Sacred Texts  Hinduism  Index  Previous  Next 


Vedic Hymns, Part I (SBE32), by Max Müller, [1891], at sacred-texts.com


The most powerful instrument that has hitherto been applied to the emendation of Vedic texts, is the metre. Metrical criticism.Metre means measure, and uniform measure, and hence its importance for critical purposes, as second only to that of grammar. If our knowledge of the metrical system of the Vedic poets rests on a sound basis, any deviations from the general rule are rightly objected to; and if by a slight alteration they can be removed, and the metre be restored, we naturally feel inclined to adopt such emendations. Two safeguards, however, are needed in this kind of conjectural criticism. We ought to be quite certain that the anomaly is impossible, and we ought to be able to explain to a certain extent

p. lxxviii

how the deviation from the original correct text could have occurred. As this subject has of late years received considerable attention, and as emendations of the Vedic texts, supported by metrical arguments, have been carried on on a very large scale, it becomes absolutely necessary to reexamine the grounds on which these emendations are supposed to rest. There are, in fact, but few hymns in which some verses or some words have not been challenged for metrical reasons, and I feel bound, therefore, at the very beginning of my translation of the Rig-veda, to express my own opinion on this subject, and to give my reasons why in so many cases I allow metrical anomalies to remain which by some of the most learned and ingenious among Vedic scholars would be pronounced intolerable.

Even if the theory of the ancient metres had not been so carefully worked out by the authors of the Prâtisâkhyas and the Anukramanîs, an independent study of the Veda would have enabled us to discover the general rules by which the Vedic poets were guided in the composition of their works. Nor would it have been difficult to show how constantly these general principles are violated by the introduction of phonetic changes which in the later Sanskrit are called the euphonic changes of Sandhi, and according to which final vowels must be joined with initial vowels, and final consonants adapted to initial consonants, until at last each sentence becomes a continuous chain of closely linked syllables.

It is far easier, as I remarked before, to discover the original and natural rhythm of the Vedic hymns by reading them in the Pada than in the Samhitâ text, and after some practice our ear becomes sufficiently schooled to tell us at once how each line ought to be pronounced. We find, on the one hand, that the rules of Sandhi, instead of being generally binding, were treated by the Vedic poets as poetical licences only; and, on the other, that a greater freedom of pronunciation was allowed even in the body of words than would be tolerated in the later Sanskrit. If a syllable was wanted to complete the metre, a semivowel might be pronounced as a vowel, many a long vowel might

p. lxxix

be protracted so as to count for two syllables, and short vowels might be inserted between certain consonants, of which no trace exists in the ordinary Sanskrit. If, on the contrary, there were too many syllables, then the rules of Sandhi were observed, or two short syllables contracted by rapid pronunciation into one; nay, in a few cases, a final m or s, it seems, was omitted. It would be a mistake to suppose that the authors of the Prâtisâkhyas were not aware of this freedom allowed or required in the pronunciation of the Vedic hymns. Though they abstained from introducing into the text changes of pronunciation which even we ourselves would never tolerate, if inserted in the texts of Homer and Plautus, in the Pâli verses of Buddha, or even in modern English poetry, the authors of the Prâtisâkhya were clearly aware that in many places one syllable had to be pronounced as two, or two as one. They were clearly aware that certain vowels, generally considered as long, had to be pronounced as short, and that in order to satisfy the demands of the metre, certain changes of pronunciation were indispensable. They knew all this, but they did not change the text. And this shows that the text, as they describe it, enjoyed even in their time a high authority, that they did not make it, but that, such as it is, with all its incongruities, it had been made before their time. In many cases, no doubt, certain syllables in the hymns of the Veda had been actually lengthened or shortened in the Samhitâ text in accordance with the metre in which they are composed. But this was done by the poets themselves, or, at all events, it was not done by the authors of the Prâtisâkhya. They simply register such changes, but they do not enjoin them, and in this we, too, should follow their example. It is, therefore, a point of some importance in the critical restoration and proper pronunciation of Vedic texts, that in the rules which we have to follow in order to satisfy the demands of the metre, we should carefully distinguish between what is sanctioned by ancient authority, and what is the result of our own observations. This I shall now proceed to do.

First, then, the authors of the Prâtisâkhya distinctly admit

p. lxxx

that, in order to uphold the rules they have themselves laid down, certain syllables are to be pronounced as two syllables. Vyûha.We read in Sûtra 527: 'In a deficient pâda the right number is to be provided for by protraction of semivowels (which were originally vowels), and of contracted vowels (which were originally two independent vowels).' It is only by this process that the short syllable which has been lengthened in the Samhitâ, viz. the sixth, or the eighth, or the tenth, can be shown to have occupied and to occupy that place where alone, according to a former rule, a short syllable is liable to be lengthened. Thus we read:

I, 161, 11. ūdvātsv̅̆a̅̄smâ̄ ăkrĭnōtănâ̄ trĭnam.

This would seem to be a verse of eleven syllables, in which the ninth syllable na has been lengthened. This, however, is against the system of the Prâtisâkhya. But if we protract the semivowel v in udvatsv, and change it back into u, which it was originally, then we gain one syllable, the whole verse has twelve syllables, na occupies the tenth place, and it now belongs to that class of cases which is included in a former Sûtra, 523.

The same applies to X, 103, 13, where we read:

pr̅̆e̅̆tâ̄ găyătâ̄ nărāh.

[paragraph continues] This is a verse of seven syllables, in which the fifth syllable is lengthened, without any authority. Let us protract pretâ by bringing it back to its original component elements pra itâ, and we get a verse of eight syllables, the sixth syllable now falls under the general observation, and is lengthened in the Samhitâ accordingly.

The same rules are repeated in a later portion of the Prâtisâkhya. Here rules had been given as to the number of syllables of which certain metres consist, and it is added (Sûtras 972, 973) that where that number is deficient, it should be completed by protracting contracted vowels, and by separating consonantal groups in which semivowels (originally vowels) occur, by means of their corresponding vowel.

The rules in both places are given in almost identically

p. lxxxi

the same words, and the only difference between the two passages is this, that, according to the former, semivowels are simply changed back into their vowels, while, according to the latter, the semivowel remains, but is separated from the preceding consonant by its corresponding vowel.

These rules therefore show clearly that the authors of the Prâtisâkhya, though they would have shrunk from altering one single letter of the authorised Samhitâ, recognised the fact that where two vowels had been contracted into one, they might yet be pronounced as two; and where a vowel before another vowel had been changed into a semivowel, it might either be pronounced as a vowel, or as a semivowel preceded by its corresponding vowel. More than these two modifications, however, the Prâtisâkhya does not allow, or, at least, does not distinctly sanction. The commentator indeed tries to show that by the wording of the Sûtras in both places, a third modification is sanctioned, viz. the vocalisation, in the body of a word, of semivowels which do not owe their origin to an original vowel. But in both places this interpretation is purely artificial. Some such rule ought to have been given, but it was not given by the authors of the Prâtisâkhya. It ought to have been given, for it is only by observing such a rule that in I, 61, 12, gōr nă pa̅̆rv̆ă̆ vĭ rădâ̄ tĭrāskâ̆, we get a verse of eleven syllables, and thus secure for dâ in radâ the eighth place, where alone the short a could be lengthened. Yet we look in vain for a rule sanctioning the change of semivowels into vowels, except where the semivowels can rightly be called kshaipra-varna (Sûtra 974), i. e. semivowels that were originally vowels. The independent (svâbhâvika) semivowels, as e. g. the v in parva, are not included; and to suppose that in Sûtra 527 these semivowels were indicated by varna is impossible, particularly if we compare the similar wording of Sûtra 974 a.

p. lxxxii

We look in vain, too, in the Prâtisâkhya for another rule according to which long vowels, even if they do not owe their origin to the coalescence of two vowels, are liable to be protracted. However, this rule, too, though never distinctly sanctioned, is observed in the Prâtisâkhya, for unless its author observed it, he could not have obtained in the verses quoted by the Prâtisâkhya the number of syllables which he ascribes to them. According to Sûtra 937, the verse, RV. X, 134, 1, is a Mahâpaṅkti, and consists of six pâdas, of eight syllables each. In order to obtain that number, we must read:

sāmrâ̄gām kārshănî̄nâ̅̆m̅̄.

We may therefore say that, without allowing any actual change in the received text of the Samhitâ, the Prâtisâkhya distinctly allows a lengthened pronunciation of certain syllables, which in the Pada text form two syllables; and we may add that, by implication, it allows the same even in cases where the Pada text also gives but one instead of two syllables. Having this authority in our favour, I do not think that we use too much liberty if we extend this modified pronunciation, recognised in so many cases by the ancient scholars of India themselves, to other cases where it seems to us required as well, in order to satisfy the metrical rules of the Veda.

Secondly, I believe it can be proved that, if not the authors of the Prâtisâkhya, those at least who constituted Shortening of long vowels.the Vedic text which was current in the ancient schools and which we now have before us, were fully aware that certain long vowels and diphthongs could be used as short. The authors of the Prâtisâkhya remark that certain changes which can take place before a short syllable only, take place likewise before the word no, although the vowel of this 'no' is by them supposed to be long. After having stated in Sûtra 523 that the eighth syllable of hendecasyllabics and dodecasyllabics, if short, is lengthened, provided a short syllable

p. lxxxiii

follows, they remark that for this purpose nah or no is treated as a short syllable:

X, 59, 4. dyŭ-bhīh hĭtāh gărĭmâ̄ sŭ nă̂h āstū, (Samh. sû no astu.)

Again, in stating that the tenth syllable of hendecasyllabics and dodecasyllabics, if short, is lengthened, provided a short syllable follows, the same exception is understood to be made in favour of nah or no, as a short syllable:

VII, 48, 4. nŭ dēvâ̄sāh vărĭvāh kārtănă nāh, (Samh. kārtănâ̄ nŏ, bhûta no, &c.)

With regard to e being shortened before a short a where, according to rule, the a should be elided, we actually find that the Samhitâ gives a instead of e in RV. VIII, 72, 5. véti stótave ambyãm, Samh. véti stótava ambyãm. (Prâtis. 177, 5.)

I do not ascribe very much weight to the authority which we may derive from these observations with regard to our own treatment of the diphthongs e and o as either long or short in the Veda, yet in answer to those who are incredulous as to the fact that the vowels e and o could ever be short in Sanskrit, an appeal to the authority of those who constituted our text, and in constituting it clearly treated o as a short vowel, may not be without weight. We may also appeal to the fact that in Pâli and Prâkrit every final o and e can be treated as either long or short a. Starting from this we may certainly extend this observation, as it has been extended by Professor Kuhn, but we must not extend it too far It is quite clear that in the same verse e and o can be used both as long and short. I give the Samhitâ text:

I, 84, 17.

ka îshate | tugyate kō bĭbhâ̄yā
ko mamsate | santam indrām kŏ āntī,
kas tokâya | ka ibhâyōtă râ̄yē
adhi bravat | tanve kō gănâ̄yā.

p. lxxxiv

But although there can be no doubt that e and o, when final, or at the end of the first member of a compound, may be treated in the Veda as anceps, there is no evidence, I believe, to show that the same licence applies to a medial or initial e or o. In IV, 45, 5, we must scan

usrâh garante prati vastōhsvĭnâ̄,

ending the verse with an epitritus tertius instead of the usual dijambus a.

Thirdly, the fact that the initial short a, if following upon a word ending in o or e, is frequently not to be elided, is clearly recognised by the authors of the Prâtisâkhya (see p. xlviii). Nay, that they wished it to be pronounced even in passages where, in accordance with the requirements of the Prâtisâkhya, it had to disappear in the Samhitâ text, we may conclude from Sûtra 978. It is there stated that no pâda should ever begin with a word that has no accent. The exceptions to this rule are few, and they are discussed in Sûtras 978-987. But if the initial a were not pronounced in I, 1, 9, sáh nah pitâ´-iva sûnáve ágne su-upâyanáh bhava, the second pâda would begin with gne, a word which, after the elision of the initial a, would be a word without an accent b.

Fourthly, the fact that other long vowels, besides e and o, may under certain circumstances be used as short in the Veda, is not merely a modern theory, but rests on no less an authority than Pânini himself.

p. lxxxv

nini says, VI, I, 127, that i, u, ri (see RV. Bh. IV, 1, 12) at the end of a pada (but not in a compound a) may remain unchanged, if a different vowel follows, and that, if long, they may be shortened. He ascribes this rule, or, more correctly, the first portion of it only, to Sâkalya, Prâtisâkhya 155 seq. b Thus kakrî atra may become kakrĭ atra or kakry atra. Madhû atra may become madhŭ atra or madhv atra. In VI, 1, 128, Pânini adds that a, i, u, ri may remain unchanged before ri, and, if long, may be shortened, and this again according to the teaching of Sâkalya, i. e. Prâtisâkhya 136 c. Hence brahmâ rishih becomes brahmă rishih or brahmarshih; kumârî risyah becomes kumârĭ risyah or kumâry risyah. This rule enables us to explain a number of passages in which the Samhitâ text either changes the final long vowel into a semivowel, or leaves it unchanged, when the vowel is a pragrihya vowel. To the first class belong such passages as I, 163, 12; IV, 38, 10, vâgî´ árvâ, Samh. vâgyárvâ; VI, 7, 3, vâgî´ agne, Samh. vâgyãgne; VI, 20, 13, pakthî´ arkaíh, Samh. pakthyãrkaíh; IV, 22, 4, sushmî´ â´ góh, Samh. sushmyâ´ góh. In these passages î is the termination of a nom. masc. of a stem ending in in. Secondly, IV, 24, 8, pátnî ákkha, Samh. pátnyákkha; IV, 34, 1, devî´ áhnâm, Samh. devyáhnâm; V, 75, 4, vâ´nîki â´-hitâ, Samh. vâ´nîkyâ´hitâ; VI, 61, 4, avitrî´ avatu, Samh. avitryãvatu. In these passages the î is the termination of feminines. In X, 15, 4, ûtî´ arvâ´k, Samh. ûtyãrvâ´k, the final î of the instrumental ûtî ought not to have been changed into a semivowel, for, though not followed by íti, it is to be treated as pragrihya; (Prâtis. 163, 5.) It is, however,

p. lxxxvi

mentioned as an exception in Sûtra 174, 9. The same applies to II, 3, 4, védî íti asyâ´m, Samh. védyasyâ´m. The pragrihya î ought not to have been changed into a semivowel, but the fact that it had been changed irregularly, was again duly registered in Sûtra 174, 5. These two pragrihya î therefore, which have really to be pronounced short, were irregularly changed in the Samhitâ into the semivowel; and as this semivowel, like all semivowels, may take vyavâya, the same object was attained as if it had been written by a short vowel. With regard to pragrihya û, no such indication is given by the Samhitâ text; but in such passages as I, 46, 13, sambhû íti sam-bhû â´ gatam, Samh. sambhû â´ gatam; V, 43, 4, bâhû´ íti ádrim, Samh. bâhû´ ádrim, the pragrihya û of the dual can be used as short, like the û of madhû atra, given as an example by the commentators of Pânini.

To Professor Kuhn, I believe, belongs the merit of having extended this rule to final â. That the â of the dual may become short, was mentioned in the Prâtisâkhya, Sûtra 309, though in none of the passages there mentioned is there any metrical necessity for this shortening (see p. lii). This being the case, it is impossible to deny that where this â is followed by a vowel, and where Sandhi between the two vowels is impossible, the final â may be treated as short. Whether it must be so treated, depends on the view which we take of the Vedic metres, and will have to be discussed hereafter.

I agree with Professor Kuhn when he scans:

VI, 63, 1. kv̆ă tyâ̄ vālgû̄ pŭrŭ-hû̄tâ̆ ādyā, (Samh. puru-hûtâdya); and not kv̆ā ty̆â̄ vālgû̄ pŭrŭhû̄tâ̄dyā, although we might quote other verses as ending with an epitritus primus.

IV, 3, 13. mâ̄ vēsāsyă pră-mĭnătāh mâ̆ â̄pēh, (Samh. mâpeh,) although the dispondeus is possible.

I, 77, 1. kăthâ̄ dâ̄sēmă āgnăyē kâ̆ āsmai, (Samh. kâsmai.)

VI, 24, 5. āryāh văsāsyă părĭ-ētâ̆ āstī.

Even in a compound like tvâ-ûta, I should shorten the first vowel, e. g.

X, 148, 1. tmănâ̄ tănâ̄ sănŭyâ̄mă tvâ̆-û̄tâ̄h,

although the passage is not mentioned by the Prâtisâkhya

p. lxxxvii

among those where a short final vowel in the eighth place is not lengthened when a short syllable follows a.

But when we come to the second pâda of a Gâyatrî, and find there a long â, and that long â not followed by a vowel, I cannot agree with Professor Kuhn, that the long â, even under such circumstances, ought to be shortened. We may scan:

V, 5, 7. vâ̄tāsyă pātmăn î̄lĭtâ̄ dāivyâ̄ hōtâ̄râ̄ mănŭshāh.

The same choriambic ending occurs even in the last pâda of a Gâyatrî, and is perfectly free from objection at the end of the other pâdas.

So, again, we may admit the shortening of au to o in sâno avye and sâno avyaye, as quoted in the Prâtisâkhya, 174 and 177, but this would not justify the shortening of au to av in Anushtubh verses, such as

V, 86, 5.

mārtâ̄yă dēvâū ădăbhâ̄,
msâ̄-ivă dēvâ̄u ārvătē,

while, with regard to the Trishtubh and Gagatî verses, our views on these metres must naturally depend on the difficulties we meet with in carrying them out systematically.

There is no reason for shortening â in

V, 5, 10. dēvâ̄nâ̄m gūhyâ̄ nâ̄mâ̄nī.

It is the second pâda of a Gâyatrî here; and we shall see that, even in the third pâda, four long syllables occur again and again.

For the same reason I cannot follow Dr. Kuhn in a number of other passages where, for the sake of the metre, he proposes to change a long â into a short one. Such passages are in the Pada text:

VI, 46, 11. dīdyăvāh tīgmă-mû̄rdhâ̄nāh, not mû̄rdhâ̆nāh.

I, 15, 6. rĭtŭnâ̄ yāgñăm â̄sâ̄thē, not â̄sâ̆thē.

V, 66, 2. sāmyăk ăsūryăm â̄sâ̄tē, not â̄sâ̆tē.

V, 67, 1. vārshīshthām kshātrăm â̄sâ̄thē, not â̄sâ̆thē. See Beiträge, vol. iii, p. 122.

I, 46, 6. tâ̄m āsmē râ̄sâ̄thâ̄m ĭshām, not râ̄sâ̆thâ̄m ĭshām.

p. lxxxviii

IV, 32, 23. băbhrû̄ yâ̄mēshŭ sōbhētē, not sōbhĕtē.

IV, 45, 3. ŭtă prĭyām mădhŭnē yūñgâ̄thâ̄m răthām, not yūñgâ̆thâ̄m răthām.

V. 74, 3. kăm ākkhă yūñgâ̄thē răthām, not yūñgâ̆thē răthām.

IV, 55, 1. dyâ̄vâ̄bhû̄mî̄ (íti) ădĭtē trâ̄sî̄thâ̄m nāh, not trâ̄sî̆thâ̄m nāh.

V, 41, 1. rĭtāsyă vâ̄ sădăsĭ trâ̄sî̄thâ̄m nāh, not trâ̄sî̆thâ̄m nāh.

I must enter the same protest against shortening other long vowels in the following verses which Professor Kuhn proposes to make metrically correct by this remedy:

I, 42, 6. hĭrānyăvâ̄sî̄māt-tămā, not vâ̄sî̆māt-tȃmā.

Here the short syllable of ganasrĭ-bhih in V, 60, 8, cannot be quoted as a precedent, for the i in ganasri, walking in companies, was never long, and could therefore not be shortened. Still less can we quote nâri-bhyah as an instance of a long î being shortened, for nâri-bhyah is derived from nârih, not from nârî, and occurs with a short i even when the metre requires a long syllable; I, 43, 6. nri̅-bhyāh nâ̄rĭ̄-bhy̆āh găvē. The fact is, that in the Rig-veda the forms nârishu and nârî-bhyah never occur, but always nârishu, nâri-bhyah; while from vâsî we never find any forms with short i, but always vâ´sîshu, vâ´sî-bhih.

Nor is there any justification for change in I, 25, 16. gâ̄vāh nă gāvyû̄tî̄h ănū, the second pâda of a Gâyatrî. Nor in V, 56, 3. ri̅kshāh nă vāh mărŭtāh sĭmî̄-vâ̄n ămāh. In most of the passages mentioned by Professor Kuhn on p. 122, this peculiarity may be observed, that the eighth syllable is short, or, at all events, may be short, when the ninth is long:

VI, 44, 21. vri̅shnē tē īndūh vrĭshăbhă ̍ pî̄pâ̄yā.

I, 73, 1. sy̆ōnă-sî̄h ătĭthīh nă ̍ prî̄nâ̄nāh.

VII, 13, 1. bhărē hăvīh nă bărhĭshĭ ̍ iprî̄nâ̄nāh.

II, 28, 7. ēnāh kri̅nvāntăm ăsŭră ̍ bhrî̄nāntī.

Before, however, we can settle the question whether in

p. lxxxix

these and other places certain vowels should be pronounced as either long or short, we must settle the more general Metre and Grammar.question, what authority we have for requiring a long or a short syllable in certain places of the Vedic metres.


Footnotes

lxxxi:a It will be seen from my edition of the Prâtisâkhya, particularly from the extracts from Uvata, given after Sûtra 974, that the idea of making two syllables out of goh, never entered Uvata's mind. M. Regnier was right, Professor Kuhn (Beiträge, vol. iv, p. 187) was wrong. Uvata, no doubt, wishes to show that original (svâbhâvika) semivowels are liable to vyûha, or at least p. lxxxii to vyavâya; but though this is true in fact, Uvata does not succeed in his attempt to prove that the rules of the Prâtisâkhya sanction it.

lxxxiii:a See Lassen, Inst. Linguæ Pracriticæ, pp. 145, 147, 151; Cowell, Vararuki, Introduction, p. xvii. Kedârabhatta says: Pâninir bhagavân prâkritalakshanam api vakti samskritâd anyat, dîrghâksharam ka kutrakid ekâm mâtrâm upaitîti. Secundum d’Alwisium commentator docet sermonem esse de litteris Sanscriticis e et o. Cf. Pischel, De Grammaticis prakriticis, 1874.

lxxxiv:a See Professor Weber's pertinent remarks in Kuhn's Beiträge, vol. iii, p. 394. I do not think that in the verses adduced by Professor Kuhn, in which final o is considered by him as an iambus or trochee, this scanning is inevitable. Thus we may scan the Samhitâ text:

I, 88, 2. rūkmō nă kītr̆āh sv̆ădhĭtî̄vâ̄n.

I, 141, 8. răthō nă ȳâ̆tăh sīkv̆ăbhīh krĭtō.

I, 174, 3. sīmhō nă dămĕ ăp̆â̄msĭ vāstōh.

VI, 24, 3. ākshō nă kākry̆ōh s̆û̄ră brĭhān

X, 3, 1. ĭnō r̆â̄gānn ărătīh sămīddhō.

This leaves but one of Professor Kuhn's examples (Beiträge, vol. iv, p. 192) unexplained: I, 191, 1. kaṅkato na kaṅkato, where iva for na would remove the difficulty.

lxxxiv:b This subject, the shortness of e and o in the Veda, has been admirably treated by Mr. Maurice Bloomfield, 'Final as before Sonants,' Baltimore, 1882. Reprinted from the American Journal of Philology, vol. iii, No. 1.

lxxxv:a There are certain compounds in which, according to Professor Kuhn, two vowels have been contracted into one short vowel. This is certainly the opinion of Hindu grammarians, also of the compiler of the Pada text. But most of them would admit of another explanation. Thus dhánvarnasah, which is divided into dhánva-arnasah, may be dhánu-arnasah (RV. V, 45, 2). Dhánarkam, divided into dhána-arkam, may have been dhána-rikam (RV. X, 46, 5). Satárkasam (RV. VII, too, 3) may be taken as satá-rikasam instead of satá-arkasam.

lxxxv:b In the Prâtisâkhya the rule which allows vowel before vowel to remain unchanged, is restricted to special passages, and in some of them the two vowels are savarna; cf. Sûtra 163.

lxxxv:c Cf. Vâgasan. Prâtisâkhya, IV, 48; Indische Studien, vol. ix, 309; vol. x, 406.

lxxxvii:a I see that Professor Kuhn, vol. iv, p. 186, has anticipated this observation in eshtau, to be read â̆-īshtaū.


Next: Part 8