāre 'sya sahāyā vita-ceta-vidūṣakādyāḥ syuh,

narmasu nipuṇāḥ kupitavadhu mānabhañjanāḥ śuddhāḥ. (3.46) uttamāḥ pithamardādyāḥ

madhyau vita-vidūṣakau, tathā śakāra-cetādyā adhamāḥ parikīrtitāḥ.

439 eṣām nāyikānukūlane pithamarda-vita-ceta-vidūṣaka-nāmānah sahāyāḥ.

440 He also gives pithamarda, vita, ceta and vidūṣaka according to Bhat, p. 177 n. 2.

Bhānudatta, Rāсамаñjarī fol. 72b 441, and the Rasaratnahāra 52–53,

which quotes Bhānu 442.

Everywhere the ceṭa is inserted between the vīṭa and the vidūṣaka, in

spite of the fact that these two belong to a higher category than he.

R. Schmidt rightly remarks in his book on erotics: “Den nur für das Drama

in Betracht kommen den ceṭa lassen wir hier unberücksichtigt . . . 443 ”.

The evidence quoted allows but one conclusion: there is no indication

to prove the existence of either court-jesters or of professional buffoons

in general and all references in literature to the vidūṣaka are ultimately

derived from the dramatic character. Huizinga, who had to base his

conclusion on very slender evidence, has here, as often, intuitively grasped

the real state of things. It is significant that his words quoted in the

beginning of this section, viz. “Outside the drama the vidūṣaka proper

does not, accordingly, seem to have developed” imply the preconceived

idea that there was a dramatic character, first and foremost, who might

have “developed” in social life but did not. There is a curious conflict

here between an intuitive insight and a certain lack of logic. How could

dramatic characters, who belong to the stage, develop to figures in social

life? It follows that the vidūṣaka whom V. S. Agrawala thinks he has

detected on a terracotta panel in the Mathura Museum 444 must, if the

identification is correct, also portray a dramatic scene.

Gonda, who disputed Huizinga’s conclusion 445, rightly recognized that

the real problem is the fact that the characteristic traits of the vidūṣaka

cannot be explained from reality and belong to the stage but he mis-

interpreted, like Konow 446, the occurrence of the vidūṣaka in the

Kāmasūtra and the Daśakumāracarita.

A few words must finally be said about the other narmasacivas, the

pīṭhamarda and the vīṭa.

As for the pīṭhamarda, this character is unknown to Bharata. It has

been suggested that he was no longer used as a dramatic figure at that

time. However that may be, not before Rudraṭa (between 825 and

850 A.D.) 447 and Dhanamjaya (between 974 and 996 A.D.) 448 is he

mentioned by theorists of dramaturgy. Bhavabhūti substitutes him for

441 teṣāṃ ca narmasacivān pīṭhamardo vīṭas tathā, ceṭo vidūṣaka iti caturdhā Bhānuno

'ditaḥ. Cf. Schmidt, Beiträge, p. 199. On the date of Bhānudatta see S. K. De,

Sanskrit Poetics I, p. 246, P. V. Kane, History of Sanskrit Poetics, p. 308.

442 See Schmidt, p. 144.

443 Beiträge, p. 146.

444 See Bhat, p. 54 with plate.

445 See Acta Orientalia 19 (1943), p. 407f. He then adds that it should not be

inferred, however, that the character of the vidūṣaka only occurs in dramas and

refers to the Daśakumāracarita and the Kāmasūtra.

446 See Konow, Das indische Drama, p. 15: ‘. . . nicht bloss im Drama . . .,

sondern auch z.B. im Kāmasāstra voll entwickelt vorlag”.

447 Kāvyalāṅkāra 12.13f., Śṛṅgāratilaka 1.39f.

448 Daśarūpaka 2.8. For further references see R. Schmidt, Beiträge zur indischen

Erotik 2, p. 142f.

232

the vidūṣaka449. Dhanika in his commentary on the Daśarūpaka450 illustrates the rules regarding the pīthamarda by a reference to Makaranda in the Mālatīmādhava and Sugrīva in the Rāmāyaṇa. For all practical purposes, however, he was unknown in theatrical art. Hence the conflicting definitions given by later theorists451. Cf. Kāmasūtra 1.4.31!

There remain two problems. Apart from Mhbh. IV, App. 21.11 crit. ed. (IV.21.33 Bombay ed.) rathinah pīṭhamardāś ca hastyārohäś ca naigamāḥ, where some kind of warrior seems to have been meant452, the word is met with in the Aupapātika Sūtra as a term for a State official453. It is not necessary for our purpose to enter more deeply into this matter.

On the other hand, if it is true that Vātsyāyana has taken over the triad pīthamarda-viṭa-vidūṣaka- from an earlier rhetorician, while the Nāṭyaśāstra does not recognize the pīthamarda as a special character, there must consequently have been a different school of theorists of dramaturgy anterior to the Kāmasūtra.

As for the viṭa, he might at first sight seem to have a better chance of having been a special character in the society of an Old Indian town. This was particularly defended by Gonda454.

To the authors of the Nāṭyaśāstra he is simply a stage character. It gives a definition (35.77) and prescribes that he is to be used in the prakaraṇa (20.53) and the bhāṇa (20.110 dhūrta-viṭa-samprayojyaḥ). As S. K. De observes455, “The viṭa, usually neglected in the serious drama,

449 See Bhat, pp. 39, 93.

450 ad Daśarūpaka 2.8.

451 Rudraṭa, Kāvyālankāra 12.13 kupitastriprasiddhakaḥ; similarly Bhānudatta, Rasamañjari and, for all three narmasacivas, Rudraṭa 12.13-15, Viśvanātha, Sāhityadarpana 3.40 and Vāgbhaṭa, Kāvyānuśāsana, p. 62. Agnipurāṇa 339.40 has pīṭhamardah samb(h)alakaḥ “matchmaker”, whereas Sāgaranandin, NLRK. 2201 explains “The king's friend in his intrigue with the courtesan is called viṭa; he who teaches her her craft is the pīṭhamarda” (vesyām prati sak(h)ā rājño viṭa ity abhidhīyate, tadvittācāryakam prāptaḥ pīṭhamardaḥ prakirtitaḥ). Similarly Halāyudha. Nilakaṇṭha's gloss rājapriyāḥ is certainly no more than a wild guess, based on nāyakasya prīya 'pi ca', which he quotes from the Medinikośa.

452 Otto Stein, Jinist Studies, p. 85f.

453 Acta Or. 19, p. 408: “Dass auch der viṭa-, “Lebemann” oder lustiger Gesell eine bekannte Figur in bestimmten Kreisen des altindischen Lebens, nicht nur eine Theaterrolle was, brauche ich nicht nachzuweisen; er wird in mehreren Werken der Sanskrit Literatur erwähnt”. Gonda refers to R. Schmidt, Beitr. zur ind. Erotik, p. 144f. and Kathāsaritsāgara 6.51ff., PW, VI, col. 1030. Schmidt's material, however, is entirely taken from dramaturgic handbooks. As for the Kathāsaritsāgara, see below.

454 Sanskrit Poetics II, p. 270 n. 25. Later theoreticians mention some kinds of plays in which a viṭa is acting, such as the durmallī or durmallikā, first mentioned, it seems, in the Śṛṅgāraprakāśa (between 1000 and 1050) and the Bhāvaprakāśana (between 1175 and 1250), then in the Nāṭakalakṣaṇaratnakośa (between 1200 and 1250), which describes it in detail (3136ff.): it consists of four acts in which successively the viṭa, the vidūṣaka, the pīṭhamarda and the nāgara appear. Similarly Sāhityadarpana 6.303-305. As an instance is quoted the play Bindumatī. According to the Nāṭakalakṣaṇaratnakośa 3149 the viṭa appears as an upanāyaka in the Prasthāna (cf. Lévi, p. 148).

233

except in Cārudatta and Mṛcchakaṭikā, appears in all his glory in the Bhāṇa, for which he is prescribed as the hero". It should be added that the viṭa occurs as a character in the Nāgananda, Act III and that the words eṣa hi Mṛdaṅgavāsulako nāma purāṇanāṭakaviṭaḥ in Śūdraka's Padmaprābṛtaka 20.5 have been interpreted as referring to the role of a viṭa in old dramas456, although the term admits of different explanations. For the viṭa as a special figure in social life, however, the evidence is virtually non-existent. In the four earliest bhāṇas of the so-called Caturbhāṇī the viṭa is already a conventional stage-character457. Therefore, they do not allow any inference about the occurrence of such a character in the actual town-life of that time.

Apte, however, rightly remarks458 that in non-dramatic Sanskrit literature viṭa means "1. A paramour; (Māl. 8.8; Śiś. 4.48). 2. A voluptuary, sensualist (BhāgP. X.1327)". This distinction, which is better than that of PW. VI, col. 1030, accounts for almost all the passages where the word occurs.

1. "Lover, paramour": Mālatīmādhava VIII.8ab

tvadvallabhaḥ kva nu tapasvijanasya hantā, kanyāviṭaḥ patir asau parirakṣatu tvām

"Where, then, is your beloved, who only kills ascetics? Let this violator of girls (commentary: kanyādūṣakaḥ), your lord, now protect you!" For pati "legal lover" (=priya) cf. Daśak. 205,15 AgashE. In Śiśupālavadha IV.48 madhukaravitāpanāmitās tarupanktīr, where there is a pun (madhu-karaviṭāḥ and madhukara-viṭaṅṅaṅnumitiṣ), madhukaraviṭā means "the bees acting as lovers". In Pūrṇabhadra's Pañcatantra, p. 210 line 19 it is said of a puṁścalī: atha pūrvaparicitaviṭagrhe gatvā "Having gone to the house of her paramour, with whom she had become acquainted before"459. In Edgerton's reconstruction the text reads dūtikayā viṭam ānayya "having sent a female messenger for her paramour" and in the same text he occurs again p. 335 line 5 asmiṁś (cā) 'ntare (saśapathāṁ) viṭenā 'bhiḥītā "meanwhile the paramour said to her", but at the end of this story this dayitajana (p. 334 line 7) is called her jāra (p. 337 line 1). This is, indeed, one of the common meanings of viṭa. It is thus used in Pūrṇabhadra, p. 223 line 19 tasya bhāryā puṁścaly aṅyāsaktamanā ajasraṁ

456 See G. H. Schokker, The Pādatāḍitaka of Śyāmilaka, pt. I, p. 43. Improbable is the interpretation by Moticandra and Agrawala, Śṛṅgār-hāṭa (Bombay 1960), p. 26, according to which the person concerned had formerly been an active actor in plays performed in the brothel but now, owing to his age, had become a mere viṭa. 457 See Schokker, p. 45 n. 137. 458 V. Sh. Apte, The practical Sanskrit-English Dictionary, 2nd ed. (1912); similarly in the new ed. by Gode and Karve, vol. III (1959), s.v. 459 Similarly Kosegarten, Textus ornatior, p. 136 line 1 and the edition Benares 1930 (The Haridas Sanskrit Series No. 13), p. 132 line 24. Less clear is the reading in the Bombay Skt. Series (ed. Bühler), IV, p. 18 line 16 tatas ca paricitāṁ kami cid viṭagrhaṁ gatvā.

234

viṭāya sakhaṁda(ghṛtān) ghṛtapūrān kṛtvā . . . prayacchati, and p. 224 line 16 atha tasyā hṛdayavallabho viṭas . . .460

2. “sensualist”: In this more general sense the word occurs, e.g., at Daśakumāracarita p. 136, 21 (ed. Agashe) vitavidheyatayā “on account of the king being in the power of the apostles of sensuality” 461 and similarly in the Pūrvapīṭhikā, p. 2,19, where the young Kāmapāla is said to be viṭa-naṭa-vāranārī-parāyaṇo durviniṁtaḥ “only caring for bon-vivants, actors and harlots, and undisciplined” ; further at Kathāsaritsāgara 6.51 viṭaprāya “a man who was a bit of a roué” 462, 6.58, where viṭa refers to persons frequenting a brothel and also 32.166 vidagdhā api vañcyante viṭavarnanayā striyaḥ “Even clever women are deceived by the tales of a rogue” (impostor, transl. Tawney). In his translation of the Rājatarangiṇī Aurel Stein translates viṭa by “roguish”, “parasites”, “wicked men”. The commentary Laghudīpikā ad Daśakumāracarita 42.8 quotes “gamyo viṭaḥ pāllaviko bhujanggaḥ” iti Bhāgurih. Viṭas did exist, consequently, but not as the stereotyped characters as which they have become known from the plays ever since the Cāruḍatta and the Caturbhāṇī. Sometimes, it is true, they are mentioned in connection with professional groups, as in the Pūrvapīṭhikā quoted above and, e.g., in Mārkaṇḍeya Purāṇa 68.26

karoti gāyatāṁ vittaṁ nṛtyatāṁ ca prayacchati bandināṁ atha sūtanāṁ viṭānāṁ lāsyapāthinām

“He bestows wealth on singers and dancers, and on minstrels, bards, sycophants (and) those who are skilled in drama” (transl. Pargiter), cf. Rājatarangiṇī 5.352 viṭa-bāndy-ādi-.

Therefore, the case of the viṭa is different from that of the vidūṣaka. He did exist in Old Indian society but, like the ceṭas “servants”, he had at the time of the Nāṭyaśāstra already become a conventional stage-character. The existence of a viṭaśāstra (quoted by the Bṛhatkathāśloka-saṁgraha 10.69) points to the same conclusion. Cf. also Śāradātanaya, p. 289,11 viṭaḥ prākṛtavādī ca prāyo . . . The evidence quoted leads to the following conclusion: Bharata mentions separately, as typical stage-characters, the sūtra-dhāra and his pāripārśvika, the viṭa, the śakāra, the vidūṣaka and the ceṭa 463. In two texts, which may be counted among the earliest of this kind of literature, viz. the Kāmasūtra and the

460 Similarly Kosegarten, p. 199, lines 8–9 and 25; ed. Benares 1930, p. 141, line 30 and p. 142, line 12. Not found in other recensions.

461 Bühler-Agashe; cf. Kane, note on p. 198, 9: “his parasites or companions in sensuality”. Meyer translates “da der Herrscher ganz von den liederlichen Lebemännern abhāngig war”.

462 Tawney-Penzer, Ocean of Stories I, p. 64.

463 35.66–80C.=24.91–104 KM (=p. 654, interpolation after 35.20). For the ceṭa see 35.80. In the KM. edition he is said to be a gandhasevaka (24.104, p. 655, line 17) “using fragrances”. He must have been different from the gāndhika “seller of perfumes” (n. 424). Ghosh, however, emends to bandhasevaka (35.80 C.) “giving service under bondage”.

Daśakumāracarita, three already occur in the fixed formula pīthamarda-viṭa-vidūṣaka. The possible theory that the description of the Nāṭyaśāstra

is older than this formula would not account for the fact that Bharata does not mention the pīthamarda. The early date of the Kāmasūtra, when

the triad had already become formulaic, excludes the idea that the pīthamarda was a late intruder in the drama. It is from the drama that

Vātsyāyana must have taken this triad, as it is highly improbable that he has created it himself. The formulaic way in which it is used in the

Kāmasūtra and general considerations lead to the conclusion that the triad must be a trace of scholastic reflexion on dramatic art, which has

found its way into the handbook on Erotics. Since the pīthamarda does not occur in the extant dramas (apart from some late attempts at a revival),

he must have belonged, like the genres Ḍima and Samavakāra to an earlier but almost forgotten stage of development of the Sanskrit drama.

If this is true, the formulaic triad must indeed, as has been suggested above, have originated in a dramaturgic school which was different

from Bharata's.

As for the vidūṣaka, the final conclusion must be that there is nothing to show that he was at any time more than a stage-character.

20. THE NĀYIKĀ AS THE LEADING LADY

In § 12 (p. 171) the tradition about the guardian deities as found in the Nāṭyaśāstra has been discussed. It was shown that this necessarily

leads to the conclusion that the nāyikā must, at an early stage of development of the Sanskrit drama, have been considered one of the

principal parts (pradhānapātrāṇi, as Abhinavagupta has it). Just as the nāyaka has Indra for his guardian deity, and the vidūṣaka the Oṃkāra,

so it is Sarasvatī who is the special patroness of the nāyikā. Since this is mentioned in the first chapter of the Nāṭyaśāstra, which is characterized

by some Vedic reminiscences and which has preserved the oldest traditions about the origin of the drama, it is a reasonable conclusion that the idea

of three special guardian deities is a heritage from the same pre-classical period to which some other traditions must be ascribed.

In the preceding study the main stress had to be laid on the ambivalent relation between the nāyaka and the vidūṣaka, which was shown to reflect

a structural antagonism between Indra and Varuṇa. For that purpose the problem of the nāyikā could be ignored as being of no direct relevance

to this antithetic relation. In the light of the theory of the identity of Indra and nāyaka, however, the relation between the nāyikā and Sarasvatī

must now be considered more closely.

In contrast with the vidūṣaka the nāyikā has never been considered to be only of secondary importance. Comparatively late dramaturgists

such as Sāgaranandin, who no longer recognize the vidūṣaka as one of the principal characters, still continue to mention the nāyikā along with

the nāyaka as the leading parts. To some extent this might be explained from her prominent role in some classical nāṭakas and prakaraṇas

(Śakuntalā, Vasantasenā, etc.). Since, however, the prototypes of the

nāyaka and the vidūṣaka could be found in Vedic mythology, there is some reason to look for a similar mythological prototype of the nāyikā. In that case the first thing that comes to mind is the old Vedic pattern of Devas and Asuras both wooing a divine maiden.

From the brāhmaṇas onwards the texts refer to a fixed type of contest between Devas and Asuras with a young goddess at stake. A well-known instance is their strife to win Vāc by calling her to their respective sides (vi-hvā-). Vāc stands in a particularly close relation to Prajāpati, the World-Father, who impersonates both the primeval Oneness and the totality of the dual cosmos. Like Prajāpati himself (ŚB. IV.5.7.2), Vāc thus stands above the duality of the differentiated cosmos. As such she is free to choose according to her own will. In ŚB. III.2.1.18-21 however, the following tale is told: “Now the gods and the Asuras, both of them sprung from Prajāpati, entered upon their father Prajāpati's inheritance: the gods came in for the Mind and the Asuras for Speech . . . (19) The gods said to the Sacrifice: “That Vāc is a woman: beckon her, and she will certainly call thee to her” . . . He accordingly beckoned her. She, however, at first disdained him from the distance . . . (20) . . . He beckoned her; but she only replied to him, as it were, by shaking her head . . . (21) . . . He beckoned her, and she called him to her . . . (23) The gods then cut her off from the Asuras . . .” (tr. Eggeling).

There is some difference between the two tales in that the first has as its central motif a contest between Devas and Asuras, both wooing Vāc, whereas in the second the gods win her stealthily. What the stories have in common is the fact that the gods owe their success to their psychological insight. In both tales the position of Vāc with regard to the two parties is characterized by her freedom to choose for herself.

For the problem here under discussion it may be of interest that in the brāhmaṇas Vāc is identified with Sarasvatī. In the Mahābhārata, where Vāc but seldom occurs as a goddess, Sarasvatī has taken over

464 See MS. III.7.3 (78,3), KS. XXIV.1 (90,10), KKS. XXXVII.2 (195.20-228,4), TS. VI.1.6.6, ŚB. III.2.4.4-6 (Devas and Gandharvas).

465 For Vāc's relation to Prajāpati cf. KS. XII.5 (167,15), XXVII.1 (137.8), KKS.XLII.1 (246.2/287,2) Prajāpatir vā idam āsīt, tasya vāg dviṭīyā 'sīt, tāṁ mithunam samabhavat . . ., PB. XX.14.2 Prajāpatir vā idam eka āsīt, tasya vāg eva svam āsīt, vāg dviṭīyā, ŚB. VI.1.1.9 vāg eva 'sya. Different ideas, according to which Vāc is secondary to mind, in ŚB. VIII.1.2.8 (cf. VS. 13.58 iyám upā́ri máttis, tásyai vāṁ mátyà), X.5.3.4. In III.2.1.18, however, Vāc is on a par with Manas (mind). Cf. J. Muir, Original Sanskrit Texts V, p. 392.

466 Cf. MS. II.3.9 (37,4), III.6.4 (64,1), KS. XII.10 (173,6), XXIII.2 (75,3), KKS. XXXV.8 (184,12/214,21), TS. VII.2.7.4-5, ŚB. III.1.4.9 and 14, III.9.1.7, V.2.2.13, V.3.4.25, IX.3.4.17, XIII.1.8.5, XIV.2.1.15 vāg vai Sarasvatī, ŚB. VII.5.1.31, XI.2.4.9 máno vai Sārasvatī, vā́k Sárasvatī, AB. III.1.10 vā́k tu Sarasvatī, III.2.10 vāg ghī Sarasvatī. Cf. Hillebrandt, Ved. Myth. II2, p. 338 n. 1.

467 See, e.g., XII.224.55 [=233 (232).25 Bomb. ed.] anādinidhānā nityā vāg utsr̥tā Svayaḿbhuva (Nilakanṭha: vedamayī divyā) and XII.231.8, where there is an identification: jihvāyāṁ vā Sarasvatīha vāg bhūtā śarīraṁ te pravekṣyati and XIV.21.13 tasmād ucchvāsam āsādya na vakṣyasi Sarasvati : (15) tasmād ucchvāsam āsādya na vāg vadati karhicit.

237

her function. She here bears the title “Mother of the Vedas” 468, and she is incidentally said to be the daughter of the All-god Brahmā, who corresponds to the Vedic Prajāpati 469.

In the light of these facts it is possible that Sarasvatī as the patroness of the nāyikā is historically related to Vāc of the brāhmaṇas 470. In any case, if the interpretation of the nāyaka and the vidūṣaka in terms of Vedic mythology is correct, only a contest between Devas and Asuras can account for the importance that was apparently attached to the part of the nāyikā. If this is true, however, the fact has to be considered that the same pattern is in later literature also found with respect to another goddess, who is much more prominent, viz. Śrī.

In the brāhmaṇas Śrī is said to be a daughter of Prajāpati 471. As such she holds the same intermediate position, between and above the two parties, that is characteristic of Vāc. This mythological trait of hers has particularly been elaborated in the Mahābhārata. Gösta Johnsen, who has given a good account of the material in IIJ. 9 (1966), p. 252ff., rightly stresses the importance of the underlying pattern. He points out 472 that “Śrī is not without a will of her own. It is she herself who decides to leave the Asura and go to Indra. She states that nobody knows her, neither Devas, nor Asuras. Nobody can be sure of possessing her”. The fickleness of the divine Rājaśrī, the impersonation of a king's good fortune, who may abandon him at any moment to support his adversary, is a well-known motif of classical poetry 473. It need hardly be added that, side by side with this deified Fortuna, the word śrī could express a more impersonal power, as in Mhbh. II.17.16 eṣa śriyam samuditām svarājñāñām grahiṣyati.

One of the most interesting instances of this role of Śrī occurs in the tale of the Churning of the Ocean, in the course of which the Devas win Śrī by churning the cosmic waters 474. The Amrtamanthana is a classical instance of the cosmogonial contest between Devas and Asuras. Since, according to the Indian tradition (p. 195), it was also the subject-matter of the first dramatic performance, it is possible that Śrī's part in this drama was important. In that case she would have been the prototype

468 XII.326.52 Śriyam Lakṣmīṁ ca kīrtiṁ ca prthivīṁ ca kakudminīm, vedānāṁ mātaram paśya matsyaṁ devīṁ Sarasvatīm. Cf. Harivaṁśa II.120.15 (=10244= App. I.35.29) vedānāṁ mātaraṁ caiva Sāvitrīṁ bhaktavatsalām.

469 See XII.330.10 rtā brahmasutā sā me satyā devi Sarasvati and cf. XII.121.23 yathoktaṁ brahmakanaye 'ti Lakṣmīr nitīḥ Sarasvati.

470 It is not quite clear what exactly M. M. Ghosh had in mind when in his note on NS. 1.96 (Translation 2 p. 13 note) he wrote that with Sarasvatī the Vedic goddess of that name was meant.

471 ŚB. XI.4.3.1 Prajāpatir vai prajāḥ srjāmāno 'tapayata. tásmāc chrāntāt tepānāc chrír úd akrāmat.

472 Op. c., p. 254, where he refers to Mhbh. XII.218.7f.

473 Cf. e.g., Mudrārākṣasa II.12, Daśak. p. 15, 7 Agashe (and p. 4, 19). See also Gonda, Aspects of early Viṣṇuism, p. 189. In Beiträge zur Geschichte der Göttin Lakṣmī by Gerda Hartmann (doctor's thesis, Kiel 1933), p. 33, there is only a casual reference to Lakṣmī's function as “Orts- und Stadtgottheit”.

474 See n. 482.

238

of the pre-classical nāyikā, the last reminiscence of whom is still preserved

in the Nāṭyaśāstra. Apart from this conjecture it may be pointed out

that Śrī's emerging from the primeval waters is mythologically equivalent

to her being sprung from Prajāpati. Originating in the undifferentiated

world she stands above the duality of the two opposed parties. Her

fickleness as a woman (as the Indian poets saw it) is, indeed, deeply rooted

in mythological notions about the undifferentiated primordial world.

In Mhbh. II.61.58ff. Vidura tells an old story of a dispute between the

Daitya Virocana and the brahmin of Angiras' race Sudhanvan about a

maiden (v. 59 kanyāhetok). In this contest Johnsen (p. 252) has recognized

an instance of epic dualism 475. In V.35.5 this samvāda is said to have been

about Keśinī (keśinyarthe). From another passage, which relates how a

fight arose between the Deva Indra and the Asura Bali 476 Johnsen rightly

infers that the kanyā and Keśinī are identical with (or should we perhaps

rather say: mythologically equivalent to?) Śrī 477.

It seems to me that Johnsen has brought to light an important mythical

pattern, which may be interpreted as follows: The Devas and Asuras

were fighting about Śrī 478 but since the divine maiden, born of the all-

encompassing totality, stood above the duality of the phenomenal world,

neither party could “know” her (XII.218.7), as she by her very nature

transcended them both. She was, indeed, free to choose her partner among

the Devas as well as the Asuras. In the story told in the Harivaṃśa Bali

vanquishes Indra (by using his vāruṇāstra “Varuṇic arrow”, which is

mightier than Indra's!). III.64.20 =14005=App. I, 42B, 2401) and Śrī

comes to him because of his sattva (III.65.9ff. = 14029ff. =App. I, 42B,

2445f.).

In the light of these facts it may not be surprising that the tradition

preserved in NŚ. 1.96, which, for several reasons, is probably the last

reminiscence of the oldest form of the Sanskrit drama, classes the nāyikā

among the three principal parts. It must be admitted, however, that of

the old pattern that has been reconstructed here no trace can be found

in the classical drama. The most important lesson, however, which can

be learnt from an attentive study of the first chapters of the Nāṭyaśāstra

is that it would be a fallacy to think that a prehistoric stage of the drama

can be reconstructed from the classical nāṭaka. In the latter the nāyaka

need not compete with an adversary to win the young girl, who is mostly

already enamoured of the hero from the first moment she has seen him.

If we look back at the older obsolete forms of drama, for which we have

to rely entirely on the brief definitions of the theorists, there is in the

Īhāmrga actually a competition between a divine nāyaka and a pratināyaka,

475 See also above, p. 20.

476 See Mhbh. XII.218 and cf. Harivaṃśa III.64.13ff. (=13998ff.=App.I, 42B. 2363).

477 In Mhbh. XII.218.7-8 Śrī says na mā Virocano veda na mā Vairocano Balī . . .

tvam mām Śakra na jāniṣe sarve devā na mām viduḥ. See p. 51 n. 164 !

478 Mhbh. XII.34.13b idam ca śrūyate Pārtha yuddhe devāsure purā, Asurā bhrātaro

jyeṣṭhā devāś cāpi yaviyasaḥ (14) teṣām api Śrinimittaṁ mahān āsit samucchrayah,

yuddhaṁ varṣasahasraṇi dvātriṁśad abhavat kila.

both well-known from tradition but the latter being an evil-doer (ayuktakṛt) by inadvertence. Although the Daśarūpaka defines the Īhāmṛga as "a contest about a woman" 479 the attempt on the part of the pratināyaka to win her by violence 480 must have excluded a free choice of the "divine maiden". She must, therefore, have been more an object of the dramatic action than a leading part.

Thus, looking for a kind of drama with a prominent nāyikā, we are again led back to the legendary samavakāra, to which type the mythical first drama, the Churning of the Ocean, is said to have belonged. The number of nāyakas prescribed for this kind amounts, it is true, to as many as twelve (see above p. 196f.). This implies that there was a contest between two parties instead of two protagonists, but each of the parties got its profit from the strife. As we have seen, Dhanika in his commentary on the Daśarūpaka adds the explanation "Just as in the Churning of the Ocean Vāsudeva etc. win Lakṣmī etc." 481. Here, indeed, the pattern is found in which the presupposed prototype of the nāyikā could take an active part in the action. After appearing at the surface of the Milk-ocean Śrī chooses of her own free will the Devas 482.

In conclusion it may be useful to summarize briefly the preceding discussion and to state which inferences would seem justified on the basis of the scarce data at our disposal. It is obvious that no well-founded history of the development of the nāyikā can be reconstructed. Particular reference can only be made to the following details. According to Indian tradition the oldest drama had the Churning of the Ocean for its subject and was a samavakāra. A samavakāra is defined by the dramaturgists as a contest for a woman and Dhanika illustrates it by referring to Vāsudeva's winning of Lakṣmī. It has generally been inferred that this definition merely describes a single specimen of this genre, a Samudramanthaṇa, which had been preserved and which must still have been known at the time when the definition of the Nāṭyaśāstra was made. From the old mythological pattern it may be inferred that in such a play Śrī/Lakṣmī must have played a prominent part because she had an independent position with respect to the two parties. It remains dubious to what extent we may rely on Dhanika's words "the winning of Lakṣmī, etc." (Lakṣmyādilābhaḥ) in assuming such a part of a nāyikā in a play for which the theory only prescribes twelve male heroes. Structurally there was a similarity and parallelism between Śrī of the epic and the Vedic

479 3.76 strīhetur iha saṃgarah, SD. 6.250 divyastrīhetukaṃ yuddhaṃ. Cf. NŚ. 20.78 divyastrīkāraṇopagatayuddhaḥ. Note v. 81 kevalaṃ amarastriyo hy asmin.

480 DR. 3.74, SD. 6.247 divyastriyaṃ anicchantīṃ apahārādine 'cchataḥ.

481 ad DR. 3.64: yathā samudramanthane Vāsudevādīnāṃ Lakṣmy-ādilābhaḥ.

482 Mhbh. I.16.3f., Matsyapurāṇa 250.2-4: prasannabhäḥ (prasannābhaḥ MP) samutpanneḥ Somaḥ śitāṃśur ujjvalah (34) Śrīḥ anantaram utpannā ghṛtāt pāṇdura-vāsinī, Surā devī samutpannā turagaḥ pāṇḍuraś tathā . . . (36) Śrīḥ Surā caiva Somaś ca turagaś ca manojavah, yato devās tato jagmur ādityapathāṃ āśritāḥ. The Matsyapurāṇa has only vv. 1-4 in common with the Mahābhārata. See also Bhāgavatapurāṇa VIII.8.8.

240

Vāc/Sarasvatī. In what way the latter goddess has become the patroness

of the nāyikā can no longer be determined. The only thing that can be

said is that the part of the pre-classical nāyikā as reconstructed here would

have fitted in very well with the general pattern which it has been the

object of this study to disclose, that is, the interplay of the two principal

male characters of the Sanskrit drama.

ADDENDA

Ad p. 125: According to Kāmasūtra 6.1.12 the nāyaka likes dramatic performances

(prekṣanaka) very much, but the "shows" or "spectacles" (prekṣā), mentioned as

diversions along with cock-fighting and gambling (1.4.25, cf. 1.4.8 and 6.1.25

prekṣanaka) need not refer to dramas. The commentary (13th cent.) explains

sajjīva-nirjīva nāṭyāprekṣādibh, cf. also ad 6.1.25). As for the 31st of the 64 kālās,

viz. nāṭakākhyāyikādarśanam (1.3.15), it apparently means "knowledge of dramas

and short stories" (parijñānam comm.). Of greater interest are the plays (prekṣanaka)

which the nāyaka, who here acts as the head of the nāgarakas, organizes, as part

of a procession (ghaṭā), in the temple of Sarasvatī (1.4.14-18). Every month or

fortnight he commissions some nāgarakas to perform the pūjā (which must primarily

have been worship of the goddess on whose special day the festival takes place)

and instructs the actors. Foreign actors are tested on this occasion. The day after

the performance the nāgarakas offer pūjā to the actors (cf. above, p. 169) and pay

them the amount due. Even though this description may be an idealized and

systematized picture of actual life, it shows that performances, sponsored by rich

citizens, did take place, on the occasion of religious festivals which were also social

events, but in temples instead of theatres.

Ad p. 208: According to Yaśodhara (13th cent.) on Kāmasūtra 1.4.33 the vidūṣaka

reviled the nāgaraka or the courtezan when they were over-bold, which he apparently

states in explanation of the term vidūṣaka: sa ca veśyāṁ nāgarakam vā kvacit

pramādyantam labdhapraṇayatvād apavadate, iti vidūṣakah.

CORRIGENDA

p. 35, line 1 from the bottom, Add : Cf. ChUp. VIII.7.8.

p. 38, line 1 read: verse 5ab.

p. 52, n. 167: delete JB.I.283.

p. 87 n. 328, p. 88 n. 333: read: N. J. Shende.

p. 87, n. 330 (line 2): delete yad.

p. 194, n. 326: Sāgaranandin, NLRK 806 quotes NŚ. 21.90.

241

Excursus: The Sādhyas

The scarce secondary literature on the Sādhyas that I know contains remarkably little information of any value. Most handbooks simply ignore them. See Weber, Ind. Stud. IX., p. 6 n. 2, Muir, Original Sanskrit Texts I, p. 10, V, p. 10ff., p. 17 n. 26, Bergaigne II, p. 76, III, p. 75, Sieg, Die Sagenstoffe des Ṛgveda, p. 15, K. Rönnow, Trita Āptya I, pp. 37f., 105f., Hillebrandt, Ved. Myth. II2, p. 407 n. 2 (“deifizierte Vorfahren”), Geldner note ad RS. I.164.50, Renou, EVP. 16, p. 149, Hopkins, Epic Myth., p. 175, Monier-Williams, Skt. Engl. Dict. The problem was correctly stated by Roth (see Whitney’s translation of AS. VII.5.1). See also, in general, on the problem of the generations of the gods Muir, OST. V, p. 16ff., Macdonell, Ved. Myth., p. 17. By far the best survey of the data has been given by S. Lévi, Doctrine du sacrifice, p. 62f. The basic facts would seem to be the following:

1. They were on earth before the Devas, but are also consistently called Devas. Cf. KS. XXIII.8 (83,12), XXVI.7 (129,19), KKS. [XXXVI.6 ex conjectura], XLI.5 (240,3/279,7) Sādhyā vai nāma devā āsan pūrve devebhyah (teṣām na kim cana svam āsīt, te ‘gnim mathitvā ‘gnau juhvata āsata), PB. XXV.8.2 Sādhyā vai nāma devebhyo devāḥ pūrva āsan, VIII.3.5, 4.1, 4.9 Sādhyā vai nāma devā āsan, MS. III.9.5 (121,1) Sādhyā vai devā āsann, átha vai tárhi nà ‘yud ‘hutir āsīt (cf. III.7.9, 10; 9.4.5), TS. VI.3.5.1 Sādhyā vai devā asmím loká āsan, nà ‘yát kim caná miṣát. The Ṛigvedic poets refer to them as púrve devāḥ “former gods”, see VII.21.7, X.109.4 (cf. 90.7), I.164.50. Cf. perhaps also AS. XI.8(10).10: ten gods devebhyah purā (“before the gods”?). It should be noted in this connection that the statement found in some modern surveys that the Sādhyas were older than the Ādityas is open to doubt. In the Yajurvedic version of the Mārtāṇḍa myth (see Karl Hoffmann, MSS. 11, p. 85ff.) only the Taittirīyas have the detail that Aditi cooked the brahmaudanam for the Sādhyas, who then must have existed in the primordial world even before the Ādityas were born (TS. VI.5.6.1ff., TB. I.1.9.1ff.). The other versions do not mention the Sādhyas, cf. MS. I.6.12 (104,10), KS. VII.15 (78,15)=GB. I.2.15, KS. XI.6 (151,5), ŚB. III.1.3.3-4. This detail is, therefore, most likely a later addition, but even if the Sādhyas did not exist in the primordial world and if they were not originally Asuras (which is, indeed, never suggested in the texts), the fact remains that the later gods were avare “younger, posterior” in respect to the Sādhyas. Cf. KS. XXIII.8 (83,13).

2. They were the first to go to heaven, where they are now. Cf. KS. PB. svargam il lokam āyan and RS. I.164.50, X.90.16 yajñéna yajñám ayajanta devás tāni dhármāṇi prathamāni āsan, té ha nákam mahimānaḥ sacanta yátra pūrve Sādhyáḥ sánti deváḥ. Somehow, accordingly, they must have

been present as Devas at the moment when the dual cosmos was created. Nowhere is it said where they came from (but see Mhbh. I.1.33).

3. That they were no ordinary Devas is shown by the fact that they sometimes stand in clear opposition to the latter, e.g., when the Sādhyas, going to heaven, try to prevent the Devas from following them (KS. XXIII.8: 83,13, a gap in KKS. XXXVI.5, not in MS. III.7.1), just as the gods try to prevent men (TS. VI.3.4.7); or when, going to heaven, they take the sacrifice and Soma with them (PB. VIII.4.1).

4. That this opposition had a structural character is shown by AS. VII.79.2cd, where the goddess of the New Moon says: "In me all the gods of the two parties, both the Sādhyas and those who recognize Indra as their leader, came together "(máyi devā ubhāye Sādhyāś cé 'ndrajyeṣṭhāḥ sámagacchanta sárve). For the "non-technical" use of sárve see p. 57, n. 190 and Renou, EVP. 4, p. 3. The word ubháye points to a binary opposition (see pp. 8 and 44 on ubháye prajāpatyàḥ). This seems to rule out the possibility of the Sādhyas standing for the Ādityas and Aṅgirasas, as is suggested by Sāyaṇa ad RS. I.164.50 (cf. AB. I.16.36-39).

5. In the system of classification the Sādhyas always belong to the centre, specifically, it seems, to the zenith. Cf. ŚB. IV.4.9 pitaraś cā 'dharāyām, Sādhyāś co 'rdhvāyām, VādhS. (Acta Or. 6, p. 130) Aṅgiraso 'dhastāt, Sādhyā upariṣṭāt, cf. ChUp. III.10.1, etc. Only conjointly with the Āptyas, it seems, are they located in the nadir. Cf. JB. II.142 Viśve Devā upariṣṭāt, Sādhyāś cā 'ptyāś cā 'dhastāt. As for AB. VIII.14.3 asyāṁ dhruvāyām madhyamāyām pratiṣṭhāyām díśi Sādhyāś cā 'ptyāś ca devāḥ dhruvā díś is here contrasted with the zenith (ūrdhvā díś). Therefore the possibility should be considered, that it is not the centre in general that is meant here but more specifically the nadir - a possibility which Gonda, Viṣṇuism and Śivaism, p. 17, too rashly denies. Cf. also the commentary ad AS. III.27.5 dhruvád dík: adhodík.

6. The Sādhyas seem to have also been associated with the idea of transcendence and, after they have gone to heaven, with what is beyond the top of the world axis. A yūpa which is thirteen cubits long is of the same measure as the Sādhyas (KS. XXVI.4: 126,13), the thirteen months of the year (ŚB. III.6.4.24) or Prajāpati (MS. III.9.2: 115,7). This number characterizes the Sādhyas as a manifestation of the totality (12+1). To the Sādhyas belongs the uppermost part of the sacrificial stake (yūpa), viz. that part which, one finger-joint long, protrudes above the top-ring (caṣāla). Cf. KS. XXVI.4 (126,20) yūvad uttamam aṅgulikāndaraṁ tāvad atirecayed, yajñaparusā saṁmitam (otherwise MS. III.9.4: 119,4 aṅguli-mātrāṁ kāryàm). Cf. for the Sādhya gods. Superabundant of the sacrifice is that which is superabundant of the sacrificial stake" (KS. XXVI.4: 127,1), "The Sādhya gods despised the sacrifice. To them, verily, belongs that which is made superabundant during the sacrifice. Superabundant during the sacrifice is what fire they sacrifice in fire. That (part) of the sacrificial stake which

243

protrudes above the caṣḍlam, that touches (spr̥śati) them" (MS. III.7.9: 89,6). This idea of atiríkta "superabundant" also explains why in JB. I. 283 the Sādhyas and Āptyas (with whom the Sādhyas in this text always form one group, the sixth and last after Vasus, Rudras, Ādityas, Viśve Devāḥ and Maruts) are associated with the metre atichandas (I.283) and pārameṣṭhyāya (II.25, III.152). The notion of atiríkta is closely connected with that of áparimita in the Brāhmaṇas. While most texts (KS. XXVI.6: 129,6, KKS. XLII.4: 239,5/278,15, TS. VI.3.4.7, cf. ŚB. III.7.1.25, ŚB. IV.4.16) content themselves with the simple statement Índrasya caṣḍlam, Sādhyāmām atiriktam, the MS. III.9.4 (119,3) has the additional detail of the Sādhyas despising the sacrifice: Índrasya caṣḍlam, yé vai devāḥ Sādhyā yajnā́m atyámanyanta tésā́m vā etád yád upā́riṣṭāc caṣḍlasyaā 'ṅgulimātrā́m kāryàṃ. This motif apparently only serves to explain why nothing but the "superabundant" (atirikta) of the sacrifice is allotted to them. Similarly TS. VI.3.4.8.

7. The only conclusion that can be drawn from the data is that the Sādhyas, unlike the Ādityas, were not originally Asuras who had gone over to the party of the Devas. They were Devas in their own right and were somehow "here", when the other (later) Devas conquered the Asuras, their most significant feat being that they were the first sacrificers. When there was nothing else to offer but Agni, they offered Agni (KS. XXVI.7: 129,19) and so ascended to heaven. It is understandable that Sylvain Lévi, p. 63, thought that they had been ousted and obliterated by the Ādityas but their place in the pantheon is so entirely different from that of the Ādityas that the latter can hardly have affected the position of the Sādhyas.

Rönnow's attempts to explain "former gods" in terms of cult changes are hardly correct, because here, as in the other cases, "former" is more likely to be a mythological than a historical notion. No well-founded interpretation can be proposed but this note may be helpful to state the problem.

GENERAL INDEX

adhipatya, 29

Aditi, 51f.

Ādityas, 32

Ādityas and Angirasas, fight between,

63ff.

Ahura Mazdā, 5, 68

altercation and tumult, 165

Angirasas, 61f.

antagonism between Indra and Varuṇa,

22f.

Aṽesta, 23

bamboo staff, 164

banner festival, 170

Baudhāyana, 97

beginning of the New Year, 11

Bergaigne, A., 5, 11, 19, 30, 33, 39, 45f.,

75

bhāratībheda, 185

bhāratīvṛtti, 185

Bhāratīya Nāṭyaśāstra, 36f.

Bhāsa, 78

Bhat, G. K., 116, 119, 207

Bradke, von. P., 5f., 62

Brahmā, 157f.

Brhaspati, 54f., 100

Cakrapāṇidatta, 191

ceremonial grouping of the Gods, 57f.

ceta, 231f.

Chambe State, 27, 84

Churning of the Ocean, 20, 105f., 107f.,

195, 238ff.

Cilappatikāram, 147

classification in mythic cosmogony, 86f.

classification of the Gods, 28f.

consecration, 166

consecration of the playhouse, 162

contest, 112

contrast between Varuṇa and Indra,

43ff.

contrasting pairs, 12f.

cosmic centre, 25

cosmic duality of heaven and earth, 39

cosmogonical fight, 11f., 162ff.

court jester, 205, 224ff.

creation myth, 10f., 105

dance, ritual character of, 126

Daśakumāracarita, 224f., 229f.

De, S. K., 118

Death, 68f., 72, 152

demon driven away from the earth, 91

demons, 76

Devas and Sādhyas, 243

Dhanamjaya, 181

drama, oldest forms of, 114

drama in social life, 123f.

drōmenon, 122f.

dual cosmos, 13

dualism, 79

dualistic structure of the cosmos, 60

dualistic world order, 8

Eggeling, 6, 27, 34f., 237

erection of poles, 159f.

euphemistic reticence, 69

evocatio, 20, 95, 99, 103

Fate, 152

fight of Devas and Asuras, 197ff.

first dramatical performance, 194f.

five-jointed jarjara, 151

fourth man, 169

Gandharva, 94f.

Geiger, B., 68

Geldner, 7, 15ff., 20, 71, 98f., 101

god of sleep, 63

gods of totality, 75ff., 101f.

gods on the divisions of the stage, 155

gold, 163

golden pitcher, 147, 164

Gonda, 105, 115f., 206f.

Grassmann, 16

Haraprasad Shastri, M., 116

Heesterman, J. C., 42f., 191, 222

Held, 105, 209

Hillebrandt, 11f. n. 29, 21, 23, 30, 32f.,

113f.

Hopkins, 36, 77, 88, 93

Horsch, 112ff.

Huizinga, J., 206, 209, 223, 232

Indra and Varuṇa, 167

Indradhvaja, 25

Indra festival, 26, 30, 136f.

Indra-pole, erecting of, 136

Indra-pole, 138f., 141

Indra's banner festival, 129f.

245

jar, 146

jar, breaking of the, 163

jar, earthen, 164

jarjara, 161f., 168

Johnsen, Gösta, 238

Kāmasūtra, 226f.

Kane, P. V., 120

Kauśika Sūtra, 136

Kauṭalya, 225f.

Kāy Kāūs, 100

Keith, 124, 137f.

Konow, 184

knowledge of cosmic mysteries, 97

legendary first performance, 128

lists of tutelary deities of the theatre, 153f.

list of gods, 147ff.

Lommel, 100

Lord of the waters, 74f.

Mahāvrata ceremony, 115

Mārtāṇḍa myth, 242

Maypole, 140f.

Meyer, J. J., 137

Mitra, 156

Mitra and Varuṇa, 7

Mitrā-Varuṇā, 63

Mitanni king Šattiwaza, 68

Mount Mandara, 108

mythical names for regions, 55ff.

nāgaloka, 82f., 85f., 89

nāgaraka, 228f.

nāndī, 170

Nārada and Mātali, 82

nāsadasīya-hymn, 13

nāyaka and vidūṣaka, 209f.

nāyikā, 236

nether world, 88f.

Norman Brown, W., 6, 105

numbers, 104

Oberhammer, G., 190f.

Oldenberg, 24f., 29, 69

Oṁkāra, 173ff., 222

order of the three groups of gods, 47f.

pātāla, 87f., 89

Patanjali, 110f.

piṭhamarda, 229ff., 232f.

Prajāpati, 102

Prakaraṇa and Nāṭaka, 211ff.

praṇava, 174f.

presents of the various gods, 144f.

primordial hill, 17

primordial world, 16f., 107

pūrvarāṅga, 122, 166, 182, 186, 189

pūrvarāṅga, two attendants of the, 190

quadripartite classification of gods, 49f.

quarrel among the Devas, 53

rainy period, 134ff.

Rāmāyaṇa, 77ff.

re-enactment of primordial myth, 34

reiteration of cosmogonical strife, 43

Renou, 5, 70, 71, 110

reorganization of the original world, 19

Sādhyas, 243

sadyahkrī ritual, 64f.

samavakāra, 195ff., 240

Śāradātanaya, 181

Sarasvatī, 236

scapegoat, 222

second creation, 15

sensualist, 235

shift of worship, 40

sleep, 31

Somadeva Sūri, 226

sons of Sagara, 82

Śrī, 20, 102, 238f.

stage, 189

stage, illumination of the, 165

status quo, 101ff.

strife between Gods and Asuras, 34

strife between two parties of Gods, 3f.

Sūtradhāra, 167

Sylvain Lévi, 75, 118

symbolical representation of the cosmos, 154f.

taboo, 176

theatre, 126f., 147ff.

Thieme, 110, 117, 200

Thomas, 83

three groups of gods, 46f.

tree, entrance of the, 132

Trigata, 177, 186, 189

two assistants, 192

underworld jar, 86

Uśanā, 93ff., 99f.

Uttarīka-episode, 83

Vāc, 237

vaihāsika, 202f.

Varuṇa and evil, 208

Varuṇa's ambiguous character, 92f.

Varuṇa's noose, 70

Varuna's world, 90f.

Vedic cosmogony, 13ff.

Vedic reviler, 208

verbal contest, 191f., 209

vidūṣaka, 193, 199ff., 205. 214ff., 217,

218, 220ff.

Viṣṇu, 38, 54, 81, 102, 106

Viśvarūpa, 101

Viśve Devas, 50ff., 58

viṭa, 233ff.

vithi, 184f.

vithyanga, 183

Vodskov, 19

war in the Mahābhārata, 212

water, 27, 85f.

Winternitz, 200, 223

world beyond the world of order, 35

world tree, 139

Yama, 61, 156

247

INDEX OF SANSKRIT WORDS

āmhas, 44, 135

akaram, 16

angahāra, 158

adyasutyā, 66

ádhipati, 27, 56

ádhipatya, 29, 80

antár, 16

apagara, 207

apa-śās, 104

abhigara, 207

abhicāra, 72

abhivādana, 167

avabhrthá, 218

ásat, 13f., 38, 81

asuravivara, 83

āji, 42

āpah, 17

ārabhati, 196

āhava, 166

indradhvajamahā, 116

upakarana, 143, 164

upajikā, 83

úpanaddha, 21

upamantrāyati, 103

upáriṣṭāt, 54

ṛtāvan, 23

ekāṣṭakā, 30

kaksyā, 155

kalaśa, 146

kālakūṭa, 107

kuṭilaka, 145

kunakhín, 220

ketu, 139

kóśa, 146

kriḍanaka, 228

kséma, 44

kṣveḍita, 166

khanḍa, 220

caṣāla, 243

cihna, 138

ceṭa, 231

jarjara, 139f.

jalpa, 190

jumbakā, 167

jyógāmāyāvin, 74

dandakāṣṭha, 145

dānavá, 17

dā́nu, 17, 85

dípsantah, 70

dīkṣita, 168

devā́suraḥ, 11

drúh, 59

dhāman, 48

dhúrti, 59

dhruvá dīk, 57

nāgaloka, 85f.

nātaka, 113

nāṭya, 113

nāndí, 125, 128, 170

nikhāta, 31

nirhūya, 103

niśācara, 91

patāká, 124

parākād, 98

parāvat, 98

párvata, 17

pātra, 146

pāpmán, 12, 72

prāvṛṣi, 136

púr ayodhyā́, 147

pūrvaranga, 114, 120

pūrve devā́h, 242

pratiṣṭhā́, 27, 89

prayatnatab, 163

prayoga, 121

prarocanā́, 184

prāśnika, 171

baṇḍa, 220

baddha, 129, 143

bandhá, 73

bṛhatī dīk, 57

bhṛṅgāra, 144, 164

manḍala, 154

mártya, 95

mukuṭa, 144

mrtyú, 12

medhā́, 41, 96

248

yajatá, 13

yajñíya, 13

yamasādana, 74

yūpa, 144

rakṣāvidhi, 36

rakṣitṛ, 56

raṅgapīṭha, 157

raṅgapūjana, 158

rāṣṭrá, 26

vayasya, 205

várunagṛhīta, 74

varunadeva, 84

varuṇālaya, 85f.

vādyavidhāna, 217

vāruṇayoga, 77

vāstu, 37

viṭa, 234

vitaṇḍā, 186, 190

viḍrava, 165

vívāc, 22f.

vivṛtam mahābilam, 83

vihavá, 103

vṛtrá, 17

vedhás, 97 n. 383

vaidheya, 206

vaira, 89

śiñjāra, 96

śailūṣá, 114

śreyas, 77

śvahsutyā, 66

sácā, 98

sát, 13f., 38, 81

sáḍas, 52

sabhā, 85

samudrá, 86

savana, 54

sáhas, 96

sāttvatī, 196

sudevá, 84

skambhá, 140f.

svápna, 31

svaráj, 26

svàṛṣāti, 112, 115, 165

249

INDEX OF SANSKRIT TEXT-PLACES

Agnipurāṇa

339. 39

227

20. 128

180

Aitareya-Brāhmaṇa [AB]

I. 16. 39

64

22. 27

187

Atharvaveda-Saṃhitā (Śaunakīya) [AS]

I. 10. 1

41

23. 151

215

I. 10. 3

71

27. 8

216

II. 3. 3-4

84

27. 13

194

II. 10. 2–8

70

32. 460

171

III. 27. 1-6

28, 56

33. 224

175

IV. 15. 12

7

35. 79

215

IV. 16. 5ab

38

35. 93

215

V. 11. 1

6

Brhad Āraṇyaka Upaniṣad

V. 13. 1

88

I. 4. 12

59

V. 24. 4

28

Chāndogya-Upaniṣad

VI. 80. 3

141

III. 1–5

59

VI. 108. 3

41

III. 6–10

59

VII. 65. 3

222

III. 11. 4–5

59

VII. 79. 2cd

243

VI. 2. 1

14

X. 5. 44

73

Daśākumāracarita (ed. Agashe)

X. 7. 29–30

141

p. 41, 1

83

X. 8. 13

14

p. 42, 8

229

XI. 8. 19

31

p. 130, 21

224f.

XII. 1. 37

20

Gopatha-Brāhmaṇa

XII. 3. 55–60

56

II. 2. 2

53n

XVI. 6. 10

70

Harivaṃśa

XVIII. 3. 25–29

58

II. 91. 25 Bomb. ed.

123

XIX. 56. 1

62

Jaiminīya-Brāhmaṇa [JB]

XIX. 56. 3

31

II. 115–116

65f.

Bhāratīya Nāṭyaśāstra [(Bh) NŚ]

II. 141–142

48

1. 51–59

143

II. 142

57

1. 59–63

144

II. 155

21

1. 96

172

II. 243

55

1. 102–107

37

III. 187–188

65f.

2. 29

124

III. 193

28

3. 62–64

157

Kādambarī [Kād.]

3. 91–93

165

p. 70, 11

225

4. 1–2

195

p. 217, 2

68

4. 269–270

126

p. 227, 2

83

5. 29–30

187

Kāṭhaka-Saṃhitā [KS]

5. 136–140

177f.

XII. 3 (164, 18)

16

5. 141

186f.

XIII. 2 (180, 15)

73

13. 138–140

214f.

XIII. 2 (181, 7)

72

13. 141

216

XVII. 19 (264, 10)

70

14. 56–59

196

XXVI. 2 (124, 4)

70

18. 179–180

180

XXXIV. 3 (37, 19)

21

18. 179b–180a

180

Kauṣītaki-Brāhmaṇa [KB]

18. 264

181

III. 7. 17ff.

17

20. 64–65

195f.

XV. 3. 5

17f.

20. 112

183

Kautilya

1. 12. 9

225

1. 21. 13

225

2. 27. 25

225

250

7. 17. 42

77

I. 151. 4

7

Kuvalayamālā

139

I. 155. 6

106

Mahābhārata [Mhbh]

I. 16. 7–9

108

I. 164. 14

29

I. 16. 12–13

105

II. 3. 4cd

50

I. 16. 18–21

107

II. 27. 10

6f., 39

I. 17. 28

90

II. 28. 7

6

I. 19. 3

82

II. 31. 1

51

I. 57. 18

133

III. 8. 8

48f.

I. 71. 5–6

94

III. 20. 5

48f.

III. 98. 3–5

91

III. 29. 14

16f.

III. 105. 18–24

82f.

IV. 2. 5

21

V. 16. 31, 33–34

29

IV. 42

22, 42f.

V. 16. 33–34

80

V. 2. 2

19

V. 96. 5–6, 9

82

V. 31. 8

98

V. 97. 1

87

V. 34. 2

96

V. 108. 12

82

V. 63. 3

7

VIII. 30. 77 crit. ed.

92

V. 108. 3

29

IX. 46. 5–12

79f.

V. 109. 8

29

XII. 91. 21–22

20f.

VI. 20. 11

95

XII. 4497 Calc. ed.

92

VI. 22. 4

7

XII. 221. 26–27

20

VI. 48

43

XIII. 140. 3ff.

91

VI. 68. 3

45

Maitrāyaṇi-Saṁhitā [MS]

VII. 10. 4

50

II. 4. 3 (40, 19)

16

VII. 35. 6

50

II. 5. 6 (55, 2)

74

VII. 35. 14

48

II. 8. 9

55

VII. 36. 2

7f.

II. 13. 21

56

VII. 59. 8

70

III. 9. 1

70

VII. 61. 5

59

Mālatīmādhava

VII. 65. 2

7

VIII. 8ab

234

VII. 82–85

43ff.

Mālavikāgnimitra

VII. 85. 3

45

I. 4

123

VII. 99. 4–5

106

IV. 15. 36 (29)

145

VIII. 7. 26

98

Mārkaṇḍeya Purāṇa

VIII. 25. 4

7

68. 26

235

VIII. 35. 1

51

Mṛcchakaṭikā

VIII. 42. 1

6

I. 42. 15

146

VIII. 43. 9

43

Pañcaviṁśa-Brāhmaṇa [PB]

VIII. 58. 2

14

XVI. 1. 1

14

VIII. 69. 12

27, 84

Rāmāyaṇa (crit. ed.)

VIII. 77. 1–2

52

VI. 14–15

85

VIII. 96. 9

7f., 15

VII. 23

89

VIII. 101. 15

51f.

Ratnāvalī

IX. 87. 3

97

II. 5. 12

145f.

IX. 97. 7

97

Rgveda-Saṁhitā [RS]

IX. 97. 41

20

I. 24. 14

6

IX. 113. 8

90

I. 51. 10–11

96

X. 14. 3–6

61

I. 54. 10

17

X. 14. 7

12

I. 83. 5

97f.

X. 22. 6

95, 98

I. 89. 10cd

50

X. 40. 7

95

I. 101. 3

30

X. 48. 11

48

I. 108. 6

103

X. 63. 4

59

I. 121. 12

96

X. 66. 1–2

26

I. 130. 9

98

X. 66. 3

50

251

X. 72. 2

14

VIII. 6. 3. 3

53n

X. 97. 16

71

XI. 1. 2. 8

35

X. 111. 5

141

XI. 1. 6. 9

40

X. 124

22ff., 32f., 38

XIII. 3. 6. 5

218

X. 124. 2

21

Taittirīya-Āraṇyaka [TĀ]

X. 124. 4a

19

I. 2. 3

219

X. 124. 4–5

15f.

Taittirīya-Brāhmaṇa [TB]

X. 124. 4, 6

20

II. 2. 9. 1

14

X. 124. 5

80

Taittirīya-Saṁhitā [TS]

X. 125. 1

50

II. 1. 4. 3

62

X. 128. 9

48

II. 4. 12. 4–5

16

X. 129. 2

14

II. 5. 1. 1

101

X. 129. 5

14

II. 5. 2. 2–3

18

X. 132. 4

7, 35f.

II. 5. 2. 5

39

X. 150. 1

51

III. 4. 5. 1

28

Śatapatha-Brāhmaṇa [ŚB]

V. 3. 4. 7, VI. 4. 8. 3

14

I. 2. 4. 8–11

34

V. 5. 10. 1–2

56

I. 6. 3. 13

18

VI. 2. 2. 1

53n

II. 1. 4. 10

35

Vājasaneyi-Saṁhitā [VS]

III. 5. 1. 13

64

XV. 10–14

55

III. 5. 1. 13–22

65f.

XXXVIII. 9

61

V. 2. 5. 17

72

Viṣṇupurāṇa

VI. 1. 1. 1

14

I. 9. 88–90

106

VIII. 4. 3. 16

27

Note ad Bhāratīya Nāṭyaśāstra :

References are, unless otherwise indicated, to the Calcutta edition (C): The

Nāṭyaśāstra Ascribed to Bharata-muni, by Manomohan Ghosh, I2 Calcutta 1967;

II (Bibl. Ind. 272A), Calcutta 1956.

B (Baroda edition) refers to M. Ramakrishna Kavi, Nāṭyaśāstra of Bharatamuni

with the Commentary Abhinavabhāratī by Abhinavaguptācārya, I2 (GOS 36) by

K. S. Ramaswami Sāstrī, Baroda 1956; II (GOS 68) by M. Ramakrishna Kavi,

Baroda 1934; IV (GOS 145) by M. R. Kavi and J. S. Pade, Baroda 1964.

KM (or KM2) refers to the 2nd ed. of Kāvyamālā 42: The Nāṭyaśāstra by Srī

Bharatamuni, ed. by Pandit Kedārnāth, Bombay 1943. KM1 refers to The Nāṭyaśāstra

of Bharata Muni, ed. by Pandit Śivadatta and Kāśīnāth Pāṇḍurang Parab,

Bombay 1894.

KSS (or K) refers to Kashi Sanskrit Series 60: Bharata-muni-praṇītam Nāṭyaśāstram

ed. by Batuk Nath Sharmā and Baldeva Upādhyāya, Benares 1929; only incidentally

quoted from other works.

Raghuvamśa (ed.): Bharat kā Nāṭyaśāstra (adhyāya 1–7, mūl, pāṭhāntar, anuvād

tathā vyākhyā). Bhāg I. Dilli-Vārāṇasī-Patnaḥ 1964.

Translation by Manomohan Ghosh: I2, Calcutta 1967; II (Bibl. Ind. 272),

Calcutta 1961.

252