An Account of the Different Existing Systems of Sanskrit Grammar c5: Early Grammatical Speculations

An Account of the Different Existing Systems of Sanskrit Grammar c5

Language: English

Early Grammatical Speculations

c5.1

2. Grammatical speculations in the Vedas.—The earliest speculations of a grammatical nature are to be met with in the later portions of the Rigveda itself; for, even if we condemn Patañjali's explanation (Mahābhāshya : Kielhorn, Vol. 1, p. 3) of चत्वारि शृङ्गा by नामाख्यातोपसर्गनिपाताः or his explanation (Ibid. p. 4 ; Rigveda viii. 69. 12) of सप्त सिन्धवः by सप्त विभक्तयः as being too subtle for the Vedic bards,¹ still passages, such as Rigveda x. 125 or Taittirīya Saṃhitā vi, 4. 7. 3, already evince the consciousness that the study of the forms of speech is of sufficient importance to be pursued by itself independently of the dealings between men and men which are rendered possible by them. It is not, however, necessary for our purpose to put together all the Vedic passages that have or can be made to have a grammatical significance. Suffice it to say that the available data do not warrant the supposition that the ‘Seers of the Mantras’ had made any considerable advance in the *science* of grammar. Indeed, it was not their business to do that. To observe the silent or violent workings of Nature and to record in fitting verse the feelings and thoughts awakened by their contemplation was enough to employ all their leisure hours. Philosophy arises only when the harmony of life is disturbed from within (or from without) so that the old child-like faith in the world and its laws becomes no longer possible; and grammar is a species of philosophy.

c5.2

¹ Compare Tantra-vārtika, Benares edition, p. 216.

c5.3

The study of grammar receives a sudden impetus when one form of speech comes into close contact with another and a different form. Thus, for example, the discovery of Sanskrit by modern Europe has created a revolution in the science of philology, just as, in ancient times, the Roman conquest of Greece and, later, the discovery of Greek after the fall of Constantinople led to equally momentous consequences in the development of thought.—The same result is also produced when in course of time there arise inevitable dialectical peculiarities *within* a language. These are either a consequence of the impact of the different races one of which conquers and dominates over the rest,² or they may be due to a change in the climatic conditions—to the people having migrated from one place to another and modified their expressions and articulations in the course of their journey. Something of this sort must have happened when the ancient Sanskrit diverged into the different forms of Prākrit, and we are probably to explain in the same way the considerable difference that is observable in the language of the Brāhmaṇas when contrasted with that of the ancient Saṃhitās.¹

c5.4

² Until the Greeks began to teach their language to the Romans, Greek grammar made little progress.

c5.5

¹ Dr. Burnell in his essay on the Aindra school of Grammarians notes, "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."

c5.6

3. Grammatical speculations in the Brāhmaṇas.—When we come to the Brāhmaṇic speculations on the nature and meaning of the utterances of the ancient sages, we find that they have already lost any living touch with the old form of the language. Old forms and old words as also old ideas had grown obsolete giving place to newer, less poetic and more practical ones.² Since, however, the Sacred Scriptures (the Vedas) were composed in the older form of the language, and since, for various reasons, it was deemed necessary to preserve intact from generation to generation the inherited stock of Vedic poetry, attention came naturally to be focussed upon the peculiarities of that form of the language, and this was the beginning of grammar proper.

c5.7

² Compare the Arctic home in the Vedas, p. 230.

c5.8

The main interest of the Brāhmaṇas, however, was sacerdotal. They busied themselves with the details of the ritual and tried to discover—or invent—a rational, that is to say, a mythological justification for every act of the priest and every element of the sacrifice. If they discussed questions of grammar or phonetics at all, they came in mainly by way of illustration, or because no other equally cogent explanation of the Saṃhitā passage in question was at hand. We cannot make much capital out of their stray and half poetic utterances.

c5.9

4. Grammatical speculations in allied works.—It was in the next period that the study of grammar *as a science* was taken in earnest. This was the period when the scattered hymns of the Vedas came to be collected into family-books and elaborate rules were framed for the regulation of the *parishads* or *charaṇas*.¹ To help students in their task there also came into being about the same time various manuals on phonetics,² which dealt with letters, accents, quantity, pronunciation, and euphonic rules. In course of time the retentive faculty came to be cultivated to an extent which is without any parallel in the history of the world. A further advance was made by the constitution of the Padapāṭha, commonly ascribed to Śākalya, which resolved the euphonic combinations and gave each word, each member of a compound, each prefix of the verb, as also each suffix or termination of the noun separately. The stock of grammatical notions familiar to this stage of development, though not very large, is already sufficient to indicate the earnestness of the search for truth.

c5.10

¹ See Max Müller's History of Ancient Indian literature, 2nd edition pp. 128, 187, &c.

c5.11

² Cp. Taittirīya Āraṇyaka, vii. 1.

c5.12

5. The predecessors of Yāska.—We are not yet certain when the art of writing came to be invented—or introduced—in Ancient India. It was certainly much earlier than what Max Müller once believed it to be.³ Whatever that period might be, it must have been prior to the production of the Prātiśākhya literature; and by this we mean not the Prātiśākhyas in their present form—which are post-Pāṇinīya and pre-suppose much of his terminology—but in some earlier form, and under whatever other names they may have been then known.¹ The contributions which these prototypes of our present Prātiśākhyas made to the science of grammar can now, in the absence of any really representative works of that class, be merely guessed at. If the nature and contents of our existing Prātiśākhya literature can safely be made the basis of any inference, we may suppose that these earlier treatises 1. classified the Vedic texts into the four forms of speech known to Yāska ; 2. framed and carefully defined some of the primitive² sañjñās or technical terms ; and 3. possibly also made some more or less crude attempts to reduce the words to their elements and explain the mode of their grammatical formation. The really creative period of this science is just this. Had there been for this period any works extant, they would have shown us Yāska in the making, as Yāska himself, to some extent, shows us Pāṇini in the making. It is a great pity, therefore, that the period should be all blank to us. Since, however, these tentative sallies of the earlier authors were not probably definite enough to constitute a *system*, and since we have here to treat of systems of Sanskrit grammar, we must next pass on to Yāska³, who, although a philologist and not a grammarian as such, can for our purpose be regarded as forming the link between the primitive Prātiśākhya type of speculation on the one hand, and the later Pāṇinīya mode of thought on the other.

c5.13

³ History of Ancient Indian Literature, p. 520. Compare on the subject Bühler's contribution to the Grundriss der Indo-Arischen Philologie, especially page 18.

c5.14

¹ Goldstücker, Pāṇini : his place in Sanskrit literature, pp. 183 and ff. ; Reprint of the same by Pāṇini office, pp. 141 and ff.

c5.15

² Primitive: those namely that Pāṇini pre-supposes and uses without explaining them. Dr. Burnell would call these the terms of the Aindra School of Grammarians.

c5.16

³ Yāska calls his own work a complement to grammar : व्याकरणस्य कात्स्न्यम् ।

c5.17

6. Yāska's Nirukta: Its date.—In a memorable passage Yāska himself roughly indicates the course of the development of Vedic studies before his time, and, reflecting the achievements made upto his days in the sciences of grammar and philology, contributes his own quota to the same. The passage has been variously interpreted, but the explanation given below may be found perhaps as acceptable as any other.¹ It mentions three distinct periods of intellectual development corresponding roughly to sections 2-5 above. Unfortunately the time of Yāska is by no means yet certain. It depends for the most part, on the date that is to be assigned to Pāṇini, between whom and this great writer at least a century, if not more, must be supposed to have elapsed in order to account properly for all the advances² in the matter

c5.18

| 1 | साक्षात्कृतधर्माण ऋषयो बभूवुः । | These are the original "Seers of Mantras". |
| | तेभ्य इमे साक्षात्कृतधर्माण उपदेशेन मन्त्रान् सम्प्रापुः । | These correspond to the authors of the Brāhmaṇic speculations; possibly also to the compilers of the family-books. |
| | उपदेशाय ग्लायन्तोऽवरे बिल्मग्रहणा येमं ग्रन्थं समाम्नासिषुः । वेदं च वेदाङ्गानि च ॥ | These are the authors of the Pāda-pāṭha, the Nighaṇṭu, and other allied works, including possibly the prototypes of our modern Prātiśākhyas. |

c5.19

c5.20

Similarly Yāska defines (rather derives) सर्वनाम as सर्वाणि नामानि यस्य । सर्वेषु सुतेषु नमति गच्छति वा । सर्वधापि । It is often used by him otherwise than as a technical term of grammar. Compare vi. 6. 8, vii. 1. 2, vii. 1. 5, &c. Again, and wording of the rules of grammar that are to be met with in the Aṣṭādhyāyī. We have dealt with the question of Pāṇini's date in another part of this essay, and if that result be accepted, Yāska must be placed about 800 to 700 before Christ.

c5.21

There are, however, a few facts which seem to militate against the view that Yāska flourished before Pāṇini. The Sūtras of Pāṇini nowhere make any provision for the formation of words like अपार्ण, which occurs in Nirukta (Bib. Ind. edition, Vol. iv. page 258 &c.). Nor did Pāṇini apparently know Yāska's explanation of अरण्यी (Rigveda x. 85.20) by अरण्यस्य पत्नी. Pāṇini must, therefore, have preceded Yāska; else how can we account for such omissions in a grammarian of the calibre of Pāṇini? The utter uselessness of these and similar negative arguments can be seen on a closer examination of the instances adduced. To obviate the last of these defects Kātyāyana¹ gives सूर्यदेवतायां चात्र वक्तव्यः as a vārtika to sūtra iv.1. 48. Kātyāyna must, therefore, have come after Yāska whose work he here presumably utilises. On the contrary, the first omission is not rectified even by Kātyāyana who gives two vārtikas (no. 7 and 8 to vi. 1.89) to explain forms like प्रावर्ण and प्रावर्ण but not अपार्ण. This would necessitate the supposition that Yāska came after Kātyāyana. A mode of argumentation which leads to such contradictory conclusions is no safe foundation for

c5.22

¹ In Kielhorn's edition vol. ii. p. 220, this is given not as a vārtika of Kātyāyana but as a part of the Mahābhāshya. In that case Yāska's explanation of अरण्यानी as अरण्यस्य पत्नी and his non-acquaintance with vārtika 1 to Sūtra iv. 1. 49 may be adduced to prove the point at issue.

c5.23

there is a great distance between Yāska's definition of निपात as उच्चावचेष्वर्थेषु निपतन्ति and his giving the meanings for each individually, and Pāṇini's classification of them into उपसर्ग when joined to verbs, गति if the root develops into a noun, and कर्मप्रवचनीय. Many more similar illustrations could be found.

c5.24

any chronological edifice, especially when the evidence for Yāska's priority to Pāṇini is so overwhelming.

c5.25

7. Nature of Yāska's work.—In form Yāska's work is a running commentary upon a list of words in five adhyāyas, known as the Nighaṇṭu. The words are all taken from the Veda; the first three adhyāyas arrange them as synonyms, the fourth is a collection of certain difficult words occurring in the Veda, while the last is a list of the names of Vedic deities. Yāska takes these words one by one (in the case of the first three adhyāyas only the more important ones), quotes Vedic passages wherein they are used, and tries to connect them with radical stems and launches into various interesting social and historical discussions in his attempts to trace the later history of these words, always giving references to any conflicting views that may have been held on the subject. Certain general reflections as to the nature and utility of the study of the Vedas, the cosmological functions of the Vedic Gods, and so forth also find their proper place in the work.

c5.26

That grammatical speculations had sufficiently advanced in the days of Yāska is evidenced even by the list of schools and individual teachers quoted or referred to in the Nirukta,¹ none of whose works have been preserved to us. Yāska already knew, what it required an Aristotle to discover subsequently, viz : the fourfold classification of words, as also the distinction between personal terminations and tense affixes on the one hand, and the primary and secondary nominal affixes on the other. Nay, he definitely formulates the theory that every noun is derived from a verbal root and meets the various objections raised against it,—a theory on which the whole system of Pāṇini is based, and which is, in fact, the postulate of modern Philology.¹

c5.27

¹ These are: आग्रयणः, आग्रायणः, आचार्याः, एके, ऐतिहासिकः, औहेवरायणः, औपमन्यवः, और्णवाभः, कात्थक्यः, कौत्सुकिः, गार्ग्यः, गालवः, धर्मशिराः, तैठिकिः, नैदानाः, नैरुक्ताः, परिव्राजकाः, पार्षदानि, मनुः, याज्ञिकाः, पूर्वे याज्ञिकाः, वार्ष्यायणिः, वैयाकरणाः, शाकटायनः, शाकपूणिः, शाकल्यः, स्थौलाष्ठीविः, हारिद्रवकम् ॥

c5.28

¹ Compare Max Müller's History of Ancient Sk. Literature, pp. 161-168.

c5.29

8. Yāska's successors.—Many valuable works on grammar subsequent to Yāska's Nirukta but anterior to Pāṇini's Aṣṭādhyāyī have been irrevocably lost to us; for, it cannot be maintained with cogency that the extremely artificial and algebraic style of the Aṣṭādhyāyī could have been completely evolved by Pāṇini himself in the absence of similar tentative works preceding his. We have got for this the evidence of Pāṇini's own sūtras, which use many technical words and formulas without having previously explained them²—an omission which, as indicated by Pāṇini at i.2.53-57, is to be accounted for on the supposition that they were too well-known or already sufficiently dealt with in other works to need any exposition at his hands.

c5.30

² Such as मत्यय, प्रथमा, द्वितीया, तृतीया, चतुर्थी, पञ्चमी, षष्ठी, सप्तमी, समास, तत्पुरुष, अव्ययीभाव, बहुव्रीहि, कृत, तद्धित, &c., occurring respectively in i. 1. 69, ii. 3. 46, ii. 3. 2, ii. 3. 18, ii. 3. 13, ii. 3. 28, ii. 3. 50, ii. 3. 36, ii. 1. 3, ii. 1. 22, ii. 1. 5, ii. 2. 23, iii. 1. 93, iv. 1, 76, and elsewhere. These could not all have been taken from the Prātiśākhya works anterior to Yāska, since some of them appear to be unknown to that author and must have come into vogue since his day. Compare also Pāṇini i. 3. 120, आङो नास्रियांम् । where Bhaṭṭoji says, आङिति वासज्ञा पाश्चाम् ।

c5.31

Some of these works must certainly have been in existence long after the time of the Mahābhāshya, since we find many quotations from them in later writers. The chief founders of grammatical schools prior to Pāṇini are, Āpiśali and Kāśakṛtsna (compare Pāṇini vi. i. 92). A rule of Āpiśali³ is given by the Kāśikā on vii. 3. 95, while elsewhere it gives us the information that the grammar of Kāśakṛtsna consisted of sūtras thrown into three Adhyāyas.¹ Kaiyyaṭa on v. 1. 21 actually gives portions of the text of both these grammarians²—and this is about all the information that we possess regarding these two ancient grammarians. To later writers like Bopadeva³ they are probably little more than mere names.

c5.32

³ आपिशलास्तुल्यम्यं सार्वधातुकासु छन्दसीति पठति

c5.33

¹ Compare the Kāśikā on v. 1. 58, and iv. 2. 65 : त्रिकं काशकृत्स्नम् । त्रिकाः काशकृत्स्नाः । Another bit of information about आपिशलि, which I owe to Professor Pathak, is that he changed the root अस् 'to be' to स. Compare अस्ति सकारमतिष्ठते, in the Mahābhāshya on i. 3. 22. Jinendrabuddhi and Śakaṭāyana (i. 4. 38) supply आपिशलिः as the subject of आतिष्ठते ।

c5.34

² आपिशलकाशकृत्स्नप्रोक्तस्वग्रन्थ इति वचनादन्यत्र प्रतिषेधाभावः ।

c5.35

³ Compare, इन्द्रश्चन्द्रः काशकृत्स्नापिशली शाकटायनः । पाणिनीयमरजैनेन्द्रा जयन्त्यष्टादशादिकाः ॥ from Bopadeva's Mugdhabodha.

c5.36

9. The so-called Aindra treatises.—The case stands a little different with Indra or Indragomin. Pāṇini nowhere mentions this name except under the general appelation of 'the easterners'. An oft-quoted passage from the fourth taraṅga of the Kathāsaritsāgara informs us that the school which Pāṇini supplanted was known as the Aindra school, and numbered among its adherents Kātyāyana alias Vararuchi, Vyāḍi, and Indradatta. Hiuen Tsang the Chinese pilgrim, and Tārānātha the Tibetan historian, both relate a similar story, the latter adding that the Cāndra vyākaraṇa agrees with Pāṇini, and the Kālāpa vyākaraṇa with the Aindra. Tārānātha also states that God Kārttikeya revealed the Aindra vyākaraṇa to Sapta-(not Sarva-)varman (compare section 64, below). Further corroborative evidence is furnished by a passage⁴ from the Taittirīya-saṃhitā (vii. 4. 7), which speaks of Indra as the first of grammarians. To all this Dr. Burnell further adds that the Tolkappiyam, one of the oldest Tamil grammars, represents itself to be full of the Aindra system, and was read in the Pāṇḍya King's assembly and there met with approval. This Tolkappiyam is closely related to Kātantra, to Kachchāyana's Pāli grammar, and to the Prātiśākhyas, all of which are to be regarded as treatises belonging to the Aindra school of grammarians. The conclusion¹ which Dr. Burnell reaches is that the 'Aindra was the oldest school of Sanskrit grammar, and that Aindra treatises were actually known to and quoted by Pāṇini and others, and that Aindra treatises still exist in the Prātiśākhyas, in the Kātantra, and in similar works, though they have been partly recast or corrected.' And again, 'the Aindra treatises belong to a system older than Pāṇini's, though there is perhaps reason to believe that not one of them is, as a whole, older than the grammar of the last.'

c5.37

⁴ वाग् वै पराच्यव्याकृताऽवदत् । ते देवा इन्द्रमब्रुवन्निमां नो वाचं व्याकुर्विति । ........ । तामिन्द्रो मध्यतोऽवक्रम्य व्याकरोत् ।

c5.38

¹ Compare his Essay on the Aindra school of grammarians, *passim*.

c5.39

That the technical terms used by the so-called Aindra treatises are connected with one another and are, further, simpler and more primitive than those of Pāṇini is quite evident; and on this ground it is not unlikely that they represent a school of grammarians prior to Pāṇini's. But since, besides the Aindra, we have at least two other schools also older than Pāṇini, it will not do to put down every one of these sañjñās as belonging to the Aindra school, seeing that we have no information regarding the sañjñās of the other two. In the present state of our knowledge, the fact that the Aindra school is nowhere quoted by name either in Pāṇini or Mahābhāṣya or Kāśikā should point to the conclusion—also endorsed by Keilhorn—that the Aindra school is post-Pāṇinīya in date, though pre-Pāṇinīya in substance. Possibly it may be no other than the Kātantra school which belongs to the early centuries of the Christian era.

c5.40

Any further details regarding the grammatical efforts earlier than Pāṇini it is not possible to give. All that we can do is, following Yāska and on the basis of references occurring in Pāṇini, Kātyāyana, Patañjali, and the earlier Prātiśākhyas and Brāhmaṇas, to frame a tabular statement of the schools and teachers with the tenets peculiar to each. A beginning towards one is made in Dr. Burnell's essay quoted before, where only the names of the teachers—some of them later than Pāṇini—are given.¹

c5.41

¹ A few instances are also collected in Indische Studien, iv. p. 76. Compare also History of Ancient Sanskrit Literature, p. 160.

Cites (5)

Aṣṭādhyāyī — Pāṇini, Sanskrit
Kṛṣṇayajurvedasaṃhitā — Sanskrit
Ṛgvedasaṃhitā — Sanskrit
Nirukta — Yāska, Sanskrit
Taittirīyāraṇyaka — Sanskrit