ALREADY in the Rigveda there are signs of considerable elaboration of ritual and of the employment of a number of priests at the sacrifice, and the later Samhitās and the Brāhmaṇas reveal to us a time when the functions of the priesthood have been definitely divided up and apportioned among sets of priests. The manual acts of the sacrifice are ascribed to the Adhvaryu priest and his assistants and are accompanied by muttered formulae, in prose or verse, styled Yajus: in addition at the greatest sacrifices, such as the Soma sacrifice, singers chant Sāmans, and reciters recite Śastras, while the Brahman priest supervises the whole performance, usually in silence. On this division of functions is based the division of the later Samhitās: the Samans are preserved in the song books of the Samaveda, the Yajuses in the Yajurveda, and the Atharvaveda is held to be connected with the Brahman priest, while the Śastras were composed of verses taken from the Rigveda. From the point of view of religion the Samaveda is mainly interesting for its form: the words which were sung were almost invariably taken from the Rigveda, but they were eked out as shown in the song-books, Gānas, with all kinds of interjections, doubtless for musical purposes, which must have converted their character in the most marked degree. The sense of the words cannot possibly have been understood in the mutilated form in which the chants were sung, and the conclusion is inevitable that their religious value lay not in the substance but in the form, so that the Sāmans have been compared, not altogether unaptly, to the revival hymns beloved by the African negro in the new world. In them doubtless the religious excitement of the priest found its fullest scope for expression.1 That this form of chant was old need not be doubted: there are clear traces in the Rigveda itself in the strophic and metrical form of certain of the hymns, that they were from the first intended for something more lively than mere recitation. With this fact accords the generally close relation of the Samaveda and the Rigveda, which renders it probable that that was of the first of the later Saṁhitās to take definite form.
The Yajurveda represents the literary fixing of the formulae used by the
1 Bloomfield, VOJ. xvii. 156 ff.; JAOS.
xxi. 50 ff. On the recensions see Caland’s ed, of the Jaiminiya Saṁhitā (1907); VOJ. xxii. 486 ff.; Oldenberg, GGA, 1908, pp. 711 ff.; Simon, VOJ. xxvii. 805 ff. For the Pūrvārcika there are the Grāmageya and Aranya, for the Uttarǎrcika the Üha and Ühya
Gānas.
- The chief metre of portions recited by the Hotṛ (as shown by internal evidence and ritual use) is Triṣţubh, without strophic form; of those used by the Udgātṛ Gayatri and Pragātha in three and two verse sets; Oldenberg, ZDMG. xxxviii. 489 ff,
17
Chap. 2]
The Later Samhitas and the Brahmanas
Adhvaryu and his assistants in their performance of the great sacrifices: clearly the actual use of such formulae must have been normal from the beginning of the sacrifice, but it was only after the collection of the Rigveda hymns that the idea of creating a similar Samhitā for the Adhvaryu became popular. This is shown unmistakably by the frequent application for the purposes of the Adhvaryu of verses from the Rigveda, in many cases without any real propriety and often with alterations deliberately planned to adapt them to their new use. Whereas the Rigveda has come down to us in but one collection, the Yajurveda 1 is preserved in two main recensions, which at comparatively early date received in India the names of the Black and the White Yajurveda. The origin of these appellations is uncertain, but later they were interpreted in such manner as to suggest that the White Yajurveda owed its name to the fact that in it the formulae of the Adhvaryu were collected separately from the explanatory remarks which accompany them in the texts of the Black Yajurveda. In making this distinction the compilers of the White Yajurveda, which has come down to us in the Samhitā called the Vājasaneyi, were merely restoring the primitive condition of the Yajurveda, which must at one time have consisted of a collection of the formulae, in prose and verse, only. But already at a comparatively early period the formulae were accompanied by explanations, called Brāhmaṇas, texts pertaining to the Brahman or sacred lore, in which the different acts of the ritual were given symbolical interpretations, the words of the texts commented on, and stories told to illustrate the sacrificial performance. Hence in the Black Yajurveda we find three complete recensions, the Taittiriya, Kāṭhaka, and Maitrāyaṇī,3 and one imperfect, the Kapiṣṭhala, in which formulae and Brāhmaṇa are closely allied, while in the case of the White Yajurveda the Brāhmaṇas are all collected in one great work, the most important of its type in Vedic literature, the Śatapatha Brāhmaṇa. Perhaps as a result of this separation, a mass of old material, partly formulae, partly Brāhmaṇa, which had not been incorporated in the Taittiriya Samhita was collected together in the Taittiriya Brāhmaṇa, which in part contains matter more recent than the Samhita, but in part has matter as old as, at any rate, the later portions of that text.
Explanations were not less required for the other Samhitas, and the Rigveda is dealt with in two Brahmaṇas, the Aitareya, and the Kauṣîtaki, the latter of which is far more concise than the former, though it covers in some respects a wider sphere. The Samaveda formed the topic of the great
1 See Keith, The Veda of the Black Yajus School (HOS. xviii and xix), 1914.
- Oldenberg, Prolegomena, pp. 290 ff. * li. 9 of this text is an obvious interpolation mentioning sub-Vedic deities such as Brahman, the four-faced, and lotusseated; Karāṭa, elephant-faced and tusked; Gaurl, mountain born; Viṣnu
2 [1.0.8. 31]
as Keśava Nārāyaṇa, &c. Similar interpolations are found in other Vedic texts, especially in the Khilas, or Apocrypha, of the Rigveda, and as a rule no mention of them is made in this work.
• Ed. and trans. Keith, HOS. xxv, 1920.
18
The Sources
2
[Part I
Pañcavinśa Brāhmaṇa1 and the Jaiminiya Brahmana, the latter of which unhappily exists only in a very imperfect text and has only in part been published, beside a large number of minor and unimportant texts styled Brāhmaṇas, of which the Saḍvinśa, a sort of supplement to the Pañcavinśa, and the Sāmavidhāna are of some value as dealing with magic practices of varied kinds.
Full as are the other Samhitās of magic rites, the Atharvaveda3 differs from them in the fact that, whereas they are essentially connected with the sacrifice, its connexion with that operation is external and mechanical. In essence it is a collection of spells for every conceivable end of human life, spells to secure success of every kind, in the assembly, in public life, to restore an exiled king, to procure health and offspring, to defeat rivals in love, to drive away diseases in every form, to win wealth and so on. But at the same time the subject-matter has been thoroughly worked over by the priesthood, and it has even in its simplest spells throughout a priestly veneer. The priests have also added many spells directly bearing on portions of their sacrificial activities, and the wedding and burial hymns appear in more elaborate forms. Theosophy qua profit-bringing is not absent, and a deliberate attempt was later made to bring the Atharvaveda into the circle of the three orthodox Vedas by the addition to the collection of book xx which contains the hymns to be used by the Brahmaṇācchańsin priest in the ritual of the Soma sacrifice. It is, however, important to note that this Veda, despite the attempts made to raise it to an equal place with the others, never succeeded in achieving this position: useful as were its spells, and much as the priests of the school of the Atharvaveda thrust themselves forward as indispensable to princes through their magic powers, there were always not lacking voices to criticize its claim to be a fourth legitimate Veda, In modern times this prejudice and recognition of the special character of the work are reflected in the suggestion that the text is actually the product of strata of society different from those of the Rigveda : Ridgeway" insists that the Atharvaveda is the record of aboriginal as opposed to Aryan religion. This view, however, cannot be pressed too far: the Atharvaveda reflects the practices of the lower side of religious life, and is closer to the common people than the highly hieratic atmosphere of much of the Rigveda: the common people, we cannot doubt, were largely influenced by aboriginal ideas through mixture with aboriginal races, but, as will be seen below, we have no criterion on which we can safely rely to decide that certain beliefs are non-Aryan and
1 See Hopkins, Trans. of the Connecticut
Acad, of Arts, xv. 20 ff.
- On the kindred, lost, Śāṭyāyana, see
Oertel, JAOS, xviii. 15.
• Trans. in Śaunaka recension by Whitney and Lanman (HOS. vii and viii); on the Paippalāda, in course of ed. in JAOS., see L. C. Barret, Studies in honor of
4
Bloomfield, pp. 1 ff.
5
Edgerton, Studies in honor of Bloomfield,
pp. 117 ff.
• Bloomfield, Atharvaveda (1899), and SBE.
xlii.
- Dramas and Dramatic Dances of non-
European Races, p. 122.
19
Chap. 2]
The Later Samhitas and the Brahmanas
aboriginal. The same problem in effect presents itself as in the case of the Homeric poems. Are we to suppose that they represent Aryan religion, and that that religion was free from admixture with the lower side of religion, which is freely revealed in the later literature of Greece and foreshadowed by the evidence of Aegean cult objects? The answer to that question given by Lang1 in the affirmative seems most improbable, though not more so than the suggestion of Gilbert Murray that the Homeric poems are the result of a process of conscious refining of older tradition. Like the Homeric poems the Rigveda does not cover the whole field of religious belief, and we have no sure ground on which to assign to the non-Aryan as opposed to the other elements in the population all the lower forms of religion.
The later Samhitās are doubtless of various date: the Samaveda must probably be reckoned as the earliest, and the Atharvaveda is certainly the youngest of all in its redaction, though it is doubtless in part old in material. Of the Yajurveda Saṁhitās the youngest is the Vājasaneyi, and the oldest perhaps the Taittiriya, but between it and the other two texts of the Black Yajurveda there is no clear distinction of time. The Brahmanas are certainly later than the formulae of the Samhitas to which they relate, and they are distinguished sharply from them both by their prose form, which is quite different from the prose of the formulae, and by the characteristics of their language, which is much less archaic than the verse or prose formulae. The order in age amongst them, and the prose portions of the Samhitas, which are essentially akin to them is doubtful; it is, however, very probable that the Aitareya in its first five books is among the oldest, that the prose parts of the Yajurveda Samhitas, and, though later, the Pañcavinśa are also old, and that the Kausitaki, Jaiminiya, and Śatapatha are the latest of the important works. For the date of the Brāhmaṇas important evidence is furnished by the development of thought: the latest portions of the texts which are of the older Brāhmaṇa style are styled Āraṇyakas, books intended by reason of the dread holiness of their contents for study in the forests, and of these certain parts which bear a more definitely philosophical aspect are styled Upaniṣads, a word apparently derived from the session of the pupils round the teacher in the process of instruction. Thus there are attached to the Brāhmaṇas of the Rigveda the Aitareya and the Kauṣitaki or Śāñkhāyana Upaniṣads, to the Taittiriya Brāhmaṇa the Taittirīya Upaniṣad, to the Śatapatha Brāhmaṇa the Bṛhadaranyaka Upanisad: the Samaveda has the Chandogya Upanisad which is the major portion of a Brāhmaṇa, and the Jaiminïya Upaniṣad Brāhmaṇa, which is one book of the Jaiminiya Brāhmaṇa, and contains in itself the Kena Upaniṣad. In the main it may be assumed that the doctrines of 1 The World of Homer (1910).
- Greek Epic (2nd ed. 1911). Even Leaf (Homer and History, ch. viii) refuses to accept the theory of expurgation.
• Keith, Aitareya Aranyaka, pp. 21 ff.; Taittiriya Saṁhitā, i. pp. clix-clxxiii ;
2*
Rigveda Brāhmaṇas, pp. 40 ff. On the Jaiminiya cf. Caland, Over en uit het Jaimintya-Brāhmaṇa (1914), pp. 5 ff., whose conclusions are dubious; see for the priority of the Kauṣitaki, Keith, BSOS. I. iv. 177.
20
The Sources
1
[Part I
these Upanisads are prior to the rise of Buddhism, which is derived logically from the system which they contain, and, as the date of the death of the Buddha may be placed with fair probability in or about the year 480 B. C., a lower terminus of 500 B. c. for the Upanisads is attained. The priority of the Brahmaṇas proper to the Upanisads is quite undoubted, and thus a lower limit of about 600 B. c. for the latest Brahmaṇas is obtained, from which may be deduced a date of about 800-700 B. C. for the Samhitās as a lower limit. The same conclusion is indicated by the facts of language: the grammarian Pāṇini, whose date can scarcely be later than 300 B. C.,1 deals with a language which is decidedly more modern than that of the Brahmaṇas to which, however, it is akin: prior to him was Yaska, whose expositions in his Nirukta of Vedic passages indicates clearly that the Rigveda was already far distant in time: earlier again than Yāska was Śakalya, by whom was produced the Pada Patha of the Rigveda, that is the text in which each word is given in its primitive form unaffected by the Sandhi of the Samhita, and earlier again than Śakalya was the making of the Samhitā Pāṭha, in which, to the utter detriment of the metre, the hiatuses which were allowed in the Rigvedic poetry are removed under the influence of the usage and grammatical theories of the day. But the Brahmaṇas as a rule ignore the Samhita text, and evidently knew only the primitive text without the latter rigid Sandhi rules, so that for them again we are forced to accept a date not later than 600 B. C.
Efforts to establish an earlier date for the Samhitas and the Brahmanas have naturally been made, and of these two may be mentioned. Jacobi 2 has insisted that the post-Vedic period may be dated from c. 800 B. c. on the strength of the fact that the end of the Vedic period is marked by the simultaneous appearance of the Sāṁkhya-Yoga and Jaina philosophies, and the latter can be carried back to c. 740 B. C., seeing that the founder of the faith was probably Parśva, whose Nirvāṇa falls 250 years before that of Mahavira, and the latter was contemporaneous with the Buddha, who died about 488 B. C. The argument is extremely unconvincing, apart altogether from our complete ignorance as to the historical character and the date of Parśva. It assumes that the Jain doctrines as we know them go back before Mahāvīra, and that they presuppose the doctrines of the Upanisads as older. Neither proposition possesses the slightest plausibility, and neither need be seriously discussed in the absence of any effort of Jacobi to support his assertion in this matter by reasoned proof.
A second line of argument is based on the war which forms the main topic of the Mahabharata; by various modes of reckoning of dynasties recorded in the Purāṇas the date of 1000 or 1100 B. c.3 is attained for the war,
Efforts to place Pāṇini much earlier are frequent, but his reference to Yavanāni, Greek writing, is difficult to reconcile with a much earlier date than about the 4th century B. c. Cf. Keith, Aitareya
Aranyaka, pp. 21 ff.; Rigveda Brāhmaṇas, p. X.
- Die Entwicklung der Gottesidee bei den
Indern (1928), pp. 24 f.
- CHI. i. 275, 306 f. An excellent reductio
The Later Samhitās and the Brāhmaṇas
21 Chap. 2]
and, as the Pandus are unknown to the Samhitas and the Brahmaṇas, it is contended that they must fall before the war of the Kurus and the Pandavas. It is difficult to appreciate the naïve credulity which accepts as having any value these late lists of kings, which are preserved to us in works dating at soonest fifteen hundred years after the alleged date of the war, and which, when they come into contact with known facts, immediately reveal themselves as without value. Thus into the dynastic list of Kosala we find that the eponymous founder of the Śakya line, the Buddha’s father, he himself, and Rāhula have all been interpolated without the slightest historical justification, and it seems puerile, in the face of these facts, to insist on regarding these lists as the basis for chronological calculations of any kind. When the conflict between the Kurus and the Pandavas took place we do not know, and the assumption that it represented a vast struggle in which all the peoples of Northern India at any rate were engaged, because in the Mahābhārata in its final form it is so represented, argues a signal forgetfulness of the powers of poetic and popular imagination, and of the history of the Roland Romance among others or of the Odysseus or Aineias legend. Hence it appears wholly unwise to seek to derive a high date for the Samhitas and Brāhmaṇas from any argument based on the date of the epic war.
Nor probably is it safe to insist that the period between the older Upanisads and Buddhism must be one of several centuries, and thus to increase the antiquity of these Upanisads, and consequently of the Brāhmaṇas and Samhitās. We have no means of estimating the rate of advance of thought in the period in question, and a further serious difficulty must be faced by those who wish to establish an early date for the Upanisads. The developed doctrines of Buddhism cannot be proved to be those of the Buddha, or to date from even the fifth century B. C., so that it is in all likelihood wiser to content ourselves with the belief, rather than the absolute assurance, that a date before 500 B. C. may reasonably be assumed for these Upanisads.3 To assert a much greater antiquity is easy and it has the advantage of increasing the interest of the study of the Upanisads, but there seems little satisfaction in beliefs which cannot be supported by any serious evidence.
A decisive argument against any early dating of the Upanisads would be available if we accepted the view often held that the Ajātaśatru who figures in the Kauṣitaki and the Bṛhadāraṇyaka Upaniṣads as king of Kāśi is
ad absurdum is found in A. C. Das, Rig-Vedic India, i. 279 ff.
1 Cf.Oldenberg, Die Lehre der Upanishaden,
pp. 288, 857, n. 185.
3
See Keith, Buddhist Philosophy, chap. i. Cf. Oltramare, La théosophie bouddhique (1928), pp. 56, 64 ff., who recognizes that Aśoka knew no canon. Max Walleser (Sprache und Heimat des Pali-Kanons, pp. 28 f.) still clings to the
alleged Aśokan date of the Kathāvatthu, a view as improbable as the theory of Pali as the language of the Sthaviras of Pāţaliputra.
- Cf. Hopkins, JAOS. xxii. 386; Rapson, Ancient India, p. 181; Keith, CHI. i, 112, 147.
• See Keith, ZDMG. lxii. 134 f. Identity is assumed in Winternitz, Gesch. d. ind. Lit. i. 484.
22
The Sources
1
[Part I
identical with the Ajātasattu of the Buddhist texts, who was contemporaneous with the Buddha. It appears to me, however, that any such identification wholly lacks justification, especially as the name is no more than an epithet and thus possesses singularly little probative value, while the king of the Buddhist texts is not king of Kaśi.
An effort has been made by Hopkins 2 to establish a more precise estimation of the period intervening between the Upanisad of the Jaiminiya and its Brāhmaṇa. The latter mentions Gauṣûkti, while the former has the same name as that of a teacher, giving after him ten recipients of the doctrine. This would give say three centuries, which he deems a not unreasonable time, in accord with the advance of the Upanisad in doctrine. The suggestion seems untenable; there is nothing whatever to prove that Gauṣūkti was a recent figure in the time of the Brāhmaṇa, nor does the fact that the Brāhmaṇa does not mention the other teachers referred to show that they existed after its composition, and, least of all, is there any evidence that we are to treat the list as representing generations. There is no evidence whatever that the record is one of teacher to youthful pupil.
3
The usual astronomical evidence has been adduced to establish the early date of the Brāhmaṇas, or at least of the statements recorded in them. As the lack of value of this evidence has been established, it is sufficient to note one point which has been held to fix definitely the date of one passage in the Śatapatha Brāhmaṇa; there, in a discussion of the time for establishing the sacred fires, the Kṛttikās are recommended as a possibility, on the score1 that they do not move from the eastern quarter, while the other Nakṣatras do move. It is really impossible to attach serious value to such an assertion, made in a passage which consists of foolish reasons for preferring one or other of the Nakṣatras; we are in the same region of popular belief as when in the Sūtra literature the existence of Dhruva, a fixed polar star, is alleged."
There are clear traces in the later Samhitās and the Brāhmaṇas of social and religious changes in the people. The centre of Vedic culture is still, as probably in the period when the main part of the Rigveda was produced, the land of the Kurus lying between the Sutlej and the Jumna, but importance now attaches also to the kindred tribe of Pañcālas, whose name
1 Vincent Smith’s dating of this prince c. 554 B. C., putting the Buddha’s death c. 546 B. C. (Oxford History of India, pp. 48, 58 n., 70), rests on a false interpretation of the inscription of Khāravela of Kalinga (see ref. in Keith, Sanskrit Drama, p. 89).
- Trans. Conn. Acad, xv, 30. * See ref. above, p. 4, n. 7.
CB. ii. 1. 2. 3. Kṛttikās must then (D. Mukhopadhyaya, The Hindu Nakṣatras (1923), pp. 41 ff.) have been on the equator, i. e. 3000 B. C.
Cf. S. B. Dikshit, IA. xxiv. 245 f.; A. C. Das (Rig-Vedic India, i. (1921)) prefers even greater antiquity. For TB. iii. 1. 1. 5, adduced by him (p. 47) from Tilak (Arctic Home, p. 2), see Keith, JRAS. 1911, pp. 794 ff.
- In favour of a late date may be adduced the mention of iron if the introduction of that can be placed c. 1000 B.C. (CHI. i. 56, 615), but this also is merely conjectural. For the question of Ayas see Vedic Index, i. 81 f., 151; ii. 235, 398.
The Later Samhitas and the Brahmanas
23
Chap. 2]
seems to signify that they were a union of five older tribes, but whose connexion in origin with the Kurus is attested by the record that they were once called Krivis. The Pañcala land stretched, eastward from Kurukṣetra, from the Merut district to Allahabad, and included the territory between the Jumna and the Ganges, called the Doab. But the Śatapatha Brāhmaṇa 2 records the advance of the Brahmanical system into Kosala and Videha, which roughly correspond with Oudh and Tirhut. The Atharvaveda knows iron and silver as well as the copper and gold of the Rigveda. The comparative frequence of mention of the elephant and the appearance of the tiger and the panther in the later Samhitas, whereas lion and wolf are conspicuous in the Rigveda, as well as the mention of rice, are clear indications of the advance of the Vedic Indians further to the east and the south. The Aśvattha (ficus religiosa) is rare in the Rigveda, but becomes common in the Atharvaveda, which also knows the Nyagrodha (ficus indica). At the same time it is clear that the system of classes became more and more complicated and the divisions were drawn more and more distinctly: the Yajurveda enumerates large numbers of special classes which in some degree at least seem to have been hereditary. The admixture of the people doubtless had proceeded very far: after the Rigveda it would be difficult to find any simple consciousness of the contrast of the colours of the Aryan and the Śūdra classes as opposed as white and black. The Rigveda, it is probable, already knew of the system by which normally the princely class, the priests, and the ordinary people were distinguished, and it knew also of the slaves made from the aborigines, but it was left to this later period to introduce a much more elaborate and fixed system of division. The Śūdras must on the one hand often have become rather serfs than slaves, when large bodies of them were reduced to subjection by the invaders, while among the ordinary people hereditary functions began to supersede the variety of choice of occupation which is evidenced by the Rigveda. To these factors of differentiation must be added the result of mixture of races and rules of intermarriage: the doctrine familiar in later texts that many classes of the people were due to mixed marriages between men and women of different classes indicates that this factor must have been of considerable importance in assisting in the development of classes into castes, the process of which, however, we have to conjecture from most inadequate material. With the development of society there doubtless took place growth in prosperity and wealth, favouring the
1 Hopkins’s suggestion (CHI. i. 254) that the Pañcālas may represent five Naga clans connected with the Kurus or Krivis (meaning “serpent’), and that none of the families is of pure Aryan blood, seems decidedly speculative.
- i. 4. 1. 10 ff.; Macdonell and Keith, Vedic
Index, ii. 288 f.
3 The reference to the Oḍras (Orissa) seen
by CHI. i. 601, in TA. ii. 1. 11, is an error, due to a hasty reading of BR. i. 1120, which really refers to Trik. (ì. e. Trikāṇḍaśeṣa) ii. 1. 11.
• Macdonell and Keith, Vedic Index, ii. 247-71. For Indo-European class distinctions cf. Feist, Kultur der Indo • germanen, pp. 291 ff.; Moulton, Early Zoroastrianism, pp. 117, 183 f.
24
The Sources
[Part I
constantly increasing elaboration of the sacrifice with its resulting exaltation of the importance of the trained priesthood, without which the offerings could not successfully be carried out. But while the tribes, in several cases at least, doubtless were more closely united, and while thus the royal power became greater, there is no proof of the growth of any large kingdoms or empires,1 nor can we say that there was much development of city life.
As sources for knowledge of the Vedic deities the later Samhitās and the Brāhmaṇas cannot be ranked high: the essential aim of the Yajurveda is the correct performance of the sacrifice, and the deities are of little consequence in comparison with the mechanism of that operation, to which is ascribed the whole control of the universe, and in the performance of which the universe is ever renewed. In the case of the Atharvaveda the position of the deities is still less important: constantly as they are introduced, their connexion with the magic spells which are the most original and essential part of that text is external merely the god, and still more his name, adds potency to the spell, and the more gods enumerated, however diverse their functions and spheres of influence, the better the result. Even where in that Samhita a deity is celebrated, the spirit is quite different from the spirit of the Rigveda: the goddess earth has a whole long hymn in a late book of the Atharva,a but the careful catalogue of all that grows on the earth and the sights and sounds upon it is recounted in a spirit quite unparalleled in the Rigveda. Hence it is not surprising that many of the minor figures of the pantheon of the Rigveda disappear, or at best sink to mere names, while on the other hand the religion shows development in two different directions. On the one hand, theosophic speculation brings into existence new and in some degree abstract deities; on the other, gods of the people receive a recognition which is not accorded to them in the Rigveda. Of the former tendency the most prominent example is the rise to high rank of Prajapati, as the creator god and the father of the gods as of men, and the exaltation to the rank of deities of such abstractions as Kāla, time ‘, Kāma,’ desire’, Rohita, the ruddy one’, perhaps an aspect of the sun, the Vratya, as the convert to the priestly faith was named, the Ucchiṣṭa, or remnant’ of the sacrificial offering, and so forth. Of the other tendency examples are to be seen in the increasing importance attached to Rudra and to Visṇu, who by the time of Megasthenes (c. 300 B. C.)3 were two of the chief gods worshipped in Northern India, and in whom we must probably see contamination of aboriginal with Aryan deities, the direct worship of snakes, perhaps induced by the experience of their terrors in India, the stress laid on the popular figures of the Apsarases and the Gandharvas, who, whatever their origin, are clearly little more in this period than fairies and sprites, and perhaps the collective view of the Asuras as a horde of evil spirits opposed in eternal, if unsuccessful, struggles to the gods in which they
1 Even AB. viii. 14, 23 shows how little real empire existed. Cf. Vedic Index, i. 19 f.; N. N. Law, Ancient Indian
Polity, pp. 13 ff. * xii. 1. * i. 29-37; L.
Chap. 2]
The Later Samhitas and the Brahmanas
25
defeat their adversaries, until by the discovery of some ritual device the gods outwit them, a conception the utility of which to the priesthood who devise the sacrifice is obvious.
On the other hand, the later Samhitas, if poor in their contribution to mythology and the higher aspects of religion, are rich in precise information regarding the ritual, and are veritable treasure houses of Indian magic. Their value in both these regards has often been under-estimated or misunderstood, doubtless through hasty preconceptions of the nature of Vedic religion based upon the theories of mythology which at one time found their chief sustenance in the Rigveda. We have here given to us for at least six, and often probably seven or eight centuries B. C., precise details of the actual carrying out of rites, accompanied in many cases by the interpretation placed by priests on the rites. In many instances these interpretations are obviously purely priestly speculation, but this is by no means always the case, and at any rate the genuineness of the practices recorded is in the majority of cases free from all doubt, as they were recorded not by students of anthropology under the influence of theories of religion, but by priests interested in the practical carrying out of the sacrifices.
It is, as in the preceding period, a question of the greatest interest to determine whether Indian religion in this period was subjected to any outside influence, and in this case the evidence for such influence, though it does not become of great importance, is nevertheless less impalpable than in the period of the Rigveda. The most important item of proof of Semitic influence is contained in the existence of the system of the Nakṣatras, * lunar mansions’, which appear in the Yajurveda Samhitas and the Atharvaveda as the stations in which the moon spends the successive nights of the periodic month. The foreign origin of the Nakṣatras 1 is suggested by the fact that they appear curiously isolated in Indian literature: the Rigveda 2 appears not to know them at all, nor to contain any hint that such a system was being developed, while they occur in China and in Arabia under conditions which render derivation from India or vice versa out of the question. That the system was derived from Babylon seems natural, but the requisite and conclusive proof of its existence there has not been brought despite the probability that it existed. The same conclusion in favour of primitive Babylonian influence is suggested by the legend of the flood which is recounted for the first time in the Śatapatha Brāhmaṇa * in connexion with the sage Manu, who rescued a fish, in return was warned by it of the danger of the flood, and in due course
4
’ Macdonell and Keith, Vedic Index, i. 409–
81; Keith, CHI. i. 148 f.
- Save in the late hymn, x. 85.
• Oldenberg, GN. 1909, pp. 544 ff.;
Whitney, Oriental and Linguistic Essays, ii. 341 ff.; Weber, Naxatra, agree in the Semitic theory of origin. Cf. Keith, CHI. i. 140. Suggestions of Semitic
influence on Indian magic occur in Henry, La magie dans l’Inde antique, pp. 98, 184.
• i. 8. 1. 1 ff. The alleged reference in AV. xix. 39. 8 is denied by Whitney, p. 961. One is possible in JB. iii. 99 (Caland, Das JB. in Auswahl, p.313). For Vend. ii, cf. Hertel, IIQ. ii, 85 ff.26
The Sources
[Part I
was towed by the fish safely over the flood to a mountain peak on which his ship grounded. It is not inconceivable that the story is of independent Indian origin, but this appears to be rather unlikely,1 and in that case Babylon seems the obvious source, though the story may have come from some other part of the Semitic area. Indeed it has been urged 2 that Indian writing was introduced via Mesopotamia about the eighth century B. C. and was based on the Phoenician script, having as its prototype writing of the character of that found on the Moabite stone,3 but this conjecture is still too uncertain to be used as a conclusive support of Semitic influences at this time. The attempt to find Sumerian influence in loha’ copper’ or ‘bronze’ is clearly inconclusive, though it has been suggested that the use of both copper and, later, iron came to India from Mesopotamia. It may be added that there is no trustworthy evidence of Egyptian influence on Indian thought in the Vedic period despite the contentions of Prof. G. Elliot Smith in his Migrations of Early Culture, Influence of Ancient Egyptian Civilization in the East and in America, and subsequent works, who would have us believe that this is the explanation of the development of Indian ideas in the sixth century B. C., ignoring the evidence of the slow emergence of the ideas of the Upanisads and Buddhism from Indian conceptions. Similarly it is unwise to demand Aryan influence on Egypt as an explanation of the rise for a brief period of the cult of Aten, however tempting it may be to connect this with the apparent worship of Śuriaś, the sun, among the Kassites, for the Egyptian phenomenon can be explained without any such hypothesis.
No specially close relation to Iran can be definitely traced in this period, though the fire cult may have been influenced by that of Iran, and Iranian influence has been seen in the development of the meaning of Asura and in the names of individual Asuras, as in the reference to incestuous unions in the Aitareya Brāhmaṇa.”
› Lindner (Festgruss an Roth, pp. 213 ff.)
defends its Aryan origin. See, however, Oldenberg, Rel, des Veda2, p. 283, n. 4. Cf. also Gomperz, Greek Thinkers, i. 95 ; Keith, JRAS. 1909, p. 590, n. 1; Gerland, Sintflut (Bonn, 1912); Winternitz, Gesch. der ind. Lit. i. 182 f., 337 ; J. G. Frazer, Ancient Stories of a Great Flood (1916).
- Būhler, Indian Studies, III, and Palaeographie (1896); CHI. i. 62. Rhys Davids (Buddhist India, p. 114) prefers a pre-Semitic Euphratean origin via Dravidian traders. For the theory of ultimate Egyptian origin see K. Sethe, GN. Gesch. Mitth. 1916, pp. 88–161; P’hil-Hist. 1917, pp. 437-7; Lehmann-
Haupt, ZDMG. Ixxiii. 51–79; Bauer, Zur Entzifferung der neuent. Sinaischrift.
- About 850 B. C.; Hall, Anc. Hist. of Near
East, p. 451.
• Feist, Kultur der Indogermanen, pp. 71, 199; von Schroeder, Arische Religion, i. 225, 233. For conjectures as to taimāta and urugūlā (AV. v. 13. 6, 8), see Bhandarkar Comm. Vol., pp. 33 f. See Chap. 29 for a conjecture as to the Babylonian origin of the cosmic character of speech (Vác, Logos).
• CIII. i. 615.
- On hrūḍu (AV.), sec Vedic Index, ii. 509.
• See below, Chap. 15, § 1.
• vii. 13, but cf. JB. ii, 113 for an Indian
rite.