A comprehensive vision was never wanting, though the search for a word for * religion/ or religious practice, as a whole was at first not very successful. The IMhmaTza- texts still struggle with the notion of the superiority of him that knows all the Vedas, and they consequently posit a sarvavidya1 which is superior to a knowledge of each of the Vedas. The most successful attempt at describing the religious literature and action as a whole is the word brahma, and, correspondingly, he who knows the religion as a whole is a brahman. Each of these words appears occasionally in the fourth place, brahma after the trayi ; brahman in company with the priests of the trayi. In a sense the brahma is a fourth Veda, but it is not co-ordinate with the other three ; it embraces and comprehends them and much else besides; it is the religious expression and religious action as a whole, and it is the learned esoteric understanding of the nature of the gods and the mystery of the sacrifice as a whole (brahma in brahmodya and brahmavadin). Needless to say, this fourth Veda, if we may so call it, has primarily no connection with the Atharvan, not even in the Atharva-sa^hita itself (XI, 8, 23 ; XV, 3, 7 ; 6, 3), nor in the Upanishads of that Veda (e.g. Nrzsi//zha- purvatapani Up. V, 2) : the claim that the Atharvan is the Brahma-veda belongs to the Atharvan ritual. In the Upanishads this brahma, still frequently contrasted with the ordinary Vedas, is taken up eagerly, extolled above all other knowledge, and in a way personified, so that it fur- nishes one of the main sources of the various conceptions which finally precipitate themselves in the pantheistic
1 Tait. Br. Ill, 10, 11, 4; Tait. Ar. X, 47; cf. Sat. Br. XIV, 6, 7, 18; 9,4. 17-
lxiV HYMNS OF THE ATHARVA-VEDA.
Brahman-Atman. The knowledge of this brahma consti-
tutes the brahmavidya, which is separated by the widest
imaginable gap from the Brahma-veda in the Atharvanic
sense ; cf. above, p. xliii.
This broader religious knowledge exists again from earliest times, not only in the abstract, but centres in persons who grasped it in its entirety, in distinction from the technically qualified priests devoted to some speciality. What the brahma is to the trayi, that the brahman is to hotar, adhvaryu, &c. Thus the important stanza, RV. X, 71, ii, depicts the activity of four priests at a jrauta-sacri- fice, the hotar (rzkam posham aste pupushvan), the udgatar (gayatra/zz gayati ^akvarishu), the adhvaryu (ya^vzasya matram vi mimite), and the brahman. The latter is de- scribed in the words, brahma vadati ^atavidyam, ' the Brah- man tells (his) innate wisdom V The association of the first three priests with the three Vedic categories, rik, saman, and ya^-u^, is expressed with a degree of clearness com- mensurate with the character of the hymn, which is in the nature of a brahmodya. But the brahman has no peculiar Veda ; certainly there is no allusion to the Atharvan. His knowledge is that of the entire Veda, the sarvavidya (Tait. Br. Ill, 10, 11, 4), religious knowledge as a whole. By means of this knowledge he is able to assume in the ritual practices the function of correcting the mistakes of the other priests, whose knowledge is more mechanical. The Brahman is as it were the stage-manager in the sacerdotal drama, the physician of the sacrifice when it is attacked by the disease of faulty execution (5at. Br. XIV, 2, 2, 19) ; he is the mind of the sacrificer (Sat. Br. XIV, 6, 1, j)2. As such he is also conversant with the mystic aspects of the divine powers, the powers of nature, and the details of the sacrifice. In the expression, brahma vadati ^atavidyam, the ' own wisdom ' is the brahma (neuter), and vadati ^ata- vidyam foreshadows the brahmodya, ' the holy, or theo-
1 Cf. RV. I, 10, 1 ; II, 1, 2 ; IV, 22, i ; VI, 38, 3- 4 J VII, 33, 14 ; X, 52, 2 ; X, 91, 10.
2 Cf. Weber, Ind. Stud. X, 135 ff. ; Hang, Brahma und die Brahmanen, p. 9 ff. ; Ludwig, Der Rigveda, III, 28 ff.
INTRODUCTION. lxV
logical mystery, or riddle V as well as the ritualist refine- ments which the Brclhma/za and Sutra-texts introduce times without end with the closely-related expression, brahma- vadino vadanti. In the non-Atharvanic Vedic texts it is never suggested that the Atharvan is the specific equip- ment, above all other things, which shapes the faculties of this all-round Vedic theologian. On the contrary, the Kaush. Br. VI, n raises the rather one-sided claim that a Rig-veda scholar is the proper Brahman 2. Vasish/^a was a celebrated Brahman and Purohita, and the qualifications for this office were said for a time (probably by the descen- dants of Vasish/^a themselves) to be especially at home in this family. But the B rah ma/za- texts declare explicitly that this is an ' iiberwundener standpunkt,' an obsolete custom : every one properly equipped may be a Brahman ; see Weber, Ind. Stud. X, 34. 35. 137. There is no original connection between Vasish/^a and the Atharvan 3, and it is not going too far to assume that the distinguished abilities demanded by the theory of this office were rare enough to admit every one that had intrinsically valid claims upon it.
How, then, did the Atharvans come to raise the plea that the Brahman must be one of themselves, and that, consequently, the Atharva-veda was the Brahma-veda? Schematically this was suggested by an obvious proportion. As the hotar, &c, is to the Rig-veda, &c, so the Brahman is to the fourth Veda, and as the Atharvan is the fourth Veda, or rather a fourth Veda, it required no too violent wrench to identify it with that other comprehensive fourth Veda, the knowledge of the brahma. Thus the Atharvan
1 See the author, Journ. Amer. Or. Soc. XV, pp. 172, 184 ff.
2 A broader view, yet one that ignores the Atharvan claim, is taken by Apastamba, in the Ya^a-paribhasha-sutra 19. There the Brahman is said to perform with all three Vedas. Only the commentator admits that the Atharvan may be included. See Max Miiller, Ancient Sanskrit Literature, p. 470 ; Zeitschr. d. Deutsch. Morgenl. Gesellsch. IX, p. xlvii ; Sacred Books, vol. xxx, p. 321. Cf. also Sat. Br. XI, 5, 8, 7, and Madhusudana's statement of the final orthodox view, Max Miiller, ib. 445 ff. ; Ind. Stud. I, 4. 14.
3 The interesting association of VasishMa with the redaction of the Atharvan, reported by Mallinatha in his comment on Kiratar^uniya X, 10, may be founded upon this very title to the office of purohita, and thus show that purohitas were naturally supposed to be Atharvavedins ; cf. above, p. lv.
lxvi HYMNS OF THE ATHARVA-VEDA.
became the Brahma-veda. The fact that there was no systematic sharply-defined provision for the Atharvanists in the scheme of the hieratic religion must have been galling at first, until this arrangement was completed to their own satisfaction. They may have, though we do not know that they did, gathered courage for this tour de force by the frequent mention in the AV. itself of the word brahma in the sense of charm, prayer, e.g. I, 10, i; 14, 44; 23, 4, &c. If this was done it was a proceeding both arbitrary and superficial: the word has in the AV. the meaning of charm only in so far and inasmuch as the hymns of that Veda happen to be charms ; the RV. employs the term freely to designate its own suktini (e.g. V, 85, 1 ; VII, 28, 1 ; 36, 1 ; X, 13, 1 ; 61, 1). One misses, too, the plural brahmam as the true Vedic type of designation for a special class of composition, on a level with rika/i, samani, ya^uwshi, atharvangirasa^, or athar- vanak (bhesha^-ani) and angirasa^ (abhi^arikam). We may also remember that the Atharvan of all Sa/^hitas contains the largest collection of theosophic hymns which deal explicitly (X, 2), or implicitly (X, 7), with Brahman and the brahma x. This may, of course, have helped to suggest that the Atharvavedin was the truly superior theo- logian. In the Upanishads the knowledge of just such theosophic relations is styled the brahmavidya. Saya^a in the Introduction to the AV., p. 4, argues that the AV. is known as Brahma-veda because it was revealed to Brahman who is called Atharvan 2. His authority, however, is Gop. Br. I, 4 ff., a text that elsewhere identifies the AV. with that bhuyish//£am brahma which was produced by the tapas (cf. AV. VIII, 10, 25), pressing to an unwarranted degree the relationship of the Atharvan texts with the sphere of the Upanishads 3 ; cf. above, p. lix.
It may be safe to assume that all these and other notions
1 Cf. also the superabundant Upanishads, composed in Atharvanic schools.
2 atharvakhyena brahma/za drzsh/atvat tannamna aya;;z vedo vyapadisyate.
3 Similarly the Vishnu- pura^a VI, 5 (Wilson's translation, vol. v, p. 210) : ' The AV. also states that there are two kinds of knowledge. By the one which is the supreme, God (akshara) is obtained ; the other is that which consists of Rik and other Vedas.'
INTRODUCTION. lxvii
flitted through the minds of the systematic theologians R , . f of the Atharvan schools as they continued the purohita to insist upon the name Brahma-veda for to t e ' their scriptures, and upon the office of Brah- man for their priests. A measure of substantiality may, however, come to their claim from another quarter at a comparatively early time, in this instance with the passive support of all Vedic schools. The matter concerns the office of the purohita, the spiritual and temporal aid of the king, his chaplain, and chancellor. One would again look in vain in the non-Atharvanic Sa^hitas, Brahma/^as, or Sutras for the direct declaration that the purohita either was, or should be, an adherent of the Atharvan. These texts do not mention the Atharvan in this connection any more than in connection with the office of the Brahman at the sacrifice. Yet it seems extremely unlikely that the knowledge of Atharvan practices should not have been considered a very valuable adjunct, if not a conditio sine qua non, of the purohiti. Purohitas, whether they are formal adherents of the AV. or not, are always engaging in Atharvanic practices, even against one another (cf. Max Muller, Ancient Sanskrit Literature, p. 486). The interests of the king and his sovereignty (kshatriya and kshatram) are too obviously dependent upon magic rites to admit the likelihood that the pretensions to this office on the part of him that knew them should have been ignored. At all periods the safety of the king, the prosperity of his people, his ascendency over hostile neighbours, must have depended upon the skill of his purohita in magic. The description, Ait. Br. VIII, 24-28, of the purohita, his func- tions, and his relation to the king, transfer the reader to the sphere and spirit of the Atharvan. The purohita secures for the king royalty, strength, emf^re, and people (VIII, 24, 7). The purohita is a fire with five flaming missiles, dangerous when not properly propitiated ; but, duly honoured, he embraces the king, protecting him with his flames as the ocean the earth (VIII, 25, 1). His people do not die young, his own life's breath does not leave him before he has reached the full limits of his life, he lives to
lxviii HYMNS OF THE ATHARVA-VEDA.
a good old age, if a Brahma/za, imbued with this know- ledge, is his purohita, the shepherd of his kingdom. The subjects of such a king are loyal and obedient (VIII, 25, %. 3). The prescriptions regarding the purohita are fol- lowed (VIII, 25) by a magic rite, called bvahvaanak pari- mara, designed to kill hostile kings, which might have found a place in the ritual of the Atharvan1. In later texts, as a matter of fact, the rule is laid down formally that the purohita should be an Atharvavedin. Thus in Gaut. XI, 15. 17 ; Yagnav. I, 312 (cf. also Manu XI, 3$) ; see p. xlviii, above. Saya/za in the Introduction to the AV., pp. 5, 6, claims outright that the office of purohita belongs to the Atharvanists (paurohitya;^ ka atharvavidai*va ka- ryam), and he is able to cite in support of his claim not only the rather hysterical dicta of the Atharvan writings, but also j-lokas from a number of Pur£/zas, the Niti^astra, &c. ; cf. above, p. lvi 2. In the Da^akumara-^arita magic rites, as well as the marriage ceremony, are in fact per- formed at the court of a king with Atharvan rites athar- va;zena (atharva/zikena) vidhina, and the statement is the more valuable as it is incidental ; see above, p. lv.