Pramāṇa · Search · About · API

On the Aindra School of Sanskrit Grammarians, their Place in the Sanskrit and Subordinate Literatures c7.s2: II. Pali

On the Aindra School of Sanskrit Grammarians, their Place in the Sanskrit and Subordinate Literatures c7.s2

Language
English

II. Pali

c7.s2.1

It has long been known that the Pali grammatical literature is very extensive. Hardy in 1848 (in the Ceylon As. Journal) gave a list of 26 such works. Some of these were first described in Westergaard’s Catalogue of the MSS. at Copenhagen, and in 1863, Mr. D’Alwis (in his “Introduction to Kaccāyana’s Grammar”) gave an admittedly imperfect list of 45 books; but several of these merely treat of particular topics, and are not regular grammars. I can only mention a few to which I am able to refer. The oldest and model grammar is that by

c7.s2.2

1. Kaccāyana. To this reference has been repeatedly made in the preceding pages. Parts of it have been edited by Mr. D’Alwis²) and Dr. Kuhn³); an imperfect edition was brought out by the late F. Mason⁴), and an admirable one (with translation) by M. Senart⁵). This last includes the

c7.s2.3

² “Introduction to Kaccāyana” 8°. Colombo, 1863.

c7.s2.4

³ “Kaccāyanappakaranae Specimen” 8°. 1869. “Kaccāyanappakaranae Specimen Alterum”, 8°. 1871.

c7.s2.5

⁴ In the Burmese character, 2 parts 8°. Toungho, 1870.

c7.s2.6

⁵ Originally in the “Journal Asiatique”, and then published separately in 1871.

c7.s2.7

a. Vutti; which may be put down as the first of the many commentaries on the sūtras. It is not quite certain whether Kaccāyana wrote these explanations or not. The tradition of the Pali grammarians is given as follows:

c7.s2.8

“Kaccāyanakato yogo,
Vutti ca Sanghanandino,
Payogo Brahmadattena,
Nyāso Vimalabuddhinā”⁶).

c7.s2.9

⁶ Quoted by Mr. D’Alwis (from the Kaccāyanabhedaṭīkā) in his “Descriptive Catalogue of Sanskrit, Pali and Sinhalese Literary Works”. Colombo, 1870. i., p. 52. Also, previously, in his “Introduction”.

c7.s2.10

“The text was made by Kaccāyana, the interpretation is of Saṅghanandi; the examples by Brahmadatta, and the nyāsa by Vimalabuddhi.” Who the author Kaccāyana was, is as yet quite undetermined; Profr. Weber has discussed¹ the little matter that there is to decide on, and considers that he is not the same as Sāriputta, an opinion in which Mr. D’Alwis now agrees. It is perhaps impossible to solve the question satisfactorily.

c7.s2.11

¹ Z. d. d. M. G. xix., pp. 649-666; reprinted in “Indische Streifen” ii., pp. 316-343. This has been translated into English.

c7.s2.12

Mr. D’Alwis considers that Pāṇini’s Grammar was the source of Kaccāyana’s², but for the reasons I have already given in the former part of this monograph, I cannot accept this view. That some details may be borrowed from Pāṇini by a subsequent redactor of the work, is possible³, but the six sūtras which are identical in both grammars appear to me to be rather from an older source than either Pāṇini or Kaccāyana; they contain nothing that is peculiar to Pāṇini, and generally agree with what has been shown to be the peculiar feature of the Aindra system. There is nothing in the sūtras to throw doubt on their antiquity, except an anubandha or so, which must have been taken from Pāṇini. Thirty-eight sūtras are, however, admittedly interpolations as the earliest commentaries allow⁴ and (as the Ceylon scholiasts admit) there are several sūtras almost identical with those of the Kātantra⁵. To me, at least, it appears probable that both Kaccāyana and the author of the Kātantra have mostly taken these last from a primitive source⁶; but further enquiry (for which I have no space here) is needed to decide the question.

c7.s2.13

² “Introduction”, p. xi. ffg.

c7.s2.14

³ e. g. the use of ṇ to indicate vṛiddhi. As Pāṇini’s anubandhas were new (see above p. 40), this must have been copied.

c7.s2.15

⁴ D’Alwis “Catalogue”, i., pp. 52-3.

c7.s2.16

⁵ Kuhn, “Specimen”, pp. 17-19.

c7.s2.17

⁶ Except ‘vā napacce’: this is Pāṇinīyan so far as the ṇ is concerned.

c7.s2.18

Another objection to the antiquity of Kaccāyana’s Grammar has been raised by Mr. D’Alwis⁷ which I must here notice. It is that Buddhaghosa (4th century) does not use (in his commentary on the three Piṭakas) Kaccāyana’s technical terms, but different ones. It is obvious that no inference can be drawn from this fact, as I have already shown that many of Kaccāyana’s terms occur in Pāṇini’s sūtras; still more occur in Patañjali, and there can, therefore, be no doubt that they belong to a system far older than the 4th century A. D. This is peculiarly the case with the only terms that Mr. D'Alwis mentions as used by Buddhaghosa, viz., the names of cases. Prathamā etc. were all known to Pāṇini, now for these Buddhaghosa uses paccatta, upayoga, etc.; but it is obvious, I think, that these terms are not really names of cases, and that they are solely intended to express the syntactic relation intended by the cases. Thus for the third case he has ‘karaṇa’, and for the fourth ‘sampadāna’ which are used in this way in the grammars¹.

c7.s2.19

⁷ “Catalogue” p. 68 ffg. It is obvious (see the article by Profr. Childers and L. C. Vijasimha in R. As. S. J. new series, V. pp. 289-302) that Buddhaghosa’s Aṭṭhakathā was a compilation from, not a translation, of older works.

c7.s2.20

¹ Kacc. iii.; 6 (sampadāna) etc. See also Profr. Childers' Dy.

c7.s2.21

I have made these remarks here, because (owing to deficient materials) I cannot attempt, even in a summary way, to criticize the present condition of a Pali text², as I propose to do (in Appendix A.) with some of the Sanskrit Grammars.

c7.s2.22

² M. Senart (p. 7) promises to treat of Kaccāyana's relation to the Sanskrit originals, and the later development of the Pali grammatical literature. Kaccāyana's Grammar is almos t certainly not the original Pali Grammar; at least, in its present form.

c7.s2.23

M. Senart promises a complete edition of the Pali grammarians, and I need, therefore, only mention briefly a few of them, which will show that the Pali type of the Aindra Grammar is as important as the Sanskrit.

c7.s2.24

Besides the commentaries (Vutti and Nyāsa) mentioned above, Mr. D'Alwis mentions: a. Dīpanī, apparently not as yet found. b. Kaccāyanabhedaṭīkā³.

c7.s2.25

³ u. s. p. 52.

c7.s2.26

Kaccāyana's Grammar has also given rise to a number of treatises:
1. Mahāsadda-(i. e. çabda)nīti; the author of this knew the Kātantra.
2. Rūpasiddhi. The author of this professes to have consulted Kaccāyana. He lived while Buddhism still existed on the continent of India⁴.
3. Bālavatāra; edited by Mr. Lee. This is according to Mr. D'Alwis an old treatise, though of uncertain date. It is very brief, and there are several commentaries on it: a. Pañcikāpadīpa, written in 1455 A. D. There was, however, an older commentary in Singhalese: b. Gaḍalādeni Saṇṇā⁵. The Balāvatāra was the source whence the earlier European scholars got their notions of Pali Grammar.
4. Payogasiddhi; of the 14th century A. D. This is an exceedingly full and valuable grammar.

c7.s2.27

⁴ Do. pp. 179-180.

c7.s2.28

⁵ Do. p. 80.

c7.s2.29

5. Ākhyātapada. This is a very valuable treatise on verbs, with full explanation of the inflections. It is, thus, not unlike the Sanskrit Mādhaviyadhātuvṛtti of Sāyaṇa.

c7.s2.30

6. Dhātumañjūsā; by Sīlavamsa, of unknown date. The author refers to previous Dhātupāṭhas. The order of the classes differs from the Sanskrit, and may perhaps be that followed by Pāṇini’s predecessors¹.

c7.s2.31

¹ I have transcripts of the whole or part of these three treatises, which I made many years ago from my honoured friend and teacher Dr. Fausböll’s copies.

c7.s2.32

Kaccāyana’s Grammar has also given rise to a modified system—Moggallāna’s Vyākaraṇa. The author who lived from 1153-1186 A. D. was an inhabitant of Anurādhapura in Ceylon. Mr. D’Alwis has described this treatise² at some length.

c7.s2.33

² u. s. pp. 188 ffg.

c7.s2.34

Grammars based on the Pali A. Singhalese Grammar. Drawn up on the plan of the Pali Grammar is that of the Singhalese language, for a knowledge of which we have to thank Mr. D’Alwis who has edited the most important and oldest treatise:

c7.s2.35

Sidat-Saṅgarāva. Colombo, 1853. In a long introduction Mr. D’Alwis has given a valuable history of Singhalese literature. He puts the date of the text between 1266 and 1410 A. D.

c7.s2.36

B. Burmese Grammar. Mgr. Bigandet appears to state³ that the Burmese Grammar is based on the Pali⁴: I have no information accessible beyond this statement, but it comes from a most competent authority. Probably the Siamese Grammar is based on the same source, and the few Javanese and Cambodian grammatical terms that I can find are from the Aindra system.

c7.s2.37

³ Advertisement to Mason’s edition of Kaccāyana—“I look upon it too, as a necessary complement of the Burmese Grammar. No one indeed can master thoroughly the Burmese language unless he possesses a fair knowledge of the Pali.”

c7.s2.38

⁴ Mason (‘Pali Grammar’, in English, p. v.) says: “The book (Kaccāyana’s Grammar) is said to have been brought to Burma, A. D. 387, by Buddhaghosa, and the Burmese translation and commentary are ascribed to him.”

c7.s2.39

The above brief list of the most important treatises, in seven different languages, which are based more or less completely on the Aindra system, will show that if the fortunes of the oldest system of Sanskrit Grammar have been rather Buddhistic (or heretical) than Brahmanical (or orthodox), yet Pāṇini’s great reform did not completely supersede the older system. This is partly to be explained by the circumstances of the time when Pāṇini appeared as an author. At that time the aim of all literary culture seems to have been victory in formal scholastic disputations, and the object of teachers seems to have been to ensure victory to their pupils by almost mechanical means; hence the sūtra style which put in order the mere heads of the successive arguments in the most concise way, and by thus rendering these notes unintelligible except to the initiated, secrecy powerfully aided their object. The peculiar relations existing in ancient India between a teacher and his pupils would effectually prevent treachery, and so a system like Pāṇini’s would long remain a secret in the possession of comparatively few persons. It is thus possible to explain how it happened that it was only after some eight hundred or a thousand years that the older treatises begin to borrow from the Pāṇinīya system¹); had it been possible, this would have occurred long before, for the Çiva-sūtras and algebraic terms of Pāṇini are too attractive to the Indian mind to be neglected, and they offer many advantages in practice.

c7.s2.40

¹ Çakaṭāyana, Vopadeva, etc. It is remarkable that an Aindra treatise (thus Pāṇiniyised) should have appeared at first the best source to Sir W. Jones and others. ("As. Res." i., p. 354.) Forster's Grammar (1810) is based on the Mugdhabodha, and probably Carey's also; the remains of a philosophical arrangement in it attracted attention rather than Pāṇini's technicalities. The earliest grammar by a European (that of P. Paulinus à St. Bartholemeo) is based on a similar source which (in a MS. note in his presentation copy to Anquetil which is now in my possession) he states he got from two Brahmans at Āṅgamala. May not Jinendra, the commentator on the Kāçikā (to Pāṇini's sūtras) and who was a Buddhist, have been the chief means of communicating Pāṇini's system to others than Brahmans?

c7.s2.41

In the history of the civilization that has spread in the course of about two thousand five hundred years from the extreme N. W. of India over the whole peninsula, and even to Indo-China and the Malay archipelago, the Aindra system of grammar fills precisely the same place that Priscian’s grammar has in the history of Latin civilization. Just as the early Catholic Missionaries in the sixteenth century, who went to Mexico, S. America and Africa, and wrote so many meritorious grammars and vocabularies of the strange languages they there discovered, but in every instance and whatever might be the character of the language, found in it the tenses and cases and parts of speech of Priscian, which, as they imagined, constituted for them a science of Universal Grammar, so the Buddhists—orthodox and heretical—who were the first Indian Missionaries, found in languages of the most opposite character, *e. g.* Tibetan and Tamil, the eight cases (resp. seven and vocative) of the Sanskrit and a fourfold division of words. Both the European and Indian Missionaries applied this system mechanically; the only grammar they both knew was to them a science, and became an imaginary universal grammar. In its beginning, Sanskrit Grammar was derived from an analysis of the facts of the language as was the Latin Grammar, but this false notion of a science of general grammar prevented, both in India and Europe, the analysis of strange languages; this true method has only been introduced in this century chiefly owing to a knowledge of Pāṇini’s system. But (if my conclusions are just), the Aindra or older Indian Grammar has had too important a place in eastern literary history to allow of its being neglected with safety in the actual progress of research. For the Aindra (or primitive Sanskrit) Grammar founded the analysis of language; Pāṇini’s systematized the results, and since then there was no real progress till German scholars took up this (to Europe) new line of research. If in any line of research, India is, here, original.

Cites (7)