FOR the latest stages of the Vedic religion on its practical side the authorities are the Śrauta and the Gṛhya Sūtras,1 which deal with different but complementary spheres, and which incidentally preserve for us a considerable amount of formulae, prose and verse, which by accident or other cause have not found a place in any of the Samhitās preserved to us. The necessity of some manuals for the actual practice of the great complicated rites of the sacrifice must have been felt from an early period, but we have not now extant any of these manuals. The Śrauta Sūtras which are now extant are all without excention later than the older Brāhmaṇas, and, while the ritual which they reveal is in general harmony with that supposed by the Brāhmaṇas, it would be idle to suppose that it actually represents it with perfect accuracy. This can be seen by one simple point: the Brahmanas often show that on questions of the exact mode of the performance of certain rites there were considerable difference of opinions in some cases the Brāhmaṇas reject definitely certain views, in others they allow varied views to stand as equally legitimate, but in the Sūtras in the great majority of such cases merely one view is laid down, the others having presumably come to be disapproved in the school in which the Sūtra arose. On the other hand, the Sūtras often give optional forms of procedure for which the Brāhmaṇas contain no hint, evidence of the development of practice in the schools. Moreover, there is clear proof that no Sūtra represents rigidly any one Samhita: even when, as is normal, a Sūtra follows generally some Samhitā it is quite ready to accept portions of its material from another.2 The Sutras, therefore, while often giving valuable confirmation and explanation of the Brahmaṇas, cannot be regarded as contemporary evidence of the practices of the Brāhmaṇas, and this conclusion based entirely on the ritual is confirmed by many lines of evidence. In addition to dealing with many rites which seem clearly elaborations and modifications of older rites, the Sutras in language are markedly more modern than the Brahmaṇas, approximating closely to the classical speech, from which they differ in the main in the use of forms of incorrect grammatical formation.3 From this fact a conclusion may fairly be drawn with regard to their chronology: it
1 The Dharma Sûtras, unquestionably later
on the whole than the Gṛhya Sūtras, are valuable as confirming the latter, but the age of new matter in them is doubtfully Vedic.
*e.g. Apastamba uses the other Samhitās as well as the Taittirfya. A direct
descent from the Brāhmaṇas is asserted by Caland, Das Śrautasūtra des Apastamba, pp. 1 ff., but this is not certain, nor very probable.
- Wackernagel, Altind. Gramm. i. pp.
xxxii ff.
28
The Sources
[Part I
can scarcely be supposed that works, not popular in character, which so flatly disregard in some points the rules of Pāṇini, should have been produced after the general acceptance of the authority of that grammarian, which falls probably in the third century B. c. at latest, and thus the period of the Sutras may be roughly set down at from 400 B. C. to 200 B. C., though neither date can be regarded as more than approximate.1
Of the two sets of Sutras the Śrauta deal with the elaborate forms of the ritual in which the presence of a priest, and usually of several, was necessary, while the Gṛhya Sūtras deal with the household ritual, most of which could be performed by the householder for himself without extrinsic aid of any kind. In all probability the literary development of the household ritual was later than that of the Śrauta ritual. It is, of course, perfectly obvious that domestic rites must be as old as any form of religion, but there is a clear difference between this fact and the question of the date of the application to the simpler rites of literary forms, and the verses which are associated with the Grhya ritual show clear traces in language and metre of not belonging to the earliest stage of Vedic poetry. On the other hand, in the case of the existing Sutras, they are compositions emanating from schools which were interested no less in the Gṛhya than in the Śrauta ritual, and the normal school manual seems to have embraced both topics. If the portions dealing with the two different topics were of different dates, the fact can hardly now be detected.2
Of the extant Sutras of the Rigveda there are two complete collections, the Aśvalayana and the Śâñkhāyana Śrauta and Grhya Sūtras: the former is undoubtedly the older, and its reputed author may be assigned with reasonable probability to about 400 B. C.3 The Samaveda has the Śrauta Sutras of Maśaka, Lāṭyāyana and Drāhyāyaṇa, and Gṛhya Sūtras by Jaimini, Gobhila, and Khadira. In the case of the Black Yajurveda Sūtras are especially frequent, including the very important Manava, the Baudhayana, Bhāradvāja, Apastamba and Hiranyakeśi, covering both the field of Śrauta and Gṛhya rites the White Yajurveda is represented by the Katyāyana Śrauta Sutra and the Paraskara Gṛhya Sūtra. The Atharvaveda has the most important in some way of all the Sutras, the Kauśika, which is invaluable as bearing a very close relation to the text of the Veda, and preserving in many cases what seem perfectly accurate accounts of the magic rites which were accom- 1 Keith, JRAS. 1909, p. 591, n. 2; Tait-
tirīya Saṁhitā, i. pp. xlv, xlvi; Rigveda Brāhmaṇas, p. 44; Hopkins (CHI. i. 249) places ApDS. in the second century B.C.
- Oldenberg, SBE, xxix and xxx. On the other hand, the Sūtras are often clearly interpolated, alluding to later customs, e. g. the lunar tithi, and the practice of marking the body with sectarian marks.
3 Keith, JRAS. 1907, p. 411; 1909, p. 591,
n. 1; Taittiriya Saṁhitā, i. pp. xlv f., clxxii ff. The author of the CCS. and ŚGS. is Suyajña, and there is a parallel to the CGS. in the Śāmbavya Gṛhya Sūtra (Oldenberg, SBE. xxix. 4 ff.; IS. xv. 4 ff.); ŚGS. v and vi are late.
⚫ Definitely late are the Vaikhānasa and Vārāha Sūtras. The Vādhúla may be earlier; for Katha texts, cf. Caland, Brāhmaṇa en Sūtra-Aanwinsten (1920).
Chap. 3]
The Later Literature
29
panied by the formulae in the text: when the Atharvavedins became desirous of assimilating their Veda in every possible manner to the three older Vedas, they invented an orthodox Śrauta Sûtra, the Vaitāna, to accompany it, and from the Vaitāna more directly and indirectly from the Kauśika is derived much of the matter of the curious work, the Gopatha Brahmana, which poses as the Brāhmaṇa of the Atharvaveda, and which, borrowing largely from the Aitareya and Śatapatha Brāhmaṇas with other texts, is in essence a pamphlet in exaltation of the Brahman priest and the Atharvaveda.1 The date of this remarkable composition is unknown: it is of course more than possible that some of its material is old, even when it is not borrowed from existing texts, for an enormous amount of Vedic literature has been lost, some within quite recent times.
Beside the Śrauta and the Gṛhya Sūtras stand the Dharma Sutras, which are more specially devoted to customary law, but which frequently contain references to religion of these the oldest and ’nost important are those of Gautama, Baudhāyana, Apastamba, and Vasiṣṭha,a but with them and still more with the later law books we pass from the ideas of Vedic religion to those of Hinduism, though the change is of course gradual and without any sharp break. The Gṛhya and Dharma Sūtras, however, are of special value as preserving for us the more domestic side of the religion practised by the ordinary householder, as opposed to the great sacrifices which were confided to the hands of the priests.
The rest of the literature is of less importance. Some value attaches to the Buddhist texts, especially such works as the Petthavatthu, which gives a fullness of view on the question of the state of the dead according to the popular belief which has every sign of age and genuine tradition. But, though these texts undoubtedly have in them much popular belief, the date of the Buddhist canon is now no longer to be placed so high as was once held, when it was believed that much of the canon really represented views prevalent in the time of the Buddha, and the use of Buddhist evidence for the Vedic period must therefore be subject to the most close scrutiny. The same consideration applies to the great epics. The redaction of the Mahabharata was not completed in all probability until the fourth century A. D. and possibly even later its earliest form cannot now be restored," and its evidential value for the period up to 500 B. C. is, therefore, of controversial character, and the
Cf. Keith, Taittiriya Saṁhitā, i. pp. clxix f.; Rigveda Brāhmaṇas, pp. x,
-
For the late date of Vasistha see Hopkins, CHI. i. 249, against Būhler, SBE. xiv. p. xvii. The Arthaśãstra, alleged to be of c. 800 B. c., is much later; Keith, JRAS. 1916, pp. 180 ff.; 1920, p. 028 ; Jolly’s ed. (Lahore, 1928).
-
See B. C. Law, The Buddhist Conception of
5
Spirits (1923).
• Franke, JPTS. 1908, pp. 1–80; VOJ. xx. 887; Keith, JRAS. 1909, p. 577; 1910, p. 216; Buddhist Philosophy, ch. i. Hopkins, The Great Epic of India (1899) ; Epic Myth., pp. 1f.; CHI. i, 258; Winternitz, Gesch, der ind. Lit. i, 896; iii. 627 ; Lévi, Bhandarkar Comm. Vol., pp. 99 f.; JA. 1915, i. 122. Cf. Dumé- zil, Le Festin d’Immortalité, pp. x, 4 ff.
30
The Sources
[Part I
Rāmāyaṇa, which may in its origin belong to the fourth century B. C., has been subjected to much later recasting.1
2
From the nature of the sources it follows that for the period up to 500 B. C. there is a continuous stream of trustworthy literary evidence, and that after that date the sources are of less value. The care taken to compose and preserve Sutras for centuries after that period shows the vitality of the Vedic religion indeed in much later times the great sacrifices of antiquity, such as the horse sacrifice, were performed by kings desirous of asserting their high prowess, and in families of priests a many of the other rites prevailed down to at least the nineteenth century. But the old order of things was greatly affected by the rise of Buddhism, which was indeed but one of many conflicting sects, but which attained under the patronage of Aśoka in the third century B. C. a leading place among religions in India. The inroads of foreigners from the north-west, which, commenced by Alexander, repelled for a time by Candragupta and his successors, became constant and effective from the second century B. C., aided in the disintegration of the religion, and materially promoted the development of that popular religion centred in the worship of Civa and Visṇu respectively, which was noted by Megasthenes as the leading feature of Indian religion when he stayed at Pāṭaliputra as the Ambassador of Seleukos to Candragupta. Moreover, it must be remembered that throughout this period the Hinduization of the people was proceeding: the process in question can still be observed at the present day in operation amongst the wild tribes, and in the period B. c. it may confidently be assumed that it was being carried on upon even a greater scale, nor is it wonderful that thus the Vedic religion should gradually lose its distinctive features and assume new forms.
5
The difficulties of applying information derived from the later texts is adequately illustrated by the case of the use of idols.4 The epic shows clearly and indubitably the use of idols of the gods, and both it and Manu mention Devalakas, persons who carry idols about, while the grammarian Pāṇini recognizes the use of the name of a god to denote his idol. On the other hand, it is perfectly clear that save in the latest stratum of the Vedic literature 7 idols are not recognized in cult, and then only in the domestic ritual. What conclusion is to be drawn from such facts? Are we to suppose that idols were
1 Keith, JRAS. 1915, pp. 318-28; Winternitz, op. cit. i. 489 ; iii. 630; Lévi, JA. 1918, i. 5 ff.
- Cf. the modern Agnihotris; Hillebrandt,
Ved. Myth., p. 54, n. 1.
- The representation of deities in human form is also probably to be ascribed to Greek influence; cf. Bloch, ZDMG. Ixii. 648-ff.; A. Foucher, The Beginnings of Buddhist Art, pp. 1 ff.; Thomas, CHI, î. 480.
• Arbman, Rudra, pp. 82 ff.
Contrast
Bloch, ZDMG. lxii. 651; Macdonell, Journ. R. Soc. Arts, 1909, p. 817.
- Hopkins, Epic Myth., pp. 72 ff.
V. 3. 99, with Patanjali ; Ludwig, Festgruss an Roth, pp. 57 ff.; Kielhorn, VOJ. i. 8 ff.; Konow, IA. xxxviii. 145 ff.; Charpentier, JRAS. 1918, pp. 671 ff. Adbhutabrāhmaṇa (Weber, Omina und Portenta, pp. 335 ff.); PGS. iii. 4. 9 are clear; cf. PGS. iii. 14. 8; Gautama, ix. 12; ApDS. i. 30. 22.
7
Chap. 3]
The Later Literature
31
really in popular use among the Vedic tribes, but were not approved by the exclusive Brahmans, to whom we owe the texts? It may be observed that when we take the ritual as a whole there is very little sign of the alleged exclusiveness of the Brahmans, whose character in this regard is assumed through the error of treating the Rigveda and the speculations of the Brahmaṇas as completely representing their views. Other causes are equally possible and more plausible. The use of idols may have been influenced by the non-Aryan population, as it gradually became assimilated; it may have used them and had fixed sanctuaries before the advent of the Aryans, whose lack of idols or sanctuaries may either have been primitive or induced by their migrations, which uprooted their local connexions. Or the use may have been a natural innovation within the Vedic circle of tribes, or introduced through contact with non-Vedic Aryans. There is no proof that the Indo-Europeans practised the use of idols, and the evidence of German religion 1 suggests that the position there as certainly in Iran a was much as in Vedic India, and it is, therefore, perhaps more plausible to believe that their employment gradually developed in India itself, though under what influences we simply do not know. This is certainly more legitimate than to suppose an idolatrous people and an exclusive priesthood. What, however, is essential is to note that Vedic religion is normally aniconic, for the interest of any religious system largely depends on what is peculiar and distinctive and not on that vast mass of beliefs which it must possess in common with other religions.
1 Cf. Helm, Altgerm. Rel. i. 216 ff., 287 f. On Greek religion see de Visser, Die nichtmenschengestaltigen Götter der Griechen, pp. 31 ff. Cf. Carnoy, Les
Indo-Européens, p. 233.
- Moulton, Early Zoroastrianism, pp. 67 f.,