Pramāṇa · Search · About · API

An Account of the Different Existing Systems of Sanskrit Grammar c6: The School of Panini

An Account of the Different Existing Systems of Sanskrit Grammar c6

Language
English

Contents

The School of Panini

c6.1

10. The School of Pāṇini.—The work which brought to a focus these tentative efforts of the early grammarians² and by its accuracy and thoroughness eclipsed all its predecessors, dominating the thoughts of generations of thinkers even to present times, is the Aṣṭādhyāyī of Pāṇini. It stands—and it will always stand as long as Sanskrit continues to be studied—as a monument at once of encyclopedic research and technical perfection. The work is also interesting in that it is probably the oldest surviving specimen of that type of literary activity which found expression in the aphoristic style.¹

c6.2

² In his sūtras Pāṇini refers to the Northern and the Eastern schools of grammarians and to the following ten individual authors: आपिशलि, काश्यप, गार्ग्य, गालव, चाक्रवर्मण, भारद्वाज, शाकटायन, शाकल्य, सेनक, and स्फोटायन. It would not be far from the truth to assume that in one way or another Pāṇini's work was an improvement upon those of his predecessors. Some of them may have confined their attention merely to the Vedic and some to the post-Vedic Literature, or, treating of both, must have given less attention to current speech and more to the scriptures. The Vedāṅga spoken of by Yāska must be such a treatise and not the Aṣṭādhyāyī.

c6.3

¹ That the sūtra-form was not new in Pāṇini's days is evident from the sūtra v.1.58 : संख्यायाः संज्ञासङ्घसूत्राध्ययनेषु ।

c6.4

11. Pāṇini's date.—The question about the age of this greatest of grammarians is by no means yet settled, or even on the way of being settled. The late Dr. Peterson was inclined to identify him with his namesake, Pāṇini the poet, quoted in Vallabhadeva's Subhāṣitāvali and elsewhere, and to place him ‘at a date much later than that ordinarily accepted,’ that is, about the beginning of the Christian era.² The identification of Pāṇini the grammarian with Pāṇini the poet was also accepted by Pischel, who however assigned to him the date cir. 500 before Christ. The question ‘how far Pāṇini will eventually have to be brought down from the date now accepted for him, or how far it may be, on the contrary, advisable to push into remoter antiquity the lyrical poetry of Northern India’ is finally left undetermined by Dr. Peterson.³

c6.5

² See his Report on the search of Sk. Mss. for 1882-83, pp. 39ff.

c6.6

³ Introduction to the Subhāṣitāvali, p. 58.

c6.7

According to this view it would appear that the two well-known references to the ākhyāyikā called Vāsavadattā occurring in the Mahābhāshya (vol. ii, p. 284) are to be taken as chronologically in touch with the celebrated romance of Subandhu, a writer of the seventh century. This will leave not even a century between Patañjali and Bhartṛihari the author of the Vākyapadīya. How in that case we are to account for the vicissitudes in the text of the Mahābhāshya as recorded in the latter work⁴ and in the Rājatarangiṇī⁵ one is at a loss to say. Since the recent discovery of Bhāsa's Svapna-Vāsavadattam, which probably was based upon an earlier epic or ākhyānaka, we are no longer required to connect Patañjali with Subandhu.

c6.8

⁴ Towards the end of Kāṇḍa ii.

c6.9

⁵ Compare 1.176 ; See also Indian Antiquary, vol. iv. p. 107.

c6.10

Weber and after him Max Müller put Pāṇini down to about 350 B. C., thereby making Pāṇini almost the contemporary of Kātyāyana the author of the vārtikas to Pāṇini’s sūtras;¹ and this opinion obtained for a time, until it was assailed by Drs. Goldstücker and Bhandarkar who have succeeded in proving that Pāṇini cannot have flourished later than B. C. 500. Goldstücker went much farther: he maintained that ‘within the whole range of Sanskrit literature, so far as it is known to us, only the Saṃhitās of the Ṛik, Sāma, and Krishna-Yajus, and among individual authors only the exegete Yāska preceded Pāṇini, and that the whole bulk of the remaining known literature is posterior to him.’² This position in an exaggerated form has been stated at length by Pandit Satyavrata Sāmaśramī, in the introduction to his Nirukta, making Yāska also a successor of Pāṇini. The date he assigns to Pāṇini is cir. 2400 before Christ.

c6.11

¹ Histoy of Ancient Sanskrit Literature, as quoted by Goldstücker in his note 91, p. 80 (Reprint, p. 60) of Pāṇini, His place &c.

c6.12

² Goldstücker, loc. cit., p. 243 (Reprint, p. 187). This view of Goldstücker, however, is not strictly accurate. Pāṇini must have known some form of the Gṛihya and the Dharma sūtras. In his sūtra iv. 4.71 Pāṇini mentions prohibited places or times for study: अध्यायिन्व्वेशकालात्. Patañjali in the Mahābhāshya (vol. ii, p. 386) explains what prohibited places (श्मशान) or times (अमावास्या or चतुर्दशी) are meant. These prohibitions are embodied in works of the Gṛihya or Dharma sūtra type, and Pāṇini must be thinking of some such works existing in his days. I owe this note to Professor Pathak.

c6.13

Conclusions of this kind it was once the fashion to brush aside as carrying the starting point of Vedic chronology much farther than there was any warrant for it. Since, however, recent researches into the antiquity of the Vedas have done much to throw a doubt over the starting point for Ancient Indian Literature accepted by Professor Max Müller and other writers, the best thing, in the absence of any positive evidence, is a suspension of judgment. In another place (pp. 6-7) we have given reasons for agreeing with Goldstücker in accepting the priority of Yāska over Pāṇini. Perhaps 700 to 600 B. C. would be as near an approximation to Pāṇini's time as, in our presnt state of knowledge, or rather want of knowledge, we are likely to get.

c6.14

¹² The view that Pāṇini cannot be placed before B. C. 350 examined.—The fact that Pāṇini in iv. 1.49 (इन्द्रवरुणभवशर्वरुद्रमृडहिमारण्ययवयवनमातुलाचार्याणामानुक्) mentions Yavanas (and the female formation Yavanānī from the stem) has led most western scholars to put down Pāṇini to a date not earlier than B.C.

c6.15

³⁵⁰ The underlying assumptions are: i. that 'Yavanas' can designate none but the Ionian Greeks, and ii. that India did not have her knowledge of 'Yavanas' prior to Alexander's invasion, B. C.

c6.16

³²⁷ Now regarding point i. the late Dr. Rājendralāl Mitra in his 'Indo-Aryans' gave ample evidence to prove that for no period of Indian history could we be quite certain that the word Yavana necessarily designated the Ionian Greeks. But even if we agree to wave this consideration for the present, point ii. is by no means a settled fact. The 'v' sound in the word 'Yavana' represents an original digamma (Ϝ) in Greek; and as the digamma was lost as early as B. C. 800, the Sanskrit word 'Yavana' must be at least as old as the ninth century before Christ. The Ionians appear in history long before B. C. 1,000 and it is not at all improbable that the Indians knew them, as well as their neighbouring races,—such as Assyrians (असुर-अक्षर-असूर्य), Scythians (शक-शकस्थानीय), Medes (मद-मेद-मदग), Persians (पारसीक), Parthians (पल्हव), etc.—perhaps centuries before Alexander's invasion. At any rate if Indian troops are

c6.17

known to have formed part of the army of Darius in the battle of Platææ (B. C. 479), India's knowledge of the Greeks can go back to the middle of the fifth century before Christ. The fact is—and scholars are just beginning to recognise it—that we have been too hasty in condemning the Pauranic accounts of the frontier tribes and races (e. g. those in the Viṣṇupurāṇa or in the Mahābhārata, Bhīṣmaparvan, Chap. xi) as purely imaginative fabrications. We have so far altogether ignored the extensive commerce and interchange of ideas that went on between the Indian Aryans and their brethren beyond the frontiers as far as the Mediterranean—and this long before B. C. 400. So much so that when other independent proofs vouch for the antiquity of an author (in the case of Pāṇini we shall discuss these proofs presently) the burden of proof rests with the person who maintains that some specific reference in that author belongs to a later and not to an earlier time, when, so far as facts go, the reference might just as well be to an earlier period.

c6.18

Nay, more. In this particular case Pāṇini's reference must certainly belong to the earlier period. Compared with Kātyāyana's knowledge about the Yavanas that of Pāṇini is very slight. Pāṇini did not know that the Yavanas had a script of their own (comp. यवनाल्लिप्याम्, Kātyāyana's vārtika 3 to iv. 1.49), or at least in his time there was no current Sanskrit word for that script. Nor was the fact that the Yavanas had a native-place and a kingdom of their own sufficiently known to Sanskrit literature, as is evidenced by Kātyāyana's vārtika कम्बोजादिभ्यो लुग्वचनं चोडाद्यर्थम् i. e. चोडकंडेरकेरल(शकयवना)यर्थम् to iv. 1.175—supposing of course that शक and यवन form a genuine part of the कम्बोजादिगण. Such slight acquaintance with the Yavanas, therefore, as Pāṇini betrays cannot have belonged to a time subsequent to Alexander's invasion.

c6.19

But there is also independent evidence to prove that Pāṇini lived before Alexander's invasion. The internal evidence which compels us to presuppose at least a couple of hundred years between Patañjali and Kātyāyana, and Kātyāyana and Pāṇini—an evidence which even Vincent Smith finds himself compelled to accept (Early Hist. 3rd. ed., p. 451, note 4)—has been indicated in note 1, page 28 below. The most important of external evidence that has been lately brought forward (by Mr. Vishvanāth Kāshināth Rājavāde in the ‘Kesari’ for 30th August 1910) is Pāṇini's mention of the town Sangala (Gr. Sángala, Sk. Saṅkala) in the sūtra सङ्कलादिभ्यश्च (iv. 2. 75). Pāṇini derives the name of the town from the proper name Saṅkala. Saṅkala is a city completed by (Prince?) Saṅkala. This city Alexander razed to the ground as a punishment for the stout resistance of its defenders (Vincent Smith, loc. cit., page 75), and Pāṇini could not have thereafter spoken of it in the manner in which he does. Pāṇini, therefore, must have lived before Alexander's invasion.

c6.20

Another independent evidence is furnished by the sūtra पर्श्वादिधौधेयादिभ्योऽण् (v. 3.117). Here the Parsus or the Persians (and the Asuras or the Assyrians) are mentioned as an आयुधजीविसंघ or an organization of mercenary fighters, similar to the Greeks of the fourth century B.C., or the Germans of the seventeenth century. The Persians were blotted out as a political power in B. C. 329, and the Assyrians in B. C. 538. Pāṇini's references to these people belong, therefore, probably to a time anterior to these dates.

c6.21

Lastly, reverting once more to Kātyāyana's vārtika to iv. 1.175, if the word शक forms a genuine part of the कम्बोजादिगण, it will be necessary to suppose that Pāṇini did not know that the Śakas or Scythians had a country or a kingdom of their own. Now the first King of the Scythians was Deioces (दिबोकस) whose date is cir. 700 B. C., and Pāṇini must have lived before B. C. 700 or at least not long after that date.

c6.22

It is of course conceded that none of these arguments are decisive taken singly. Alternative suppositions could be made to explain away some of these facts. Thus Pāṇini may conceivably mention the city of Sangala even after its destruction by Alexander. The Persians and the Assyrians might have turned into mercenary soldiers after the loss of their independence. And in the case of the कम्बोजाल्लुक् sūtra, since Patañjali in his gloss on Kātyāyana's vārtika does not mention the Śakas or the Yavanas, the two words may not possibly form a genuine part of Kātyāyana's addition, and consequently no cogent argument could be based on that circumstance,—waving the alternative possibility of Pāṇini having at times made mistakes. Finally, it is not altogether impossible that the sūtras on which our arguments for Pāṇini's antiquity are based, were taken over by Pāṇini bodily from some of his predecessors, just as, contrariwise, the sūtras from which his modernity is inferred (especially the word यवन in sūtra iv. 1.149) were later interpolations. But in that way anything is possible and we would be reduced to speechlessness.

c6.23

The upshot of all this is that there is nothing in Pāṇini's Aṣṭādhyāyī that is inconsistent with his having flourished in the seventh century B. C., and this negative conclusion is all that I am content to reach for the present, leaving the burden of proof with those who wish to maintain the contrary.

c6.24

13. Known facts about Pāṇini's life.—As differing from himself Pāṇini mentions (v. 3. 80, vi. 2. 74, etc.) a school of Eastern grammarians, and in later literature he is also known by the name Śālāturīya¹ which is probably derived from his native place. Cunningham has identified Śālātura with the present Lahaur in the Yusufzai valley. In the days of Hiuen Tsang the valley was known as Udyāna and Śālātura was a prosperous town. To-day it is an obscure deserted village in the North-western Frontier Province, near Attock. In his Mahābhāshya¹ Patañjali gives another bit of biographical information about Pāṇini whom he calles दाक्षीपुत्र. Dākshī then was Pāṇini's mother. The Kathāsaritsāgara (taraṅga 4) makes Pāṇini a contemporary of Kātyāyana and Vyāḍi and Indradatta, along with whom he studied at the house of उपाध्याय वर्ष. Not succeeding in his studies Pāṇini practised penance and received from God Śiva the fourteen pratyāhāra sūtras. The story about his death from a tiger² as recorded in Pañchatantra, if based on fact, may or may not refer to our Pāṇini. And this is about all that we know of Pāṇini's personality.

c6.25

¹ शालातुरीयशकटायनचन्द्रगोमी &c, from गणरत्नमहोदधि stanza 2.

c6.26

¹ सर्वे सर्वपदादेशा दाक्षीपुत्रस्य पाणिनेः। Kielhorn's ed. vol. i. p. 75.

c6.27

² सिंहो व्याकरणस्य कर्तुरहरत् प्राणान् प्रियान् पाणिनेः। Tantra ii, stanza 33.

c6.28

14. Character of Pāṇini's work.—Pāṇini's work consists of nearly four thousand sūtras thrown into eight adhyāyas of four pādas each : hence its name Ashṭādhyāyī. The text of the sūtras has come down to us almost intact. A doubt exists as to the genuineness of only five³ of these sūtras, and that is because they are given in the Mahābhāshya as vārtikas to the sūtras just preceding them. When we say that the text has been preserved intact, it is not meant that it is exactly as we find it in any of our current editions. The late Dr. Kielhorn drew attention⁴ to the tendency to regard as sūtra what is given as vārtika, and vice versa, has created some confusion in the exact enumeration of the sūtras. The whole matter needs to be critically studied. Compare Goldstücker page 29 (Reprint, p. 21), note 28.

c6.29

³ Namely, two between iv. 3.131 and 132 and v. 1.36, vi. 1.62, and vi. 1.100,—the last three being given in the Mahābhāshya as vārtikas to the sūtras immediately preceding. The

c6.30

⁴ Indian Antiquary, volume xvi, page 179.

c6.31

fact that the text of the sūtras has not received from the editors all the care that is necessary. All that we mean is that with sufficient pains we can restore from the vārtikas and the Mahābhāshya the exact words as they were used by Pāṇini himself. Changes have been suggested in more than one place by more than one writer, but they were not actually made until after the times of Chandragomin, the Kāśīkākāras, and subsequent writers.

c6.32

Pāṇini has discussed his entire subject in a manner which is very simple in outline, could we but once grasp it, but which has proved very complex in execution. We may conceive of it in some such way as the following.

c6.33

Analysing language—and this is what vyākaraṇa literally means—the first element we reach is a sentence, which again consists of a verb in the various tenses and moods, and a number of substantives in case-relations to each other. [The indeclinables we do not count for the present; they are put in towards the end of 1.4.] Now the forms of verbs that we meet in sentences seem to be made up of an original root-stem and a number of pratyayas or endings, and it is these endings that give the verbs their several modal and temporal significances. These endings, we further notice, group themselves into two sets, and some roots take invariably only one of them, others both, while a number of others change from one to the other under certain circumstances. At the outset then, and to get rid of extra complexity, we dispose of these so-called Atmane-pada and Parasmai-pada prakriyās (i. 3).

c6.34

Turning pari passu to the other element of the sentence, having defined a case-relation (i. 4), we notice that there are often in a sentence sustantives without any case termination at all. We explain these as the members of a whole which we technically call a samāsa or a compound. The formation and the varieties of these must first be explained (ii. 1 and 2), before we actually treat of the kārakas or case-relations (ii. 3).

c6.35

Taking up the verbs where we left them, we next, after a few preliminary definitions and other cognate matters (ii. 4 end), deal at length with the formation and the uses of the various tenses and moods; and, while we are still on the subject, we explain what are usually known as verbal derivatives, that is to say, those elements of sentences which, although by reason of their case-endings they may seem to belong to the category of substantives, do yet bear a very close affinity in meaning and formation to the root stems from which they are derived (iii. 1--4).

c6.36

Now we are free to concentrate ourselves on the noun-element of the sentence. The Nairuktas or Etymologists seem to assert that all these nouns are derived from the root-stems, which were the ultimate factors that we reached in our examination of the verb-element of the sentence. Let us examine this theory.

c6.37

To simplify matters we must, in the first place, dispose of a large number of nouns which are derived from other nouns by the addition of the so-called taddhita affixes (iv.1.76--v.4). Then it is that we reach the substantive divested of all external wrappings. But may not there be some changes in the very body of the nouns which we can explain? It is only when we have done that (vi.4--vii.4) that we are at liberty to style the residual as 'अव्युत्पन्नानि प्रातिपदिकानि,'—unless, of course, we intend to step outside the rôle of a mere grammarian, as distinguished from a philologist, and try to trace even these back to some more primitive verb-stems. Pāṇini has made his contribution to philology in the form of the Uṇādi-sūtras (see below, § 16).

c6.38

This gives us the complete programme of the Aṣṭādhyāyī, and if Pāṇini seems to depart from this in places it is more for convenience of treatment than for anything else. He begins, as was quite appropriate, with a few definitions and canons of interpretation (i. 1 and 2), and he always takes care to introduce such definitions wherever they are required. Some minor topics usually found included in systematic treatises on grammar, such as the Svara-prakaraṇa (vi. 2) or the Strī-pratyayas, Pāṇini has attempted to put into the places where they would most fit in, the only prominent exception to the above rule being the Sandhi-prakaraṇa, which may conceivably have as well been placed elsewhere than where it occurs (vi, 1 and viii. 2-4), and which in any case need not have been cut into two halves separated from one another by the whole matter of nearly two chapters. His system of pratyāhāras and his anxiety to secure a maximum of brevity are perhaps responsible for this lapse in regular logical sequence. But barring these paltry exceptions there is no doubt that Pāṇini has succeeded remarkably well in welding the whole incongruous mass of grammatical matter into a regular and a consistent whole.¹

c6.39

¹ I do not wish to conceal the fact that the above topical scheme for the whole of the Aṣṭādhyāyī will be found wanting, if tried in details. It would seem as if Pāṇini was working alternately upon the two main aspects of his problem: the nouns and the verbs; and the present arrangement of the sūtras in the Aṣṭādhyāyī is the result of attempting to dovetail the two into a coherent whole, involving in the process many an addition and omission and transposition. It may even be that some sections of the sūtras are post-Pāṇinīya interpolations, just as, contrariwise, other sections of the sūtras Pāṇini may have bodily taken over from some earlier

c6.40

15. Technical devices used by Pāṇini.—The difficulty in understanding Pāṇini comes from the very circumstance which Pāṇini himself perhaps considered as his real advance over all his predecessors, namely his attempt to economise expression where conceivably he could do so without being misunderstood. Why Pāṇini should have elected to strain all his nerves to bring about a result which a student of grammar is often likely to regard as the curse of his lot is more than what we can say. His object may have been to give his students aids to memory, or the sūtra-style may have arisen, as suggested by Goldstücker, in the scarcity of the material for writing. In any case we have reasons to assume that the sūtras from the earliest times were accompanied by a traditional explanation of them.

c6.41

Let us for a moment dwell a little longer on this point and note the various means whereby Pāṇini attempted to secure terseness and brevity of expression. The foremost amongst the devices used was of course that of the pratyāhāras or elliptical statements, and of the anubandhas or significant endings. The first was effected by means of the fourteen Śiva-sūtras, which, according to tradition, were revealed to him by God Śiva himself by sounding his tabor. As to the second, although the anubandhas used by Pāṇini are peculiar to himself, the device does not appear to have been his invention. The practice already existed, and Pāṇini only utilised it to its utmost limits.¹

c6.42

¹ Compare Mahābhāṣya on vii, 1. 18 : अथवा पूर्वसूत्रनिर्देशोऽयम् । पूर्वसूत्रेषु येऽनुबन्धा न तैरित्का-र्याणि क्रियन्ते ।

c6.43

The formation of gaṇas, by which are meant lists of words which undergo similar grammatical changes, also tended towards the same result. Some of these gaṇas are complete and some ākṛti-gaṇas, that is to say, gaṇas which do not exhaustively enumerate all the words of a

c6.44

grammars. But for the intrinsic difficulty of the task and for the fact that we have no extant authority earlier than the Mahābhāṣya, which knows the Aṣṭādhyāyī in practically the same form in which we have it now,—here would be a splendid problem in textual criticism.

c6.45

class, but rather give merely a few leading types. Pāṇini in his sūtras gives only the first word of a gaṇa and they have hence been considerably tampered with since his times. So, although we cannot be certain whether any one word now found in the Gaṇapāṭha existed in Pāṇini's day, still the bulk of our present Gaṇapāṭha may safely be considered as coming from the hands of the grammarian himself.

c6.46

The next device to secure brevity was the invention of peculiar technical symbols such as घ, षष्, ङुक्, लु, लुप &c. Some of these may have been known to Pāṇini from his predecessors, while others were probably of his own creation. Patañjali distinctly tells us that ङि, ङु and घ were known to him already.¹

c6.47

¹ Mahābhāṣya on i. 2. 53, and Kaiyyaṭa in the same place.

c6.48

In the framing of the sūtras Pāṇini always scrupulously omitted all such words as may be conveniently supplied from sense or from preceding sūtras. The technical name for this process is anuvṛtti, and to secure it he has made some of his sūtras adhikāra-sūtras,² that is to say, sūtras which have to be repeated, wholly or in part, each time any of the sūtras dominated by it are to be interpreted. Lastly, in portions of the Aṣṭādhyāyī he has so arranged the sūtras that where two sūtras appear equally applicable, that which comes earlier in the order of the Aṣṭādhyāyī must obtain precedence over the one which comes later.³

c6.49

² Pāṇini shows that a particular sūtra is an adhikāra sūtra by i. the word अतः followed by a word in the ablative case occurring in a subsequent sūtra to which the adhikāra is to continue; as in i. 4. 56; 2. स्वरितत्व—e. g. i. 2. 48, where ङि has it; 3. giving a numerical value to some mute letter added to the sūtra; e. g. ङ (=2) is supposed to be added to v. 1. 30 to show the extent of the adhikāra; and 4. व्याप्तिमात्र—तो विशेषप्रतिपत्तिः ।

c6.50

³ Pāṇini viii. 2. 1—पूर्वत्रासिद्धम् ।

c6.51

There is yet one more device serving the same end which remains to be mentioned and of which so much was made in later grammatical speculations: namely, the use of the paribhāṣās or canons of interpretation. Some of them are enunciated by Pāṇini himself, but a larger number he found already current in his day, and so used them tacitly, and the task reserved for later grammarians was to discover what facts in Pāṇini's sūtras imply the use of what particular paribhāṣās.¹

c6.52

¹ For the distinction between the परिभाषासूत्र and the ज्ञापकसूत्र and the whole question of Pāṇini's use of paribhāṣās see Goldstücker, pp. 106-118 (Reprint, pp. 81-90).

16. Treatises accessory to Pāṇini's Aṣṭādhyāyī

c6.s1.1

In addition to the Aṣṭādhyāyī, Pāṇini put together a Dhātupāṭha or list of roots, a Gaṇapāṭha or list of words which behave alike grammatically, and Uṇādi-sūtras in some form or other. Regarding the first, Pāṇini mentions in the sūtras themselves all the ten classes and even some of their sub-divisions just as they occur in the Dhātupāṭha.² The anubandhas of the Dhātupāṭha, further, have the same significance³ as those of the Aṣṭādhyāyī. These facts tend to establish Pāṇini's authorship of the Dhātupāṭha. We have already spoken (p. 23 above) about the Gaṇapāṭha, which also in the main belongs to Pāṇini.

c6.s1.2

² Compare i. 3. 1; ii. 4. 72 and 75; iii. 1. 25, 55, 69, 73, 77, 78, 79, 81; iii. 3. 104; vi. 1. 15; vii. 1. 59; vii. 2. 45; &c.

c6.s1.3

³ Westergaard's Radices Linguae Sanscritae, pp. 342, 343.

c6.s1.4

The question as to the authorship of the Uṇādi-sūtras cannot be so easily settled. They are commonly supposed to be the work of Śakaṭāyana on the basis of statements found in the Nirukta⁴ and the Mahābhāṣya,⁵ according to which Śakaṭāyana agreed with the नैरुक्तसमय in deriving all nouns from roots. Since, however, no work of Śākaṭāyana has come down to us, and since the Śabdānuśāsana which now passes under his name is a comparatively late production (see below, § 52), we cannot say whether this ancient Śākaṭāyana left behind him any work in justification of the views which he doubtless held.

c6.s1.5

⁴ Nirukta i. 4. 1 : नामान्याख्याता-जानीति शाकटायनो नैरुक्तसमयश्च ।

c6.s1.6

⁵ Kielhorn, vol. ii. p. 131 : नाम च धातुजमाह निरुक्ते व्याकरणे शाकटायन च तोकम् ।

c6.s1.7

On the other hand the Uṇādi-sūtras exhibit unmistakable marks of Pāṇini's system. They use sañjñās such as ह्रस्व, दीर्घ, प्लुत, उदात्त, उपधा, लोप, संप्रसारण, and अभ्यास in the same sense in which Pāṇini uses them. The anubandhas of the Uṇādis are also similar to Pāṇini's. This raises a strong presumption that the Uṇādi-sūtras are the work of Pāṇini himself; and it is further corroborated by the fact that Kātyāyana in more than one place takes objection to the technical application of a rule in the Aṣṭādhyāyī urging that it does not hold good in the case of particular Uṇādi-sūtras—an objection which could not have been urged unless Kātyāyana regarded Pāṇini to be the author of the Uṇādis; for, Pāṇini was not to be expected to frame rules that would hold good in other people's works.¹ There is no reason why we should not accept this conclusion.

c6.s1.8

¹ Examples are vii. 3. 50, vii. 4. 13, viii. 2. 78, and viii. 3. 59. In most of these cases Kātyāyana has the remark उणादीनां प्रतिषेधो वक्तव्यः or words to this effect. Patañjali's defence of Panini is throughout grounded on the fact that उणादयो व्युत्पन्नानि प्रातिपदिकानि ।

c6.s1.9

We cannot, however, assign all the Uṇādi-sūtras to Pāṇini's authorship, seeing that in some places their teaching runs counter to the Aṣṭādhyāyī.² The probable view, as suggested by Goldstücker,³ is that the Uṇādi list was first drawn up by Pāṇini, but that it was afterwards modified or corrected by Kātyāyana. The extent of the changes introduced by the author of the Vārtikas must have been so great as to credit him, in popular tradition, with their sole authorship. Thus Vimalasarasvati,¹ a writer not later than the fourteenth century A. D., and Durgasimha² who belongs to the early centuries of the Christian era, both assign the authorship of the Uṇādi-sūtras to Vararuchi alias Kātyāyana. The poet Māgha, however, seems to look upon the Uṇādis as belonging to Pāṇini,³ though his words are not quite explicit.

c6.s1.10

² Thus, Uṇādi-sūtra iv. 226 goes against Pāṇini vi. 2. 139.

c6.s1.11

³ Pāṇini, his place &c., pp. 170 (Reprint, 130) and 181 (Reprint, 139).

c6.s1.12

¹ In the रूपमाला, the India Office Ms. of which is dated 1381 A. D., we find : उणादिसूत्री-करणाय वररुचिना पृथगेव सूत्राणि प्रणीतानि । तद्यथा । कृवापाजि &c.

c6.s1.13

² He begins his com. on the कृत् section of the Kātantra with the verse : वृक्षादिवक्षमी रूढाः कृतिना न कृताः कृतः । कात्यायनेन ते सृष्टा विशुद्धमतिबुद्धये ॥ The krits in this school also include the Uṇādis, as will be seen later.

c6.s1.14

³ Śiśupālavadha xix, 75, and Mallinātha's commentary upon the same.

c6.s1.15

The other works appended to Pāṇini's system probably do not come from him. The Phiṭ-sūtras are, by unanimous testimony, the work of Śāntanavāchārya, a writer much later than Pāṇini.⁴ The Śikshā bears on the face of it the stamp of modernness, notwithstanding the fact that a verse from it has found its way into the Mahābhāshya;⁵ and the same is true of the Liṅgānuśāsana. Regarding the Paribhāshās, in addition to those given by Pāṇini in his Ashṭādhyāyī there may have been others current in Pāṇini's time and tacitly employed by him; but no ancient collection of them has come down to us. The Paribhāshās are usually assigned to the authorship of Vyāḍi who comes between Pāṇini and Patañjali.

c6.s1.16

⁴ Compare नागोजिभट्ट on फिट्सूत्र ii. 21, where he remarks—यद्वा फिट्सूत्राणि पाणिन्यपेक्षया आधुनिककर्तृकाणीति परत्वं बोध्यम् ।

c6.s1.17

⁵ Mahābhāshya, vol. i. p. 2—दुष्टः शब्दः &c. = शिक्षा, stanza 52—मन्त्रा हीनः &c. This stanza, however, forms a genuine part of the Mahābhāshya, seeing that it is commented upon by भर्तृहरि in his महाभाष्यटीका, Kielhorn, vol. ii, preface, p. 13, and is quoted by कुमारिल in the Tantravārtika, Benares ed., p. 233.

c6.s1.18

Between Pāṇini and the next great grammarian, Kātyāyana, came many authors, who attempted, more or less successfully, to emend or justify Pāṇini's rules, and some of the metrical vārtikas found in the Mahābhāshya probably belong to these predecessors of Kātyāyana. We must needs assume this, unless we are ready to suppose that the considerable interval of time¹ that exists between Pāṇini and Kātyāyana was altogether barren of grammatical speculations. Whoever these predecessors were, as our knowledge about their works is next to nothing, we must now pass on to Kātyāyana himself.

c6.s1.19

¹ Goldstücker proves this by showing that 1. grammatical forms current in Pāṇini's time are obsolete in that of Kātyāyana. 2. So also the meanings of words. 3. Words acquire in Kātyāyana's time significances which they had not in Pāṇini's. 4. Literature known to Kātyāyana was unknown to Pāṇini. 5. Writers contemporary with or little separated in time from Pāṇini are looked upon by Kātyāyana as very ancient, e.g. Yājñavalkya; on his last point the Kāśikā remarks: याज्ञवल्क्यादयो हि न चिरकाला इत्याख्यानेषु वार्ता. For fuller particulars see Goldstücker on Pāṇini, pp. 122-157 (Reprint, pp. 94-120).

c6.s1.20

17. Kātyāyana: His date.—The Kathāsaritsāgara makes Kātyāyana the contemporary of Pāṇini, or more accurately, the senior of the two; and had not this tradition been to this extent accepted by so great an authority as Max Müller, we might have explained this on the analogy of a row of columns seen in perspective, where the columns which are farthest from us look nearest to each other, for the simple reason that we cannot discern any marks in the interspaces. We must be prepared however to give up this view and presuppose between Pāṇini and Kātyāyana that much time which the nature of the changes in the forms of language above indicated will reasonably require; and unless we assume that language and customs were in an extraordinarily volatile condition in ancient times, about two to three centuries would not by any means be too great an interval that we can suppose to have elapsed between them. In the present state of our knowledge we cannot therefore, unfortunately, arrive at a greater approximation than 500-350 B. C., nearer to the latter limit if the relation of Kātyāyana with the Nandas mentioned in Kathāsaritsāgara has any basis in fact.

c6.s1.21

18. Nature of Kātyāyana's work.— Kātyāyana's work, the vārtikas, are meant to correct, modify, or supplement the rules of Pāṇini wherever they were or had become partially or totally inapplicable. There are two works¹ of his which aim at this object. The earlier² is the Vājasaneyi Prātiśākhya, a work dealing with the grammar and orthography of the Vājasaneyi-Saṃhitā. Being limited by the nature of his subject to Vedic forms of language only, Kātyāyana has herein given his criticisms on such of the sūtras of Pāṇini as fell within his province. Taking up the suggestion which dawned upon him probably in the course of his Prātiśākhya, Kātyāyana next subjected Pāṇini's Aṣṭādhyāyī to a searching criticism. Since here his object was not to explain Pāṇini but find faults in his grammar, he has left unnoticed many sūtras that to him appeared valid. Of the nearly 4,000 sūtras Kātyāyana noticed over 1,500 in about 4,000 vārtikas. We must add to these the considerable number of cases where Kātyāyana has criticised Pāṇini's rules in his Prātiśākhya. Some of these criticisms he repeats as vārtikas, generally saying there what he had to say in a more correct form.¹

c6.s1.22

¹ Kātyāyana is credited with the authorship of a third work in sūtra style, the Kātyāyana Śrauta-sūtras (published in the Chaukhamba Sanskrit series), but it has nothing to do with grammar. It might have given Kātyāyana practice in writing sūtras, but that is all.

c6.s1.23

² That the Vājasaneyi-Prātiśākhya is posterior to and based upon Pāṇini is clear from the fact i. that many of the sūtras there given are identical with those of Pāṇini. ii. The pratyāhāras and anubandhas are in most cases those of Pāṇini. iii. Where there are changes they are improvements upon Pāṇini, such improvements as Kātyāyana later embodied with occasional changes for the better in his vārtikas. See Goldstücker, Pāṇini, pp. 199 (Reprint, pp. 153) and the following.

c6.s1.24

¹ For Pāṇini's— अदर्शनं लोपः १-१-६० तस्मादित्युत्तरस्य १-१-६७ मुखनासिकावचनोऽनुनासिकः १-१-८ Kātyāyana in the Prātiśākhya has— वर्णस्यादर्शनं लोपः १-१४१ तस्मादित्युत्तरस्यादेः १-३५ मुखनासिकाकरणोऽनुनासिकः १-७५

c6.s1.25

Kātyāyana has not merely stated his doubts and objections in regard to some of Pāṇini's rules, but in most cases has shown how they can be solved or removed.² At the same time he always takes care to prove his propositions, and when suggesting an alternative course, he always tells us that he does so. Notwithstanding this there are, according to Patañjali's showing, a good many cases where his criticisms are misplaced, or are the result of misunderstanding Pāṇini.

c6.s1.26

² Usually by phrases such as उक्तं वा. Compare Indian Antiquary, volume v, Note 2 on the Mahābhāshya, where Kielhorn discusses the whole subject.

c6.s1.27

Some of the vārtikas are written in prose, while others are thrown into a metrical form. In a vast number of cases Kātyāyana has clearly indicated the rules of Pāṇini to which his remarks refer by repeating the sūtras *verbatim*,³ or with slight changes,⁴ or by taking its most important⁵ or introductory⁶ word. Cross references to his own vārtikas he gives by उक्तं शेषे, उक्तं वा, or उक्तं पूर्वेण.⁷

c6.s1.28

³ Vārtika 1 to sūtra ii. 1. 33 ;

c6.s1.29

⁴ Vārtika 1 to sūtra iii. 1. 84 ;

c6.s1.30

⁵ Vārtika 1 to sūtra v. 2. 47 ;

c6.s1.31

⁶ Vārtika 1 to sūtra vi. 4. 14 ;

c6.s1.32

⁷ Vārtika 2 to sūtra iii. 4. 79 ; —to give but one instance of each.

c6.s1.33

Kātyāyana, in that he meant to write a criticism on Pāṇini was compelled to adhere to the latter's terminology. Notwithstanding this fact he has used स्वर for अच्,

c6.s1.34

| For Pāṇini's— | Kātyāyana in the Prātiśākhya has— |
| अदर्शनं लोपः १-१-६० | वर्णस्यादर्शनं लोपः १-१४१ |
| तस्मादित्युत्तरस्य १-१-६७ | तस्मादित्युत्तरस्यादेः १-३५ |
| मुखनासिकावचनोऽनुनासिकः १-१-८ | मुखनासिकाकरणोऽनुनासिकः १-७५ |

c6.s1.35

व्यञ्जन for हल्, समानाक्षर for अच्, अवन्ती and अवतनी for ल् and लुक्. This fact, together with the statement in the Kathāsaritsāgara¹ to the effect that he was a follower of the Aindra school, makes it probable that he belonged to a school, of grammar different from Pāṇini’s. Patañjali distinctly calls him a ‘Southerner’.²

c6.s1.36

¹ Taraṅga iv. and elsewhere: तेन प्रणम्येन्द्रं तदुक्तव्याकरणं भुवि ।

c6.s1.37

² Mahābhāshya, vol. 1, p. 8, line 2: नियतलिङ्गा दाक्षिणात्याः ।

c6.s1.38

19. Vārtikakāras before and after Kātyāyana.—As observed before (p. 28), Kātyāyana had several predecessors from whose works he may have taken many suggestions. In his Prātiśākhya he refers to Śākaṭāyana³ and Śakalya,⁴ names already quoted by Pāṇini; while in the vārtikas he refers by name to Vājapyāyana,⁵ Vyāḍi,⁶ and Paushkarasādi,⁷ and designates a number of others under the general appellation of एके, केचित्, and so forth.⁸ Some of these latter must have been scholars who, like Kātyāyana himself, subjected the wording of the sūtras of Pāṇini to a critical examination. Vyāḍi we know, was the author of an extensive work called Saṅgraha, referred to in the Mahābhāshya⁹ which is in fact based upon it.

c6.s1.39

³ iii. 8: प्रत्ययसवर्णे सुवि शाकटायनः ।

c6.s1.40

⁴ iii. 9: अधिकारं शाकल्यः शपसेषु ।

c6.s1.41

⁵ Vartika 35 to i. 2. 64.

c6.s1.42

⁶ Vartika 45 to i. 2. 64.

c6.s1.43

⁷ Vartika 3 to viii. 4. 48.

c6.s1.44

⁸ Vartika 4 to ii. 1, 1, &c.

c6.s1.45

⁹ Vol. i. p. 6, line 2; The Vākya-padīya describes the Mahābhāshya as सङ्ग्रहप्रतिकञ्चुक.

c6.s1.46

Kātyāyana was followed in his task by a vast number of writers. The names of some of these are preserved for us by Patañjali.¹⁰ To that list we must add the author or authors of the metrical vārtikas (over 250) that are quoted in the Mahābhāshya. Some of these belong to Patañjali himself, others probably to Kātyāyana, while still others, to either the predecessors or successors of Kātyāyana.¹¹ That some of them at least presuppose Kātyāyana is proved by kārikā 1 on Pāṇini iii. 2.118, which quotes one of his vārtikas. Unfortunately none of these successors of Kātyāyana are known to us otherwise than through quotations made by Patañjali in his Mahābhāshya. We must therefore next pass on to Patañjali, with whom ends the first period in the history of the Pāṇinīya school.

c6.s1.47

¹⁰ Namely, भारद्वाजीय, सौनाग, कुणरवाद्य, वाष्व, सौर्यभगवत् and कुणि.

c6.s1.48

¹¹ The question as to the authorship of these श्लोकवार्तिकs is discussed in the Indian Antiquary vol. v, Note 4 on the Mahābhāshya.

c6.s1.49

20. Patañjali: His date and personal history.—The date of Patañjali the author of the Mahābhāshya is not subject to as vague a guess-work as that of Kātyāyana or Pāṇini. At one time scholars were inclined to make him a contemporary of Christ, but Dr. Bhandarkar has fought through the pages of the Indian Antiquary for an earlier date; and it has been now accepted by scholars all round, and formed, in fact, until the recent discovery of the Kautilīya, the one definite landmark in the history of ancient Indian Literature, by a reference to which the dates of Patañjali's predecessors and successors could be approximately determined. The main arguments for assigning him to 150 B. C. are these: i. The instance इह पुष्यमित्रं याजयामः in such a context that the event must have occurred within the lifetime of Patañjali. ii. Similarly the instances अरुणद्यवनः साकेतं and अरुणद्यवनो माध्यमिकाम्, which refer to a siege by Menander. iii. As a collateral evidence, the mention of a financial expedient of the Mauryas.¹

c6.s1.50

¹ The references are: Indian Antiquary i. 299-302; ii. 57, 69, 94, 206-10, 238, and 362; xv. 80-84; xvi. 156, 172; and Goldstücker, pp. 228-38 (Reprint, pp. 175-183).

c6.s1.51

Regarding the personal history of Patañjali very little is known. He was a contemporary of Pushpamitra and probably much honoured by him for his learning. It is usual to suppose that the epithets Gonardīya and Goṇikāputra used in the Mahābhāshya² are his own other names.

c6.s1.52

² Vol. i. pp. 78, 91, 336, &c.

c6.s1.53

derived from his native place and the name of his mother, but it has been shown by Rājendralāl Mitra¹ and Dr. Kielhorn² that they are distinct authors, and as such they are quoted by so early a writer as Vātsyāyana the author of the Kāma-sūtra.³ The best account of Patañjali's time, if not of his person, is to be found in the Mahābhāshya itself; and a detailed exposition of the religious, historical, geographical, social, and literary data as resulting from the contents of that work is to be found in the Indische Studien, xiii. pp. 293-502.

c6.s1.54

¹ Journal of the Asiatic Society of Bengal, vol. lii. p. 269.

c6.s1.55

² Indian Antiquary xiv, p. 40.

c6.s1.56

³ See Kāma-sūtra, p. 67 (Kāvya-mālā edition).—According to गोणिकापुत्र the नायिकाs are divided into four classes, while गोणदीव divides them into eight.

c6.s1.57

We have stated that Patañjali was not the first to deal with Kātyāyana in the same way in which the latter dealt with Pāṇini. Patañjali was perhaps the most successful if not also the last of the number. Besides giving his ishtis (desiderata) on Pāṇini's sūtras, wherever Kātyāyana had omitted to give vārtikas, his chief aim was to vindicate Pāṇini against the often unmerited attacks of Kātyāyana; and in this he has achieved a remarkable success, although in some places he overdoes his defence and becomes decidedly unfair to Kātyāyana. The style of his work is unparalleled in the whole range of Sanskrit Literature, only the Śarīra-bhāshya of Śaṅkara being worthy of a mention by its side.

c6.s1.58

Regarding the text of the Mahābhāshya the traditions recorded in the Rājatarangiṇī⁴ and in the Vākyapadīya⁵ state that it had become so hopelessly corrupt in the time of king Abhimanyu of Kāśmīr that only one authentic Ms. of it existed throughout India, from which all subsequent copies of it have been derived. The work, like Pāṇini's Ashṭādhyāyī, is divided into eight adhyāyas of four pādas each, each pāda being further subdivided into from one to nine āhnikas. The Mahābhāshya does not notice all the sūtras of Pāṇini, but only such as were noticed by Kātyāyana, as also such others as Patañjali himself considered incomplete and capable of improvement. Whether the remaining were likewise commented upon by Patañjali or not is more than what we can say.¹

c6.s1.59

⁴ Vide note 5 on p. 13 above.

c6.s1.60

⁵ Kāṇḍa ii, stanzas 484-90.

c6.s1.61

¹ A fanciful explanation of the fact that some of Pāṇini's sūtras are not to be found in the Mahābhāshya is given in the Patañjala-charita (Kāvya-mālā, No. 51), where it is said that some of the leaves of the originally complete copy of the Mahābhāshya were blown away by the wind and others got disarranged. Another account makes a monkey वदव-कयशातलोल: responsible for the accident.

c6.s1.62

21. Patañjali's Mahābhāshya as marking the end of the first period in the history of the Pāṇinīya school.—Pāṇini, Kātyāyana, and Patañjali are traditionally known as the "three sages," muni-trayam, who gave the law to the science of grammar. Each took for his study the whole field of the living language, and the contribution made by each to the stock of inherited knowledge and ideas is quite considerable. Patañjali's Mahābhāshya for a time marked the highest point in the development of the science of grammar. So far as grammatical speculations go, the next three or four centuries—which coincided with the bloom of the classical Prākrit literature and which also witnessed the Scythian invasions on a large scale—are a perfect blank to us; and our next leap from Patañjali should be to Chandragomin, the founder of the Cāndra school.

c6.s1.63

22. Chandragomin and his work.—Chandragomin² was a close student of Pāṇini, Kātyāyana, and Patañjali, and for his work he utilized all their labours, trying in several places, in the light of the changes that had come over Sanskrit since the days of the author of the Mahābhāshya to improve upon them in the form as well as the matter of their sūtras and vārtikas and ishṭis. Chandragomin was a Bauddha, and one of his objects in writing a new grammar must have been to supply, for the benefit of members of his Church, a grammar that would be free from the traditional Brahmanical element. The more orthodox grammarians, however, were not willing to accept his innovations. They accordingly tried to invent new maxims of interpretation, tending to show, after a very diligent analysis of the works of the three great sages, that such defects as Chandragomin and others tried to find in the Pāṇinīya grammar were in it already implicitly provided for. This procedure was no doubt unhistorical, but so was that of Kātyāyana or of Patañjali. As yet we cannot fix upon any great leading names,¹ but the traditional elaboration of the system of jñāpakas and Paribhāshās must be referred to the time somewhere between 470 (the date of Chandragomin) and 650 (the date of one of the authors of the Kāśikā).

c6.s1.64

² For a more detailed account of him see §§ 42 and following.

c6.s1.65

¹ Unless it be those of वैयाकरण, सौभरि, and हर्यक्ष mentioned in the Vākyapadīya, Kāṇḍa second, stanza 487.

23. The Kāśikā of Jayāditya and Vāmana

c6.s2.1

Itsing, the Chinese pilgrim, speaks of Jayāditya of Kāśmīr as the author of a grammatical work called vṛtti-sūtra, which it is usual to identify with the Kāśikā, a joint work of Jayāditya and Vāmana. Itsing tells us that Jayāditya died about A. D. 660; and if the above identification is correct,² this gives us the date of the Kāśikā.

c6.s2.2

² Itsing's account of the वृत्तिसूत्र by जयदित्य may not after all refer to the काशिका. He speaks of a com. on the वृत्तिसूत्र by Patañjali and writes as if जयदित्य completed the वृत्ति himself. Even so, however, we cannot bring the Kāśikā any earlier than 650 A. D., seeing that on iv. 3. 88 it mentions the Vākyapadīya by name. Jayāditya then appears to be

c6.s2.3

The Kāśikā was once believed to be the work of one author variously called Vāmana, Jayāditya, or Vāmana-Jayāditya. It has now been found out that they are two distinct persons. Bhattoji Dikshita clearly distinguishes between their views,¹ and the concurrent testimony of Mss. from all parts of India assigns to Jayāditya the authorship of the first five chapters of it, while the last three belong to Vāmana, who probably came soon after Jayāditya and certainly before the time of Jinendrabuddhi, who comments upon the whole work.²

c6.s2.4

¹ Compare the प्रौढमनोरमा on Pāṇini v. 1. 42: एतत् सर्वं यथोद्दिष्टमस्येनोक्तम् । वामनस्तवाह &c.

c6.s2.5

² On the question of the different authorship of the Kāśikā see Dr. Bhandarkar's Report for 1883-84, p. 58.

c6.s2.6

Regarding the personality of the authors of the Kāśikā little definite is known. Neither of them begins his work with any maṅgala, both exhibit an unorthodox tendency to introduce changes into the wording of the sūtras, and Jayāditya at any rate refers on i. 1. 36, with evident satisfaction, to the work of the Lokāyatikas.³ These reasons tend to show that the author or authors were Bauddhas. It is supposed that Jayāditya is to be identified with king Jayāpīla of Kāśmīr, whose minister, as mentioned by Kalhaṇa, was a person named Vāmana.⁴ This may not be strictly accurate. Dr. Bühler believed that the author was a native of Kāśmīr.

c6.s2.7

³ See Bāla Śāstrī's edition of the Kāśikā, p. 62—चार्वाक बुद्धिः । तत्सम्बन्धादाचार्योऽपि चार्वाकः । स लोकायते शास्त्रे पदार्थान् नयते । उपपत्तिभिः स्थिरीकृत्य शिष्येभ्यः प्रापयति । युक्तिभिः स्थाप्यमानाः सम्भाविताः पूजिता भवन्ति ।

c6.s2.8

⁴ Dr. Bühler's Report for 1875-76, p. 73.

c6.s2.9

at least a contemporary of Bhartṛhari the author of the Vākyapadīya. Vāmana who probably wrote the last three chapters of the Kāśikā came soon after Jayāditya, and Jinendrabuddhi, the author of the Nyāsa on the Kāśikā came probably before 750, seeing that he is quoted by so early an author as Bhāmaha. Compare also J. B. B. A. S. for 1909, p. 94; Indian Antiquary, xli, pp. 232-237 and xlii, pp. 253-264.

c6.s2.10

The Kāśikā is a running commentary on Pāṇini's Aṣṭādhyāyī, and its merit consists in the lucid manner in which it has explained the sūtras of Pāṇini, clearly indicating all the anuvrittis and giving numerous illustrations for each rule. Sometimes the Kāśikā gives us information which we could not possibly have obtained from any other source. Thus on sūtra vii.3.95 it gives us a rule of Āpiśali,¹ the grammarian who preceded Pāṇini and whose work must consequently have been known to the authors of the Kāśikā. On sūtra vii. 2. 17 it gives us a vārtika of the Saunāgas other than those quoted in the Mahābhāshya. These facts, however scanty by themselves, corroborate the tradition of the existence of a vast number of grammarians prior and subsequent to the time of Kātyāyana.

c6.s2.11

¹ See above, page 9 note 3.

24. The indebtedness of the Kāśikā to Chandragomin

c6.s3.1

The object of the Kāśikā was to embody in the Pāṇinīya system all the improvements that were made by Chandragomin. As the result of an exhaustive analysis of the text of Pāṇini's sūtras as given in the Kāśikā-vritti Dr. Kielhorn² sums up his conclusions thus : "The text of the Aṣṭādhyāyī as given in the Kāśikā differs in the case of 58 rules from the text known to Kātyāyana and Patañjali. Ten of these 58 rules are altogether fresh additions; nine are a result of separating (by yoga-vibhāga) the original 8 sūtras into 17. In 19 cases new words have been inserted into the original sūtras, while in the rest there are other changes in the wording &c. of the sūtras."

c6.s3.2

² See Indian Antiquary vol, xvi, pp. 179 and following.

c6.s3.3

Some of these changes had been already suggested by Kātyāyana or Patañjali, especially in the matter of yoga-vibhāga. The additional words also were mostly taken from the vārtikas or from the notes in the Mahābhāshya, as well as from some of the added rules. Most of the new matter found in the Kāśikā can, however, be traced to Chandragomin, from whose work he diligently draws his material without anywhere acknowledging his sources.¹ This fact, as before pointed out, settles 470 A. D. as the upper limit for the date of the Kāśikā.

c6.s3.4

¹ Thus on iv. 2. 138 Kāśikā gives the vārtika वेणुकादिभ्यश्छण् वक्तव्यः which is Chandra sūtra iii. 2.61; the kārika on v. 4.77 in the Kāśikā embodies sūtras iv. 4. 72 and 73, of Chandra, the Kāśikā further remarking तदेतत् सर्वं योगविभागं कृत्वा साधयन्ति; Panini's sūtra viii. 3. 118, सदेः परस्य लिटि, Chandra changes into सदिश्ञ्च्जेलिटि (vi. 4.98), following herein a vārtika of Kātyāyana (सदो लिटि मतिषेधे स्वञ्च्जेपसङ्ख्यानम्) ; while Kāśikā reads the sūtra itself in conformity with the Chandra vyākaraṇa. Many more similar instances are given by Lieblich in his edition of the Chandra vyākaraṇa.

First 132 of 213 passages. Read on

Cites (46)