Pramāṇa · Search · About · API

The History of Indian Literature c9: Second Period. Sanskrit Literature

The History of Indian Literature c9

Language
English

Contents

Second Period. Sanskrit Literature

c9.2

SECOND PERIOD.

c9.4

HAVING thus followed the first period of Indian literature, in its several divisions, down to its close, we now turn to its second period, the so-called Sanskṛit literature. Here, however, as our time is limited, we cannot enter so much into detail as we have hitherto done, and we must therefore content ourselves with a general survey. In the case of the Vedic literature, details were especially essential, both because no full account of it had yet been given, and because the various works still lie, for the most part, shut up in the manuscripts; whereas the Sanskṛit literature has already been repeatedly handled, partially at least, and the principal works belonging to it are generally accessible.

c9.5

Our first task, naturally, is to fix the distinction between the second period and the first. This is, in part, one of age, in part, one of subject-matter. The former distinction is marked by the language and by direct data; the latter by the nature of the subject-matter itself, as well as by the method of treating it.

c9.6

As regards the language, in the first place, in so far as it grounds a distinction in point of age between the two periods of Indian literature, its special characteristics in the second period, although apparently slight, are yet, in reality, so significant that it appropriately furnishes the name for the period; whereas the earlier one receives its designation from the works composing it.

c9.7

Among the various dialects of the different Indo-Aryan tribes, a greater unity had in the course of time been established after their immigration into India, as the natural result of their intermingling in their new homes, and of their combination into larger communities. The grammaticaln∗ study, moreover, which by degrees became necessary for the interpretation of the ancient texts, and which grew up in connection therewith, had had the effect of substantially fixing the usage; so that a generally recognised language, known as the *bháshá*, had arisen, that, namely, in which the Bráhmaṇas and Sútras are composed.n† Now the greater the advance made by the study of grammar, the more stringent and precise its precepts and rules became, and all the more difficult it was for those who did not occupy themselves specially therewith to keep in constant accord with grammatical accuracy. The more the language of the grammatically educated gained on the one hand in purity, and in being purged of everything not strictly regular, the more foreign did it become on the other hand to the usage of the majority of the people, who were without grammatical training. In this way a refined language gradually disconnected itself from the vernacular, as more and more the exclusive property of the higher classes of the people;n‡ the estrangement between the two growing more and more marked, as the popular dialect in its turn underwent further development. This took place mainly under the influence of those aboriginal inhabitants who had been received into the Brahmanic community; who, it is true, little by little exchanged their own language for that of their conquerors, but not without importing into the latter a large number of new words and of phonetic changes, and, in particular, very materially modifying the pronunciation. This last was all the more necessary, as the numerous accumulations of consonants in the Aryan *bháshá* presented exceeding difficulties to the natives; and it was all the easier, as there had evidently prevailed within the language itself from an early period a tendency to clear away these troublesome encumbrances of speech,—a tendency to which, indeed, the study of grammar imposed a limit, so far as the educated portion of the Aryans was concerned, but which certainly maintained itself, and by the very nature of the case continued to spread amongst the people at large. This tendency was naturally furthered by the native inhabitants, particularly as they acquired the language not from those who were conversant with grammar, but from intercourse and association with the general body of the people. In this way there gradually arose new vernaculars, proceeding directly from the common *bháshá*,[^asterisk] and distinguished from it mainly by the assimilation of consonants, and by

c9.8

n∗ * Respecting the use of the verb *vyákri* in a grammatical signification, Sāyaṇa in his introduction to the Ṛik (p. 35. 22 ed. Müller) adduces a legend from a Bráhmaṇa, which represents Indra as the oldest grammarian. (See Lassen, *I. AK.*, ii. 475.) [The legend is taken from the TS. vi. 4. 7. 3. All that is there stated, indeed, is that *vách* was *vyákṛitá* by Indra; manifestly, however, the later myths which do actually set up Indra as the oldest grammarian connect themselves with this passage.]

c9.9

n† † *Bháshika-svara* in Kátyáyana, Śrauta-Sútra, i. 8. 17, is expressly interpreted as *bráhmaṇa-svara*; see *Váj. Saṃh. Specimen*, ii. 196. 197. [*I. St.*, x. 428–429, 437.] Yáska repeatedly opposes *bhásháyám* and *anvadhyáyam* (i.e., ‘in the Veda reading,’ ‘in the text of the hymns’) to each other; similarly, the Prátiśákhya-Sútras employ the words *bháshá* and *bháshya* as opposed to *chhandas* and *veda*, *i.e.*, *saṃhitá* (see above, pp. 57, 103, 144). The way in which the word *bháshya* is used in the Gṛihya Sutra of Śáṅkháyana, namely, in contradistinction to *Sútra*, shows that its meaning had already by this time become essentially modified, and become restricted, precisely as it is in Páṇini, to the extra-Vedic, so to say, profane literature. (The Áśvaláyana-Gṛihya gives instead of *bháshya*, in the corresponding passage, *bhárata-mahábhárata-dharma*.) [This is incorrect; rather, in the passage in question, these words follow the word *bháshya*; see the note on this point at p. 56.] In the same way, in the Nir. xiii. 9, *mantra*, *kalpa*, *bráhmaṇa*, and the *vyávahárikí* (sc. *bháshá*) are opposed to each other (and also *Ṛik*, *Yajus*, *Sáman*, and the *vyávahárikí*).

c9.10

n‡ ‡ Ought the passage cited in Nir. xiii. 9 from a Bráhmaṇa [cf. Káṭh. xiv. 5], to the effect that the Brahmans spoke both tongues, that of the gods as well as that of men, to be taken in this connection? or has this reference merely to a conception resembling the Homeric one?

c9.11

[^asterisk]: And therefore specially so called down even to modern times; whereas the grammatically refined *bháshá* afterwards lost this title, and substituted for it the name *Saṃskṛita-bháshá*, 'the cultivated speech.' The name *Prákṛita-bháshá*, which was at the same time applied to the popular dialects, is derived from the word *prakṛiti*, 'nature,' 'origin,' and probably describes these as the 'natural,' 'original' continuations of the ancient *bháshá*: or does *prákṛita* here signify 'having a *prakṛiti* or origin,' *i.e.*, 'derived'? [Out of the signification 'original,' 'lying at the root of' (*prakṛiti-bhúta*), 'unmodified,' arose that of 'normal,' then that of 'ordinary,' '*communis*,' '*vulgaris*,' and lastly, that of 'proceeding in common from.' The term directly opposed to it is not *saṃskṛita*, but *vaikṛita*; see, *e.g.*, Ath. Pariś. 49. 1, "*varṇán púrvaṃ vyákhyásyámaḥ prákṛitá ye cha vaikṛitáḥ.*"] The earliest instances as yet known of the name *Saṃskṛit* as a designation of the language occur in the Mṛichhakaṭí (p. 44. 2, ed. Stenzler), and in Varáha-Mihira's Bṛihat-Saṃhitá, 85. 3. The following passages also of the Rámáyaṇa are doubtless to be understood in this sense, viz., v. 18. 19, 29. 17, 34 (82. 3), vi. 104. 2. Páṇini is familiar with the word *Saṃskṛita*, but does not use it in this sense; though the Páṇiníyá-Śikshá does so employ it (v. 3), in contradistinction to *prákṛita*.

c9.12

the curtailment or loss of terminations. Not unfrequently, however, they present older forms of these than are found in the written language, partly because the latter has rigorously eliminated all forms in any way irregular or obsolete, but partly also, no doubt, from the circumstance that grammar was cultivated principally in the north or north-west of India, and consequently adapted itself specially to the usage there prevailing. And in some respects (*e.g.*, in the instr. plur. of words in *a*?)¹⁹⁷ this usage may have attained a more developed phase than appears to have been the case in India Proper,∗ since the language was not there hampered in its independent growth by any external influence; whereas the Aryans who had passed into India maintained their speech upon the same internal level on which it stood at the time of the immigration,† however considerable were the external modifications which it underwent.

c9.13

¹⁹⁷ This example is not quite pertinent, as the instr. plur. in *-áis* is of very ancient date, being reflected not only in Zend, but also in Slavonic and Lithuanian; see Bopp, *Vergl. Gram.*, i. 156^2 (159^3).

c9.14

∗ The difference in usage between the Eastern and Western forms of speech is once touched upon in the Bráhmaṇa of the White Yajus, where it is said that the Váhíkas style Agni *Bhava*, while the Práchyas, on the contrary, call him *Śarva*. Yáska (ii. 2) opposes the Kambojas (the Persa-Aryans?) to the Áryas (the Indo-Aryans?), stating that the latter, for instance, possess derivatives only of the root *śu*, whereas the Kambojas possess it also as a verb. (Grammarians of the Kambojas are hardly to be thought of here, as Roth, *Zur Lit.*, p. 67, supposes.) Yáska further opposes the Práchyas and the Udíchyas, and the same is done by Páṇini. According to the Bráhmaṇa, the Udíchyas were most conversant with grammar [see *I. St.*, i. 153, ii. 309, 310, xiii. 363, ff. Burnell's identification of the Kambojas here, and in the other earlier passages where they are mentioned, with Cambodia in Farther India, see his *Elements of South Indian Palæography*, pp. 31, 32, 94, is clearly a mistake. For the time of the Páḷi Abhidhánappadípiká (*v.* Childers, *Páli Dict.*) this identification may perhaps be correct; but the older Páḷi texts, and even the inscriptions of Piyadasi (*e.g.*, most distinctly the facsimile of the Khálsi inscription in Cunningham's *Archæological Survey*, i. 247, pl. xli., line 7), introduce the Kambojas in connection with the Yavanas; and this of itself determines that the two belonged geographically to the same region in the north-west of India; see *I. Str.*, ii. 321. In addition to this we have the name Kabujiya = Καμβύσης, and therewith all the various references to this latter name, which point to a very wide ramification of it throughout Irán; see *I. Str.*, ii. 493. To Farther India the name Kamboja evidently found its way only in later times, like the names Ayodhyá, Indraprastha, Irávatí, Champá; though it certainly remains strange that this lot should have fallen precisely to it. Perhaps causes connected with Buddhism may have helped to bring this about. See on this point the Jenaer *Literaturzeitung*, 1875, p. 418; *Indian Antiquary*, iv. 244.]

c9.15

† Much as the Germans did, who in the middle ages emigrated to Transylvania.

c9.16

The second period of Indian literature, then, commences with the epoch when the separation of the language of the educated classes—of the written language—from the popular dialects was an accomplished fact. It is in the former alone that the literature is presented to us. Not till after the lapse of time did the vernaculars also in their turn produce literatures of their own,—in the first instance under the influence of the Buddhist religion, which addressed itself to the people as such, and whose scriptures and records, therefore, were originally, as for the most part they still are, composed in the popular idiom. The epoch in question cannot at present be precisely determined; yet we may with reasonable certainty infer the existence of the written language also, at a time when we are in a position to point to the existence of popular dialects; and with respect to these we possess historical evidence of a rare order, in those rock-inscriptions, of identical purport, which have been discovered at Girnar in the Gujarát peninsula, at Dhauli in Orissa, and at Kapur di Giri¹⁹⁸ in Kabul. J. Prinsep, who was the first to decipher them, and Lassen, refer them to the time of the Buddhist king Aśoka, who reigned from B.C. 259; but, according to the most recent investigations on the subject—by Wilson, in the "Journal of the Royal Asiatic Society," xii., 1850 (p. 95 of the separate impression)—they were engraved "at some period subsequent to B.C. 205,"∗ and are still, therefore, of uncertain date. However this question may be settled, it in any case results with tolerable certainty that these popular dialects were in existence in the third century B.C. But this is by no means to be set down as the limit for the commencement of their growth; on the contrary, the form in which they are presented to us sufficiently shows that a very considerable period must have elapsed since their separation from the ancient *bháshá*. This separation must therefore have taken place comparatively early, and indeed we find allusions to these vernaculars here and there in the Bráhmaṇas themselves.∗

c9.17

¹⁹⁸ This name ought probably to be written *Kapardigiri*? See my paper on the Satruṃjaya Máhátmya, p. 118. In these inscriptions, moreover, we have a text, similar in purport, presented to us in three distinct dialects. See further on this subject Burnouf's admirable discussion of these inscriptions in his *Lotus de la bonne Loi*, p. 652, ff. (1852); *I. St.*, iii. 467, ff. (1855); and Kern, *De Gedenkstukken van Aśoka den Buddhist* (1873, particularly p. 32 ff., 45 ff.).

c9.18

∗ And that not much later; as is vouched for by the names of the Greek kings therein mentioned—Alexander, Antigonus, Magas, Ptolemy, Antiochus. These cannot, it is true, be regarded as contemporaneous with the inscriptions; but their notoriety in India can hardly have been of such long duration that the inscriptions can have been composed long after their time. See Wilson, *l. c.*

c9.19

∗ Thus in the second part of the Aitareya-Bráhmaṇa the Śyáparṇas, a clan (?) of the western Salvas, are mentioned as "*pútáyai vácho vaditáraḥ*," 'speaking a filthy tongue;' and in the Pañchaviṅśa-Bráhmaṇa, the Vrátyas are found fault with for their debased language. The Asuras are similarly censured in the Śatapatha-Bráhmaṇa (iii. 2. 1. 24), where, at the same time, the Brahmans are warned against such forms of speech; "*tasmád bráhmaṇo na mlechhet.*"—I may remark here in passing that M. Müller, in his edition of the Ṛik, in Sáyaṇa's introduction, p. 36. 21, erroneously writes *helayo* as one word: it stands for *he 'layo*,—the Asura corruption of the battle-cry *he 'rayo* (*arayo*): according to the Śatapatha-Bráhmaṇa, it even took the form *he 'lavo*.

c9.20.1

The direct data, attesting the posteriority of the second period of Indian literature, consist in these facts: first, that its opening phases everywhere presuppose the Vedic period as entirely closed; next, that its oldest portions are regularly based upon the Vedic literature; and, lastly, that the relations of life have now all arrived at a stage of development of which, in the first period, we can only trace the germs and beginning. Thus, in particular, divine worship is now centred on a triad of divinities, Brahman, Vishṇu, and Śiva; the two latter of whom, again, in course of time, have the supremacy severally allotted to them, under various forms, according to the different sects that grew up for this purpose. It is by no means implied that individual portions of the earlier period may not run on into the later; on the contrary, I have frequently endeavoured in the preceding pages to show that such is the case. For the rest, the connection between the two periods is, on the whole, somewhat loose: it is closest as regards those branches of literature which had already attained a definite stage of progress in the first period, and which merely continued to develop further in the second,—Grammar, namely, and Philosophy. In regard to those branches, on the contrary, which are a more independent growth of the second period, the difficulty of connecting them with the earlier age is very great. We have here a distinct gap which it is altogether impossible to fill up.

c9.20.2

The reason of this lies simply in the fact, that owing to the difficulty of preserving literary works, the fortunate successor almost always wholly supplanted the predecessor it surpassed: the latter thus became superfluous, and was consequently put aside, no longer committed to memory, no longer copied. In all these branches therefore—unless some other influence has supervened—we are in possession only of those master-works in which each attained its culminating point, and which in later times served as the classical models upon which the modern literature was formed, itself more or less destitute of native productive energy. This fact has been already adduced as having proved equally fatal in the case of the more ancient Brâhmaṇa literature, &c.; there, much to the same extent as here, it exercised its lamentable, though natural influence. In the Vedic literature also, that is to say, in its Śákhás, we find the best analogy for another kindred point, namely, that some of the principal works of this period are extant in several—generally two—recensions. But along with this a further circumstance has to be noted, which, in consequence of the great care expended upon the sacred literature, has comparatively slight application to it, namely, that the mutual relation of the manuscripts is of itself such as to render any certain restoration of an original text for the most part hopeless.

c9.20.3

It is only in cases where ancient commentaries exist that the text is in some degree certain, for the time at least to which these commentaries belong. This is evidently owing to the fact that these works were originally preserved by oral tradition; their consignment to writing only took place later, and possibly in different localities at the same time, so that discrepancies of all sorts were inevitable. But besides these variations there are many alterations and additions which are obviously of a wholly arbitrary nature, partly made intentionally, and partly due to the mistakes of transcribers. In reference to this latter point, in particular, the fact must not be lost sight of that, in consequence of the destructive influence of the climate, copies had to be renewed very frequently. As a rule, the more ancient Indian manuscripts are only from three to four hundred years old; hardly any will be found to date more than five hundred years back.198a Little or nothing, therefore, can here be effected by means of so-called diplomatic criticism. We cannot even depend upon a text as it appears in quotations, such quotations being generally made from memory,—a practice which, of course, unavoidably entails mistakes and alterations.

c9.21

198a Regarding the age, manner of preparation, material, and condition of text of Indian MSS., see Ráj. Lála Mitra's excellent report, dated 15th February 1875, on the searches instituted by him in native libraries down to the end of the previous year, which is appended to No. IX. of his *Notices of Sanskrit MSS.* Quite recently some Devanágarí MSS. of Jaina texts, written on broad palm-leaves, have been discovered by Bühler, which date two centuries earlier than any previously known. A facsimile of one of these MSS. in Bühler's possession, the Ávaśyaka-Sútra, dated *Saṃvat* 1189 (A.D. 1132), is annexed to the above-mentioned report: "it is the oldest Sanskrit MS. that has come to notice," Ráj. L. Mitra, *Notices*, iii. 68 (1874). But a letter from Dr. Rost (19th October 1875) intimates that in one of the Sanskrit MSS. that have lately arrived in Cambridge from Nepál, he has read the date 128 of the Nepál era, *i.e.*, A.D. 1008. Further confirmation of this, of course, still remains to be given.

c9.22

The distinction in point of subject-matter between the first and second periods consists mainly in the circumstance that in the former the various subjects are only handled in their details, and almost solely in their relation to the sacrifice, whereas in the latter they are discussed in their general relations. In short, it is not so much a practical, as rather a scientific, a poetical, and artistic want that is here satisfied. The difference in the form under which the two periods present themselves is in keeping with this. In the former, a simple and compact prose had gradually been developed, but in the latter this form is abandoned, and a rhythmic one adopted in its stead, which is employed exclusively, even for strictly scientific exposition. The only exception to this occurs in the grammatical and philosophical Sútras; and these again are characterised by a form of expression so condensed and technical that it cannot fittingly be termed prose. Apart from this, we have only fragments of prose, occurring in stories which are now and then found cited in the great epic; and further, in the fable literature and in the drama; but they are uniformly interwoven with rhythmical portions. It is only in the Buddhist legends that a prose style has been retained, the language of which, however, is a very peculiar one, and is, moreover, restricted to a definite field. In fact, as the result of this neglect, prose-writing was completely arrested in the course of its development, and declined altogether. Anything more clumsy than the prose of the later Indian romances, and of the Indian commentaries, can hardly be; and the same may be said of the prose of the inscriptions.

c9.23

This point must not be left out of view, when we now proceed to speak of a classification of the Sanskrit literature into works of Poetry, works of Science and Art, and works relating to Law, Custom, and Worship. All alike appear in a poetic form, and by ‘Poetry’ accordingly in this classification we understand merely what is usually styled *belles-lettres*, though certainly with an important modification of this sense. For while, upon the one hand, the poetic form has been extended to all branches of the literature, upon the other, as a set-off to this, a good deal of practical prose has entered into the poetry itself, imparting to it the character of poetry ‘with a purpose.’ Of the epic poetry this is especially true.

c9.24.1

It has long been customary to place the Epic Poetry at the head of Sanskrit literature; and to this custom we here conform, although its existing monuments cannot justly pretend to pass as more ancient than, for example, Pāṇini’s grammar, or the law-book which bears the name of Manu. We have to divide the epic poetry into two distinct groups: the *Itihāsa-Purāṇas* and the *Kāvyas*. We have already more than once met with the name Itihāsa-Purāṇa in the later Brāhmaṇas, namely, in the second part of the Śatapatha-Brāhmaṇa, in the Taittirīya-Āraṇyaka, and in the Chhāndogyopanishad. We have seen that the commentators uniformly understand these expressions to apply to the legendary passages in the Brāhmaṇas themselves, and not to separate works; and also that, from a passage in the thirteenth book of the Śatapatha-Brāhmaṇa, it results with tolerable certainty that distinct works of this description cannot then have existed, inasmuch as the division into *parvans*, which is usual in the extant writings of this class, is there expressly attributed to other works, and is not employed in reference to these Itihāsa-Purāṇas themselves.

c9.24.2

On the other hand, in the Sarpa-vidyā (‘serpent-knowledge’) and the Devajana-vidyá (‘genealogies of the gods')—to which, in the passage in question, the distribution into *parvans*, that is to say, existence in a distinct form, is expressly assigned—we have in all probability to recognise mythological accounts, which from their nature might very well be regarded as precursors of the epic. We have likewise already specified as forerunners of the epic poetry, those myths and legends which are found interspersed throughout the Bráhmaṇas, here and there, too, in rhythmic form,∗ or which lived on elsewhere in the tradition regarding the origin of the songs of the Ṛik. Indeed, a few short prose legends of this sort have been actually preserved here and there in the epic itself. The Gáthás also—stanzas in the Bráhmaṇas, extolling individual deeds of prowess—have already been cited in the like connection: they were sung to the accompaniment of the lute, and were composed in honour either of the prince of the day or of the pious kings of old (see *I. St.*, i. 187). As regards the extant epic—the *Mahá-Bhárata*—specially, we have already pointed out the mention in the Taittiríya-Áraṇyaka, of Vyása Páráśarya¹⁹⁹ and Vaiśampáyana,²⁰⁰ who are given in the poem itself as its original authors; and we have also remarked (p.

c9.24.3

143) that the family of the Paráśaras is represented with especial frequency in the *vanśas* of the White Yajus.[^asterisk] We also find repeated allusions in the Bráhmaṇas to a Naimishíya sacrifice, and, on the authority of the Mahá-Bhárata itself, it was at such a sacrifice that the *second* recitation of the epic took place in presence of a Śaunaka. But, as has likewise been remarked above [pp. 34, 45], these two sacrifices must be kept distinct, and indeed there is no mention in the Bráhmaṇas of a Śaunaka as participating in the former. Nay, several such sacrifices may have taken place in the Naimisha forest [see p. 34]; or it is possible even that the statement as to the recitation in question may have no more foundation than the desire to give a peculiar consecration to the work. For it is utterly absurd to suppose that Vyása Páráśarya and Vaiśampáyana—teachers mentioned for the first time in the Taittiríya-Áraṇyaka—could have been anterior to the sacrifice referred to in the Bráhmaṇas. The mention of the “Bhárata” and of the “Mahá-Bhárata” itself in the Gṛihya-Sútras of Áśvaláyana [and Śáṅkháyana] we have characterised [p. 58] as an interpolation or else an indication that these Sútras are of very late date. In Páṇini the word “Mahá-Bhárata” does indeed occur; not, however, as denoting the epic of this name, but as an appellative to designate any individual of special distinction among the Bháratas, like Mahá-Jábála,-Hailihila (see *I. St.*, ii. 73).

c9.24.4

Still, we do find names mentioned in Páṇini which belong specially to the story of the Mahá-Bhárata—namely, Yudhishṭhira, Hástinapura, Vásudeva, Arjuna,dagger Andhaka-Vṛishṇayas, Droṇa (?); so that the legend must in any case have been current in his day, possibly even in a poetical shape; however surprising it may be that the name Páṇḍu[^double-dagger] is never mentioned by him. The earliest direct

c9.25

∗ As, for instance, the story of Hariśchandra in the second part of the *Aitareya-Bráhmaṇa*.

c9.26

¹⁹⁹ Vyása Páráśarya is likewise mentioned in the *vaṅśa* of the Sáma-vidhána-Bráhmaṇa, as the disciple of Vishvaksena, and preceptor of Jaimini; see *I. St.*, iv. 377.—The Mahábháshya, again, not only contains frequent allusions to the legend of the Mahá-Bhárata, and even metrical quotations that connect themselves with it, but it also contains the name of Śuka Vaiyásaki; and from this it is clear that there was then already extant a poetical version of the Mahá-Bhárata story; see *I. St.*, xiii. 357. Among the prior births of Buddha is one (No. 436 in Westergaard's *Catalogus*, p. 40), bearing the name Kaṇha-Dípáyana, *i.e.*, Kṛishṇa-Dvaipáyana!

c9.27

²⁰⁰ Vaiśampáyana appears elsewhere frequently, but always in special relation to the transmission of the Yajur-Veda. By Páṇini, it is true (iv. 3. 104), he is simply cited generally as a Vedic teacher, but the Mahábháshya, commenting on this passage, describes him as the teacher of Kaṭha and Kalápin. In the Calcutta Scholium, again, we find further particulars (from what source? cf. Táránátha on *Siddh. Kaum.*, i. 590), according to which (see *I. St.*, xiii. 440) nine Vedic schools, and among them two belonging to the Sáma-Veda, trace their origin to him. In the Ṛig-Gṛihya he is evidently regarded (see above, pp. 57, 58), after the manner of the Vishṇu-Puráṇa, as the special representative of the Yajur-Veda; and so he appears in the Anukr. of the Átreyí school, at the head of its list of teachers, specially as the preceptor of Yáska Paiṅgi.

c9.28

dagger † A worshipper of Vásudeva, or of Arjuna, is styled ‘Vásudevaka,’ ‘Árjunaka.’ Or is Arjuna here still a name of Indra? [From the context he is to be understood as a Kshatriya; see on this, *I. St.*, xiii. 349, ff.; *Ind. Antiq.* iv. 246.] [^double-dagger]: ‡ This name only occurs in the Mahá-Bhárata and in the works resting upon it. Yet the Buddhists mention a mountain tribe of Páṇḍavas, as alike the foes of the Śákyas (*i.e.*, the Kośalas) and of the inhabitants of Ujjayiní; see Schiefner, *Leben des Śákyamuni*, pp. 4, 40 (in the latter passage they appear to be connected with Takshaśilá?), and, further, Lassen, *I. AK.*, ii. 100, ff.; Foucaux, *Rgya Cher Rol Pa*, pp. 228, 229 (25, 26).

c9.29

[^asterisk]: * This renders Lassen’s reference (*I. AK.*, i. 629) of the name Páráśarya to the astronomer or chronologer Paráśara, highly questionable.

c9.30

evidence of the existence of an epic, with the contents of the Mahá-Bhárata, comes to us from the rhetor Dion Chrysostom, who flourished in the second half of the first century A.D.; and it appears fairly probable that the information in question was then quite new, and was derived from mariners who had penetrated as far as the extreme south of India, as I have pointed out in the *Indische Studien*, ii. 161–165.∗ Since Megasthenes says nothing of this epic, it is not an improbable hypothesis that its origin is to be placed in the interval between his time and that of Chrysostom; for what ignorant† sailors took note of would hardly have escaped his observation; more especially if what he narrates of Herakles and his daughter Pandaia has reference really to Kṛishṇa and his sister, the wife of Arjuna, if, that is to say, the Páṇḍu legend was already actually current in his time. With respect to this latter legend, which forms the subject of the Mahá-Bhárata, we have already remarked, that although there occur, in the Yajus especially, various names and particulars having an intimate connection with it, yet on the other hand these are presented to us in essentially different relations. Thus the Kuru-Pañchálas in particular, whose internecine feud is deemed by Lassen to be the leading and central feature of the Mahá-Bhárata, appear in the Yajus on the most friendly and peaceful footing: Arjuna again, the chief hero of the Páṇḍus, is still, in the Vájasaneyi-Saṃhitá and the Śatapatha-Bráhmaṇa, a name of Indra:‡ and lastly, Janamejaya Páriksbita, who in the Mahá-Bhárata is the great-grandson of Arjuna, appears, in the last part of the Śatapatha-Bráhmaṇa, to be still fresh in the memory of the people, with the rise and downfall of himself and his house. I have also already expressed the conjecture that it is perhaps in the deeds and downfall of this Janamejaya that we have to look for the original plot of the story of the Mahá-Bhárata;[^asterisk] and, on the other hand, that, as in the epics of other nations, and notably in the Persian Epos, so too in the Mahá-Bhárata, the myths relating to the gods became linked with the popular legend. But so completely have the two been interwoven that the unravelling of the respective elements must ever remain an impossibility. One thing, however, is clearly discernible in the Mahá-Bhárata, that it has as its basis a war waged on the soil of Hindustán between Aryan tribes, and therefore belonging probably to a time when their settlement in India, and the subjugation and brahmanisation of the native inhabitants, had already been accomplished. But what it was that gave rise to the conflict—whether disputes as to territory, or it may be religious dissensions—cannot now be determined.—Of the Mahá-Bhárata in its extant form, only about one-fourth (some 20,000 *ślokas* or so) relates to this conflict and the myths that have been associated with it;²⁰¹ while the elements composing the remaining three-fourths do not belong to it at all, and have only the loosest possible connection therewith, as well as with each other. These later additions are of two kinds. Some are of an epic character, and are due to the endeavour to unite here, as in a single focus, all the ancient legends it was possible to muster,—and amongst them, as a matter of fact, are not a few that are tolerably antique even in respect of form. Others are of purely didactic import, and have been inserted with the view of imparting to the military caste, for which the work was mainly intended, all possible instruction as to its duties, and especially as to the reverence due to the priesthood. Even at the portion which is recognisable as the original basis—that relating to the war—many generations must have laboured before the text attained to an approximately settled shape. It is noteworthy that it is precisely in this part that repeated allusion is made to the Yavanas, Śakas, Pahlavas,201a and other peoples; and that

c9.31

∗ \* It is not, however, necessary to suppose, as I did, *l. c.*, that they brought this intelligence from the south of India itself: they might have picked it up at some other part of their voyage.

c9.32

† That they were so appears from their statement as to the Great Bear, *l. c.*

c9.33

‡ In the thirteenth book of the Śatapatha-Bráhmaṇa, Indra also bears the name Dharma, which in the Mahá-Bhárata is especially associated with Yudhishṭhira himself, though only in the forms *dharma-rája*, *dharma-putra*, &c.

c9.34

²⁰¹ And even of this, two-thirds will have to be sifted out as not original, since in the introduction to the work (i. 81) the express intimation is still preserved that it previously consisted of 8800 *ślokas* only.

c9.35

201a In connection with the word *Pahlava*, Th. Nöldeke, in a communication dated 3d November 1875, mentions a point which, if confirmed, will prove of the highest importance for determining the date of composition of the Mahá-Bhárata and of the Rámáyaṇa (see my Essay on it, pp. 22, 25), as well as of Manu (see x. 44). According to this, there exists considerable doubt whether the word *Pahlav*, which is the basis of *Pahlava*, and which Olshausen (v. sup., p. 4, note) regards as having arisen out of the name of the *Parthavas*, Parthians, can have originated earlier than the first century A.D. This weakening of *th* to *h* is not found, in the case of the word *Mithra*, for example, before the commencement of our era (in the MIIPO on the coins of the Indo-Scythians, Lassen, *I. AK.*, ii. 837, and in *Meherdates* in Tacitus). As the name of a people, the word *Pahlav* became early foreign to the Persians, learned reminiscences excepted: in the Pahlaví texts themselves, for instance, it does not occur. The period when it passed over to the Indians, therefore, would have to be fixed for about the 2d–4th century A.D.; and we should have to understand by it, not directly the Persians, who are called Párasikas, rather, but specially the Arsacidan Parthians. and as the old friend of Yudhishṭhira's father; see *I. St.*, v. 152.—In the name of the Yavana prince Kaserumant, we appear to have a reflex of the title of the Roman Cæsars; see *Ind. Skiz.*, pp. 88, 91; cf. L. Feer on the *Kesarí-námasamgrámaḥ* of the *Avadána-Śataka* in the *Séances de l'Acad. des Inscr.* (1871), pp. 47, 56, 60.

c9.36

[^asterisk]: * Which of course stands in glaring contradiction to the statement that the Mahá-Bhárata was recited in his presence.

c9.37

these, moreover, appear as taking an actual part in the conflict—a circumstance which necessarily presupposes that at the time when these passages were written, collisions with the Greeks, &c., had already happened.²⁰² But as to the period when the final redaction of the entire work in its present shape took place, no approach even to a direct conjecture is in the meantime possible;²⁰³ but at any rate, it must have been some centuries after the commencement of our era.* An interesting discovery has recently been made in the island of Bali, near Java, of the Kavi translation of several *parvans* of the Mahá-Bhárata, which in extent appear to vary considerably from their Indian form.²⁰⁴ A special comparison of the two would not be without importance for the criticism of the Mahá-Bhárata. For the rest, in consequence of the utter medley it presents of passages of widely different dates, the work, in general, is only to be used with extreme caution. It has been published at Calcutta,²⁰⁵ together with the *Hari-vaṅśa*, a poem which passes as a supplement to it.*—Respecting the *Jaimini-Bhárata*, which is ascribed, not to Vyása and Vaiśampáyana, but to Jaimini, we have as yet no very precise information: the one book of it with which I am acquainted is wholly different from the corresponding book of the ordinary Mahá-Bhárata.†

c9.38

²⁰² Of especial interest in this connection is the statement in ii. 578, 579, where the Yavana prince Bhagadatta (Apollodotus (?), according to von Gutschmid's conjecture; reg. after B.C. 160) appears as sovereign of Maru (Marwar) and Naraka, as ruling, Varuṇa-like, the west,

c9.39

²⁰³ With regard to the existence, so early as the time of the Mahábháshya, of a poetical version of the Mahá-Bhárata legend, see *I. St.*, xiii. 356 ff. "Still this does not in the smallest degree prove the existence of the work in a form at all resembling the shape in which we now have it; and as the final result, we do not advance materially beyond the passage in Dion Chrysostom (*I. St.*, ii. 161 ff.), relating to the 'Indian Homer.' For the statements of the Greek writer themselves evidently date from an earlier time; and although not necessarily derived, as Lassen supposes, from Megasthenes himself, yet they at any rate take us back to a period pretty nearly coincident with that of the Bháshya."

c9.40

∗ We have a most significant illustration of the gradual growth of the Mahá-Bhárata in an episode commented upon by Śaṃkara, which by the time of Nílakaṇṭha (*i.e.*, in the course of 6 or 7 centuries) had become expanded by a whole chapter of 47 *ślokas*; see my *Catal. of the Sanskrit MSS. in the Berlin Lib.*, p. 108.

c9.41

²⁰⁴ See the observations, following R. Friederich's account, in *I. St.*, ii. 136 ff.

c9.42

²⁰⁵ 1834-39 in four vols.; recently also at Bombay (1863) with the commentary of Nílakaṇṭha. Hippolyte Fauche's incomplete French translation (1863-72, ten vols.) can only pass for a translation in a very qualified sense; see as to this *I. Str.*, ii. 410 ff. Individual portions of the work have been frequently handled: *e.g.*, Pavie has translated nine pieces (Paris, 1844) and Foucaux eleven (Paris, 1862). Bopp, it is well known, early made the finest episodes accessible, beginning with the *Nala* (London, 1819), whereby he at the same time laid the foundation of Sanskrit philology in Europe. For the criticism of the Mahá-Bhárata, the ground was broken and important results achieved by Lassen in his *Indische Alterthumskunde* (vol. i. 1847). For the contents of the work, see Monier Williams's *Indian Epic Poetry* (1863), and *Indian Wisdom* (1875). the 7th century; see *I. Str.*, i. 380. A French translation by A. Langlois appeared in 1834.] † See my *Catal. of the Sanskrit MSS. in the Berl. Lib.*, pp. 111-118: according to Wilson (*Mack. Coll.*, ii. 1), this book would appear to be the only one in existence; see also Weigle in *Z. D. M. G.*, ii. 278. [This book, the *áśvamedhikam parva*, was printed at Bombay in 1863; according to its concluding statements as they appear in this edition, Jaimini's work embraced the entire epos; but up to the present, apart from this 13th book, nothing further is known of it; see as to this my paper in the *Monatsberichte der Berl. Acad.*, 1869, p. 10 ff. A Kanárese translation of this book is assigned to the beginning of the 13th century (*ibid.*, pp. 13, 35); quite recently, however, by Kittel, in his Preface to Nágavarma's *Prosody*, pp. vi. lxxi., it has been relegated to the middle of the 18th (!) century. The peculiar colouring of the Krĭshṇa sect, which pervades the whole book, is noteworthy; Christian legendary matter and other Western influences are unmistakably present; *Monatsb.*, *l. c.*, p. 37 ff. A good part of the contents has been communicated by

c9.43

∗ In Albírúní's time, the 11th century, it passed as a leading authority; see *Journ. Asiat.*, Aug. 1844, p. 130. [Subandhu, author of the Vásavadattá, had it before him, in Talboys Wheeler in his *History of India*, vol. i. (1867), where, too, there is a general sketch of the contents of the Mahá-Bhárata itself; see *I. Str.*, ii. 392.—It remains further to mention the re-cast of the Mahá-Bhárata by the Jaina Amarachandra, which is extant under the title *Bála-Bhárata*,—in 44 *sargas* of 6550 *anushṭubh* verses,—and which appeared in the Benares *Paṇḍit* (1869 ff.), edited by Vechana Rámaśástrin. This work belongs probably to the 11th century, see *Z. D. M. G.*, xxvii. 170.

c9.44

Side by side with the Itihása we find the *Puráṇa* mentioned in the Bráhmaṇas, as the designation of those cosmogonic inquiries which occur there so frequently, and which relate to the '*agra*' or '*beginning*' of things. When in course of time distinct works bearing this name arose, the signification of the term was extended; and these works came to comprehend also the history of the created world, and of the families of its gods and heroes, as well as the doctrine of its various dissolutions and renovations in accordance with the theory of the mundane periods (*yugas*). As a rule, five such topics are given as forming their subject (see Lassen, *I. AK.*, i. 479), whence the epithet *Pañcha-lakshaṇa*, which is cited in Amara's lexicon as a synonym of Puráṇa. These works have perished, and those that have come down to us in their stead under the name of Puráṇas are the productions of a later time, and belong all of them to the last thousand years or so. They are written (cf. Lassen, *l. c.*) in the interests of, and for the purpose of recommending, the Śiva and Vishṇu sects; and not one of them corresponds exactly, a few correspond slightly, and others do not correspond at all, with the description of the ancient Puráṇas preserved to us in the Scholiasts of Amara, and also here and there in the works themselves. "For the old narratives, which are in part abridged, in part omitted altogether, have been substituted theological and philosophical doctrines, ritual and ascetic precepts, and especially legends recommending a particular divinity or certain shrines" (Lassen, *I. AK.*, i. 481). Yet they have unquestionably preserved much of the matter of these older works; and accordingly it is not uncommon to meet with lengthy passages, similarly worded, in several of them at the same time. Generally speaking, as regards the traditions of primitive times, they closely follow the Mahá-Bhárata as their authority; but they likewise advert, though uniformly in a prophetic tone, to the historic lines of kings. Here, however, they come into the most violent conflict, not only with each other, but with chronology in general, so that their historical value in this respect is extremely small. Their number is considerable, amounting to eighteen, and is doubled if we reckon the so-called *Upapuráṇas*, in which the epic character has been thrust still more into the background, while the ritual element has come quite to the front. Up to this time only one single Puráṇa, the Bhágavata-Puráṇa, has been published—the greater part of it at least—edited [and translated] by Burnouf: but of the others we have excellent notices in Wilson's translation of the Vishṇu-Puráṇa.²⁴⁶

c9.45

²⁴⁶ As also in the separate analyses of various Puráṇas, now collected in vol. i. of Wilson's *Essays on Sanskrit Literature* (ed. Rost, 1864). Above all, we have here to mention, further, the minute accounts given of the Puráṇas by Aufrecht in his *Catal. Cod. Sansc. Bibl. Bodl.*, pp. 7-87. The *Vishṇu-Puráṇa* has been recently published at Bombay, with the commentary of Ratnagarbhabhaṭṭa (1867); Wilson's translation of it has been republished, edited by Fitzedward Hall in five vols. (1864-1870), with material additions and corrections. There are now also several editions of the *Bhágavata-Puráṇa*; amongst them, one with the comm. of Śrídharasvámin (Bombay, 1860). The *Márkaṇḍeya-Puráṇa* has been edited in the *Bibl. Indica* by K. M. Banerjea (1855-1862); and the *Agni-Puráṇa* is now appearing in the same series (begun 1870; caps. 1-214 thus far). An impression of the *Kalki-Puráṇa* appeared at Calcutta in 1873; and lithographed editions of the *Liṅga-Puráṇa* (1858) and of portions of the *Padma*, *Skanda*, *Garuḍa*, *Brahmavaivarta*, and other Puráṇas have appeared at Bombay; see *I. Str.*, ii. 245 ff., 301 ff.

c9.46.1

As the second group of Epic Poetry we designated the *Kávyas*, which are ascribed to certain definite poets (*kavis*); whereas the Itihásas and Puráṇas are attributed to a mythical personage, Vyása, who is simply Διασκευή (Redaction) personified.∗ At the head of these poems stands the *Rámáyaṇa* of Válmíki, whose name we found cited among the teachers of the Taittiríya-Prátiśákhya.† In respect of language, this work is closely related to the war-portion of the Mahá-Bhárata, although in individual cases, where the poet displays his full elegance, it bears plainly enough on its surface, in rhyme and metre, the traces of a later date. In regard to contents, on the contrary, the difference between it and this portion of the Mahá-Bhárata is an important one. In the latter human interest everywhere preponderates, and a number of well-defined personages are introduced, to whom the possibility of historical existence cannot be denied, and who were only at a later stage associated with the myths about the gods. But in the Rámáyaṇa we find ourselves from the very outset in the region of allegory; and we only move upon historical ground in so far as the allegory is applied to an historical fact, namely, to the spread of Aryan civilisation towards the south, more especially to Ceylon. The characters are not real historic figures, but merely personifications of certain occurrences and situations.

c9.46.2

Sítá, in the first place, whose abduction by a giant demon, and subsequent recovery by her husband Ráma, constitute the plot of the entire poem, is but the field-furrow, to which we find divine honours paid in the songs of the Ṛik, and still more in the Gṛihya ritual. She accordingly represents Aryan husbandry, which has to be protected by Ráma—whom I regard as originally identical with Balaráma "halabhṛit," "the plough-bearer," though the two were afterwards separated—against the attacks of the predatory aborigines. These latter appear as demons and giants; whereas those natives who were well disposed towards the Aryan civilisation are represented as monkeys,—a comparison which was doubtless not exactly intended to be flattering, and which rests on the striking ugliness of the Indian aborigines as compared with the Aryan race. Now this allegorical form of the Rámáyaṇa certainly indicates, *à priori*, that this poem is later than the war-part of the Mahá-Bhárata; and we might fairly assume, further, that the historical events upon which the two works are respectively based stand to each other in a similar relation. For the colonisation of Southern India could hardly begin until the settlement of Hindustán by the Aryans had been completed, and the feuds that arose there had been fought out. It is not, however, altogether necessary to suppose the latter;

c9.46.3

and the warfare at least which forms the basis of the Mahá-Bhárata might have been waged concurrently with expeditions of other Aryan tribes to the south. Whether it was really the Kośalas, as whose chief Ráma appears in the Rámáyaṇa, who effected the colonisation of the south,∗ as stated in the poem; or whether the poet merely was a Kośala, who claimed this honour for his people and royal house, is a point upon which it is not yet possible to form a judgment. He actually represents Sítá as the daughter of Janaka, king of the Videhas, a tribe contiguous to the Kośalas, and renowned for his piety. The scanty knowledge of South India displayed in the Rámáyaṇa has been urged as proving its antiquity; since in the Mahá-Bhárata this region appears as far more advanced in civilisation, and as enjoying ample direct communication with the rest of India. But in this circumstance I can only see evidence of one of two things: either that the poet did not possess the best geographical knowledge; whereas many generations have worked at the Mahá-Bhárata, and made it their aim to magnify the importance of the conflict by grouping round it as many elements as possible: or else—and this is the point I would particularly emphasise—that the poet rightly apprehended and performed the task he had set himself, and so did not mix up later conditions, although familiar to him, with the earlier state of things.

c9.46.4

The whole plan of the Rámáyaṇa favours the assumption that we have here to do with the work, the poetical creation, of one man. Considering the extent of the work, which now numbers some 24,000 *ślokas*, this is saying a great deal; and before epic poetry could have attained to such a degree of perfection, it must already have passed through many phases of development.† Still, it is by no means implied that the poem was of these dimensions from the first: here, too, many parts are certainly later additions; for example, all those portions in which Ráma is represented as an incarnation of Vishṇu, all the episodes in the first book, the whole of the seventh book, &c. The poem was originally handed down orally, and was not fixed in writing until afterwards, precisely like the Mahá-Bhárata. But here we encounter the further peculiar circumstance—which has not yet been shown to apply, in the same way at all events, to the latter work—namely, that the text has come down to us in several distinct recensions, which, while they agree for the most part as to contents, yet either follow a different arrangement, or else vary throughout, and often materially, in the expression. This is hardly to be explained save on the theory that this fixing of the text in writing took place independently in different localities. We possess a complete edition of the text by G.

First 52 of 487 passages. Read on

Linked texts (86)