Nakamura On Bhartrhari

Author(s): Ashok Aklujkar
Publisher: Ashok Aklujkar
Catalog link: https://jainqq.org/explore/269330/1

Page #1
NAKAMURA ON BHARTRHARI by ASHOK AKLUJKAR Vancouver 1.1 An article by Professor Hajime Nakamura, "Bhartshari The Scholar", was published in the fourth volume of the Indo-Iranian Journal (1960: 282-305). That article was a revised translation of a part of Nakamura's * Kotoba no Keijijogaku, which is regarded by many scholars to be Naka mura's important contribution to the study of Vedanta in general and to Bhartshari studies in particular. Naturally I was very surprised to find in it, as I shall presently demonstrate, a large number of inaccurate translations, remarks, conclusions, and comparisons. My purpose in demonstrating what I consider to be Nakamura's mistakes is, of course, purely that of sastra-suddhi "purification of a branch of learning"; the positive aspects of the present article, namely the correct translations of some of the key verses in Bhartphari's Trikandit (TK in abbreviation) and a correct understanding of Bhartshari's position, are more important in my view than the refutation of the contents of Nakamura's article.2 1 (a) It is generally believed that the title of the work to which I refer as the Trikandi is Vakyapadiya. In a recent article (Aklujkar, 1969: 547-555), I have argued that Vak yapadiya was originally the title of only the first two books of Bhartshari's magnum opus and that Trikandi is the only ancient name that can refer to the work under study as a whole. (b) Some scholars advocate the view that the composition referred to as Vrtti (as V in abbreviation) is not Bhartphari's work and that it is much later than the verses (karika) which alone form the genuine TK. I see absolutely no reason to subscribe to this view. In my paper, "Authorship of the Vak yapadiya-yrtti", read at the annual meeting of the American Oriental Society (1969) (to appear in Wiener Zeit schrift fur die Kunde Sud- und Ostasiens, vol. xvi, May 1972), I have exposed the weaknesses of the arguments on which this view is based, and I have shown with unmistakable internal evidence that the traditional ascription of the V to Bhartshari cannot be doubted by any unprejudiced mind. 2 (a) My practice in the following is to quote Nakamura's words first, then to reproduce the Sanskrit original, and then to translate the Sanskrit original as I think it should be translated. In general, I do not adduce any additional evidence or refer to the ancient commentators when it is quite obvious that my translation follows the Sanskrit original more closely than Nakamura's. It should also be noted that I italicize all Sanskrit words for the sake of uniformity. Nakamura's practice, which is justifiable

Page #2
162 ASHOK AKLUJKAR 2.1 Incorrect translations: P. 283.16-17, fn. 5: "... so does the smrti[go on forever without interruption] though it was composed by scholars." sistair nibadhyamana tu na vyavacchidyate smstih // (TK 1.146/1.136cd): "But the smrti is not lost (literally, cut off) if it is composed [again and again) by the elite." The verse (karika) expresses the view of the Mimamsakas (Punyaraja (? see fn. 2 above), p. 58.4-8; V pp. 223.6-224.4, p. 226.13; cf. 5.1 below). According to them, the sruti is eternal and hence it does not require the assistance of the elite for its survival, but the smrti is preserved only by the elite. 2.2 P. 283.17-18, fn. 6: "The words of the sruti were composed a long, long time ago." sastram atidure vyavasthitam // (TK 3.14.74, p. 487/p. 39). Here we come across a classic instance of how Nakamura takes Bhartshari's words out of context (cf. 2.6, 12, 18 below). The entire verse runs thus: sabda yatha vibhajyante bhagair iva vikalpitaih, anvakhyeyas tatha, sastram atidure vyavasthitam // It refers to Bhartshari's well-emphasized thesis that the segments of linguistic units (sentences, words, and phonemes, which correspond respectively to Sanskrit vakya, pada, and varna) set up in grammar are convenient fictions, and that, for this reason, the science from a different point of view, is not to italicize those words which, although of Sanskrit origin, have now become a common stock of the English language. (b) Although it has been convincingly argued as early as 1930 (Charudeva Shastri 1930: 636-644, 1934: Skt. Intro. pp. 18-26; Ramakrishna Kavi 1930: 235-241; Kunhan Raja 1936: 285-298), that the Brahma-kanda commentary published in the Benares Sanskrit Series (Nos. 11 19, and 24; 1887) and ascribed to Punyaraja is an abridgement of Bharthari's V (see fn. 1b above) on the karikas of the first book of the Vakyapadiya, Nakamura refers to that commentary as Punyaraja's work. He also seems to have not known the two volumes of Helaraja's commentary Prakirnaka-prakasa published under the editorship respectively of Samba-siva Sastri and Ravi Varma in 1935 and 1942 (see the bibliography at the end for more particulars). Those volumes give a better text of the Prakirnaka-prakasa than the volume published in the Benares Sanskrit Series (Nos. 95, 102, 130, 160-164; 1905-37), which Nakamura has used. However, it should be noted that my criticism of Nakamura's article is entirely independent of the literature that could have become available to Nakamura but was obviously unavailable. If I refer to the editions later than the ones used by him, it is only to facilitate the locating of almost identical passages in the different editions. 3 In the place of Nakamura's abbreviation Vakyap. (for Vakyapadiya), I use the abbreviation TK (for Trikandi, see fn. 1 above). Also, I change his Roman numerals to Arabic numerals, thus writing 1 for I and 2 for II. References according to more recent editions (see the bibliography at the end) are given after the references according to Nakamura are specified. 4 (a) pade na varna vidyante varnesy avayava na ca / vak yat padanam atyantam praviveko na kascana // (TK 1.73) "There are no phonemes in a word, and there are no parts in phonemes. There is absolutely no separate existence whatsoever of the words from the sentence." iha nirbhagesy apurvaparesy abhedyesu varna-pada-vak yesu ... (V 1.85 p. 151.4); V 1.88 pp. 154.5-155.2; TK 2.13a and V; TK 2.30 and V. For morreferences to the statements of this point in the TK and its commentaries, see not.

Page #3
NAKAMURA ON BHARTRHARI 163 of grammar is far removed from the permanent, unitary linguistic units, which constitute the linguistic reality.5 As is obvious from the context as well as Helaraja's crystal-clear comments, the verse has nothing to do with the sruti or the Veda. It should be translated as follows: "Linguistic units should be derived in grammar in accordance with how they are segmented by the imagined parts, (for) the science (of grammar] remains far away [from the linguistic units]." 2.3 Pp. 283.30-284.4, fn. 8: "... the composition of the smrti writings are multifarious: some are clear as to their aim and purpose (for instance the writings on medicine); some are vague and difficult to follow (as for example the writings on what foods should and should not be eaten). But all the smrti writings were propounded by the Vedic scholars (Vedavid) following the special characteristics (linga) of the Vedas." smstayo bahurupas ca drstadsta-prayojanah / tam evasritya lingebhyo veda-vidbhih prakasitah | (TK 1.7). Here drsta-prayojana and adrsta-prayojana should have been translated respectively as "having mundane purposes" and "having non-mundane (otherworldly) purposes", and linga as "indications"; see V 1.7, pp. 28.4-29.4; Visabha, p. 28.16-17, 22-23; V 1.11, p. 40.6; Tripadi,6 p.35.20. It is a thesis accepted by many schools of Indian philosophy that the smrtis must have some supporting indications in the Vedas in order to qualify as authoritative sources of knowledge; cf. V 1.7, p. 30.1-2. 2.4 P. 284.16-20, fns. 16-17; "... the sruti perishes only to grow like 4.19b-d of my Harvard dissertation referred to in the bibliography. (b) The following passages establish that, on the level of analysis, Bhartphari assumes the sentence to be made of words and that, on that level, he thinks of the word in itself as identical with its counterpart in a sentence: V 1.24-26 p. 65.4-5; Vrsabha, p. 79.21-22; TK 2.10cd; BSS p. 91.14-15, p. 132.17-20, p. 143.19-21; vyaptimams ca laghus caiva vyavaharah padasrayah / loke sastre ca karyartham vibhagenaiva kalpitah//(TK 2.344) "The linguistic activity centering around words, which allows generalization and achieves economy (of effort), is brought about in ordinary life and in the science (of grammar) for the sake of the [desired] purpose only through the analysis (of sentences]." BSS pp. 167.15168.3, p. 186.17; TK 3.1.1, Hela. p. 2.3. (c) For a similar consideration applied to the word and the sub-word units, see yatha pade vibhajyante prakyti-pratyayadayah / (TK 2.10ab); V 2.164-165; TK 3.1.1d, Hela. p. 2.3-4, 7-8. 5 TK 2.226-238 and V; especially, sastresu prakriya-bhedair avidyaivopavarnyatel anagama-vikalpa tu svayam vidyopavartate // (TK 2.233) "With different analyses and derivations, it is only avidya (knowledge of a unitary entity as non-unitary) that is put forth in (all) sciences; vidya, which is free of agama (see 3.2 below) and intellectal constructs, approaches on its own (that is, is not acquired as a direct product of the study of sciences)." 6 Modern scholars generally give the name of Bhartshari's commentary on the Mahabhasya as Mahabhasya-dipika. In a forthcoming article, I shall argue that the evidence for this name is very scanty and that the commentary should be referred to by the name Tripadi (The Adyar Library Bulletin, 1971).

Page #4
164 ASHOK AKLUJKAR seeds (bija) out of the Vedas again. It is our fixed belief that the agama was produced out of some source or other. When all the agama [sic] have perished, the three Vedas will remain like seeds to give new birth to the agama."? tani[=veda-vakyani]purvagamesu vicchinnesu anyesu pranetrsu agamantaranusamdhane bijavad avatisthante. (Punyaraja(?) on TK 1.134/ 1.124, pp. 51.20-52.1) na jaty akartTkam kascid agamam pratipadyate bijam sarvagamapaye trayy evadauo vyavasthita // (TK 1.134/1.124). The correct translation of Punyaraja's (?) remark is: "After the earlier schools of thought (agama; see 3.2 below) have perished, they (the sentences of the Veda) remain like seeds for the bringing into currency of new schools of thought on the part of different (new) proponents (of schools]." Also, verse 1.134 should be translated thus: "No one (no proponent of a school of thought) is ever of the view that (his) school is non-manmade (eternal.) In the beginning (see fn. 9 above) when all schools of thought have perished, only the Vedic Triad remains as a seed (for the future schools of thought]." 2.5 P. 286.17-20, fn. 22: "Therefore the excellent scholars began to teach the significance of words according to the Vedas which were not created by man, and, according to the smrti, which laid down the rules (on the usage of words)." tasmad akstakam sastram smrtim valo sanibandhanam / asrityarabhyate sistaih sabdanam anusasanam || (TK 1.43) "Therefore the instruction of linguistic units is begun by the elite by depending on the non-manmade science (sruti) and (see fn. 10 above) the smoti which has support [in the sruti or in the conduct of the elite]." Cf. Visabha, p. 99.19-22. 2.6 P. 286.29-30, fn. 25: ... [grammar is] the highest among the studies of the three Vedas which are made up of many branches of study." aneka-tirtha-bhedayas trayya vacah param padam || (TK 1.144/1.134). If Nakamura had taken into consideration the first line of this verse, the word vacah, or Punyaraja's(?) commentary, he would have noticed that the verse does not say anything about the three Vedas. What Bhartshari here speaks about are the three levels, aspects, or phases of language ? It is impossible to reconcile these lines of Nakamura's article with the remarks in the preceding lines of the same paragraph. In particular, it is not clear what distinction of reference he sees between sruti and Veda, and how he separates agama from the three Vedas once (p. 284.11-13) he has remarked that agama includes sruti. 8 Subramania Iyer's edition of the unabridged V (see fn. 2b above) has the following reading: tany agamantaranam pranetTsu vicchinnesy agamantaranusamdhane bijavad anutisthante. However, the purport of the passage is not affected by this reading. * The reading of Subramania Iyer's edition is evato. 10 Vrsabha reads ca, which Subramania Iyer rightly accepts as the correct reading.

Page #5
NAKAMURA ON BHARTRHARI 165 (vak), namely vaikhari, madhyama, and pasyanti. Each of these levels becomes different in numerous (aneka; in fact, innumerable; see V p. 214.2, p. 215.3, p. 220.4) ways because of the difference in loci (tirtha). The science of grammar is the great abode of such three-fold language. 2.7 P. 287.1-3, fn. 26: "As the species of the meanings of all words are subject to their forms (akrti), the study of grammar is the original source of all studies (vidya) in the world." yathartha-jatayah sarvah sabdakrti-nibandhanah | tathaiva loke vidyanam esa vidya parayanam || (TK 1.15) "Just as all meaning-universals are tied to the universals of linguistic forms, so also this science [grammar] is the ultimate path (or merging point) of [all] sciences in the world." In certain contexts Bhartrhari does not distinguish between jati and akrti; cf. Tripadi p. 3.17-20, p. 17.21-22, p. 18.11-25; V 1.23 pp. 52.1-57.2, p. 58.5-7; Sreekrishna Sarma, 1957: 56-57. For the point of the first line in this verse, see TK 3.1.6. It is also important to note that Bhartrhari does not assign the highest position to grammar for any partisan or sectarian reasons; grammar is the science of sciences for the simple reason that exact expression and correct understanding depend on it. See Vrsabha, p. 49.24; V pp. 48.649.5; Punyaraja (?), p. 8.14-17. 2.8 P. 287.3-5, fns. 27-28 (cp. p. 288.17-22): "It is the foundation on which all other studies are based and can be likened to a purifying fire." "Being a purifying fire to all study, it shines brightly over all bracnhes of learning." pavitram sarva-vidyanam adhividyam prakasate || (TK1.14cd) "A purifier (remover of errors) of all lores [through the teaching of what expressions are best suited for the intended meanings], it (the lore of grammar) shines at the top of [all] lores." See conclusion of 2.7 above. 2.9 P. 288.11-12, fn. 35: "... The rules of this sacred study have been handed down in an unbroken chain to the educated." avicchedena sistanam idam smrti-nibandhanam || (TK 1.143/1.133). This passage can be translated in three different ways, but certainly not in the way in which Nakamura translates it. From the words etad and tad in TK 1.144/1.134 and 1.145/1.135, which stand for vyakarana, it is clear (cf. Vrsabha, p. 212.20) that the word idam in this passage also refers to vyakarana. The three possible and justifiable translations, therefore, are: (a) "Because of the unbroken chain of the elite, this (grammar) has its roots in the preserved [reliable] knowledge." sistanam avicchedena idam smrti-nibandhanam. Cf. V. p. 213.1-2. (b) "For the elite, this (grammar) is always a bond [called] smrti." sistanam idam avicchedena smrtinibandhanam (=smrti-rupam nibandhanam). Cf. V p. 212.8-9. (c) "This composing of the smrti on the part of the elite is without break (that is,

Page #6
166 ASHOK AKLUJKAR the elite compose the smrti called vyakarana again and again)." sistanam idam smsti-nibandhanam avicchedena. Cf. V p. 213.1. 2.10 P. 289.10-11, fn. 42: "The knowledge of the inner self cleans dirt from the body, from speech, from the mind and can be compared to a healing process." kaya-vag-buddhi-visaya ye malah samavasthita) / cikitsa-laksanadhyatma-sastrais tesam vieuddhayah || (TK 1.148/1.138) "Those which are impurities of the body, the speech, and the intellect are removed by the sciences (respectively) of medicine, grammar, and yoga." Nakamura has somehow failed to notice the yathasamkhya "respectively" construction in the verse. He also seems to be unaware of the fact that Sanskrit authors many times refer to the system of their present concern with the word laksana; cf. V 1.12 p. 43.6. 2.11 P. 290.17-19, fn. 45: "Various divine powers such as demons, pitys and vampires which exist outside the realm of perception and inference (employed by the ordinary man) are the result of certain processes such as asceticism." pratyaksam anumanam ca vyatikramya vyavasthitah / raksah-pit?-pisacanam karmaja eva siddhayah // (TK 1.36) "The extraordinary capacities of the demons, manes, and goblins, [capacities] which remain beyond perception and inference, can only be results of actions [in the previous lives)." The verse is intended, as is obvious from the verses 1.35, 37, 38, and 39, to point out that there can be extraordinary means of knowledge, means which are not exhausted by perception, inference, and testimony; cf. Vrsabha, p. 93.20. Its context and wording do not in any way support Nakamura's (p. 290.19-23) interpretation of it. 2.12 P. 290.23-26, fn. 46: "There is no ordinary linga to prove the existence of the universal and the form, the essence and the species. Therefore some other source (i.e. that of the sacred books) is used." samanyam akstir bhavo jatir ity atra laukikam / lingam na sambhavaty eva tenanyat parigrhyate // (TK 3.14.323 p. 605 / 3.14.320 p. 143). Here we come across one more (see 2.2 above) instance of how Nakamura interprets the TK verses out of context. The present verse is intended to restate Patanjali's (on Panini 4.1.3) observation that the grammarians cannot afford to confuse the ordinary meaning "sex" of the word linga with the meaning "gender", that they cannot accept determination of gender by sex. As an illustration of this point, attention is drawn to the words samanya, akyti, bhava, and jati. All these words denote the universal, but samanya is a neuter word, bhava a masculine, and akrti a feminine. As one and the same entity, which is the universal in this instance, cannot actually possess mutually contrary properties resembling the sex distinc

Page #7
NAKAMURA ON BHARTkHARI 167 tions, it is clear that the genders of words are not related to the sex (laukika linga) distinctions. The grammarians, therefore, accept alaukika (anyat) linga; that is, they understand the word linga in a specific, technical sense as denotative of some non-actual distinctions expressed by the words. 2.13 P. 291.1-2, fn. 47: "... the meaning of things is indeed determined by the sacred scriptures." taya hy artho vidhiyate // (TK 1.130d/1.120d). Here Nakamura overlooks the fact that sruti can mean "sound", "word", or "formal aspect of a linguistic unit as opposed to its semantic aspect", 11 and that artha can mean "entity" or "thing" as in 2.17 below. What Bhartshari points out in this line is that words give rise to entities lacking physical existence (e.g. sasa-visana or Pegasus), that words shape, fashion, or determine entities. This is clear even from Punyaraja's (?) commentary (p. 50.1-4) which Nakamura claims to have followed here. 2.14 P. 291.18-22, fn. 52: "Therefore as the spirit (in the doctrine of subjectivity) is not harmed by anything), so will those who follow the ways of agama handed down to us without interruption meet with no refutation from the schools of logic (hetu-vada)." caitanyam iva yas cayam avicchedena vartate / agamas tam upasino hetu-vadair na badhyate || (TK 1.41) "One who adheres to this agama (see 3.2 below) which exists without break like sentience, is not moved from his position by the talk of [logical] reasons as in the case of sentience." The phrase caitanyam iva serves two purposes in the present verse. It points out that the continuity of agama is like the continuity of sentience, beginningless and natural (V p. 98.3; Vssabha, p. 98.14), and that just as, with the help of reasoning, one cannot talk a man into believing that he does not exist, so also one cannot, with logical reasoning as means, make an agama-advocate give up his reliance on agama (Punyaraja (?), p. 19.5-7; V p. 98.3-8). 2.15 P. 295.20-21, fn. 69: "When a person does not doubt his knowledge even as he does not doubt his senses, how can anyone persuade him?" yo yasya svam iva jnanam darsanam nabhisankate / sthitam pratyaksa-pakse tam katham anyo nivarayet || (TK 1.39) "How can the other person (using reasoning] prevent X [from accepting as truths those matters which lie beyond ordinary cognition] who does not doubt the [extraordinary] perception on the part of Y as if it were his own expe 11 For example, see: V 1.9 p. 37.3; V 1.12 p. 42.2; V 1.34 p. 92.3; V 1.44 p. 101.6, p. 102.4; TK 1.46; TK 1.60 V p. 119.7; V 1.62 p. 120.9; V 1.68-69 p. 131.2, p. 132.3, 5; V 1.74 p. 139.8-9; TK 1.87 V p. 153.8; V 1.92 p. 158.2; V 1.93 p. 159.4; V 1.102 p. 168.3; V 1.107 p. 175.3; V 1.115 p. 189.3; V 1.116 p. 191.2; TK 1.120 V p. 197.4, 7; TK 1.121 V pp. 198.3-199.1; V 1.134 p. 213.6; V 1.143 p. 232.3. See also 3.7 below.

Page #8
168 ASHOK AKLUJKAR rience, and [thus) stands on the side of perception (that is, thinks of the matters being disputed as self-perceived truths if Y says that he cognized them)?" In this verse Bharthari points out that inference or logical reasoning cannot be an effective tool in changing the views of a person who firmly believes that he has received his information from a highly reliable person gifted with extraordinary capacities of cognition. 2.16 P. 298.24-27, fn. 77: "Though all men are able through the agama to understand things the fruit of which is yet to be experienced, what is stated in the agama can be interpreted in the contrary meaning." sarvo 'drsta-phalan arthan agamat pratipadyate / viparitam ca sarvatra sakyate vaktum agame || (TK 1.142/1.132) "All come to know the matters of non-mundane consequences [only] through agama (see 3.2 below). And it is possible to voice opposite views [as regards the consequences) in the case of all agama." Cf. Punyaraja(?), p. 55.4-5; V p. 212.1-4. 2.17 P. 301.8-11, fn. 89: "Therefore the wise observe correctly with the mind even those things which can be directly perceived by the senses. As intuition should be true knowledge, we should not present the things which we perceive." tasmat pratyaksam apy artham vidyan ikseta yuktitah / na darsanasya pramanyad drsyam artham prakalpayet // (TK 2.143) "Therefore, a wise man should examine with reasoning even that thing which is perceptible. He should not form his ideas about a perceptible thing believing in the validity of his perception." 2.18 Fn. 96: "If the words were founded upon the darsana (=pratyaksa) there would be no need for inference. Indeed when the darsana is established why should be need for inference?" na calam anumanaya sabdo darsana-purvakah12 / siddhe hi darsane kim syad anumana-prayojanam // (TK 3.13.12 p. 436 / p. 142). Here again (see 2.2 above) we notice Nakamura's utter disregard of the contexts of Bharthari's remarks. As Helaraja explains, the real purpose of the present verse is to reiterate Patanjali's (on Panini 4.1.3) objection anyonya-samsrayam tv etad directed against those who hold that actual distinctions corresponding to the distinctions in genders exist in the entities (see 2.12 above). A more accurate translation would be as follows: "The word (having a distinction in the form of its gender) is not capable (i.e. does not serve as a valid tool) of inference [that an actual distinction corresponding to the genderdistinction must exist in the denoted entity] if it is preceded by a perception [of that actual distinction]. For what purpose is there for (employing] 12 Samba-siva Sastri's edition accepts the reading sabdo 'darsana-purvakah, that is, with an avagraha or akaralopa in between the two words. But Helaraja's commentary does not seem to support that reading.

Page #9
NAKAMURA ON BHARTkHARI 169 an inference when a perception (of the entity to be inferred] has already] materialized?"13 3.1 Incorrect remarks: P. 284.6: "Frequently however the term sastra refers to the Vedas in a broad sense, ..." In the verses TK 1.43 and 1.137/1.127, to which Nakamura refers as evidence, we find that the words akstaka sastra and veda-sastra denote the Vedas as a phrase and as a compound expression. It is evident then that sastra alone does not stand for the Vedas as far as Bhartshari is concerned. Nakamura's identification of that word with the sruti or Veda results info wrong translations and wrong understanding of Bhartshari's statements on p. 302.4-10. 3.2 P. 284.11-12: "But in the Vakyapadiya the term agama refers to the books of the pure orthodox Brahmanic schools ..." This statement is 'contradicted by Nakamura's footnotes 14 and 15 which show that agama can mean "something which has been handed down from ancient times." and that "it can stand for books such as those of the Vaisesika school." There is no point, therefore, in taking the statement seriously, but I may seize the opportunity provided by it to explain Bharthari's concept of agama.14 A study of the occurrences of the word agama shows that it possesses a host of related meanings which can be divided into at least three strata or groups. The most general meaning can be said to be "inherited lore", "conceptual inheritance", "knowledge that one inherits because of birth in a particular biological species or linguistic community", or "lore of the forefathers".15 Next, in those contexts in which awareness and preservation of knowledge is assumed, come the meanings: "tradition", 16"traditionally inherited body of knowledge",17 "traditional instruction of 13 This section concludes the demonstration of the eighteen glaring mistakes of translation that are found in Nakamura's article. Of relatively less importance are the following errors: (a) dharma = laws (p. 291.24). In the context of TK 1.30, dharma means "ritually beneficial action", "conduct leading to favorable consequences", or "merit, punya". (b) rupa = colour; tangible, material contents of the sacred books (p. 299.12-16, fn. 80). In the context of TK 1.137/1.127, the word rupa quite obviously means "form (of a word or expression)". (c) apta-vacana = Intuition of the Brahman (p. 302.29). I do not know how Nakamura derives this meaning from the etymological and generally accepted meaning "testimony, statement(s) of a trustworthy person". 14 The following exposition is in all essentials a reproduction of $ 0.13 of my Harvard dissertation referred to in the bibliography. 15 TK 2.118ab, 151. 16 sistanam yad idam avicchinna-paramparyam smaranam sa agamah. (Vrsabha, p. 81.19-20); V 1.27 p. 82.1, 4. 17 TK 1.30 V p. 86.2, 4.

Page #10
170 ASHOK AKLUJKAR conduct", 18 "body of knowledge and views preserved by the adherents of a system",19 "traditionally preserved view", 20 "sources of testimony", "authoritative literature", 21 and "imbibed or internalized results of the activity of systematization". 22 Most particularly, in sufficiently transparent contexts, agama denotes the sruti and smrtis.23 3.3 P. 287.6-7: "Sometimes it (the grammar] is likened to a light in the darkness (tamasi jyotis)." Obviously Nakamura has in mind TK 1.18cd: yad asminn eva tamasi jyotih suddham vivartate //. But from TK 1.17-22, it is evident that the line describes brahman, not vyakarana. Cf. Vrsabha, pp. 50.19-51.10, p. 51.21-24. - 3.4 P. 288.9-10, fn. 34: "According to Bhartshari the sacred study of grammar aimed at continuity (nitya) and correctness of language." TK 1.29, to which Nakamura here refers as evidence in support of his remark, runs thus: nanarthikam iman kascid vyavastham kartum arhati | tasman nibadhyate nitya24 sadhutva-visaya smrtih //. How can nitya be translated as "continuity" or be construed as an adjective of language? It is obviously an adjective qualifying smrti. 3.5 P. 288.22-24, fn. 38: "Bhartshari illustrates this point that there cannot be an understanding of the real nature of words without grammar] by means of the simile of the mirror which reflects the images of various objects." As pointed out in 3.3 above, TK 1.20 is about brahman, not grammar. 3.6 P. 288.30-32, p. 301.23 (cp. p. 301.30-31): "... [Bhartshari) came to the conclusion that words were component parts of the Brahman and therefore on the same level as it." This is a gross misrepresentation of Bhartshari's metaphysics, but considerations of space prevent me from including here a proof to that effect. For my understanding of Bhartshari's position, I refer the reader to my Harvard dissertation (1970). 3.7 P. 292.4-6, fn. 56: "He [Bhartshari] holds, too, that something which may be quite difficult to exist in actual life must nevertheless be acknowledged if it is laid down in the sacred agama." atyantam atatha-bhute nimitte sruty-upasrayat drsyate 'lata-cakradau vastv-akara-nirupana // 18 anadir agrhyamana-karana acaropadesa agama ity agama-laksanam. (Vrsabha, p. 82.15). 19 TK 1.26d; V 1.74 p. 139.7; TK 1.124 V p. 203.7-9; V 1.125 p. 204.3-4; TK 2.482484, 486-487. 20 V 1.108 p. 178.1; TK 3.7.164ab; TK 3.8.23. 21 V1.11 p. 41.1; V 1.127 p. 205.6, p. 206.1-2; V 1.130 p. 210.5; TK 1.132 V p. 212.3-4. 22 V 1.24-26 p. 65.3; V 2.235; TK 3.11.6, 9. 28 paramparyenavicchinna upadesa agamah sruti-laksanah smrti-laksanas ca. (Tripadi p. 8.17-18); TK 1.41 V p. 98.5. 24 The correct reading is sistaih. See V 1.29 pp. 84.6-85.3; Vtsabha, p. 84.17-25.

Page #11
NAKAMURA ON BHARTkHARI 171 (TK 1.131/1.121). Here again Nakamura misses the contextually appropriate meaning of the word sruti (see 2.13, fn. 11 above). The verse means: "[Even] when the cause [of the coming into currency of a denotative form, that is, the referent] is of an altogether different nature (that is, does not possess the nature of a spatially continuous thing) (then also), because of the (very] recourse to the denotative form, a cognition of the form of a thing is found in the case of "firebrand-wheel", etc."25 4.1 Incorrect conclusions: Although the central theme of Nakamura's article is to show how Bhartshari preferred testimony to logical reasoning (see 4.2 below), he drops hints throughout his article to the effect that somehow Bhartshari's advocacy of agama was half-hearted and superficial (p. 282.16-21, p. 285.33-34, p. 290.9-10, p. 299.22-24, p. 300.1-3, 5-7, p. 302.13-16). This is not only contradictory but also lacks justification in the form of supporting evidence from Bhartshari's writings. 4.2 Nakamura (p. 290.11-13, 32-33, p. 291.9-11, p. 292.2-10, p. 295.2-6) depicts Bhartrhari as a philosopher who had declared a total war against inference or logical reasoning. From the preceding sections it is clear that Nakamura's inability to interpret Bhartshari's remarks accurately has led him to this wrong view of a pioneer Indian philosopher.26 What Bhartshari actually does27 is to draw attention to the lack of guarantee about the validity of an inferential cognition and to point out the limitations of inference. An inferential cognition may turn out to be true or false. 28 Its unreliability is borne out by the perpetual differences of opinion among philosophers.29 The fact that inference is governed by three factors - perception,30 inherited lore, 31 and language32 - provides many 25 Besides the incorrect remarks pointed out in these sections, Nakamura's article contains a number of unsubstantiated remarks. For example, he does not furnish any evidence in support of the statements appearing at the following places: pp. 285.35286.2; p. 291.12-15; pp. 296.28-297.25. 26 Nakamura (p. 287.1-16; p. 298.8-15; p. 302.21-24; p. 303.3-4) criticizes Bhartshari for refuting inference with the help of inference. As there is no evidence of any wholesale refutation of inference by Bhartphari, it follows that this criticism is unwarranted. 27 The following portion of the present section is reproduced from $8 1.8-11 of my Harvard dissertation referred to in the bibliography. 28 ko hy anavasthita-sadharmya-vaidharmyesu nityam alabdha-niscayesu purusa-tarkesu visvasah. (V 1.30 p. 86.2-3) "How can one trust the inferences drawn by [ordinary] persons -- [inferences) in which the similarity and dissimilarity (of entities) are uncertain and about which one always lacks certitude?" V 1.38 p. 96.3. 29 yatnenanumito'py arthah kusalair anumatsbhih / abhiyuktatarair anyair anyathaivopapadyate // (TK 1.34) "A point inferred with effort by skilled logicians is accounted for quite differently by other [logicians) of still greater skill." V pp. 90.9-92.3; BSS p. 285.5-6. 30 How perception figures in inference (fn. 31a below) and how it can occasionally

Page #12
172 ASHOK AKLUJKAR grounds for possible error; none of these factors assures a valid knowledge of reality. Moreover, it is often found that an event or result predicted through inference does not take place because of some change in the environment.33 Further, there are branches and stages of philosophical inquiry in which the unrestricted validity of inference cannot be accepted. In deciding non-mundane matters like whether a specific kind of knowledge or conduct causes merit or demerit, inference cannot be given a blank cheque. Such questions arise if at all one is ready to accept the authority of some scriptural lore; if one's logical instincts forbid one to believe in any tradition, these problems do not exist.34 Once, however, lead to an invalid inference is pointed out in the following remark: ... pakadyanumanarthani kriya-visesesu pratiniyatany angani vipralambharthany api kaiscit kathancid upadiyante (V 1.34 p. 92.4-5) "Some people somehow take to (start the performance of) those components (of actions] which are regularly related to specific actions and which are intended for (to lead someone to the inference of actions like cooking even for the purpose of deceiving (someone)". In the light of the explanation given by Vrsabha (p. 92.24-27), the purport of this remark is as follows: To deceive someone, say, a mendicant, one may start performing actions which are usually performed in the process of cooking. Thinking that he will get a sumptuous meal, the mendicant may stay around only to find that he has been cleverly deceived. Thus, his inference based on perception makes him a laughing stock. 31 (a) ... tad-avirodhad evanumana-siddhih. yad vanumite 'pi tavan na nisciyate yavat pratyaksagamabhyam na pratitam iti (Vrsabha, p. 92.23-24) "An inference becomes valid only when it does not conflict with them (namely, perception and agama; for the latter see 3.2 above). Or, even when [an entity] is inferred, a definite conclusion (concerning it] is not reached as long as [it] is not borne out by perception and agama." (b) It seems that, according to Bhartshari, one acquires most of the agama without one's noticing it. In its most general sense, "inherited lore" (see 3.2 above) agama pervades language entirely; in fact, it is nothing but language understood in an abstract sense, as a totality of signs and sign-combinations, as pasyanti; and, in its particular sense, "tradition" or "traditional thought of a philosophical school", it affects language considerably (V 1.24-26 p. 65.3-4; V 1.121 p. 199.3-4; Visabha, p. 199.16-17; TK 2.134-137; BSS pp. 138.8-139.11; TK 2.235-237 and V, TK 3.11.6, 9). As the first language that one learns, at least, is internalized without one's being conscious of it, so also are the inherited lore and traditional thought contained in it. Inferences of logicians, therefore, cannot be totally independent of either. 82 (a) Bhartphari (TK 1.129 V p. 209.2-3) points out that inference means determining what a linguistic unit can mean in the case of those thinkers who use inference only to interpret scriptural lore. (b) For the influence of language on inferences, see fn. 31b above. 88 avastha-desa-kalanam bhedad bhinnasu saktisu / bhavanam anumanena prasiddhir atidurlabha || (TK 1.32) "It is extremely difficult to establish [the existence, or nature, of] things on the strength of inference, as the capacities (of the things) become different in different states, regions, and times." V pp. 88.5-89.6; TK 1.33 V p. 90.3-6. 84 V 1.27 p. 82.3-4; nagamad ste dharmas tarkena vyavatisphate / (TK 1.30b) "Merit (favorable non-mundane result) cannot be determined with logical reasoning, without agama." adrstarthanam karmanam phala-niyame sva-bhava-samvid agama-pratibaddha (V 1.30 p. 86.1-2) "In determining the fruits of actions intended for non-mundane results, the knowledge of the nature (of the actions, that is, the knowledge whether the

Page #13
NAKAMURA ON BHARTRHARI 173 some authority is accepted, inference can properly be used only for the purposes of interpretation and justification; it cannot be used to oppose or uproot the authoritative source. It is always possible to deduce through some inference a conclusion contradictory or detrimental to that of a scripture, 35 but so doing, one contradicts one's own initial acceptance of the validity of the scripture. For this reason, in transcendental matters, inference is to be restricted to the interpretation and justification of the accepted authority.36 Just as it is incapable of contradicting scriptural lore, for the two do not belong to the same sphere, in the same way it is incapable of contradicting or invalidating a cognition derived from extraordinary perception.37 Furthermore, it cannot be used to question the nature of the final postulates of a theory or system. There is a limit to which one can answer the question, "Why is it so?" At some level, it is inevitable that one will be forced to say, "Well, this is the very nature of the thing(s); it cannot be otherwise." We cannot go on questioning the postulates of a theory indefinitely with a "why". If, for example, an atomist (Vaisesika) were asked, "Why is it that the smallest particles combine in a particular way?" he would point out that specific particles attract certain other particles only. But this is just another way of phrasing the point in question; it does not answer the "why". 38 5.1 Incorrect comparisons: As I have demonstrated so far, there are serious flaws in Nakamura's understanding and statement of Bhartshari's views. Consequently, his comparisons are of little value. I shall touch upon only one. Nakamura (p. 283.25-27) remarks that Bhartphari's position on the eternality of the Vedas is "precisely the same as" that of the Mimamsakas. But in fact there is an important difference. According to Bhartshari, after the dissolution of the universe, the Veda remains in a actions are capable of non-mundane results and, if they are capable, what type of non-mundane result they produce) is tied to (is derived only from) agama." TK 1.132ab. 35 See TK 1.142cd/132cd in 2:16 above. 36 (a) tasmad agamam kancit pramanikrtya vyavasthite tasmin ya kacid upapattir ucyamana pratipattavi upodbalakatvam labhate (V 1.132 p. 212.4-5) "Therefore, any piece of logical reasoning which is stated after an agama has been accepted and well established, serves to strengthen one's understanding (or acceptance, or determination to practice the teaching [of the agama)." (b) Bharthari realizes the fact that some persons misinterpret agama intentionally (V 1.31 p. 88.1-2). 87 atindriyan asamvedyan pasyanty arsena caksusa / ye bhavan vacanam tesam nanumanena badhyate // (TK 1.38) "The statement of those who, with their seer's eye, see entities that are beyond senses (and) beyond feeling, is not invalidated by inference." V pp. 95.6-96.5. 88 sarve hi vadino duram api gatva sva-bhavam na vyativartante. (V 1.30 p. 86.1); Vrsabha, p. 86.7-14; TK 3.1.95, Hela. p. 95.11-20; cf. V 1.127 pp. 205.8-206.1.

Page #14
174 ASHOK AKLUJKAR unitary, subtle form which is not different from the language-principle, sabda-tattva-brahman.39 It again appears as a diversity with main and subsidiary divisions after the creation of the universe. The Mimansakas, on the other hand, hold that even the divisions of the Veda are eternal, that it remains with divisions even in the event of the dissolution of the universe. See TK 1.5cd, V. pp. 24.2-26.5; V 1.6 p. 27.8-10; V 1.136 pp. 223.6-224.3; V 1.137 pp. 226.1-227.7. BIBLIOGRAPHY Khw.ca) Tripadi, Idge, Mass. (Unbaris Trikandi". Aklujkar, Ashok, 1969. "Two Textual Studies of Bhartphari", Journal of the American Oriental Society (New Haven), 89.547-563. - 1970. "The Philosophy of Bhartshari's Trikandi". Ph. D. dissertation, Harvard University. Cambridge, Mass. [Unpublished.] Bhartphari. (a) Tripadi. See fn. 6 above. Mahabhasya-dipika of Bhartshari, Part I. (Eds.) K. V. Abhyankar and V. P. Limaye) (= Bhandarkar Oriental Research Institute. Postgraduate and Research Department Series, No. 8) (Poona, 1967). I have received the forms of the second part. The particulars of publication are the same. (b) Trikandi. See fn. 1 above. (i) kanda 1. (Ed.) K. A. Subramania Iyer, Vakyapadiya of Bhartshari with the Vrtti, and the Paddhati of Vrsabhadeva (= Deccan College Monograph Series, No. 32) (Poona, 1966). (ii) kanda 2. (Ed.) Gangadhara Sastri Manavalli, Vakyapadiyam ... Sri-Bharthari- ... viracitam Sri-Punyaraja-krtaprakasakhya-tikayutam (=Benares Sanskrit Series, Nos. 11, 19 and 24) (Benares, 1887). The numbering of the Vak yapadiya verses in this edition is quite often incorrect. My references to the second kanda assume a correction of those inaccuracies. Unfortunately the Vrtti 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 Vrtti. See Aklujkar, 1969: 555-561. (iii) kanda 3. With Helaraja's commentary Prakirnaka-prakasa. Samuddesas 1-7: (Ed.) K. A. Subramania Iyer (= Deccan College Monograph Series, No. 21) (Poona, 1963). Samuddesas 8-13: (Ed.) K. Sambasiva Sastri (=Trivandrum Sanskrit Series, No. CXVI) (Trivandrum, 1935). Samuddesa 14: (Ed.) L. A. Ravi Varma (= University of Travancore Sanskrit Series, No. CXLVIII) (Trivandrum, 1942). BSS. See Bhartshari (b) (ii) above. Charudeva Shastri, 1930. "Bharthari: a Critical Study with Special Reference to the Vakyapadiya and its Commentaries", Proceedings and Transactions of the Fifth Indian Oriental Conference, Vol. I (Lahore), pp. 630-665. - 1934. (Ed.) Vak yapadiyam Prathamam Kandam (Lahore, Ramlal Kapoor Trust). Helaraja. See Bharthari (b) (iii) above. Kunhan Raja, C., 1936. "I-tsing and Bhartphari's Vakyapadiya", S. Krishnaswami Aiyangar Comm. Volume (Madras), pp. 285-298. Punyaraja. See fn. 2b and Bhartphari (b) (ii) above. 89 V1.5 p. 24.2-4 (cp. V 1.134 p. 216.1-2, p. 219.6-7), p. 25.2, p. 26.5; V 1.6 p. 27.8-10; TK 1.9 V p. 37.3-8; TK 1.10a V pp. 38.3-39.8; TK 1.11a V p. 40.1; TK 1.112 V p. 184.2; chandah-sabdena param vak-tattvam aha. (Vrsabha, p. 183.21-22); TK 1.137ab V p. 227.3-4.

Page #15
NAKAMURA ON BHARTkHARI 175 Ramakrishna Kavi, M., 1930. "The Discovery of the Author's Vrtti on the Vak yapa diya", Journal of the Andhra Historical Research Society, 4.235-241. Ravi Varma. See Bharthari (b) (iii) above. Samba-siva Sastri. See Bhartshari (b) (iii) above. Sreekrishna Sarma, E. R., 1957. "The Words Akrti and Jati in the Mahabhasya", The Adyar Library Bulletin (Madras), 21.54-65. Subramania Iyer, K. A. See Bartshari (b) (i) and (iii) above. Trikandi, TK. See Bharthari (b) above. Tripadi. See Bhartshari (a) above. Vrsabha. See Bharthari (b) (i) above. Vrtti, V. See Bhartshari (b) (i) and (ii) above.