Authorship Of Vakyapadiya Vrtti

Author(s): Ashok Aklujkar
Publisher: Ashok Aklujkar
Catalog link: https://jainqq.org/explore/269221/1

Page #1
THE AUTHORSHIP OF THE VAKYAPADIYA-VRTTI* By Ashok Aklujkar, Vancouver 1.1 It has been a long tradition in India to ascribe the Vstti (V in abbreviation) of the first two kandas of the Trikandi to Bharthari and to accept it as an integral part of the Vakyapadiya?. This tradition is * I am grateful to Dr. MADELEINE BIARDEAU for promptly making available to me a copy of her edition of the Brahma-kanda, and to Dr. GEORGE L. HART, the authorities of the Adyar Library, Prof. K. V. ABHYANKAR, ACARYA LIMAYE, PANDITA V. B. BHAGWAT, Dr. S. D. JOSHI, and Prof. WILHELM RAU for making the Vakya-kanda-vstti available to me in microfilms and transcripts. The kernel of the present article constituted a paper read at the one hundred and seventy-ninth annual meeting of the American Oriental Society in New York (March 1969). 1 (a) Vakyapadiya was originally the title of only the first two kandas of Bharthari's magnum opus; the entire work consisting of three kandas was called Trikandi in the older tradition (AKLUJKAR 1969: 547-555). (b) The V of the first two kandas only is available to us and it alone seems to have been accepted in the tradition as a genuine, inseparable constituent of the Vakyapadiya. See "The Extent of Bhartshari's Vstti" (AKLUJKAR, forthcoming). (c) It has been convincingly shown by C. SHASTRI (1930: 636-644, 1934: Skt. Intro,: -1826; cf. R. KAVI 1930: 235--241; KUNHAN RAJA 1936: 285--298) that the commentary on the first kanda published under the name of Punyaraja in the Benares Sanskrit Series by MANAVALLI (1887) is in fact an abridgement of Bhartphari's V of the same kanda. The unabridged V was edited for the first time by C. SHASTRI (1934), and it is his text that has essentially been followed in the editions by R. SARMA (1963), BIARDEAU (1964a), BHAGWAT (1965), and S. IYER (1966a). BIARDEAU and S. IYER (1964a, 1965) have translated the V into French and English respectively. But, as fns. 16, 19, 22--24, and 29 below indicate and as my forthcoming studies of the V and of BIARDEAU's interpretation will substantiate, much improvement is desirable both in the Sanskrit text and the translations. (d) Two Sanskrit commentaries on the Brahma-kanda-vitti are available, an old one by Vssabha and a modern one by R. SARMA (1963). The first has been edited by C. SHASTRI (1934) in excerpts and by S. IYER

Page #2
182 ASHOK AKLUJKAR preserved in the manuscripts2 as well as in all the ancient commentaries of the Trikandi. It is also supported by the writings of Abhinava-gupta, Indu, Utpala(-deva), Kamala-sila, Jayanta-bhatta, Jnanasri-bhadra, (1966 a) in full. In its former version it finds its way into most of R. SARMA's commentary Ambakartri, the other-original-parts of the Ambakartri being mostly unhelpful for an exact understanding of Bhartrhari's position and for the solution of textual difficulties. (e) For the state of the Vakya-kanda-vrtti studies, see AKLUJKAR 1969: 555-561. (f) That the Vs of both the kandas are written by one and the same author has found unanimous acceptance, although, as far, as I know, no demonstration of the identity of the authors is available in print. I have noticed more than fifty instances of similarity of diction, style, views, and sources between the two Vs, which put their common authorship beyond doubt as far as I am concerned. 2 (a) In the manuscripts containing both the karikas and the V, the following colophon is found at the end of the Brahma-kanda with insignificant scribal variations: iti sri-hari-vrsabha-mahavaiyakarana-viracite vakyapadiye agama-samuccayo nama brahma-kanddam samaptam (C. SHASTRI 1930: 635; S. IYER 1965: xviii; and the concluding pages of the Sanskrit text in the editions referred to in fn. 1c above). In at least three manuscripts consisting only of the karikas of the three kandas, the colophon is: iti sribhagavad-bhartrhari-vrsabha-mahavaiyakarana-pada-viracite vakyapadiye prakirnaka-namani pada-kande vrtti-samuddesas caturdasah (C. SHASTRI 1930: 635; S. IYER 1965: xviii; RAU 1962: 387-388, 390, 392, 1964: 194). Now, there is no justification for supposing that the person designated by harivrsabha-mahavaiyakarana of the first colophon is different from the person designated by bhagavad-bhartrhari-vrsabha-mahavaiyakarana-pada of the second colophon, for (i) hari as a shorter form of bhartrhari occurs quite frequently in later literature, and (ii) the terms bhagavat and mahavaiyakarana-pada are merely more expressive of respect than the term mahavaiyakarana. Thus, the karikas as well as the composition including both the karikas and the V have been ascribed to one person Hari-vrsabha in the manuscripts. That this Hari-vrsabha is none other than Bhartrhari or Hari is evident from the alternative form Bhartrhari-vrsabha and also from (b) below. (b) Two explanations have so far been put forward as to why Hari or Bhartrhari is called Hari-vrsabha or Bhartrhari-vrsabha in certain colophons. According to C. SHASTRI (1930: 635) and S. IYER (1965: xviii), the member -vrsabha has been placed in the compound designation merely as expressive of respect (puja-vacana), which is justified in the light of fairly common Sanskrit expressions like purusa-vrsabha. But R. KAVI (1930: 235 fn. 2) thinks that -vrsabha was appended to Bhartrhari's name because he staunchly maintained the doctrine of sabda-brahman denoted by the word sabdavrsabha in his work (1.123 and its V). Going against the first explanation is the fact that vrsabha as expressive of respect is redundant in the light of mahavaiyakarana or bhagavad-mahavaiyakarana-pada. Detrimental to the second explanation is the fact that the soul of human beings, which in its

Page #3
The Authorship of the Vakyapadiya-Vrtti 183 Nagesa, Prabha-candra, Bhatta-narayana-kantha, Bhoja, Mandanamisra, Malla-vadin (senior), Rama-kantha, Vadi-deva-suri, Simha-surigamin, Somananda and Skandasvamin-Mahesvara, and perhaps of Abhaya-deva, I-ching, Karnakagomin, Dinnaga, Dharma-kirti, Mammata, and Suresvara. Furthermore, overwhelming similarities of thought, diction, and sources can be shown to exist between the V and the Tripadis on the one hand, and the V and the verses of the Trikandi on the other purest essential aspect is sabda-brahman of Bhartshari (see AKLUJKAR 1970), was described as vrsabha from at least as early a time as that of the Mahabhasya (KIELHORN's edition, Vol. I p. 3.20--22), and hence the idea of applying the term vrsabha to the supreme reality in the philosophy of the grammarians could not have been considered to be original with Bhartshari. Besides, no reference to Bhartphari as Hari-vrsabha has so far been found in the grammatical and philosophical works of the later period. There is room, therefore, for a third explanation: (i) Among the manuscripts which contain only the karikas of the Trikandi, the term -vrsabha is found in the manuscripts of the southern recension only (RAU 1962: 387-388, 390, 392, 396, 1964: 194195). (ii) A study of the known karika-vstti manuscripts, which invariably contain the term -vrsabha in their colophons, reveals that all those manuscripts must have had a common (most probably, south Indian) origin (see my forthcoming textual study of the V). Thus, the use of Hari-vrsabha for Bhartphari can be said to be a peculiarity of the southern recension only. Now, all the available manuscripts of Vrsabha's commentary on the karikas and V of the first kanda are also found to belong to a common south Indian source (S. I kuu 1966 a: ix, xv--xvi), which indicates that Vrsabha's Tika was once widely used in south India. Moreover, there are some readings in the Brahma-kanda-vrtti which could not have been a result of any factor other than the use of Vrsabha's commentary to understand the V (e. g. atma-tattvam in V 1.5; see my forthcoming textual study of the V for more on this point). We can, therefore, assume that, in the past, there were some manuscripts of the V with Vrsabha's commentary in south India. A common colophon of such manuscripts must have contained the words hari (or bharthari) and vrsabha which may be said to have found their way together through some inadvertance on the part of the scribes, first, in those manuscripts which consist of the karikas and the V, and next, in those manuscripts of the southern recension which contain only the karikas. 3 Bhartshari's commentary on the Mahabhasya, published under the inaccurate title Mahabhasya-dipika. See AKLUJKAR, "Mahabhasya-dipika or Tripadi ?" to be published in the Adyar Library Bulletin, Madras. 4 The evidence in favor of Bhartshari's authorship of the V that I have collected from Bhartshari's ancient commentators and the authors mentioned in this section is too extensive to be included here without obscuring the main lines of my argument. Also, the parallels that I have noticed between (a) the V and the Tripadi, and (b) the V and the Trikandi are so numerous as to deserve separate treatments. I shall, therefore, reserve these

Page #4
184 ASHOK AKLUJKAR 1.2 Under these circumstances, any doubt about Bhartphari's authorship of the V may seem highly improbable; but, today, all scholars who are interested in the Trikandi, as far as I know, entertain such a doubt. Their uncertainty of opinion usually begins when they realize that the V occasionally gives two or more interpretations of one verse (karika). Then this uncertainty is deepened either by the occurrence in the V of the word tatra-bhavat, which, in the usage of some ancient authors, serves as Bhartshari's epithet, in a manner indicating reference to a person other than the author (S. IYER 1965: xxxi--xxxii), or by a feeling that some divergence of views exists between the V and the karika-text (BIARDEAU 1964a: 5--21 (summarized by S. IYER 1965: xxxiii-xxxiv), 1964 b: 260). But doubtful as they may be, no scholar except MADELEINE BIARDEAU is known to me who has declared the traditional authorship of the V to be illfounded and incorrect. BIARDEAU has gone beyond the range of uncertainty about the validity of the tradition and reached the conclusion that the V cannot be a work of Bhartphari, that it must have been written by Hari-vrsabha sometime after Kumarila, and that the tradition accepted it as Bhartphari's work through a confusion of names. 1.3 The purpose of the present article is to refute this conclusion. Not only do I uphold the validity of the traditional ascription, but I also maintain that the V is an inseparable part of the Vakyapadiya and that it is wrong to think of the Vakyapadiya as a work consisting of karikas only. Now, there are two ways of establishing this thesis, one negative matters for exhaustive presentations in the near future and, in the meantime, merely draw the attention of scholars to some of my published sources where the relevant pieces of evidence are already pointed out to some extent: C. SHASTRI 1930: 634638, 644-645, 1934: fns. to various V passages (especially, pp. 3, 5, 87, 103, 115, 126, 128), 1941 (?): fns. to various V passages (especially, V 2.28-29); FRAUWALLNER 1933: 237 (Although the parallel noticed in this article remains valid, FRAUWALLNER, for good reasons, changes his view on the relative chronology of Bhartphari and Dinnaga in his 1959 and 1961 publications); K. M. SARMA 1940: 2-4, 1942: 405409; NAKAMURA 1955: 130; SWAMINATHAN 1963: 66--70; Y. MIMANSAKA samvat 2020: 347; BIARDEAU 1964b: 260 fn. 4 conclusion; S. IYER 1965: xvii-xxix, 1966b: 28-30; ABHYANKAR-LIMAYE 1965: 210--216, 240, 297-301 (cp. NAKAMURA 1955: 125, V 1.1), 312 (cp. NAKAMURA 1955: 126, V 1.1 verse tha), 328--329 (cp. V 2.315--316), 333 (cp. V 2.64), 352--357; LIMAYE 1966: 228--229; JAMBUVIJAYA 1966: 40 (V 1.52), 128, tippanani p. 57 (cp. V 1.140) 192, tippanani p. 68 (V 1.8), 197 (V 1.2), 239, tippanani p. 77 (V 1.9), 241, tippanani pp. 76-77 fns. 3-5 (V 1.1); AKLUJKAR 1969: 557-561.

Page #5
The Authorship of the Vakyapadiya-Vrtti 185 and another positive. The negative way consists in challenging BIARDEAU's method of solving the problem of authorship, in pointing out the difficulties to which her conclusion leads, and in demonstrating that the divergences which she notices between the views and use of terms in the V on the one hand, and in the karikas on the other, are superfluous and that some of her interpretations are inaccurate. I have followed this way in a forthcoming sequel article, and hence it would be proper to devote the present article to a positive demonstration of Bhartphari's authorship of the V. 2.1 It would not be out of place, however, to preface this presentation of new positive evidence by a consideration of two really relevant facts noticed by S. IYER (1965: xxx-xxxii) as going against the thesis of identical authorship for the karikas and the V. The first is that of the alternative explanations in the V of the expressions in the karikas. The instances falling in this category, as far as I could find out, are the following: V 1.10, 12, 13, 44; 2.20--21, 39. In the case of all these instances, a careful reading of the Trikandi and its ancient commentaries will reveal that what seem to be different explanations are also statements of different theses and thoughts acceptable to Bhartshari on different levels and in different contexts 5; there is no question of prefer 5 (a) With jnanatmakatvad va purusasyaiva samskara-hetavah (vidyabhedah) of V 1.10, compare atha va atma-samskarartham cadhyeyam vyakaranam of Tripadi p. 5.17; cf. AKLUJKAR 1970: $ $ 1.37-38. (b) V 1.12 clarifies the ways in which vak can become differentiated. For Bhartshari's acceptance of those ways from different points of view, see AKLUJKAR 1970: $$ 4.1-10, 21, 5.21-28. (c) For Bhartphari's acceptance of the theses expressed in V 1.13, see (i) AKLUJKAR 1970: $ $ 5.15-17, 22; V 1.129 p. 209.1-2; Trikandi 2.429 and V, 2.434 and V, 3.14.197, 570; (ii) AKLUJKAR 1970: $ 1.4, n 1.5, $ 6.4, n 6.21-25; (iii) AKLUJKAR 1970: $$ 4.22, 5.3; (iv) AKLUJKAR 1970: $ 6.4; (v) AKLUJKAR 1970: $ 5.19; (vi) AKLUJKAR 1970: $ 5.15. (d) That there is no incompatibility in the different meanings of the word upadana is clear from--V 1.44 itself, where the conditions of the acceptance of each meaning are specified with such words as vyutpatti-paksa and sva-rupa-padarthakesu. See also V 1.47 p. 105.3--5, V 1.50 p. 109.3-4, Trikandi 2.128, BSS p. 136, Trikandi 2.175, BSS p. 151, V 2.262. (e) V 2.20--21 interprets the two karikas in the light of both the sabda-vyakti-paksa and the sabdakrti-paksa. For Bhartshari's acceptance of these paksas, see AKLUJKAR 1970: $ 4.18. (f) V 2.39 first explains the karika as stating that the primacy of the sentence is noticed even in the explanatory remarks (vigraha-vakya) of grammarians and then as stating that, whether there is one word or more, it is only the sentence which is used for communication. That both these interpretations of the karika express points acceptable to BhartThari is evident from AKLUJKAR 1970: 88 4.22, 5.3. 1970.ceptance ofrys in which : $$ 1.37

Page #6
186 ASHOK AKLUJKAR ring one explanation or statement to the exclusion of the other, for there is no contradiction between them, the differences between them being more apparent than real. Thus, the verses under consideration can be said to have been deliberately and ingeniously composed in a sufficiently vague or general manner, and the V can be said to have been given the task of clarifying how the statement in a verse holds good on more than one level and in more than one context. In other words, the instances of alternative explanations are also the instances of skilful and careful composition in verse, of welding into one a number of statements. Another fact to be noticed in this connection is that quite frequently the alternative explanations are introduced by the word apara. One can show with unmistakable evidence that apara does not always mean simply "someone else" or "different"; it is often connotative of approvale, as if it is intended to be dissolved as na vidyate paro yasmat. Moreover, Bhartshari's zest for, and skill in, varied interpretation are apparent to anyone who reads his commentary on the Mahabhasya. Those two aspects of his genius are perfectly understandable in the light of what we learn from Trikandi 2.478-487. Bhartphari's works, it seems, marked the culmination of a movement that was devoted to the revival of Mahabhasya studies, in particular, to the unraveling of the various nyaya-bijas, sakhas, nyaya-prasthana-margas, vartmans, and agamas that were hidden in the apparently simple statements of the Mahabhasya. Therefore, the alternative explanations in the V and the multifaceted composition of certain karikas should not come as a surprise to us. The alternative explanations do not indicate that the author of the V is uncertain about the meaning of a verse; he is, in fact, pre-equipped to deal with a verse in more than one way, for he himself has invested certain verses with convenient ambiguity and has thus made them amenable to different interpretations? 2.2 The objection to Bhartshari's authorship of the V based on the epithet tatra-bhavat can also be easily answered. It is true that Abhinavagupta (see S. IYER 1965: xxii; ABHYANKAR-LIMAYE 1965: 211-213, & THIEME 1956: 15; AKLUJKAR 1970: $ 0.7; Sankara's bhasya on the Brahma-sutras 1.1.25, 27, 1.2.12, 1.3.1, 13. ? It should be noted that my treatment of the objection based on the alternative interpretations is different from that of S. IYER (1965: xxxv), who, in fact, does not offer any satisfactory reply in the comment: "What looks like two different ways of explaining the same text may be nothing more than the recording of the opinion of another on the the [sic; same?] subject, unconnected with the text." The alternative explanations are, on S. IYER'S (1965: xxx) own admission, connected with the karika-text.

Page #7
The Authorship of the Vakyapadiya-Vrtti 187 271--272, 274), Indu (see ABHYANKAR-LIMAYE 1965: 329), Bhattanarayana-kantha (see NAKAMURA 1955: 130), and Helaraja (3.1.1 p. 7.3, 3.9.62 p. 72.7) seem to associate the epithet rather closely with Bhartphari. But, as S. IYER (1965: xxxvi), who first noticed the possible objection to Bhartrhari's authorship of the V arising out of the occurrence of the epithet in the V, himself observes, tatra-bhavat can very well be a reference to a person other than Bhartphari. Afterall, tatra-bhavat is a very general honorific in Sanskrit at least from the time of the Mahabhasya (KIELHORN's edition, Vol. 1 p. 3.5, 11.12, 22.23, 117.23; Vol. 2 p. 233.13, 254.17--19, 265.23--24, 275.21, 314.13, 405.16-18; Vol. 3 p. 174.10). Candrananda (see JAMBUVIJAYA 1961: 68) uses it in what is obviously not a reference to Bhartshari. Even Abhinava-gupta does not restrict its usage to Bhartshari; he uses it also to refer to the authors of the Samkhya-karikas and the Yoga-sutras (see MASSON-PATWARDHAN 1969: 114, 125 fns. 1--2). The Vakya-kanda commentary & (p. 284.19 20, possibly p. 286.3) applies it also to Bhartshari's teacher Vasurata. Moreover, tatra-bhavat does not occur only once in the V (2.42) as S. IYER (1965: xxxi, xxxvi) seems to think; it occurs twice (V 2.444), and possibly thrice (V 2.450)'. A close study of those occurrences reveals that it refers to an author (or a group of authors; see fn. 9b above) later than Patanjali 10 whose views lend support to, or agree with, I amy reis restricte che taomlication of the 8 This commentary, published in the Benares Sanskrit Series (MANA. VALLI 1887), is usually ascribed to Punyaraja. In a forthcoming article, I wish to point out that a good case can be made for Helaraja's authorship of it. If my guess is correct, then it follows that Helaraja, like Abhinavagupta, does not restrict the application of the honorific tatra-bhavat to Bhartphari, that he uses it also in referring to Vasurata. . (a) There is a gap in the manuscript before nto manyante of V 2.450. A comparison with V 2.444, where the expression tatra-bhavanto manyante occurs, shows that the gap can be best filled by supplying the letters tatrabhava. One does not have any justification to supply the letters iha-bhava in this instance for two reasons: (i) The expression iha-bhavantah does not oceur anywhere in the extant portions of the V. (ii) All occurrences of the expression iha-bhavantah in the Tripadi are immediately followed by tv ahuh, and not by manyante. (b) The honorific tatra-bhavat is used in the singular number in V 2.42, whereas it is used in the plural in V 2.444 and V 2.450. The available evidence is not sufficient to determine if the singular number had a specific connotation in the use of this honorific. 10 (a) Note atah "therefore" in V 2.42: sa cayam vakya-padayor adhikyayor bhedo bhasya (Patanjali on Panini 2.3.46, 50) evopavyakhyatah. atas ca tatra-bhavan aha ... Also note that in V 2.444 tatra-bhavat is used only after a reference is made to Patanjali's Bhasya on Panini 8.1.28 and 3.1.67. (b) From the evidence recorded in (a) above it is clear that the guesses of Y. MIMAMSAKA (samvat 2020: 334) and ABHYANKAR-LIMAYE (1965: 440)

Page #8
188 ASHOK AKLUJKAR Bhartphari's viewg11. In the latter aspect it can be said to contrast with the expression iha-bhavat, which occurs five times in the Tripadi (ABHYANKAR-LIMAYE edition 1967-70: 51.22, 58.4, 86.2, 204.24, 207.3) in the statements of those views about which Bhartshari does not seem to be enthusiastic12. Its presence in the V, therefore, cannot be thought of as a definite indication of the V not having been written by Bhartshari, and our solution of the problem of authorship must not depend on it alone, especially in the light of the facts pointed out above (1.1, 2.1) and below (3.1-8). 3.1 As was indicated in 1.3, it is not my objective in the present article to point out only the weaknesses of the arguments advanced either by BIARDEAU or by any other scholar. Instead of offering a negative defence of the traditional evidence, I propose to bring to light a new kind of evidence that will clinch the issue in favor of the tradition. So far those scholars who are inclined to accept Bhartphari's authorship of the V have concentrated only on the external evidence such as the testimony of authors later than Bhartphari. Strangely enough, no serious effort 13 has been made to examine the internal evidence. My principal objective in this paper, therefore, is to present this latter kind of evidence. However, I shall not here make a case out of the overwhelming similarities of diction, thought, and sources that are noticed in the V, karikas, and Tripadi, for, although it may seem strange to a man of my view to overlook such an impressive body of parallels, BIARDEAU and those who share her view will always be logically free to accept the possible alternative that the parallels are a result either of borrowings from, or of an that tatra-bhavat of V 2.42 is a reference either to Patanjali or to Vyali cannot be correct. Moreover, Bhartshari's explicit references to Patanjali and Vyadi are characterized respectively by the words bhasya (or curni) and samgraha, not by tatra-bhavat. 11 (a) With the view of tatra-bhavat in V 2.42, compare V 1.24-26 p. 77, Vrsabha pp. 78.2579.22, Trikandi 3.14.6 and Helaraja thereon pp. 45. (b) From BSS pp. 70--71 and Trikandi 2.112 (which answers the objection raised in 2.76), it is evident that the view ascribed to tatra-bhavat in V 2.444 meets Bharthari's approval. (c) Confirmation of the view expressed in V 2.450 is found in 2.68 and its V, and in Helaraja 3.1.5 p. 15.12--13. See also AKLUJKAR 1970: $ 6.4, n 6.24. 12 For the occurrences of iha-bhavat in other works, see Y. MIMAMSAKA samvat 2020: 304 and OBERHAMMER 1960: 80 fn. 22. 13 C. SHASTRI (1930: 636) and K. M. SARMA (1942: 408) offer some inconclusive internal evidence, while the internal evidence recorded by R. KAVI (1930: 238) is inaccurate.

Page #9
The Authorship of the Vakyapadiya-Vrtti 189 intensive study of Bhartphari on the part of the author of the V. I shall, therefore, mainly focus on what may be described as syntactical and compositional kind of internal evidence. It alone will serve to disprove BIARDEAU's (1964b: 260.10--12) contention that neither the first nor the second kana evinces any need for the V in between the karikas, that the karika-text is complete in itself14. 3.2 First of all, let me point out that the author of the V, at one place, clearly indicated that he wrote the karikas too. The passage I have in mind is as follows15: ajasra-vrttir yah sabdah suksmatvan nopalabhyate / vyajanad vayur iva sa sva-nimittat praciyate // 1.108 // athapara agamo 'nugamyate. suksmo vayu-samnicaya ivantar bahis ca sarva-murtinam dhvanir avasthitah. sa cainam akasa iti pratipadyate. sa, yathaiva tu sarvatra paramanu-sambhave samhatatvad vyajanabhighatena vayur asraya-sthanat pravibhajyamanah kriyabhir avisyate tathaiva, dhvanih sva-nimittair abhivyakta-pracita-vikriya-rupah srotra-pradesam prapta upalabhyate samskaroti ca. tasya prane ca ya saktir ya ca buddhau vyavasthita / vivartamana sthanesu saisa bhedam prapadyate || 1.109 // darsana-bheda 16 evaite. nayam anantarah pracaya-dharma dhvanir iha 14 It is important to remember in this connection that the Trikandi karikas do not belong to the sutra-form of literature. Syntactical incompleteness cannot, therefore, be a regular feature of them. Hence, if it is shown, as has been attempted in the following pages, that the V brings about a syntactical completion of certain karikas, then it follows that the karikas anticipate the V, that Bhartshari expects his reader to read the karikas in conjunction with the V. 16. To understand the difficult Sanskrit passages quoted in this and the following sections the existing translations of the Brahma-kanda, however unsatisfactory they may be at places, should be utilized, those of BIARDEAU (1964 a) and S. IYER (1965) for the karikas and the V, and that of Santibhiksu Sastri (1963) for the karikas only. It is regrettable that a good edition and translation of the Vakya-kanda are even now the great desiderata of BhartThari studies. To keep my presentation uncomplicated, I have refrained from introducing any translations in the present article, although all ideas and expressions that are of vital importance for the topic under discussion have been invariably clarified. 16 All the editions of the Brahma-kanda-vstti read paksa-bheda; but commentator Vrsabha (p. 179.6--7) preserves the reading accepted here. His is a more appropriate reading, for while paksa usually means "alternative" darsana means "point of view, theory, view of reality," and verses 1.107-109 are definitely not a statement of alternatives. Also, see V 1.45 p. 103.6, Tripadi p. 19.1.

Page #10
190 ASHOK AKLUJKAR sloke nirdisyate. sabdas tu purva-prakrtah pravada-bhedair anvakhyayate. ... The last two sentences of this passage mean: "It is not this immediately preceding dhvani, having the property of growth (= becoming gross), which is mentioned in this verse; but the sabda which was formerly taken up for discussion with reference to various views is being explained (or subsequently mentioned)." Now, the immediately preceding verse, 1.108, does not contain the word dhvani. Only the V on that verse mentions dhvani. But the author of the V evidently suspected that some reader may misunderstand tasya in verse 1.109 as a pronoun standing for dhvanih in V 1.10817. Obviously, he looked upon both the V and the karika to be his own words intended to be read consecutively. . 3.3 My first piece of the syntactical kind of evidence reveals that, at least once 18 in the first kanda, a karika is incomplete without the y and vice versa. Thus, we read in the case of verse 1.92: sphotesu 19 bhagavatsv api tesv eva rupa-bhedo dhvaneh kramat / nirbhagesv abhyupayo va bhaga-bheda-prakalpanam // 17 tasyeti sarva-namnanantarasya dhvaneh pracaya-dharminah pratyavamarsasanka na karaniya [ity aha] ... kasya tarhi tasya ity anena sambandha ity aha sabdas tu iti. (Vrsabha pp. 178.26-179.6) 18 It is possible that the V supplied padani to verse 2.55. Unfortunately, however, the guess cannot be confirmed, for the V of that verse is missing in the only available, incomplete, and highly corrupt manuscript of the Vakya-kanda-vitti (AKLUJKAR 1969: 555-556). 19 In all the editions of the Brahma-kanda-vstti, sphotesu is printed as the last word in the V of verse 1.91: ... tathainam arvag-darsananam pratipattinam vakya-8va-rupa-grahana-purvakena vakyartha-grahanena pradhanena prayuktanam niyatopaye sadhye tasminn arthe niyata-kramaparinama-bhagakara-pratyavabhasa-matra-yukta buddhayah pravartante sphotesu // 91 //. This is an error on the part of the editors not only for the reasons mentioned in 3.3 but also for some additional ones. If the cited sentence or clause is read in this manner, its interpreter must understand that there are two loci or objects for the cognitions (buddhi) of which it speaks; one locus is expressed by niyatopaye sadhye tasminn arthe and the other is expressed by sphotesu. The interpreter then fails to understand why the first expression should be in the singular and the second in the plural or why there should be a long intervening phrase between the two. In fact, he fails to understand why there should be two loci at all. Furthermore, a sentence ending with pravartante sphotesu goes against the stylistic peculiarity of the V that a verb form usually occurs at the end of a sentence. Failure to realize this editorial mistake of C. SHASTRI (1934: 91) has led BIARDEAU (1964a: 133) and S. IYER (1965: 90) to offer confusing translations, and R. SARMA (1963: 167) to paraphrase the words of the original in an irritatingly superfluous manner.

Page #11
The Authorship of the Vakyapadiya-Vitti 191 Here sphotesu in the V will clearly remain dangling if it is not understood to be a part of what follows. The karika will also be incomplete and out of step with the preceding karika (see fn. 19 above) if sphotesu is not read into it. Thus, here we have one phrase with a locative absolute construction tesv eva sphotesu bhagavatsv api which spreads over both the V and the verse, the subject of which appears only in the V, and the pronoun in which will have no antecedent if the V is missing. 3.4 The third piece of internal evidence is found in verse 1.65 and its V: evam ca krtvoccaryamanasya tasyabhidheya-bhavena yah sabdah samavasthitah/ tasyapy uccarane rupam anyat tasmad vivicyate II, Here the syntactical relation between uccaryamanasya and tasya if too obvious to need any comment. What is more important is that, is the phrase evam ca krtvoccaryamanasya were missing, we would not have either understood the relation of the present karika with the preceding karika or we would have done so only by overlooking a grave syntactic anomaly nowhere to be noticed in the well-preserved portions of Bhartshari's writings. In the preceding karika, which is gunah prakarsa-hetur yah svatantrenopadisyate / tasyasritad gunad eva praktstatvam pratiyate // 1.64 II, the pronoun tasya stands for gunah. In the absence of the V phrase connecting the two karikas, tasya in the present karika would also have referred back to gunah and the karika would not have conveyed any consistent meaning whatsoever. 3.5 This leads us to the next finding that, in about ten 20 instances, the V is joined to the following karika in a manner which is different from the manner in 'which Sanskrit commentators, while commenting on somebody else's work, introduce the words of that work. The instances I have in mind are as follows: (a)... tatra vyddhyadayah sabdah sva-rupadhisthanah svenarthenarthavantah sva-rupena sabdantara-sva-rupany upajighsksantah svaranunasikya-bhinnair akaradibhir, adaicchabdadibhih pratyayitaih, sam 20 I am not sure whether one should include V 2.28-29 in this group.

Page #12
192 ASHOK AKLUJKAR bandham yena prakarena pratipadyante, tenaiva prakarena 21 duravadharatve 'pi bhedasya 22 agni-sabdas tathaivayam agni-sabda-nibandhanah/ agni-drutyaiti sambandham agni-sabdabhidheyaya // 1.60 // (b) 1.65 and its V. See 3.4. (c) yatha vinayendriya-dharma evayam prakrta-caksusam, durad arupa-matropalabdhau 23 vrksadin hastyadivat prati padyante, taddesavasthita eva pranidhanabhyasat kramena punar yathavayavam [yathavad ?] upalabhante, vyaktalokac ca desat sahasa mandasamnivista-prakasan apavarakadin praviaya rajjvadin sarpadivat pratipadya tathaiva pranidhanabhyasac caksusi prakrtisthe yathavad upalabhante 24 vyajyamane tatha25 vakye vakyabhivyakti-hetubhih / bhagavagraha-rupena purvam buddhih pravartate // 1.90 // (d) 1.92 and its V. See 3.3. (e) devadattadinam ca [vi]bhagabhimatanam artha-yogabhyupagame, 21 There is no indication of the phrase tenaiva prakarena in Vrsabha's commentary. In view of tathaiva in the karika that follows, it does introduce redundancy in the construction proposed by me. But this fact does not indicate that the author of the V is not the same as that of the karika-text. In all well-known works of the karika-vstti type, an effort is made to give as much syntactical independence as is possible to each type of text. This is most probably so because the ancient writers were aware of the tendency of (and convenience in) singling out the metrical karikatext for the purposes of study and memorisation. Connecting the drstanta with the darstantika separately in verse and prose portions is a result of such considerations. Compare the V on karikas 1.89-90; 2.20--21, 185-186, 216217, 275-278, 298-299, 300--301, 420-421, 462-463. Unfortunately the V on the karikas 2.8--9, 90--92, 93--94, in which also the dpstanta and the darstantika spread over more than one verse, is lost in the manuscript. 22 All editions indicate the end of V 1.59 after bhedasya. Overlooking the fact that the preceding words do not form a complete sentence. As a consequence, BIARDEAU (1964a: 101) and S. IYER (1965: 64) are forced to supply some words in their translations. R. SARMA (1963: 121) seems to have sensed the syntactical difficulty; but he does not point it out explicitly. 23 The editions read akyti-matropalabdhau; but Vrsabha's (p. 155 fn. 3) reading, which is adopted here, is more appropriate both contextually and stylistically. See V 1.81 p. 148.1; V 1.102 p. 168.2 (Vrsabha p. 168.14); V 1.116 p. 191.1. 24 According to the editors of the available editions, upalabhante marks the end of a syntactically complete sentence. That is why they put a full stop after it. As a result, both BIARDEAU (1964 a : 131) and S. IYER (1965: 89) overlook the initial word yatha in their translations. . 25 The word tatha is primarily connected with yathaiva in verse 1.89 and only secondarily with yatha in V 1.89. See fn. 21 above.

Page #13
The Authorship of the Vakyapadiya-Vrtti 193 samanye 'vasthitanam padanam visese 'vasthanam ity etasmin satyatamsena 26 parigrhyamane samanyarthas tirobhuto na visese 'vatisthate / upattasya kutas tyago, nivrttah kvavatisthatam // 2.15 // (f) nitya-vadinam tv 27 anagantukam anatikrantam anupacayapasauat prakasaka-prakasyatvam karya-karana-rupata / antarmatratmanas tasya sabda-tattvasya sarvada || 2.32 // (g) tatraivam sastremaparyudaste visesantara-sahacarini dravyatve asamnidhau pratinidhir ma bhun nityasya karmanah / kamyasya va pravsttasya lopa ity upapadyate // 2.70 // (h) nirjnate ca bhede prasiddhartha-viparyasa-nimittam yac ca dTsyate | yas tasmal laksyate bhedas tam asatyam pracaksate || 2.289 || (i) ye tutsargapavadayor eka-vakyatvam icchanti tesam niyamah pratisedhas ca vidhi-sesas tatha sati / dvitiye yo lug akhyatas tacchesam alukam viduh // 2.350 // It should be noted that in these instances the introductory words of the V are not simply of the form tasmat, api ca, apara aha, etc. as they are in some other portions of the V. Nor do they simply explain the background of a verse, although they are found to play that role in other introductions of the V (see 3.8). What we have here are words which are syntactically connected with the verses, which are absolutely essential to understand the connections between verses, and which, in most cases, supplement the statements in the verses significantly. The author of the V clearly does not follow the path which Sanskrit commentators commenting on somebody else's work follow. 3.6 Now we come to the compositional kind of evidence. In verses 1.24--26, eight topics which form the subject matter of the Vakyapadiya are enumerated. They are: (a) meanings determined through analysis; (b) meanings of stable character; (c) linguistic forms or units that are to be analyzed; (d) linguistic forms or units that figure in grammatical derivations; (e) cause-and-effect relation; (f) fitness or capability relation; (g) relation that leads to merit; and (h) relation that causes cognition: 26 My transcript of the Vakya-kanda-vitti manuscript reads satyatomsena. C. SHASTRI (1941 [?]) reads satyekantena. 27 I do not understand the significance of nitya-vadinam tu in this specific context. The preceding verse or portion of the V does not seem to be a statement of the view of the anitya-vadin. 13 WZKSA XVI

Page #14
194 ASHOK AKLUJKAR apoddhara-padartha ye, ye carthah sthita-laksanah / anvakhyeyas ca ye sabda, ye capi pratipadakah || 24 || karya-karana-bhavena yogya-bhavena ca sthitah ! dharme ye pratyaye cangam sambandhah sadhvasadhusu || 25 || te lingais ca sva-sabdais ca sastre 'sminn upavarnitah / smrtyartham anugamyante kecid eva yathagamam // 26 // After this enumeration, one would naturally expect verse 27 to begin a discussion of the first topic, namely the apoddhara-padartha. Instead, one finds it initiating the discussion of the seventh topic which is the dharma-sambandha and adharma-sambandha respectively of the sadhu and asadhu linguistic forms. The only satisfactory reason for skipping over the first six topics in the karikas is that those, and only those, are mainly covered in the V of verses 1.24--26. The seventh topic alone remains to be explained 28 and the twenty-seventh verse takes it up for discussion. 3.7 The next piece of evidence indicates even more strongly that the karikas anticipate the V. It is as follows: vitarkitah pura buddhya kvacid arthe nivesitah / karamebhyo vivrttena dhvanina so 'nugrhyate || 1:47 || ... avikriya-dharmakam hi sabda-tattvam dhvanim vikriya-dharmanam anu vikriyate. tac ca suksme vyapini dhvanau karana-vyaparena praciyamane, sthulenabhra-samghatavad upalabhyena nadatmana prapta-vivarte 29 tad-vivartanukarenatyantam avivartamanam vivartamanam iva glhyate. 28 That there is cognition of meaning because sabda and artha are related is such a basic, common sense fact (Vrsabha, p. 81.18-19) that Bhartshari has not been forced to devote much space to discuss it. Also, the discussions of the three remaining relations are indirectly a discussion of the pratyayanga relation. 29 The editions read prapta-vivartena. But then the sentence seems to be syntactically anomalous; the relation of the locative absolute construction suksme ... praciyamane to the succeeding portion of the sentence is not clear; further, there is ambiguity as to what the component tad. in the compound tad-vivartanukarena refers to-to dhvani or to nadatman. Most probably it refers to dhvani, for the segment nadatmana prapta-vivartena tad-vivartanukarena with tad- referring to nadatman would be a very awkward way of saying what could be said simply with nadatma-vivartanukarena. Moreover, in the very preceding sentence, BhartThari says ... dhvanim vikriya-dharmanam anu vikriyate. It is almost certain then that tad- refers to dhvani. This point aids us further in guessing what the original text of the V could have been. The vivarta mentioned in the compound praptavivarta must then be the vivarta of dhvani and the compound as a whole must qualify dhvani; that is, its form must have been prapta-vivarte which

Page #15
The Authorship of the Vakyapadiya-Vrtti 195 nadasya krama-janmatvan na purvo na paras ca sah / akramah krama-rupena bhedavan iva jayate // 1.48 // kramavata hi vyaparenopasamhriyamana-pracaya-rupo nadah sapratibandhabhyanujnaya vrttya sphotam avadyotayati. ... pratibimbam yathanyatra sthitam toya-kriya-vasat / tat-pravrttim ivanveti sa dharmah sphota-nadayoh || 1.49 // In this passage, we observe a transfer from dhvani in verse 1.47 to nada in verse 1.48. These two words are not synonymous for the sabda-vyakti-vadin, whose view is put forward in this passage and in verses 1.94--101. According to him, dhvani is subtle and pervading, whereas nada is the gross and perceptible form of dhvani. He maintains that dhvani manifests the sphota through the intermediacy or instrumentality of nada. In switching over to nada, verse 1.48 must be said to assume an indication of this theoretical subtlety in the V of verse 1.47. Otherwise, Bharthari's choice of terms becomes pointless; the juxtaposition of sphota and nada in verses 1.48-49, 97 and, rather indirectly, in 1.101 does not contrast significantly with the juxtaposition of sphota and dhvani in verses 1.75, 77, 81, and 93 30 which seem to put forward the view of the sabdakyti-vadin. The second transfer that we notice in the passage cited above is from sah in verses 1.47 and 48 (= buddhisthah sabdah in verse 1.46) to sphota- in verse 1.49. It also cannot be accounted for unless V 1.48 introduces the term sphota which is not found in any of the preceding karikas. 3.8 Furthermore, the first two kandas contain many instances in which a plausible and straight-forward interpretation of a karika is made possible only by the V. In about thirty-six cases, the fact that the karika considers and answers an objection or a query, and the nature of that objection or query are known only from the V: 1.76, 95--100, 132, 142; 2.14, 23, 25, 36-37, 46, 48, 51, 62, 68, 159, 193--196, 200--201, 221-226, 333, 340, 363, 365. That the karika switches over to the consideration of an alternative or a different view is made known only by agrees with dhvanau. This emendation extends the locative absolute construction up to prapta-vivarte and the syntactic anomaly is removed. It seems that the copyists were led to add -na after -vivarte by the frequent occurrence of n and na in this sentence. 30 In verse 1.84, which forms a part of the group of verses (1.7593) mainly stating the process of sphota-manifestation according to the sabdakrti-vadins, nada and dhvani occur side by side. This is probably due to the exigency of the meter. Vssabha (p. 150.7, 9), who usually does not explain the meanings of common expressions, comments specifically in this case: nadaih iti dhvanibhih.

Page #16
196 ASHOK AKLUJKAR the V in about forty-two cases: 1.73, 94, 104, 108-109, 137; 2.19-22, 41-42, 49, 60-61, 66, 183--184, 256, 261262, 269, 285-286, 315-- 316, 328-329, 331, 350--352, 360, 395, 409, 415, 438, 440-441, 445446, 455, 459, 461-463, 473. Moreover, the relevance, background, role in a particular context (say, as analogy or example), and serving as the starting point of a new topic would never be known exactly, if the V were not available to guide us, in the case of at least 25 karikas: 1.28, 63-64, 76, 78-80, 84; 2.15-16, 41, 59, 64, 70, 76-87, 164--165, 197--198, 205, 239, 272, 298--299, 304-313, 353, 372, 417. It is the responsibility of those who hold that the available V is not an integral part of the Vakyapadiya, to demonstrate that each and every verse mentioned above can be interpreted satisfactorily without the aid of the V or of any commentary following the V. In my opinion, the karikas obviously need supplementation to be understood properly and hence clearly evince the author's plan to write a gloss on them 31. As the present V accomplishes the desired supplementation and as it is unanimously held to be an integral part of the Vakyapadiya in a continuous, old, and impressively documented tradition, it must be the gloss written by Bhartphari. To say that the original gloss of Bharthari was lost and a new one written by some later author took its place is unwarranted, is not borne out by any reliable piece of evidence, and amounts to nothing but a desparate attempt to seek refuge in the subterfuge of a remote possibility. BIBLIOGRAPHIE ABHYANKAR, K. V. and LIMAYE, V. P. 1965. (Eds.) Vakyapadiya of Bhartphari. University of Poona Sanskrit and Prakrit Series Vol. II. Poona. 1967. (Eds.) Mahabhasya-dipika of Bhartshari. Part I. Bhandarkar Oriental Research Institute. Post-graduate and Research Department Series No. 8. Poona. 1969/70. Part II of the same. My references are to the printed forms kindly supplied by the editors. AKLUJKAR, ASHOK. 1969. "Two Textual Studies of Bhartrhari." JAOS 89. 547-563. New Haven. - 1970. The Philosophy of BhartThari's Trikandi. Ph. D. dissertation. Harvard University. Cambridge, Mass. [Unpublished). BHAGWAT, V. B. 1965. (Ed. and transl.) Srimad Bhartphari-viracitam Svopajna-tika-samanvitam Vakyapadiyam (Brahma-Kandam). Grantha Samsodhana Prakasana Mandala, publication no. 2. Tilak Maharashtra Vidyapeeth. Poona. (Marathi). 31 In a future article, I shall consider the verses of the third kanda from this point of view.

Page #17
The Authorship of the Vakyapadiya-Vrtti 197 BIARDEAU, MADELEINE. 1964 a. (Ed. and transl.) Vakyapadiya Brahma. kanda avec la Vstti de Harivrsabha. Publications de l'Institut de Civilisa tion Indienne. Serie IN-80. Fascicule 24. Paris. - 1964 b. Theorie de la Connaissance et Philosophie de la Parole dans le Brahmanisme classique. Mouton and Co. The Hague. BSS. See MANAVALLI. CHARUDEVA SHASTRI. 1930. "Bhartshari: a Critical Study with Special Reference to the Vakyapadiya and its Commentaries." Proceedings and Transactions of the Fifth Indian Oriental Conference. Vol. I. Pp. 630-- 665. Lahore 1934. (Ed.) Vakyapadiyam Prathamam Kandam. Ramlal Kapoor Trust. Lahore. 1941 (?). (Ed.) Vakyapadiyam Dvitiyam Kandam. Ramlal Kapoor Trust. Lahore. [Incomplete; see AKLUJKAR 1969: 555). FRAUWALLNER, ERICH. 1933. "Dignaga und anderes." Festschrift Moriz Winternitz. Leipzig. - 1959. "Dignaga, sein Werk und seine Entwicklung." WZKSO 3.83-164. Vienna. - 1961. "Landmarks in the History of Indian Logic." WZKSO 5.125 148. Vienna. Helaraja. Prakirnaka-prakasa: (a) Samuddesas 1-7: (Ed.) SUBRAMANIA IYER, K. A. Deccan College Monograph Series No. 21. 1963. Poona. (b) Samuddesas 8-13: (Ed.) SAMBA-SIVA SASTRI, K. Trivandrum Sanskrit Series No. CXVI. 1935. Trivandrum. (c) Samuddesa 14: (Ed.) RAVI VARMA, L. A. University of Travancore Sanskrit Series No. CXLVIII. 1942. Trivandrum. JAMBUVIJAYA, MUNI. 1961. (Ed.) Vaisesika-sutra of Kanada with the Commentary of Candrananda. Gaekwad's Oriental Series No. 136. Baroda. - 1966. (Ed.) Dvadasara-naya-cakra of Malla-vadin with the Commentary of Simha-suri-ganin. Jain Atmananda Sabha. Bhavnagar. K. M. SARMA. 1940. "Bhartphari Not a Buddhist: Evidence from Nearer Home." PO 5.1-5. Poona. - 1942. "Gleanings from the Commentaries on the Vakyapadiya." ABORI 23.405-412. Poona. KIELHORN, F. 1880. (Ed.) The Vyakarana-mahabhasya of Patanjali. 3 Vols. Department of Public Instruction, Government of Bombay. Second edition 1892. Third edition 1962--1970: Bhandarkar Oriental Research Institute, Poona. KUNHAN RAJA, C. 1936. "I-tsing and Bhartphari's Vakyapadiya." S. Kri shnaswami Aiyangar Commemoration Volume. Pp. 285--298. Madras. LIMAYE, V. P. 1966. "Dhyanagraha-kara or Dhyana-kara: a Pre-Bhartphari Grammarian." VIJ 4.228-229. Hoshiarpur. MANAVALLI, GANGADHARA SASTRI. 1887. (Ed.) Vakyapadiyam... Sri Bhartphari. ... viracitam Sri-Punyaraja-krta-prakasakhya-tika-yutam. Benares Sanskrit Series Nos. 11, 19, and 24. Benares. MASSON, J. L. and PATWARDHAN, M. V. 1969. Santarasa and Abhinava gupta's Philosophy of Aesthetics. Bhandarkar Oriental Series No. 9. Bhandarkar Oriental Research Institute, Poona. NAKAMURA, HAJIME. 1955. "Tibetan Citations of Bhartphari's Verses and

Page #18
198 ASHOK AKLUJKAR the Problem of His Date." Studies in Indology and Buddhology Presented in Honour of Professor Susumu Yamaguchi on the Occasion of His Sixtieth Birthday. Pp. 122-136. Kyoto. OBERHAMMER, G. 1960. "The Authorship of the Sastitantram." WZKSO 4.71-89. Vienna. RAGHUNATHA SARMA. 1963. (Ed. and comment.) Vakyapadiya [Brahma kandawith the Commentary Ambakartri. Sarasvati Bhavana Grantha mala No. 91. Varanasi. RAMAKRISHNA KAVI, M. 1930. "The Discovery of the Author's Vstti on the Vakyapadiya." JAHRS 4.235--241. Rajahmundry. RAU, WILHELM. 1962. "Uber sechs Handschriften des Vakyapadiya." Oriens, 15.374398. Wiesbaden. - 1964. "Handschriften des Vakyapadiya." Oriens, 17.182-198. Wiesbaden. SADHU RAM. 1952. "Bharthari's Date." JGJARI 9.135_151. Allahabad. SANTI-BHIKSU SASTRI. 1963. (Ed. and transl.) "Agamasamuccaya * Alias Vakyapadiya-Brahmakanda of Bhartshari." Wissenschaftliche Zeitschrift der Karl-Marx-Universitat Leipzig 12.1914-228. Gesellschafts und Sprachwissenschaftliche Reihe. Heft 1. Leipzig. SUBRAMANIA IYER, K. A. 1965. (Transl.) The Vakyapadiya of Bhartshari with the Vrtti. Chapter I. Deccan College Building Centenary and Silver Jubilee Series No. 26. Poona. 1966a. (Ed.) Vakyapadiya of Bhartshari with the Vrtti, and the Paddhati of Vrsabhadeva. Deccan College Monograph Series No. 32. Poona. - 1966 b. (Ed. and transl.) The Sphota-siddhi of Mandana Misra. Deccan College Building Centenary Series No. 25. Poona. SWAMINATHAN, V. 1963. "Bhartshari's Authorship of the Commentary on the Mahabhasya." ALB 27.59-70. Madras. THIEME, PAUL. 1956. "Panini and the Paniniyas." JAOS 76.1--23. New Haven. Trikandi. (a) Kanda 1: see $. IYER 1966 a. (b) Kanda 2: see MANAVALLI 1887. The numbering of the Vakyapadiya verses in this edition is quite often inaccurate. My references to the second kanda assume a correction of those inaccuracies. Unfortunately the V of the second kanda is yet to be published in its entirety. I have, therefore, been forced to refer to my hand-written copy of that V. See AKLUJKAR 1969: 555-556. (c) Kanda 3: see Helaraja. Tripadi. See ABHYANKAR-LIMAYE 1967-1970. Vrsabha. See S. IYER 1966 a. Vstti. See S. IYER 1966 a, C. SHASTRI 1941 (?), and Trikandi. YUDHISTHIRA MIMAMSAKA. samvat 2020. Samskrta Vyakarana-sastra ka Itihasa. Vol. I. Revised edition. Bharatiya Pracya-vidya Pratisthana. Ajmer.