1. Sanskrit Grammatical Literature: A Survey

The origin of Sanskrit grammar is shrouded in mystery. The firstever mention of it by name is found in the Gopatha Brahmana¹ in which various grammatical terms also occur. Earlier, the urge to analyse speech, which is the basis of all grammatical literature, was alluded to in the Taittiriya Samhita.² One of its oft-quoted passage relates the myth of how the gods went to Indra and requested him to split up speech. Indra obliged them by parting it in the middle, thus splitting it up. Panini: History & Tradition and The earliest extant systematic treatment of grammar is Panini's Astadhyayi, with its 3,995 sutras (formulas), divided into eight adhyayas (chapters) of four padas (quarters) each. There were, however, grammarians before Panini, for the great grammarian himself mentions by name many of his predecessors³ and hints at the existence of many more.4 Thus he testifies to considerable grammatical activity having taken place before him. This fact is also borne out by the various older technical terms he uses, and also the discrepancies and the promiscuous use of some expressions in his work. Panini deals with both Sanskrit and Vedic grammar and mainly Vedic accent, though he deals more fully with Sanskrit grammar. According to tradition, Panini lived in the fifth century B.C., although critics like Keith would like to place him about 350 B.C.

He was a native of Salatura, near Attock, now in Pakistan. Hiuen Tsang records that he saw a statue there to his memory. Panini, as his name would suggest, was the son of Panina, although the Bhavisya Purana' would have us believe that his father's name was Samana. Panini's mother, Daksi, was the sister of Vyadi who composed Sangraha, said to have contained a hundred thousand verses on Sanskrit grammar and more particularly, grammatical philosophy. Tradition regards Panini as the pupil of the sage Varsa who was the brother of the sage Upavarsa. A legend ascribes Panini's death to an encounter with a lion. Panini lived after Yaska, the author of the Nirukta (a work on Vedic etymology), and according to some texts, was a contemporary of Mahapadma Nanda in the fifth century B.C.10 Highly scientific and precise in his treatment, which won him well-deserved praise,11 Panini was greatly concerned with the economy of words. To effect this economy he adopted many devices in which, in the words of Keith, 'the cases are used pregnantly, verbs are omitted, leading rules are understood to govern others which follow; above all algebraic formulae replace real words.' The whole scheme of his work covering the eight adhyayas, as described by Keith, comprised the treatment of 'technical terms and rules of interpretation (i), nouns in composition and case relations (ii), the adding of suffixes to roots (iii) and to nouns (iv, v), accent and changes of sound in word formation (vi, vii) and the word in the sentence (viii). But this scheme is constantly interrupted, rules being interpolated illogically because it was convenient to do, or because space could thus be saved, for the whole book is dominated by the aim to be as brief as possible.'12 Katyayana, Patanjali, and Bhartrhari Many grammarians followed Panini during the next two centuries, but their works are no longer extant; we know of them because their names and quotations from their works are found in Patanjali's Mahabhasya. Some of those names are: Katyayana, Bharadvaja, Sunaga, Vyaghrabhuti and Vaiyaghrapadya. All these grammarians wrote varttikas (aphorisms) on CC-0. Prof. Satya Vrat Shastri Collection, New Seman Panini's work

x f Sanskrit Grammatical Literature: A Survey 15 Among them, Katyayana wrote varttikas on 1,245 of Panini's sutras and these were incorporated and commented upon by Patanjali in his Mahabhasya. Patanjali is believed to be an incarnation of the Serpent Sesa, who is Visnu's resting place. He may be said to belong to the second century B.C., a contention that is supported by the fact that he refers in his Mahabhasya 13 to the Mauryas, to Pusyamitra or the Sunga dynasty, 14 and to a Greek invader, identified as Menander. 15 In addition to his comments upon Katyayana's varttikas, Patanjali deals with some of the sutras in Panini's work not taken up by Katyayana, explaining and justifying them and occasionally rejecting them. Patanjali's Mahabhasya is one of the most important treatises on Sanskrit grammar; it influenced later grammatical works to a very great extent. It is written in a pleasant and lively conversational style, while the proverbial expressions which occur in it and its references to matters of everyday life serve both to enliven the discussion and to provide valuable hints regarding the conditions of life and thought in Patanjali's time. According to a tradition recorded by Bhartrhari 16 and by Kalhana (twelfth century), 17 the study of the Mahabhasya at one time fell upon bad days; it was, however, later revived by scholars such as Candracarya (fifth century A.D.). There are numerous vrttis (commentaries) on this work, and a good number of them are still in manuscript form. One commentary is Pradipa, written by the pre-thirteenth century Kashmirian scholar, Kaiyata; the seventeenth century critic Nagesa wrote a commentary on Pradipa which he called Uddyota. Bhartrhari's commentary was called the Mahabhasya-dipika. Helaraja, however, referred to it as Tripadi, 18 suggesting that it covered only the first three padas of the first adhyaya. Its only available manuscript, now in Berlin, is but a fragment; it goes up to the fifty-third sutra of the first pada of the first adhyaya. The three great grammarians we have so far referred to, Panini, Katyayana and Patanjali, are called collectively the munitraya (the three sages). After them came Bharthari, although

his date is very uncertain. He is usually assigned a date between the sixth and seventh centuries A.D., and according to the Chinese traveller, I-tsing, he died about A.D. 615. Some scholars, however, place him in the fifth century between A.D. 450 and 500,19 while others place him in the third century, or even earlier. 20 Bharthari is the author of two works, the Mahabhasya-dipika, already mentioned, and the Vakyapadiya, a grammaticophilosophical work in three kandas (sections) called the Brahmakanda (dealing with Supreme Logos), the Vakya-kanda (dealing with sentences), and the Pada-kanda (dealing with words), the last being styled the Prakirnaka-kanda (miscellaneous section). Since it consists of these three books, the Vakyapadiya also carries the alternative name of Trikandi (the three-sectioned book). Altogether it has 1,966 karikas (comment in metrical form). Of these, 1,323 are found in the Pada-kanda divided into fourteen samuddesas (chapters). A commentary on the first kanda was written by Bhartrhari himself, while commentaries were written on the second and the third kanda by Punyaraja and Helaraja. An unidentified later commentator, probably of the North, condensed and simplifed Bharthari's own commentary, while Vrsabhadeva, probably hailing from the South, wrote the Paddhati in which Bhartrhari's commentary was explained at length. The Astadhyayi: Commentaries and Arrangements The first extant vrtti on Panini's Astadhyayi is the Kasikavrtti , the Banaras commentary, written jointly by Vamana and Jayaditya, who are usually regarded as having lived in the seventh century. The Kasika-vrtti is presumed to be a Buddhist work on account of the complete absence in it of mangala (benedictory verse) and also because of the liberty with which it handles the text, for it shows as many as fifty-eight variations from the original. The writers are also credited with the authorship of an independent treatise, the Vrtti-sutra. Among other prominent commentaries on the Astadhyayi is the Bhaga-vrtti by Bhartrhari or Vimalamati, although this work is now no longer available.

Sanskrit Grammatical Literature: A Survey 17 There is also the Bhasa-vrtti by Purusottamadeva (sixteenth century A.D.) who drew inspiration from both the Kasika and the Bhaga-vrtti. The Durghata-vrtti, a work on some selected sutras from the Astadhyayi, was written by the Buddhist scholar, Saranadeva; he mentions the date of his work as Saka era 21 1095, which is A.D. 1172. The name of this work is derived from the fact that it seeks to offer justification for durghatas, i.e. points which are normally difficult to justify by grammar. The last of the important commentaries on the Astadhyayi is the Sutraprakasa by the well-known sixteenth century South Indian writer Appaya Diksita. The Kasika-vrtti has two important commentaries. In the seventh century was written Nyasa or Kasika-vivarana-panjika by Jinendrabuddhi, 22 and in the eleventh century was written Haradatta's Pada-manjari.23 In about the eleventh century, the Astadhyayi was given a new form by the Buddhist scholar, Dharmakirti. He rearranged some of the useful sutras topicwise. In about the fourteenth century 24 another Buddhist scholar, Vimalasarasvati, did precisely the same thing in a work called Rupamala. About a century later, the Andhra scholar Ramacandra, in his Prakriya-kaumudi, followed the same pattern but extended the scope of his work by including some of the sutras left out by his predecessors. Two commentaries were written on this work: the Prakriya-prakasa by Sesakrsna (sixteenth century A.D.), the teacher of the famous Bhattoji Diksita; and the Prasada by Vitthala (sixteenth century A.D.). Next came the Siddhanta-kaumudi by Bhattoji Diksita (A.D. 1600-1650), which, in arrangement, closely followed Prakriyakaumudi and the earlier works, but differed from them in that it incorporated all of Panini's sutras. The Sidhanta-Kaumudi is the c most popular extant manual on Sanskrit grammar. Bhattoji Diksita himself wrote a commentary on this work, which he called Praudha-manorama and his grandson Hari Diksita wrote a commentary on it called Sabda-ratna. His principal aim in writing Sabda-ratna was to refute what he regarded as unfair criticism CC-0 Prof. Satya Vrat Shastri of the Praudha-manorama, which occurred in the Praudha-

IS Grammar and Linguistics manoramakucamardini by Panditaraja Jagannatha, a contemporary of his father Bhanuji Diksita . Another commentary on the Siddhanta-kaumudi was written by Nagoji Bhatta (eighteenth century A.D.), a pupil of Hari Diksita. He wrote his commentary in two versions, the longer being called (Brhat) Sabdendu-sekhara, and the shorter one Lagu-sabdendu-sekhara. Nagoji Bhatta wrote another work in three versions, the Vaiyakarana-siddhanta-manjusa, short version the Laghu-Manjusa and the still shorter version the Paramalaghumanjusa. This work was an independent treatise on grammatical philosophy and semantics in the Navya-nyaya style (neo-logical style of argumentation). The author based his discussion mainly on Patanjali's Mahabhasya and on Bhartrhari's Vakyapadiya, but not unoften he differed from them, showing remarkable originality. Nagoji Bhatta also wrote Paribhasendusekhara, a gloss on the paribhasas (grammatical dicta). This was in line with such treatises as the Paribhasa-vrtti by Siraeva. A work by Bhattoji Diksita which deserves special mention is the Sabdakaustubha, although now it is available only in two fragments; one fragment comprises the first portion of the work; from the beginning to the end of the second pada of the third adhyaya; the other fragment contains the fourth adhyaya. The Sabda-kaustubha is an independent commentary on Panini's Astadhyayi and is based primarily on Patanjajli's Mahabhasya, as the author himself expressly states 25 and as Haradatta Misra also states in his Pada-manjari. The Sabda-kaustubha was written & earlier than the Sidhania-kaumudi, for it is referred to there, and it is in this work that the true greatness of Bhattoji Diksita as an original thinker is revealed. He summarized the main conclusions of his work in seventy-four karikas. Kaundabhatta, Bhattoji Diksita's nephew, embodied and expounded these seventy-four karikas in his Vaiyakarana-bhusana; the shorter, Vaiyakatranabhusana-sara; and the still shorter, Laghu-vaiyakarana-bhusanasara. Like the later work, Nagoji Bhatta's Manjusa, Kaundabhatta's work deals with grammatical philosophy and semantics. Kaundabhatta lived in the middle of the seventeenth century. Prof. Satya Vrat Shastri Collection, New Delhi. Digitized by S 3 Foundation USA

Sanskrit Grammatical Literature: A Survey 19 In the first quarter of the eighteenth century, Varadaraja, Bhattoji Diksita's pupil, evolved from Siddhanta-kaumudi two school manuals, Madhya-kaumudi and Laghu-kaumudi. These two manuals, especially the shorter one, Laghu-kaumudi, are very popular in tols (traditional Sanskrit schools) and pathasalas (primary schools) even to this day. Among the works which are ancillary to Panini's sutras is the Unadi-sutra which is presupposed by Panini; it is ascribed to Sakatayana. A later work, Dhatu-patha, which is in substance that of Panini, was the source of inspiration for three works: Dhatupradipa by Maitreyaraksita (A.D. 1165); Daiva by Deva on which Krsnalilasuka Muni wrote a commentary, Purusakara; and Madhaviyadhatu-vrtti by Sayana (thirteenth century A.D.). Ganaratnamahodadhi by Vardhamana (1140) does not seem to be based on the Ganapatha of Panini, for Panini's work itself has not been handed down in its authentic form, since it has additions and alterations made in the light of the Ganapathas of other grammatical treatises. A post-Patanjali writer, Santanava, composed the Phit-sutras, which dealt with the rules of Vedic and classical accent. Post-Paninian Schools The Katantra School: The earliest of the post-Paninian schools is the Katantra (little treatise). It is also known by two other names, Kaumara and Kalapaka according to two traditions associated with its origin. The author of the Katantra, Sarvavarman, is said to have propitiated Lord Siva who in turn ordered Kumara-Karttikeya, his son, to give instruction to Sarvavarman, thus the work came to be called Kaumara. KumaraKarttrikeya is said to have inscribed it in the first instance on the tail (kalapa) of his peacock and thus the work came to be called Kalapaka, or because of the incorporation into it of some parts from a bigger treatise, 26 an obvious reference to brevity, for it is the shortest extant grammar. The Katantra was composed by Sarvavarman in about the first century ADD 2 dfors ar Satavahana X

king, 28 as a tradition recorded in an old text would have us believe.29 A vrtti on it was written by Durgasimha in the eighth century; and on the vrtti a nyasa (an elaborate commentary ) called Sisyahita, was written by Ugrabhuti in the eleventh century. A number of other commentaries on the Katantra were also written. The Katantra also appears in Tibetan translation with a supplement and Durgasimha's commentary. The Katantra had a considerable influence on the Pali grammarian Kaccayana and also on the Dravidian grammarians, and is still popular in Bengal. The Candra Vyakarana School: The Candra Vyakarana by the Buddhist scholar Candragomin (fifth century) is a work longer than Katantra but shorter than the Astadhyayi, its length being three-fourths of the length of the Astadhyayi. It was once popular in the Buddhist countries of Kashmir, Tibet, and Nepal. According X to internal evidence, it was composed in about A.D. 470, the reference in the text being to the victory of a Jarta king over the Hunas. 'Jarta' is taken to be a corruption for 'Gupta', and the king was, most probably, Skandagupta. From external evidence, however, its date of composition seems to be A.D. 600, the date mentioned in the accounts of the Chinese travellers. Candragomin, who is also called Candracarya, mentions in the beginning of his own vrtti on Candra Vyakarana and that the special characteristics of his grammar are brevity, lucidity, and comprehensiveness. He was also versed in Patanjali's Mahabhasya, and was linked with the revival of its study, as Bhartrhari and Kalhana (12 th cen.) testify. The Jainendra Vyakarana School: Although the Jainendra Vyakarana is considered to be the work of Jina Mahavira himself, it was actually composed by Pujyapada Devanandin (sixth century). The colophons in the manuscript itself testify to this. It is a sort of condensation of the works of Panini, Katyayana, and Patanjali; it has a recast, meant for beginners, called Panca-vastu. Two recensions of the Jainendra Vyakarana are extant, the northern and the southern. There is wide divergence between the two texts, for the northern recension has about three thousand sutras while the southern has three thousand seven hundred. There

Sanskrit Grammatical Literature: A Survey 21 are also many variations in expression in the sutras. There are two vrttis on the Jainendra Vyakarana: the Mahavrtti by Abhayanandin and the Laghu-jainendra by Mahacandra. There is also a nyasa, the Sabdambhoja-bhaskara, by Prabhacandra (A.D. 1075-1125). The Sakatayana Vyakarana School: The Sabdanusana by Palyakirti (ninth century) now goes by the name of Sakatayana Vyakarana. It was composed during the reign of the Rastrakuta King Amoghavarsa I (A.D. 814-77). The evidence for this is furnished by the fact that he gave the name Amogha to the extensive commentary he wrote on his own work, and also by the actual mention of the name of the king in one of the illustrations.30 Prabhacandra wrote a Nyasa on the Amogha-vrtti. *Yaksavarman wrote a commentary, Cintamani on the Sakatayana Vyakarana, in which he alluded to its all-comprehensive nature; the sutras, he said, included what in other grammars would be conveyed by istis (grammatical principles) or by upasamkhyanas (additional grammatical rules). pr The Siddha-Hemancandra School: The SiddhaHemacandra, or simply the Haima Vyakarana, was based on the Sakatayana Vyakarana and was written by Hemacandra Suri for King Jayasimha Siddharaja. The king procured from Kashmir eight older works for Hemacandra Suri's use. The Haima Vyakarana is a grammar of Prakrit as well as of Sanskrit; its 3,566 sutras, constituting the first seven adhyayas, deal with Prakrit Sans grammar. The work is a good manual, practical in arrangement and terminology-an aspect in which it agrees mainly with the Katantra. It omits Vedic grammar and accent. Hemacandra Suri wrote his work in two versions called Laghvi, the shorter, and Brhati, the longer. He also wrote a vrtti on his work, and an extensive treatise called the Brhannyasa. The Saravati-kanthabharana School: The most extensive of the grammars is the Sarasvati-kanthabharana written by the Paramara King Bhoja (eleventh century A.D.). The total number of sutras in it is 6,421, which is 2,426 more than even the Astadhyayi has. This is because, included in the very sutras are

the unadis (the siffix un etc.), the paribhasas, and the ganas (groups of words). The first seven adhyayas of the work are devoted to Sanskrit grammar, while the eighth deals with Vedic grammar and accent. Three commentaries have been written on the Sarasvati-kanthabharana: Hrdaya-harini by Dandanatha Narayana Bhatta; Purusakara by Krsnalilasuka Muni; and Ratnadarpana by Ramasimhadeva. Other works Grammars continued to be written in later centuries too, but they could at best find only local acceptance. Of these, the following four are of some importance: The Samksiptasara: Kramadisvara wrote the Samksipta-sara after 1150. In its first seven adhyayas it deals with Sanskrit grammar, and in the eighth with Prakrit grammar. This work is also known as the Jaumara after its redactor Jumaranandin. In the colophons of many of the manuscripts Jumaranandin is styled Maharajadhiraja. Goyicandra Autthasanika, a later writer, appended supplements to the grammar and wrote commentaries on its sutras, unadis, and the paribhasas. The Samksipta-sara, however, was popular only in the western part of Bengal. The Mugdhabodha: Among the works written by Vopadeva (thirteenth century), who flourished under King Mahadeva of Devagiri, was the Mughabodha, a small manual on grammar. Many commentaries have been written on it, the best known among them being the one by Durgadasa Vidyavagisa (seventeenth century A.D.). The Mugdhabodha attained great popularity in Bengal, and it is still in use there. Two other works by Vopadeva are the Kavi-kalpadruma, a work on roots, and a commentary on this called Kamadhenu. The Supadma Vyakarana: Written in 1375 by Padmanabha (fourteenth century A.D.), the Supadma Vyakarana was popular in the eastern part of Bengal. Five commentaries have been written on it, including the Panjika by the author himself and the Supadma-makaranda by Visnu Misra. The Sarasvata Vyakarana: This grammar is traditionally ascribed to Anubhutisvarupacarya, although it might actually have

Sanskrit Grammatical Literature: A Survey 23 been composed by Narendracarya, Anubhutisvarupacarya being merely a prakriyakara. As Vopadeva does not mention this work, it was probably written after him but before the Mohammedan ruler Ghiasuddhin Khilji (A.D. 1469-1500), since one of his ministers, Punyaraja, wrote an extensive commentary on it called Prakriya. This grammar has been widely commented upon; in fact, it has as many as eighteen commentaries and two recasts. The Linganusasanas: Of some grammatical importance are the treatises on gender known as the Linganusasanas. Some of them are ascribed to Panini, Vararuci, Sakatayana, and Hemacandra (twelfth century). There are two about whose date and authorship there is no dispute. These are by Harsadeva (A.D. 606-47) and by Vamana (A.D. 800). 300 REFERENCES 1. I. 24. 2. VI. 4.7 3. Panini's reference to earlier acaryas in IV. 1.17; IV.1.157; VII. etc. 4. Apisali (VI. 1.92), Kasyapa (VIII. 4.67), Gargya (VII. 3.99; VIII. 3.20; VIII. 4.67), Galava (VI. 3.61; VII. 1.74; VIII. 4.67), Cakravarmana (VI. 1.130), Bharadvaja (VII. 2.64), Sakatayana (III. 4.3; VIII. 3.18; VIII. 4.50), Sakalya (I. 1.16; VI. 1.127; VIII. 3.19; VIII. 4.51), Sphotayana (VI. 1.123). 5. Shastri, M. D. The Relation of Panini's Technical Devices to his Predecessors', Proceedings and Transactions of the All India Oriental Conference, 4 th Session, Vol. II, 1928, pp. 465-74. 6. Pawate, I. S. The Structure of the Astadhyayi, Chapters VIII and IX, pp. 93, 109. 7. II. 31.2. 8. Bhartrhari, Mahabhasyadipika, MS., p.30; Punyaraja, Vakyapadiyatika, Benares Sanskrit Series, Benares, p. 383; Nagesa. Mahabhasya Pradipoddyota, Nirnaya Sagar Press, Bombay, 1917, Vol. p. 55. 9. Vide: simho vyakaranasya kartur aharat pranan priyan panineh, Pancatantra, II. 36. 10. Aryamanjusrimulakalpa, p. 427; Kathasaritsagara, I. 4.

. Cf. itipanini. tatpanini. Paninisabdo loke prakasate... and akumaram yasah panineh.-Kasika under the sutras II. 1.6, 13. 12. Keith, A. B. HSL, pp. 423-24.9 13. V. 3.69 r 14. III. 1.26 15. III.2.3 16. Vakyapadiya, II. 4.89. 17. Rajatarangini, I. 176. 18. Being published serially in the Annals of the Bhandarkar Oriental Research Institute, Poona. 19. The Vakyapadiya of Bhartrhari, University of Poona Sanskrit and Prakrit Series, Vol. II. 1965, Introduction, p. xiii. 20. Sadhu Ram, 'The Date of Bhartrari', Journal of the Ganganatha h Jha Research Institute, Allababad, Vol. IX, Part II, 1952. 21. An era founded by a Saka king Salivahana approximately A.D. 78. 22. Shrish Chandra Chakravarti assigns Nyasa to A.D. 725-50. See Introduction to his edition of the Nyasa, Varendra Research Society, Rajashahi, 1913, p. 26. 23. ibid., Yudhisthira Mimamsaka places Haradatta Misra in A.D. 1058, vide Samskrta Vyakarana Sastra ka Itihasa, Second edition, Samvat 2020, p.473. On the authority of the Bhavisya Purana, Jacobi believes that Haradatta died by about A.D. 878, JBBRAS., Vol. XXIII, p. 31. 24. K. P. Trivedi takes Dharmakirti and Vimalasarasvati to be contemporaries. According to him the Rupavatara and the Rupamala were composed about the same time; vide Introduction to his edition of the Prakriyakaumudi, Bombay Sanskrit and Prakrit Series, No. LXXIII, 1925, p. xxxiii. 25. phanibhasitabhasyabdheh sabdakaustubham uddhare, verse 3. 26. Vide Hemacandra: Brhattantrat kala apibati, Yudhisthira Mimamsaka, Samskrta Vyakarana Sastra ka Itihasa, Second edition, Samvat 2020, p. 502. 27. Winternitz, HIL, Vol. III, Part II, p. 439. 28. The king was not well versed in Sanskrit. To instruct him within six months, Savarvarman composed Katantra grammar. 29. Kathasaritsagara, I. 6-7. 30. adahad amoghavarso'ratin.