वेदे विश्वासटिप्पन्यः

THE oldest and most important of the sources for Indian religion is the collection of 1,028 hymns known as the Rigveda Samhita, which has been handed down to us in the Śakala recension. Preserved in its early stages by oral tradition and long regarded as too sacred to be reduced to writing, the text affords abundant internal proof of the general accuracy with which it was preserved. Moreover, an invaluable form of control exists in the texts of the other Vedas, the Yajurveda in its different recensions, the Samaveda, and the Atharvaveda, all of which contain much of the matter of the Rigveda. The older view, that in these texts might be found traces of earlier forms of the verses of the Rigveda, has not borne close examination and comparison in detail with a very few possible exceptions the variations which are found in these texts from the Rigveda can be unhesitatingly classed as products either of an inferior tradition on the one hand or of deliberate alteration on the other. Similarly the efforts which have been made by Hillebrandt 2 to prove that, in a stage earlier than that recorded, the Rigveda was a definitely practical collection of hymns, arranged according to their connexion with the crificial ritual, must be pronounced to have failed. Whereas all the other amhitas, except the Atharvaveda, which occupies a peculiar position, are definitely in their non-Brahmana portions manuals of the chants and formulae

-Brāhmaṇa used by the priests in the ritual, the Rigveda is not a practical but a historical hand-book. It must represent a collection of hymns made by unknown hands at a time when for some unrecorded reason it was felt desirable to preserve the religious poetry current among the Vedic tribes.

The collection must have been made from a considerable area of country, for it contains hymns emanating from very varied families. Tradition ascribes to books ii to vii as authors the seers Gautama, Viśvāmitra, Vāmadeva, Atri, Bharadvaja, and Vasistha, but this view cannot be taken quite literally the hymns themselves reveal abundant evidence that, for the most part at least, they were not composed by these personages, but by men claiming to be of the families bearing their names, and the family character of the hymns in these books is in the main clear. With these six books must be

1 Oldenberg, Prolegomena, pp. 289 ff.; cf.

Bloomfield, Rig-Veda Repetitions, p. 406; Brune, Zur Textkritik der dem Sāmaveda mit dem VIII Maṇḍala des

1 [H.0.8. 31]

Ṛgveda gemeinsamen Stellen (1909).

• ZDMG. xl. 708; GGA. 1889, pp. 418 M.

• Oldenberg, GGA. 1907, pp. 211 ff.; Keith,

JRAS. 1908, pp. 224–9.

The Sources

[Part I

classed the groups of hymns ascribed to different families or authors in book i, 51-191, and this may have been the extent of the oldest collection made, though it is perhaps more likely that i. 51-191 were collected later than ii-vii. The earlier portion of book i and the whole of book viii are ascribed to seers of the Kanva family; it would appear that these two separate collections were at some time added, the one in front of, and the other after, the existing Samhitā, but which addition was first made there is no clear ground to show. When the collection had reached the compass of seven or eight books, another, now the ninth, was created by extracting from the other books all the hymns addressed to Soma Pavamāna, that is the Soma as it was poured through the filter, which were then united into one group : for this change no reason is obvious.2 This did not, however, end the history of the collection: at a time when the nine books had already taken form, a tenth was added, consisting on the whole of more recent hymns. The late character of this book can be established by a number of proofs. Its extent, 191 hymns, has obviously been brought up to that of book i; the language shows development in different aspects: hiatus becomes rarer, old words like the particle sīm disappear, new words and forms are found, and the metre shows affinities with the metre of the later Samhitās. The same result is indicated by the new features in religion which appear: the Dawn, the most poetic of Indian goddesses, all but disappears, Varuna, most moral and spiritual of Vedic deities, loses in position, while Indra, the Indian god par excellence, and Agni, the priestly god of fire, retain all their importance. The more or less abstract conception of the All-gods increases in importance, and real abstract deities appear in Faith and Wrath. The growth of religious thought is also shown by the occurrence of philosophical and cosmogonic hymns, and in imitation of the hymns to the gods the wedding and funeral services are now provided with elaborate hymnologies in place of the more simple formulae which were doubtless earlier in use. The book shows also the employment of hymns for spells and incantations, and here again we must doubtless see the application to the lower side of life of the instrument devised primarily to placate the high gods. The advance in religion is paralleled by the advance in society: in this book we for the first time meet the fully developed system of the four castes or classes, Brahman, warrior, clansman, and Śūdra.

It is naturally tempting to seek to carry the process of dissection further, and to discover the different ages of the several portions of the Rigveda.

1 That viii had no claim to age was indicated

by Hopkins, JAOS. xvii. 28 ff., and is confirmed by the evidence of repetitions, Bloomfield, Rig-Veda Repetitions, pp. 640 ft.

*ix is usually superior to viii (Bloomfield, p. 644). As new books were added, the Soma Pavamāns hymns were added to what is now ix. Its position as ix, how-

ever, suggests that i-viii already existed.

Oldenberg, Prolegomena, pp. 263 ff.

• Contrast Bloomfield, op. cit., pp. 21, 649, who holds that x. 14. 14 and 15. 14 are later than i. 15. 9; 108. 12, but neither case is convincing. The wedding hymn is clearly late; x. 85. 18 echoes i. 108. 1, based on vii. 61. 1.

Chap. 11

The Rigveda and the Aryans

Within the groups it is often possible to prove introduction of later material by the violation of the rules of order adopted by the compilers of the collection,1 but differences of age among the groups themselves and between individual hymns, which are not marked out as foreign to the groups in which they are found, cannot yet be established. The most elaborate attempt made of late to find strata in the Rigveda is that of Prof. Arnold, who by the test of metre divides the collection into five layers which cut sharply across the traditional grouping; but his criteria are clearly unsound,3 and depend on a purely hypothetical reconstruction of the metrical history of the hymns, to which objection can be taken on many grounds. Moreover, the results thus attained render any intelligible account of the development of Vedic religion impossible: the hymns to Dawn are certainly the most beautiful and least sacerdotal of all those of the Rigveda, and, for this reason and because in the later cult Dawn has but a small place, it is natural to assign them to the earliest period of Indian hymnology. The same conclusion is also indicated by the fact that this view alone harmonizes with the probable movements of the Vedic Indians. There can be little doubt that the bulk of the hymns cannot have been produced, as was formerly thought, in the Punjab, where the phenomena of the rains are poor and uninteresting and could not have given rise to the remarkable stress laid on these natural features by the Vedic poets in their conceptions of Indra and the Maruts. We must seek for the main home of the Vedic Indian in the country afterwards famous as Kurukṣetra, between the rivers Sarasvati, now Sarsuti, and Dṛṣadvatī, probably the modern Chitang, and in the region of Ambāla, and the oldest hymns only, those to Dawn, can reasonably be supposed to have been composed while the invaders were still in the land of the five rivers. But Prof. Arnold is forced by his metrical tests to ascribe the importance of the Dawn and of the deities, sky and earth, which, like the Dawn, seem among the oldest, to a secondary state of the Vedic religion, when Dawn and sky and earth were not revered for themselves, but because of their connexion with the fire ritual, dawn being the time of sacrifice and the fire serving as a pillar to join heaven and earth, but yet to keep them asunder, while in the earliest period he sets Indra, the warrior god.

If we cannot hope to reach any assured results as regards the different strata in the Rigveda itself, it remains to be seen what date can be ascribed to the Rigveda as a whole. The Saṁhitā is absolutely lacking in reference to any historical event which we can date. It had indeed been sought time after time to demonstrate the contrary, but no such attempt has yet approached plausibility. Ludwig in an elaborate examination of the question decided

Oldenberg, Prolegomena (1888) and

Rgveda-Noten (1909-12).

Vedic Metre (1905). Contrast Bloomfield, Rig-Veda Repetitions, pp. 585 ff., 640, 687, for the use of grammatical, lexical, metrical, ritual, sense, and other considerations as to repeated passages to

decide relative dates.

Keith, JRAS. 1906, pp. 486–90, 718-22 ;

1912, pp. 726-9.

• Hopkins, JAOS. xix. 19; cf. Keith, CHI.

i. 80 ff.

‘Proc. Bohem. Acad. 1885.

The Sources

[Part I

that from the mention of two eclipses in the Rigveda could be deduced a date of the eleventh century B. C. for the hymns in which these phenomena were mentioned, but this suggestion has been totally disproved by Whitney.1 An alleged reference to the capture of Babylon by Aryan tribes which might be brought into connexion with the advent of the Kassite dynasty at Babylon in the eighteenth century B. C. is a wild guess of Brunnhofer, which it is quite impossible seriously to consider. Much more substantial are the arguments adduced by Prof. Jacobi who sees traces of evidence that the Rigveda goes back as far as the third millennium B. C. He thinks that the Rigveda shows that the winter solstice took place in the month Phālguna, and on the ground of the precession of the equinoxes this must mean that the observation thus recorded was made in the third millennium B. C. This view, which rests ôn the interpretation of a very doubtful passage in the Rigveda, he supports by the fact that in the Gṛhya Sūtras, or manuals of domestic ritual, of much later date, the ceremonial of the wedding includes an injunction to the wife to look at the star called Dhruva, ‘fixed’, and this can only have originated at the time when a Draconis was in the vicinity of the pole, there being no other star which could be called fixed at any period coincident with the probable age of the Rigveda: further he contends that the fact that Kṛttikās, the Pleiades, are placed at the head of the list of twenty-seven or twenty-eight Nakṣatras, ’lunar mansions,’ in the Yajurveda and Atharvaveda Samhitās means that Kṛttikās marked the vernal equinox when the list was compiled, and this date fell in the third millennium B. C. The first of these arguments seems clearly to be based on a misunderstanding of the Rigvedic passage in question; " the argument from the pole star assumes an accuracy in the demands of the primitive Indian wedding ritual which is wholly unnatural; and the assumption that the Kṛttikās coincided with the vernal equinox is most improbable, if we are to regard the Nakṣatras as an Indian invention, since the equinoxes play otherwise no part in early Indian ideas, and if, as is far more probable, the Nakṣatras were borrowed from some other nation, then the period when Kṛttikās were chosen as the head is without relevance to the date of Indian literature.?

1 JAOS. xiii. pp. Ixi-lxvi.

2 Iran und Turan, p. 221.

› Festgruss an Roth, pp. 68 ff.; GN. 1894,

p. 110; ZDMG. xlix. 218 ff. ; 1. 69 ff.; JRAS. 1909, pp. 721–6; 1910, pp. 456– 64.

✦ x. 85. 18; AV. xiv. 1. 13.

• A somewhat similar view is found in B. G. Tilak’s The Orion (1898) and The Arctic Home in the Vedas. Cf. Būhler, IA. xxiii. 288 ff. Contrast A. C. Das, Rig-Vedic India, i, 356 ff.

• The argument involves (1) the deduction from RV. vii. 103. 9 that the year began

with the summer solstice, and (2) from x. 85. 13 that the marriage of the sun in the Phalgunis must fall at the beginning of the year, i. e. the summer solstice. Both views are most implausible; în vii. 108, 9 that dvådaśásya means ’ year ‘is practically certain, and thus ruins the whole structure of conjecture.

Oldenberg, ZDMG. xlviii. 629 ff.; xlix. 470 ff.; 1. 450 ff.; JRAS. 1909, pp. 1090–5; GN. 1909, pp. 544 ff.; Thibaut, IA. xxiv. 85 ff.; Whitney, JAOS. xvi. pp. lxxxi ff.; Keith, JRAS. 1909,

Chap. 1]

The Rigveda and the Aryans

We are compelled therefore to content ourselves in the main with internal evidence. There is, however, one point of interest arising from the discoveries at Boghaz-Köi,1 where among gods invoked by the King of the Mitanni are found names suggestive of the gods Mitra, Varuṇa, Indra, and the Nāsatyas, i. e. the two Aśvins, who often bear that name or epithet in the Rigveda. The existence of these gods seems, therefore, established for a period which may be placed about 1400 B.C., but unfortunately there is nothing in the record to show decisively whether these gods are to be regarded as the gods of an Aryan people, no clear separation of Iranian and Indian yet having taken place, or of the proto-Iranians, or of the proto-Indians. From the names of kings of the Mitanni preserved in the Tell-el-Amarna letters, it has been deduced that there were proto-Iranian elements among the Mitanni, and this possibility is not to be denied, though it is at least certain that the people were not a pure Aryan race. But in view of its uncertain value no direct light can be thrown on the age of the Rigveda either in its earliest or its latest form. A priori it is clear that the gods must have existed before the hymns, and there is nothing special about the grouping of the gods as found at Boghaz-Köi which would justify us in holding that the pantheon had by that time assumed the definite form which it takes in the Rigveda, and that the Rigveda must then

pp. 1095-1100; 1910, pp. 465–8; Macdonell and Keith, Vedic Index, i. 420-81; Keith, Taittiriya Samhita, i. pp. clix ff.; JRAS. 1917, pp. 135 ff. ; Lehmann-Haupt, ZDMG. Ixiii. 717. 1 Winckler, MDOG. Dec. 1907; Jacobi JRAS. 1909, pp. 721-6; Meyer, SBA. 1908, pp. 14 ff.; KZ. xlii. 16 ff.; Gesch. des Alt3 II. i. p. 652; Keith, Bhandarkar Comm. Volume, pp. 81–92 ; Winternitz, Gesch, der ind. Lit, iii, 621 f. 2 Bloomfield, AJP. xxv. 8; Hall, Anc. Hist. of Near East, pp. 201, 381. According to Winckler (OL. xiii. 291 ff.) the Aryan element bore the name Charri; the Susian version of Darius’s inscriptions has Harriya for Aryan; their followers are named marianni, in which may be seen the Vedic marya, with suffix ana; cf. Leumann, Zur nordarischen Sprache und Literatur, pp. 5 ff.

The theory is carefully criticized by W. E. Clark, Am. Journ. Sem. Lang. xxxiii. 261–82. Since then much evidence has been accumulated, without decisive result. Hrozný (Keilschrifttexte aus Boghazkởi and Völker und Sprachen des alten Chatti-Landes), agreeing in considerable measure with E. Forrer, Die acht Sprachen der Bog-

hazkoi-Inschriften (1919), holds that the Chatti were non-Indo-Europeans ruled by persons of quasi-Indo-European origin, speaking a centum language, while the Charri were also non-European under rulers speaking a satem language of Indian, not Iranian type (Jensen, Indische Zahlwörter in keilschrifthittitischen Texten, Berlin, 1919). The efforts from the numerals and names to establish Indian rulers, as opposed to Iranian or Aryan, are not convincing, as we simply have no evidence of early Iranian, and the process of restoring the Avesta to its true form, undertaken by Andreas and Wackernagel, is still unfinished. The form Assara Mazāś of Assurbanipal’s record, pointing to a much earlier borrowing, possibly during the Kassite dominion in Babylon, is clearly not Indian, and, while it may be Aryan, it is possible that it is proto-Iranian ; cf. Hommel, PSBA. 1899, pp. 127, 138 f.; Geographie und Geschichte des alten Orients, p. 204; Moulton, Early Zoroastrianism, pp. 428 f.; Konow, JRAS. 1911, pp. 41 ff., argues that the Mitanni names are early Iranian. Cf. CAH. i. 311 f., 469, 553; Forrer, ZDMG. lxxvi. 250 ff.6

The Sources

[Part I

have existed. Still less importance attaches to the occurrence of a name like Śuriaś (perhaps Surya, the sun) in Kassite records, which leads Meyer to the conclusion that the Kaśśū were originally settled in Media and driven west by Aryan tribes.1

The internal evidence is more satisfactory, if less definite. It is practically certain that the Rigveda was to all intents and purposes complete before the other Samhitas came into being, and it is certainly anterior to the whole of the other literature of India, which presupposes it and takes it as given. With this fact accords its language, which is much more archaic than the language of the other early literature of India, and its metre, which has only emerged from the simplest form in which the number of syllables in each line was the sole mark of differentiation of verse from prose. For reasons which are given in the next chapter it is impossible to suppose that the later Samhitās date substantially if at all after 800 B. C., and this may probably be taken as the lowest possible date for the completion of the Rigveda. The real difficulty arises in deciding how much farther back the collection is to be carried, and in this regard it is probably necessary to beware of exaggeration. There are many references in the Rigveda to former poets, and unquestionably, as we have seen, there was a distinct development of language and thought during the period of its production. But to allow too extended a time for the process of development and decline-for it is clear that the end of the period saw the passing from favour of original composition of hymns-is unnecessary, and there are two distinct grounds against adopting any such view. In the first place the poets never attain any very great command of their material, whether in language or metre, though in certain cases poetic results are attained by simple means. To the end the structure of the sentences remains naïve and simple, and, when the poet seeks to compass more elaborate thought, his power of expression seriously fails him: it can hardly be supposed that in a period of many centuries the Vedic poet’s control over his instruments of expression would not have risen superior to the difficulties which faced him. In the second place, if the Rigveda is put as far back as 1500 B. C., it becomes very difficult to explain the extremely close parallelism between the speech of the Avesta and that of the Rigveda, especially if the traditional date (660-583

In CHI. i. 65 ff. the case for a western home of the Indo-Europeans is stated by Dr. P. Giles, who would clearly place the invasion of India after 1500 B.C. The eastern theory is defended by Feist, Indogermunen und Germanen (1919). Carnoy (Les Indo-Européens, pp. 55 ff.) decides for the Dnieper region. On Kassite names see Bloomfield, AJP. XXV. 1-14. The western home is supported by von Schroeder, Arische Religion, i. 214 ff. Hirt (Die Indo-

germanen, i. 22) places the migrations late. Cf. Ipsen, IF. xli. 174 ff.

2 Bloomfield (Religion of the Veda, p. 20) prefers 2000 B. c. for the beginnings. But in Rig-Veda Repetitions (pp. 20, 21) he stresses the absence of archetype hymns and the epigonal character of the collection (cf. JAOS. xxix. 287), and in the earlier work he accepts 1600 B. c. în lieu of 1400 B. c. as the Mitanni date. › Macdonell, Hymns from the Rigveda (1922),

pp. 17 f.

Chap. 1]

The Rigveda and the Aryans

B.C.) of Zoroaster is accepted.1 It is possible to diminish the force of this objection by postulating an earlier epoch for Zoroaster; but, even so, it is very doubtful whether the prophet can be carried far enough back to make any earlier date than 1200 B. C. or 1800 B.c.3 for the Rigveda reasonably probable. If we seek to ascribe a higher date than this, we must recognize that we are dealing with conjectures for which no very substantial evidence can be adduced.

A very serious difficulty, it must be added, presents itself in the way of the early dating of the Rigveda in the shape of the fact that it seems very dubious whether we can place at all early the period of the dispersal of the Indo- Europeans or of the Indo-Iranians. If the Rigveda belongs to even 2000 B. C. we must assume that the Indo-Iranians parted at some date decidedly before that epoch, and there certainly seems every reason, arguing from general probabilities, not to place the entrance of the Aryans into India substantially before 1600 B. C., and the process was probably one of long duration and slow accomplishment.

A proof of the long connexion of the Indians and the Iranians before the latter settled definitely in India is seen by Hillebrandt in certain names in the Rigveda, which incidentally in this view aid us in assigning an earlier date to certain hymns at least of the sixth book, composed in Arachosia. This view involves the identification of the Panis not with mythological figures but with the Parnians, of the name Parthava with the Parthians, the Dasas with the Dahae, the river Sarasvati with the Iranian Harahvaiti, the Hariyūpīya with the Iryab or Haliāb, a tributary of the Krumu, and the Arjïkīya with a name connected with Arsakes, while Bṛbu Takṣan, the enemy of the Panis, is brought into connexion with the later city of Takṣaśilā, which may represent an eastern settlement of a tribe originally situated further to the west. Against this view, however, there are two serious objections. The identifications are all of the most dubious character, and, even if they were genuine, it would be difficult to make out any chronological result from them, seeing that other possibly Iranian names occur in other books of the Rigveda, such as Sṛñjaya and Pārāvata in several books, Drṛbhika in book ii, Sṛbinda in viii,? Parśu and Tirindira in viii.

1 Jackson, Zoroaster, pp. 150 ff.; Práśek,

Gesch. der Meder, i. 204 ff.; West, SBE. xlvii. p. xxviii; cf. Hertel, IIQ. i. 7 ff.

Moulton, Early Zoroastrianism, p. 18; Geldner, Enc. Brit. xxi. 246; xxviii. 1041; Bartholomae, Altir. Wörterbuch, p. 1675; Keith, JRAS. 1915, pp. 798, 799; Peters, JAOS. xxxi. 378; Jackson, CHI. i. 823.

• Macdonell, Hymns from the Rigveda, p. 7.

• Cf. Morgan, Les premières civilisations, pp. 264 ff., 314; J. L. Myres, The Dawn of History, pp. 189 ff.; Kennedy, JRAS.

1909, p. 1119.

Ved. Myth. i. 83 ff.; iii. 268; (Kl. Ausg.), pp. 95, 114, 191 f.; GGA. 1894, pp. 648 ff.

• See Macdonell and Keith, Vedic Index, i. 29, 349, 357, 450, 504 f., 518 f., 521 f.; ii. 470; Keith, CHI. i. 80 f. The obvious possibilities of mere parallelism of name between India and Iran seem sometimes ignored (cf. Jackson, CHI. i. 322). Parśavas is certainly not a proper name in x. 33. 2.

› Brunnhofer, Iran und Turan, p. 122.

The Sources

[Part I

It is, however, an interesting question 1 how far there can be traced in the Rigveda evidence of closer connexions with Iran on the part of some of the families of poets, even if, as is doubtless proper, we reject the suggestions of Hillebrandt which would make certain of the hymns of the Rigveda (book vi) a product of the time when the seers dwelt beyond Indian limits proper. Here again, however, we must be contented in large measure with a negative result. Thus it has been suggested that we are to see special closeness of connexion in the case of book viii, on the strength of the occurrence of such names as the Gomati, Suvāstu, Asikni, Paruṣṇi, and the hostile aspect in which the Gandharva is viewed,2 contrary to the usual honourable position occupied by that spirit. Arjika or Arjikiya is also cited as pointing to some Iranian locality. This, however, as has been said, is uncertain, and the most certain indication of Iranian influence, the form titaū with an unparalleled hiatus, is found in book x, other alleged instances of such influence being most dubious.

It may, however, be noted that from the Iranian side the suggestion has been made that the Tir Yaśt represents an Indian phenomenon, the breaking of the south-west monsoon, which has no Iranian parallel. Hope Moulton 5 connected this view with the appearance of gods, whom he regarded as Indian, among the Mitanni, thus arriving at the conjecture of a movement back out of India on the part of tribes which had become dissatisfied with conditions there, but carried traditions with them. Indo-Iranian relations might account for the phenomenon adequately, but the whole matter is too conjectural to yield any assured result.

The Rigveda is not, therefore, among the oldest literary monuments of the world viewed merely from the point of date, but its extent, which is comparable with that of the Iliad and the Odyssey put together, and the practically exclusively religious character of its contents, make it unique in its revelation of the religion of the Vedic tribes. Of the condition of life of these tribes comparatively little is made known to us, but there is enough to show that the people were divided up among small kingdoms, under hereditary princes, often engaged in war among themselves and still more often involved in conflicts with the dark skins ‘, over whom they seem normally to have been victorious, perhaps as a result of the body armour which they wore, and the spears and battle-axes of metal-copper, or later iron-which, with the bow, formed their chief weapons in war. They were not merely a pastoral but also an agricultural people, but there is no clear trace of a town life: the forts, which both they and the aborigines owned, were doubtless nothing more than places of refuge, with ramparts of mud or wood, used both in time 1 Cf. Hillebrandt, Aus alten und Neuindien

(1922), pp. 8 ff.; Hopkins, JAOS. xvii. 78 ff. See also below, Part II, Chap. 15, § 1 as to Asura.

RV. viii. 1. 11. In Iran his parallel is a demon, but the suggestion is pro-

blematic.

• Macdonell and Keith, Vedic Index, i, 62 f.;

Stein, Bhandarkar Comm. Vol., p. 27.

• Cf. Wackernagel, Altind. Gramm. i. § 87.

1 (b).

• Early Zoroastrianism, pp. 25 f., 436 f.

Chap. 1]

The Rigveda and the Aryans

་

of war and in time of flood.’ The richness in gold, which is characteristic of the age, may be compared with the wealth of the Aegean civilization of Crete. but there is no trace of the artistic spirit of the Aegean pre-Hellenic people. Nor is there any sign that large kingdoms had yet appeared: confederations of tribes, such as that of the famous five peoples, Anus, Pūrus, Druhyus, Turvaśas, and Yadus, might exist, and we hear even of a battle of ten kings, but these were clearly not lasting federations, but loose unions for war. On the other hand a great homogeneity of culture and religion among the tribes seems to result from the evidence of the Rigveda and to attest the definite and distinctive character of the Vedic people as distinct from the tribes of aborigines.

•

The language of the Veda is essentially akin to Iranian as seen in the Avesta, and more remotely to the other tongues which make up the Indo- European family. From this fact, and from the picture of strife against peoples of dark colour in the Rigveda, has been deduced the theory that the Vedic Indians formed a body of invading tribes which broke into India from the north-west and carried with them a distinctive culture and religion, which they developed in a special manner under the influence of the new climatic conditions in which they found themselves in Northern India, and of intermixture of blood through marriage with the aboriginal population. Of the latter fact there are probably clear traces already in the language of the Rigveda, which contains in the cerebral letters a series in the main unknown to other cognate languages and most plausibly to be ascribed to the deterioration of sounds in the mouths of generations of mixed blood. Moreover, all analogy is distinctly in favour of an early process of admixture. Complete destruction by invaders of pre-existing peoples is a comparatively rare phenomenon and connotes a bloodthirsty spirit among the invaders which is not suggested by anything in the Rigveda.

An alternative hypothesis has, however, been freely urged of late, which would see in the Aryan speech of the Rigveda no proof of real invasion of a people, and would, therefore, refer the religion of that Samhitā not to Aryans but to the aborigines, presumably the Dravidians, who are clearly the most important of the early inhabitants of India.

1 Cf. Feist, Kultur der Indogermanen, pp. 144-6; von Schroeder, Arische Religion, i. 247; Macdonell and Keith, Vedic Index, i. 539 f.; Hopkins, Trans. Conn. Acad. xv. 32.

Zimmer, Altindisches Leben (1879); Macdonell and Keith, Vedic Index (1912); Keith, CHI. i. 77 ff.; Kennedy, JRAS. 1919, pp. 498 ff.; 1920, pp. 81 ff.
Reminiscences of an older non-Indian home (seen, e. g., by Weber, Ind. Stud.1. 161 ff., and B. G. Tilak, The Arctic Home in the Vedas) may be safely

With this theory may be con-

regarded as purely speculative.

Cf. Wackernagel, Altind. Gramm. 1. § 144 and p. xxii; Macdonell, Ved. Gramm., p. 88. Objections to the view of aboriginal influence are suggested but not proved by Michelson, JAOS. xxxiii. 145-9. Cf. Keith, CHI. i. 109 f. ; G. W. Brown, Studies in honor of Bloomfield, pp. 75 ff.; Petersen, JAOS, xxxii. 414 f.
Srinivas Iyengar, Life in Ancient India, pp. 6 ff.; G. Slater, The Dravidian Element in Indian Culture (1928).

The Sources

[Part I

nected the view suggested by Hall1 that the Sumerians were originally Dravidians who developed their civilization in the valley of the Indus, and thence introduced it to the half nomadic Semites, teaching them the arts of writing, of town-dwelling, and of building in stone. The Aryans who invaded India were then civilized by the Dravidians, just as, according to the prevailing theory, the Aryans of Greece owed their civilization to the Aegean race. The fatal difficulty from the point of view of proof presented by this theory is that there is not available any evidence by which it can even be made plausible. If the Sumerians were originally Dravidians, and attained a high civilization in the Indus valley, it is remarkable that no trace of this high civilization is to be found in India, which, as far as we know, first attained the art of writing from Semites not before 800 B.C., and which commenced building in stone and town-dwelling long after the age of the Rigveda. No traces of the stone buildings which presumably the Sumerians erected in the Indus valley have been discovered, and Dravidian civilization is first known to us as a historic fact

many centuries after the latest date to which the Rigveda can be ascribed. The ascription to the Dravidians of the civilization of the Rigveda, therefore, remains a mere hypothesis, and one which is difficult to maintain in view of the clear opposition of the white and the dark races made in the Rigveda, where the white shows throughout its contempt for the black. Moreover, there is one very definite piece of evidence which suggests that the invaders were conscious, not merely of racial, but also of religious differences between themselves and the aborigines. In two passages are mentioned phallus-worshippers and in both cases with abhorrence: it is certain that the Dravidians in historical times were addicted to this form of fetishism, and it is as probable as anything can be that the phallus-worshippers opposed by the singers were aborigines. But it is of course obvious that, with the admixture of races which was inevitable, the admixture of religion was certain to follow, and traces

1 Anc. Ilist. of Near East, pp. 173, 174.

The facial aspect of Gudea in his statues seems to me wholly un-Dravidian. Rapson (CHI, i. 48) accepting a connexion derives the Dravidians from Western Asia. A. C. Das (Rig-Vedic India, i. 208 ff.) believes in Aryo- Dravidian influence on the Sumerians, holding that the Punjab was the Aryan home even in the Miocene epoch, and peopling Egypt with Dravido-Aryans. E.g. Hall, Aegean Archaeology (1915);

Evans, JHS. xxxiii. 277 ff. There is some exaggeration in this view; an Aryan infiltration may have preceded the Achaean, as suggested by Kretschmer, Glotta, i. 21 ff.; Keith, JRAS. 1912, pp. 473, 474.

• RV. vii. 21. 5 x. 99.8; von Schroeder,

VOJ. ix. 237. (These passages are erroneously cited by Dr. Farnell (Cults of the Greek States, v. 8) as applicable to Vedic religion.) That RV. x. 101 and ix. 112 imply ritual use of the phallus is certainly implausible. It is Śiva who is specially connected with the phallus from the epic onwards; Vaiśravana (Kubera) and Iśāna (Rudra-Śiva) are worshipped for the bridegroom, a fact which Hopkins (CHI. i. 283) interprets as pointing to their phallic nature (PGS. i. 8. 2; ŚGS. i. 11. 7). RV. viii. 1. 34 has no reference to cult; cf. Hertel, VOJ. xxv. 172 ff. For a Greek parallel, ef. Keith, JHS. xxxvii, 238.

• Contrast A. C. Das, Rig-Vedic India, i. 267 f.; Gūntert, Weltkönig, pp. 805 ff.

Chap. 1]

The Rigveda and the Aryans

of such influence which are scanty in the Rigveda can be seen in greater abundance in the later texts.

It has been assumed that the Dravidians may be reckoned as the aboriginal population encountered by the Aryan invaders, and, though this cannot strictly be proved, it is rendered extremely probable by the existence of a people of Dravidian speech, the Brāhuis, in Baluchistan, whether we regard them in origin-now they are greatly mixed and un-Dravidian in type-as an advanced guard of a Dravidian movement from India, or as the remnant of an older population, left behind on the Dravidian advance from Western or Central’ Asia into India. It is, however, possible that the aborigines met by the Aryans included members of the pre-Dravidians who are still found as jungle tribes, and who are by some authorities brought into relation with the Veddahs of Ceylon and the Sakai and Semang of the Malay Peninsula; the term ’noseless’ applied to their opponents by some Aryan invaders at least is held to accord better with the appearance of pre-Dravidian than with that of Dravidian tribes. The argument is not decisive, but there is no reason to doubt that both pre-Dravidians and Dravidians may have been encountered by the Aryans. Whether Muṇḍā-speaking tribes were among their enemies it is idle to enquire, for we know even less of Mundā movements than of Dravidian; their physical appearance is now very much that of Dravidians, though their language proves to have affinity with the Mon-Khmer languages of Assam and Burma as well as with other forms of Austric speech scattered over the Pacific.

Physical evidence of the present day suggests that about the longitude of Sirhind there sets in a distinct change of type in Northern India, and the type to the west of the line has been characterized as Indo-Aryan, that to the east as Aryo-Dravidian, the first including the areas of Kashmir, the Punjab to the longitude of Ambāla, and Rājputana, the latter the eastern border of the Punjab, the United Provinces, and Bihar. Taken in conjunction with the grouping of modern vernaculars, this distinction has been made the basis of a theory which asserts that the Aryan invasion of India took place in two distinct movements of very different character; the one was carried out by tribes which entered India through the passes of the Hindu Kush, passing through South Afghanistan, and the valleys of the Kabul, Kurram, and Gumal rivers, and settling in the N.W. Frontier Province and the Punjab. These tribes were accompanied by their wives and families, a fact which is held to explain the predominantly Indo-Aryan character of the population west of Sirhind. On the other hand the second invasion was by the difficult way of Gilgit and Chitral, and was carried out by men unaccompanied by women, who, therefore, had to form alliances on a wholesale scale with the

1 Imp. Gazetteer, i, 292 ff.; Rapson, CHI.

i. 40 ff.

Thurston, The Madras Presidency, pp. 124f. Cf. the Niṣādas of the Vedic texts; Chanda, The Indo-Aryan Races, i. 4 ff.,

who holds they originally were Muṇḍā speakers; Kennedy, JRAS. 1919, pp. 501 ff.; Thurston, Castes and Tribes of Southern India, i. pp. xx ff.; A. C. Das, Rig-Vedic India, i. 99 ff.; CAH. i. 27 f.

The Sources

[Part I

Dravidians, whence the changed type. The argument from ethnology is clearly unsatisfactory; in the first place it is impossible to ignore the fact that there is still doubt whether the Indo-Europeans were Nordic blonds 1 or Mediterranean brunettes or Alpine brachycephalics or a mixed race; that the north-west of India has been the scene of prolonged and repeated inroads; and that the present racial types are, therefore, very poor evidence for the racial types of 1200 B. C., not to mention 8000 B. C. Secondly, it is simple to explain the change as due merely to the fact that about the longitude of Sirhind the Dravidians were established in larger numbers and that the progress of the Aryans became seriously hampered; they had to convert rather than conquer, and the racial type is, therefore, naturally a compromise. The evidence from language is clearly of even less value. The facts of the later dialectic differences can be wholly and satisfactorily explained3 by the inevitable mode of propagation of linguistic influence; from the centre of that influence, the middle country of the Brahmaṇa period, linguistic influence was exerted in a manner which necessarily became more and more feeble in proportion to the distance of the peoples affected from the centre; hence the phenomena of outer and inner languages are explained without recourse to the speculation which introduces invaders over an almost impossible route, and, what is far worse, demands that we should recognize a sharp break between the civilization of the Rigveda and that of the Brāhmaṇas, assigning the former to the Punjab, and the latter to the middle country. The literature of the Vedic period shows emphatically no break of any kind in culture; it displays instead evidence of the advance of the Vedic civilization from the Punjab to the middle country, in an orderly progress, which conforms precisely to what would a priori be expected.

The religion of the Rigveda is, therefore, the product of Aryans who must have been affected considerably by their new environment and whose blood must have been becoming more and more intermingled by intermarriage ; but it is only proper to recognize that we really do not know, and have no means of ascertaining, how far the people at the period of the Rigveda can be styled Aryo-Dravidian, rather than Indo-Aryan. For this reason it is hopeless to seek to estimate the relative contributions of Aryan and Dravidian to the intellectual product of the Brahmans, for we have insufficient knowledge of what was true Aryan, and we know facts regarding Dravidian thought only

1 Von Schroeder, Arische Religion, i. 174 ff.

The modern conditions are fully reviewed in Sir A. Baines’s Ethnography. * Grierson, Imp. Gazetteer, i. 857 ff.; Risley (The Peoples of India, p. 55) renders the theory untenable by placing the first invaders originally in Arachosia and Seistan.

See Rapson, CHI. i. 50; cf. Keith, ibid. p. 119; Kennedy, JRAS. 1919, pp. 526 ff., who, however, errs in saying

that the Bharatas found the speech of the Pūrus barbarous, for mṛdhravác refers to hostile speech (Vedic Index, i. 471), nor is it at all clear that the Bharatas were late comers.

• Assertions of Dravidian predominance (Crooke, North-Western Provinces of India, p. 60) can neither be proved nor disproved, but the prevailing of Aryan speech must be remembered.

Chap. 1]

The Rigveda and the Aryans

long after it had been affected by the Aryan invasion. Here as often confession of ignorance is preferable to the idle affectation of knowledge.

•

As the Rigveda is of so recent a period, it is natural to ask whether the religion which it contains has not traces of influence by the cultures of the great nations of the East and above all of Babylonia. The answer to this question cannot be given with any certainty as in the affirmative: the only cogent proof of the borrowing of deities by one people from another, in cases where the borrowing is not formally recorded, is afforded by the appropriation of the name and the similarity of character of the gods: mere similarity is wholly insufficient, unless the conception formed of the particular divinity is of so special a kind that parallelism is not a reasonable explanation. In the case of the Rigveda and of the later Vedic texts no such instance of borrowing is hinted at, and no case is known in which the similarity of name even suggests that a god has been taken over from another neople, so that at most we are left to rely on the argument from similarity of character. Strength would doubtless be given to such arguments if the language of the Rigveda could be proved to contain loan-words from Semitic sources, but the only two which have with any probability been alleged, the word manā,1 apparently meaning ́ ornament’ and described as golden, which is often equated with the Babylonian Mina, and the word paraśu, axe, are too isolated to prove anything at all. Aśśur cannot reasonably be connected with Asura either as source or result and it is impossible to prove that the year of 360 days of the Rigveda is to be derived from the Babylonian year, and still less that the sacred number seven is adopted under Babylonian influence for an Aryan nine."

While the religion of the Rigveda seems to stand free of foreign elements, it cannot be assumed that the version presented to us in that collection is at all a complete record of the religion of the period of the composition of the hymns. It contains the poetry used by the priests in the sacrifices to the high gods, but not, with rare exceptions, the lower religious or magical beliefs. Even, however, of the hieratic views it gives no complete account: the collectors of the hymns in the main were interested in the Soma ritual, and the great majority of the hymns deal with some form or other of that rite: the animal sacrifice is hardly noticed, save in the case of the most important and rare sacrifice, that of the horse. Moreover it cannot be doubted that much of

1 RV. viii. 78. 2; Macdonell and Keith, Vedic Index, ii. 128, 129. The alleged borrowing of the war chariot from Babylon is wholly dubious and in any case is probably pre-Aryan; Schroeder, Arische Religion, i. 288. For guesses, see Brunnhofer, Arische Urzeit, pp. 89 f., 415; B. G. Tilak, Bhandarkar Comm. Vol., pp. 29 ff.

von

Wackernagel, Altind. Gramm. i. p. xxii; Kretschmer, Gesch. d. griech. Sprache, p. 106; Feist (Kultur der Indoger-

manen, p. 214) suggests possible borrowing from a third source by Babylon and India. But see Macdonell and Keith, Vedic Index, ii. 128 f.

• Cf. Thomas, JRAS. 1916, p. 364, with Chadwick in Moulton, Early Zoroastrianism, p. 81.

Keith, JRAS. 1916, p. 355. Cf. Meyer,

Gesch. des AU♪ I. ii. p. 913.

• Von Schroeder, Arische Religion, ii. 426 ff. Contrast Hopkins, Origin of Religion,

pp. 291 f.

The Sources

[Part I

the poetry is highly artificial, the expression not of naïve faith but of refined speculation: there is much-usually empty-mysticism, and phrase making, the work of competing poets without religious inspiration. On the other hand there are numerous hymns which are perfectly simple in thought and even in diction, and part of the obscurity of the poetry is due merely to the fact that it is rich in references to myths, which are, as is inevitable in hymns, only alluded to and not set out in detail. Such references are of comparatively little importance in the consideration of Vedic religion, of which it is possible to obtain definite views irrespective of the exact force to be ascribed to obscure myths.

The accusation, however, which is often made against the Rigveda of being purely sacerdotal cannot be accepted, for it contains enough matter in its later portions to show that the compilers were perfectly familiar with the popular religion of the day. Thus we have hymns intended to act as spells against vermin, or the disease Yakṣma,2 to bring back the life of one apparently dead,3 to destroy enemies, to procure children, to destroy the demon who kills offspring," to induce sleep, and even to oust a co-wife from a husband’s affections. Most of these hymns occur in book x, which preserves also the marriage hymn," a piece of priestly ingenuity, and the funeral hymns.10 These with four or five gnomic hymns," some philosophic and cosmogonic speculations,12 and some hymns, or portions of hymns, in praise of generous patrons of the priests relieve the monotony of the collection, and help to obviate the wholly erroneous view that the early religion of India consisted merely in the invocation of high gods. But the real extent of the popular religion and much of the hieratic must be sought for in the later Samhitās, and above all in the Atharvaveda.

The limitations of the Rigveda have been ascribed by Hillebrandt 14 to the existence in the period of that text of a ritual distinction of fundamental importance, that between the Devayana, the period when the gods are worshipped, and the Pitṛyāṇa, the period when the Fathers are revered. The former is the time when the sun is in the constellations in the north, and the moon in those to the south, while the reverse is the time of the Pitṛyāṇa, the distinction being marked in the mythology by the flight of the god Agni, possibly a reflection of the disappearance of the sun in the darkness of winter. The Rigveda, on this view, would represent the worship of the Devayāna; its exclusive character would be merely apparent. Unfortunately the sugges-

1i. 191.

x. 168.

x. 58; 60. 7-12.

蠢

x. 166.

孱 x. 183.

x. 162.

7 V. 55.

8 x. 145 ; cf. x. 159.

X. 85.

10 x. 14-18.

11 ix. 112; x. 34, 71, 117.

x. 81, 82, 90, 121, 129; i. 164, which, like

viii. 29, is a riddle hymn.

13 Macdonell, Sansk. Lit., pp. 120 ff.

1 Ved. Myth. iii. 67, 71, 204, 285, 299 ; (Kl. Ausg.), pp. 20, 50, 170, 177. Contrast Oldenberg, Rel. des Veda3, p. 11, n. 1.

Chap. 1]

The Rigveda and the Aryans

tion is open to two fatal objections. It is not in accord with tradition which does not thus connect the Rigveda with the ceremonies of the Devayana or the Uttarayana,1 with which Hillebrandt, without warrant, identifies that term, and, as a matter of fact, the Rigveda does contain, along with other matter not appropriate to its supposed purpose, a most important section of hymns dealing with the worship of the Fathers. We find, indeed, once more that only as a historical rather than a liturgical collection is the condition of the Rigveda logically explicable.

The form of the collection is entirely metrical, and it is matter of pure conjecture that in some cases the verses preserved represent merely one side of an ancient form of composition in which verses inserted in prose expressed the chief emotional points in conversation or narrative, or in the alternative that some hymns represent dramas in nuce. Neither hypothesis appears to have much plausibility, but for the purposes of the history of Vedic religion the question possesses no great importance.a

For this term see Macdonell and Keith,

Vedic Index, i. 529; ii. 467.

? See ref. in Keith, Sanskrit Drama, chap. i.

For Hertel’s theory of the origin of the Rigveda see Appendix A.