Pramāṇa · Search · About · API

On the Aindra School of Sanskrit Grammarians, their Place in the Sanskrit and Subordinate Literatures c6: III

On the Aindra School of Sanskrit Grammarians, their Place in the Sanskrit and Subordinate Literatures c6

Language
English

III

c6.1

It is now necessary to consider if it be possible to trace in any way the innovations Pāṇini made in the treatment of Sanskrit Grammar; it appears to me that this can be done successfully to a very considerable extent, though a difference of opinion must always exist as to a few details.

c6.2

There can be no question that Pāṇini’s Grammar made an epoch in Indian literature; his name occurs everywhere, his treatise soon superseded all others, and has exercised the ingenuity of a countless number of followers. Certainly, for near two thousand years, Pāṇini’s word has been law in India on all questions of grammar. It is evident that for it to have gained such a position of authority in so conservative a country as India, it must have been vastly superior in the eyes of the Brahmans to all the numerous treatises which must have been in existence before Pāṇini’s time: the Tibetan writers¹ (whatever may have been their authority for saying so) cannot be far wrong in making out that Pāṇini’s work was the first complete and systematic treatise on Grammar. It is also tolerably plain that the circumstances under which this book was composed differed greatly from the older stage of Indian history; without some contact with foreign peoples and bitter disputes among religious sects at home, such highly developed enquiry into language as Pāṇini’s treatise displays, is contrary to all experience. Until Greeks began to teach their language to the Romans, grammar made but little progress², and the origin of Arabic and Hebrew Grammar is due to the contact of the Semitic races of Arabia with Persians, Syrians and other foreigners. The latest and most probable conclusions as to the date of Pāṇini point to such a period in Indian events, when Buddhism had become a prominent, if not the most prominent, religion in Northern India, and an intercourse with the Persians and Greeks had begun which soon became of vast importance owing to the long continuance of the Bactrian kingdom. The influence of Buddhism on grammatical studies has been lately pointed out by Mr. Sayce; he says³: “it is very possible that the Sanskrit grammarians were excited to their work by the native dialects, which had been quickened into activity and raised to the level of respectability by the spread of Buddhism.” This remark is, I believe, new, but it is perfectly just. Observations of dialectic variations of pronunciation and of forms of words occur very early in the Sanskrit grammatical literature, and if they are not to be found in Pāṇini’s Grammar it is evident that the system of the work precludes insertion of such matter. There are several notices of provincialisms in the Mahābhāshya which have been collected by Profr. Weber¹), and the Mīmāmsā sūtras attributed to Jaimini also shew that foreign or local words were early noticed even in the Vedas²). Pāṇini’s sūtras mention several foreign kingdoms, though it is not, perhaps, altogether safe to assume that some are not interpolations. Thus circumstances would lead, at the period when Pāṇini lived, to a far more comprehensive and philosophical study of language than had as yet obtained, remarkable in every way though the results already arrived at were; and, as there is much reason to believe that Sanskrit was, at that time, rapidly becoming extinct, this must also have led to increased zeal in observing and classifying facts. The history of grammatical studies in many other countries points irresistably to this conclusion.

c6.3

¹ Tāranātha (by Schiefner) pp. 43 and 54 respectively.

c6.4

² Teuffel, “Geschichte der Römischen Literatur”, § 50.

c6.5

³ “Principles of Comparative Philology”, p. 246.

c6.6

¹ “Indische Studien”, xiii., pp. 365-6.

c6.7

² See Jaimini-s i., 3, 10 and Kumārila Bhaṭṭa’s C. on this passage. The ṭīkās notice some such details (Haug, “Ved. Accent”, p. 64). For the Indian philological views regarding foreign languages see my “Elements of S. Indian Palæography” p. 85. In this respect the Indian grammarians were, perhaps, weakest. See Profr. Max Müller’s A. S. L. (2nd. ed.) pp. 117-8 on the early observation of dialectical and other peculiarities; also his note in Z. d. d. M. G. Vol. vii. on the same subject.

c6.8

Pāṇini’s innovations come (as far as I have been able to trace them) under four heads: (1) the invention of the Çivasūtras and of the use of pratyāhāras for groups of terminations; also the extensive use of gaṇas, if not their invention; (2) the invention of a new system of anubandhas; (3) the invention or more exact use of several technical terms; (4) the use of a technical syntax and a more elaborate system of sūtras.

c6.9

1. The entire absence of the Çivasūtras as part of the technical system of the Aindra treatises has been already repeatedly noticed³), as also the absence of such pratyāhāras as tiṅ, sup, taṅ⁴) which are formed on the same principle, i. e., by giving the first and last letter of a series of letters or terms. I may add here that the Sanskrit grammarians have always regarded the Çivasūtras as the essence or revealed part of his grammar. Thus Nāgoji says: “teshām anāditvād eshām Pāṇinikṛitatvāt"—for those (Çivasūtras) are without origin, these (the rest of the grammar) were composed by Pāṇini. So the original edition of Pāṇini¹ edited under Colebrooke's superintendence begins with the verse:

c6.10

³ pp. 23, 24, 29.

c6.11

⁴ pp. 29-30.

c6.12

¹ Calcutta, Ç. 1731. p. 1.

c6.13

Yenā 'xarasamāmnāyam adhigamya Maheçvarāt |
Kṛtsnam vyākaraṇam proktam tasmai Pāṇinaye namaḥ ||

c6.14

This is from the Pāṇinīya-çixā (Yajurv. recension, edited by Weber, p. 359), and, therefore, is probably of a very respectable age. But I think that Pāṇini's authorship is also proved by his explanation of the use of these sūtras (i., 1, 71)—“ādir antyena sahe 'tā”. If he were not the first to use these sūtras, his explanation, as here given, is superfluous and contrary to his own expressed intention.

c6.15

I have remarked above on the use of gaṇas in the various grammatical treatises which I have compared.

c6.16

2. The invention of a new system of anubandhas. This is stated by Patañjali in a passage which was first pointed out by the late Profr. Goldstücker². This passage is as follows; it occurs in the discussion on the sūtra vii., 1, 18 (“auṅa āpaḥ”)—“athavā pūrvasūtranirdeço 'yam | pūrvasūtreshu ca ye 'nubandhā na tair ihe 'tkāryāṇi kriyante”³. Patañjali here states that there were grammatical sūtras by predecessors of Pāṇini in which anubandhas were used, but that they had no effect in Pāṇini's treatise. This statement is of great value, for it settles a question which would be otherwise left undetermined by the existing Aindra treatises. Anubandhas do not occur in the Tolkāppiyam: indeed, it is impossible to see what use they could be in a grammar of one of the so-called Dravidian languages. Nor is there any place for them in the Prātiçākhyas. It would be useless to consider the Kātantra and Kaccāyana's Grammar which furnish a few, as these do not throw any light on the subject; the latter treatise expressly states that “technical terms (used) by other (grammarians) are used (here)”⁴. It would thus be very hazardous to draw any inferences from a comparison of these with the anubandhas used by Pāṇini, for we have not a standard such as the Tolkāppiyam furnishes in respect of the technical terms, and the few that occur in the Kātantra are again different. The differences between Sanskrit and Pali would naturally lead to a difference of treatment in this respect; nor is there any reason to believe that the old grammarians who preceded Pāṇini were more uniform in respect of the anubandhas they used than in their use of technical terms. The fact remains that Pāṇini invented a new system, and it is obvious that the principle on which he selected the letters he uses as prefixes and suffixes, was the avoidance, if possible, of unnecessary confusion; he has not, however, always succeeded, but his display of ingenuity is very remarkable.

c6.17

² “Pāṇini's Place”, p. 181 (Sep. imp.).

c6.18

³ Benares ed. f. 57.

c6.19

⁴ ed. Senart, p. 10 (sūtra. i., 1, 9). The sūtra appears to include anubandhas as well as technical terms such as are mentioned by the commentary.

c6.20

3. The invention or more exact use of several technical terms.

c6.21

The technical terms used by Pāṇini are of two kinds: *a)* a word used as a type of a class; or *b)* a purely conventional symbol (kṛtrimāḥ saṃjñāḥ as Patañjali calls them).

c6.22

Of the first (*a*) class the following are examples:—kṛt, as a name of the affixes by which primary nouns are formed; nadī, a name for certain feminine nouns in -ī and -ū; strī, as a name of feminine nouns; gotra, as a general name for descendants of a man; saṅkhyā, as a name of numerals. The origin of these is evident; except the two last mentioned, they appear to be the first words of original gaṇas. Such terms we find in the Kātantra¹) and it is probable that they formed part of the older treatises.

c6.23

¹ e. g. Agni (ii., 1, 8); nadī (ii., 1, 9 etc.); çraddhā (ii., 1, 10), etc.

c6.24

The terms of the second class are remarkable; examples in Pāṇini are—gha (= -tara and -tama); ghi (= -i and -u); ghu (= √da, √dhā, etc.); ṭi (=the last vowel in a word and following consonant or consonants); bha used as a name of what Bopp termed the weakest cases; it—as a name of anubandhas.

c6.25

Of these ghu, ṭi and bha are certainly older than Pāṇini, if we are to believe Patañjali, as the late Profr. Goldstücker pointed out. Similar symbols occur in the Kātantra²) and still more in Kaccāyana’s grammar³) though (as might be expected) not identical with Pāṇini’s. These symbols must, therefore, have been used in the older treatises. Here again we have no standard of comparison, for such terms are wanting in the Tolkāppiyam, but it seems likely that there were many such symbols in use, for we find others in the Vājasaneya Prātiçākhya⁴). Pāṇini probably made a selection, and used those more accurately (as Goldstücker suggested); others, again, such as ḷu, luk, etc. are probably his own invention. It would be interesting to discuss all the symbols we find used in the earlier and later grammatical treatises, but this laborious task would only show the growth of method in Sanskrit Grammar, and would throw no light on the point to which I have devoted this monograph.

c6.26

² e. g. Ghuṭ (ii., 1, 3. for the sarvanāmasthāna cases); dhuṭ (ii., 1, 18); nu (ii., 2, 11) etc.

c6.27

³ Ga, a name of the vocative (ii., 1, 6); gha, final ā (ii., 1, 9); jha (ii., 1, 7); pa (ii., 1, 8); la, etc.

c6.28

⁴ e. g. sim, jit, mud, dhi as names of classes of letters.

c6.29

The origin of these symbols is an interesting question, but involved in much obscurity. Some are, no doubt, merely arbitrary symbols, but others are, it is certain, the abbreviations of real words. Thus Dr. Bühler has suggested that ‘it’ is for ‘iti’, a derivation that at once commends itself. It is certain that the symbols used in the Phulla (resp. Pushpa) sūtras (which teach the formation of the Sāma verses as recited) are merely mutilated words; thus, *e. g.* bha is for sto-bha; tara for rathan-tara; hārādi for prati-hārādi¹).

c6.30

¹ As was first pointed out by Profr. Weber, “Ind. Studien” i., p. 46.

c6.31

Thus, in respect of symbolic and other technical terms, the only possible conclusion is, that Pāṇini used some of those invented by his predecessors, made them more precise, and added new terms of his own. He extended this system, originally only applied to details, to the entire grammar; his system banishes all traces of the philosophical analysis of the language, and presents a purely conventional representation in which phonetics and etymology, originally kept separate, are completely fused.

c6.32

4. The use of a more elaborate and technical syntax and a more elaborate system of sūtras.

c6.33

Here it is possible to arrive at a more exact conclusion as to what innovations Pāṇini made²). It has been well remarked by Regnier that the tendency in the Sanskrit style of composition is to exaggerate synthesis³), and this peculiarity is carried to excess in the sūtras of Pāṇini who has not only exhausted the natural resources of the copious system of inflections presented by Sanskrit, but has added new and peculiar meanings to some of the cases—such as the ablative and locative¹), and with him even the arrangement of words is full of meaning²). But there is nothing of this in the works that I have classed together as belonging to the Aindra School, which are all written in a concise but hardly technical style; the sūtras are aphorisms and are sometimes obscure, but they are not algebraical formulas like the sūtras of Pāṇini. Again in the sūtras of Pāṇini we find a syntactical concatenation which is almost wanting in the older sūtras, *e. g.* a sūtra is often of a word or words in the genitive case, but this is required by a nominative in a sūtra, often far before it.

c6.34

² Profr. Max Müller has distinguished the several styles of composition which can be found in the earlier Sanskrit (Vedic) works, in his “Ancient Sanskrit Literature”.

c6.35

³ “Étude sur l’ idiome des Védas” (1855) p. vii. As this book is not common I shall give a few of M. Regnier's excellent remarks: “Cette gradation toujours croissante de la synthèse finit par aboutir à un dernier degré, qui est vraiment, ce semble, la dernière et infranchissable limite de la puissance de combinaison au moyen du discours. Je me hâte de dire, et on le comprend avant que je le dise, que ce dernier excès appartient à une manière de langage tout artificielle. L’ultrasynthétisme peut être naturel et populaire; les idiomes de certains peuples sauvages nous en donnent la preuve; mais, dans ces idiomes, cette fusion, qu’on peut appeler confusion, tient à ce qu’ils ne savent pas décomposer, discerner les éléments de la pensée; tandis que, dans l’excès de combinaison dont je parle, c’est avec des matériaux bien distincts, des mots que la grammaire et le lexique présentent comme autant d’unités separées,

c6.36

¹ "Tasmād ity uttarasya" i., 1, 67; "tasminn iti nirdishṭe pūrvasya" i., 1, 66.

c6.37

² *e. g.* "Sthāne 'ntaratamaḥ" i., 1, 50; "yathāsaṅkhyam anudeçaḥ samānām" i., 3, 10.

c6.38

It is hardly necessary to remark that the sūtra style was in use before Pāṇini and that Patañjali mentions grammatical sūtras anterior to him, but he refined on this style of composition to a great extent. Thus (as grammarians tell us) he omitted the paribhāṣās³) or rules respecting the logical method of his treatise, and which always form a part of Indian scientific treatises. The almost incredible complications of the Pāṇinīya paribhāṣās are now—thanks to Profr. Kielhorn—no longer an insuperable difficulty to students. These sūtras chiefly determine conflicts of rules and similar matters, and present the greatest contrast to the simple method of the older treatises which have only the simple artifice of division into adhikaraṇas or topics.

c6.39

³ Profr. Kielhorn's "Paribhāshenduçekhara" p. i.

c6.40

This very brief survey of Pāṇini’s technical system enables us to add a few new details in regard to the system of his predecessors, which cannot be gathered from the existing treatises that represent now the earlier

c6.41

qu'on parvient à former sciemment et à dessein cette chaîne de ténèbres dont parle, dans un autre sens, le livre de la Sagesse, *catena tenebrarum*, et à enfermer la pensée dans les liens d'une longue nuit, *vinculis longae noctis compedire* (ch. xvii., v. 2 et 17), ou des esprits européens, habitués à la forme de la pensée moderne, ont souvent bien de la peine à porter la lumière et à trouver leur voie. J'ai dit qu'une telle manière d'écrire ne pouvait être qu'artificielle, et elle l'est en effect. C'est dans la prose philosophique et technique que nous la rencontrons. Je ne veux pas parler de l'algèbre des grammairiens, des formules de Pāṇini: il y a dans la construction de ces axiomes une grande puissance d'esprit, et la langue y montre une incroyable aptitude à la concision; mais ce n'est point là du style, c'est comme je le disais, de l'algèbre. La forme de discours et de construction que j'ai en vue ici, je la trouve dans l'exposition et dans la dialectique des grammairiens, des interprètes."

c6.42

stages of Sanskrit Grammar, but which do not contradict the results arrived at by an examination of those works. So far, then, the results at which I have arrived are confirmed from this point of view.

c6.43

It has long been recognized that Pāṇini's date is of the utmost importance in Indian literary history, and it has been discussed with great care¹. The result, as now accepted, is that he lived in the 4th century B. C.; I cannot see that there is any reason why he should not be placed nearly a century later, which would remove some difficulties that the earlier date presents. But, though his name will always be great, we do know but little about him, and there is no probability that we shall ever know more. That he was born at Çālātura in the extreme north-west of India is certain; his mother's name was Dākṣī², or perhaps this only indicates her race³. His life was probably uneventful, though tradition says that he met with a violent death, being killed by a lion⁴.

c6.44

¹ The latest are Profr. Goldstücker's "Pāṇini's Place", and Profr. Weber's strictures on it in Vol. V. of his "Indische Studien", pp. 1-176 (1862).

c6.45

² Patañjali on vii., 1, 18.

c6.46

³ The "Samyamīmāmalikā" by a Çankara (Tanjore MS. No. 10. 468) has (çl. 5-6). . . . . . . Granthakartā ca sūtrakṛt (5) Çālāturiyako Dākṣīputraḥ Pāṇinir Āhikaḥ |

c6.47

⁴ Pañcatantra ed. Kosegarten ii., 34 (=ii., 32 of the Bombay edition, but not in the oldest recension): Simho vyākaraṇasya kartur aharat prāṇān priyān Pāṇineḥ.

c6.48

Whatever merits may be fairly attributed to his predecessors, it cannot be disputed that Pāṇini's work was stupendous in respect of Sanskrit, and he made a good beginning of a grammar of the Vedic language. For two thousand years no one in India has been able to do better. It is thus no matter for surprise that his life should be the subject to which (like that of his still greater countryman Buddha)⁵ the common-places of mythology have attached themselves. In the story of the idle stupid boy who suddenly became the greatest scholar that India ever saw, one can recognize the "Boots" or "Aschenputtel" of our nursery tales⁶—he sits idle and despised at home, till the time comes for his real nature to show itself. That he got supernatural assistance is also in accordance with the same old stories, and in this particular case is entirely in unison with the Buddhist and Vedantist notions; that Çiva gave this aid, seems to have arisen out of the popular name for the first 14 sūtras being misunderstood. The only fact in the legend is that Pāṇini’s grammar superseded those of his predecessors; and these, with reference to the Vedic legend already¹ given, became an ‘Indra Grammar’. The reason why this name was given to treatises by his predecessors is not far to seek. Like all Indian sciences, grammar began out of a study of the texts used for sacrificial purposes, and when by imperceptible degrees it became a rudimentary science, a divine origin was, as a matter of course, attributed to it. Taught orally and writing being almost unknown, the primitive systems could not preserve enough individuality to bear the names of their authors, and were in the same position as the Brāhmaṇas. When writing became more general, Pāṇini’s genius could thus render his name lasting; the names of most of his predecessors could only be recollected in connection with details in which they differed, or by their discoveries.

c6.49

⁵ See Senart's "Essai sur la légende du Buddha" in the "Journal Asiatique" vii. ième série, t. ii., pp. 113 ffg.

c6.50

⁶ Grimm "Kinder- und Hausmährchen", iii., pp. 37-9. Dasent "Tales from the Norse" passim cfr. also the Psalm cli. (in some old rituals): "Pusillus eram inter fratres meos, et adolescentior in domo patris mei" etc. (of David) — "Ps. Mozarabicum" (ed. 1775) lxv., a.

c6.51

¹ p. 6, note 4.

Cites (10)