Indian Semantic Analysis: The nirvacana Tradition.

SCHARF, PETER M.

Indian Semantic Analysis: The nirvacana Tradition. By EIVIND KAHRS. University of Cambridge Oriental Publications, vol. 55. Cambridge: CAMBRIDGE UNIVERSITY PRESS, 1998. Pp. xv + 302. $69.95.

Kahrs presents the first major contribution to the study of the Nirukta tradition since the publication in 1982 of Vijayapala's edition of the Niruktaslokavarttika. He undertakes a thorough analysis of the syntax of the explanations of words in Yaska's

Nirukta in order to determine precisely what information they were meant to convey and how they conveyed it.

After announcing (p. 9) in chapter one, "An outline of strategies" (pp. 1-12), that he intends to "let the texts speak for themselves" to reveal the nature of "one consistent means of interpretation" called nirvacana, he determines in chapter two, "Nirvacanasastra" (pp. 13-54), that the nirvacana technique explains the reason a nominal word means what it does by referring to the action in which the thing referred to participates (or to which it is somehow related) and frequently also to the specific mode of its participation in that action. He (p. 27) distinguishes nirvacana from the etymology of historical and comparative linguistics, because it is synchronic while the latter is diachronic; from the vyurpatti of Vyakarana, because it is primarily concerned with explanation of meaning while the latter is primarily concerned with the derivation of speech forms (see also pp. 167-68); and from the explanation of the aitihasikas, because it prefers allegorical explanation while the latter traces the reason a wor d means what it does to specific events in past time. In chapter 3, "Praxis: Saiva Kashmir" (pp. 55-97), Kahrs examines the use of non-technical nirvacana analysis of the name bhairava by Abhinavagupta, Ksemaraja, and Jayaratha in Tantric Saiva scriptures from the tenth to thirteenth centuries in Kasmir. They use it as a mode of argumentation against dualists to instill in shared texts and traditions their non-dualist interpretation of Bhairava as the absolute deity, simultaneously both transcendent and immanent (pp. 72-73).

Chapters four and five form the main bulk of the book. In chapter 4, "The Universe of Yaska" (pp. 98-174), Kahrs explores the relation sustained between a stem and its corresponding explanatory expression in the more technical formulations of Yaska, by careful examination of the contents of expressions of various types. He (p. 168) concludes that an explanation in -anat, using the ablative of an action noun, gives the reason that the nominal word means what it does rather than directly stating the root from which the word derives. While such an analysis usually establishes a phonetic link as well, the semantic content is primary (p. 101). Similarly, the expression etas-mad eva, using the ablative of the demonstrative pronoun, refers back to such a reason. An explanation in the form of an analytic sentence followed by iii, he argues, likewise states a reason. In contrast, he determines that explanatory forms in -eh and teh are not ablatives but genitives of the verbal root citation forms in -i and -ti, thereb y correcting the errors of Skold, Sarup, and S. Varma. In an explanation in the form of an analytic sentence followed by iti satah (or satyah, satam), the genitive of the present active participle of the verb as 'be' refers to the denoted object characterized by the expression followed by iti (pp. 161, 167). Kahrs concludes (pp. 158, 171) that the pronoun in the expression iti apy asya bhavati, following the synonym of a previously analyzed word, refers either to the object denoted or to the verbal root from which the form is derived.

Kahrs closes chapter four with the claim that the genitives in the Nirukta are substitutional genitives comparable to what Sanskrit grammarians call a sthanasasthi (p. 173). In chapter 5, "Substitution" (pp. 175-267), he undertakes a thorough explication of Patanjali's Mahabhasya, together with its commentaries, on the sutra introducing the sthanasasthi, Panini 1.1.49, sasthi sthaneyoga. The book ends with a short sixth chapter, "Epilogue" (pp.. 268-79), in which Kahrs summarizes his conclusion that the explanatory expressions used to elucidate words in the Nirukta provide paradigmatic expressions of meaning of which the explained words are substitutes. He attaches a bibliography (pp. 280-94) and an index (pp. 295-302).

The analysis of the syntax of the Nirukta undertaken by Kahrs furthers the understanding of Nirukta syntax. His explication of the Mahabhasya on 1.1.49 is thorough and helpful. He accurately portrays the section's conclusions that such a substi-tutional genitive in Paninian grammar indicates that a speech form occurs in place of another speech form and that this "place" has been explained by commentators as the meaning (artha) of the speech form, the possible occurrence (prasanga) of the speech form (p. 267), or the time (kala) of its occurrence (p. 266). However, the argument he puts forth that the Nirukta's use of the genitive is a substitutional one and that "ultimately all nirvacanas are to be understood according to a substitutional model" (p. 174) draws a long bow. In a separate paper analyzing the syntax of the genitive in the Astadhayi I have shown that such an argument is invalid and misguided. The genitive in Yaska's Nirukta is not comparable to the substitutional genitive in Panini's Astadhyayi. H ence, it is not correct to understand the genitive in the Nirukta to signify that Yaska's explanatory expressions are paradigmatic semantic placeholders replaced by the explained words.

An incongruity presents itself in the comparison of Kuhrs's conclusions concerning the syntax of explanations in iti satah and explanations in iti apy asya bhavati. He concludes that in the former the genitive of the present active participle refers to the denoted object, whereas in the latter the pronoun refers either to the object denoted or to the verbal root from which the form is derived. It seems to me that it would be consistent in all examples of the latter to understand that the pronoun refers to the object denoted, as does the participle in the former, not to the verbal root or original speech form. Hence, the genitive in such cases is not a substitutional genitive used in relation to a synonym in the nominative (a conclusion drawn in support of his unwarranted general conclusion that genitives in the Nirukta are substitutional genitives); rather, the genitive in both expressions iti satah and iti apy asya bhavati exhibits the word-meaning relation, the denoter-denoted relation. In these instances a word such as naman or namadheya in the nominative should be supplied to complete the sense. The speech form to be explained is a designation of the object under discussion referred to by the present active participle sat of the verb 'to be' or the pronoun idam in the genitive. This analysis is suggested by the immediately preceding use of the term namadheya with the genitive in Nirukta 1.6 dyuh iti ahnah namadheyam; dyotate iti satah, and Nirukta 2.7 bhuri iti bahunah namadheyam; prabhavati iti satah, and by the immediately preceding use of the term naman with the understood genitive in the sasthitatpurusa compounds rupanama in Nirukta 2.9 vavrih iti rupanama; vrnoti iti satah and karmanama in Nirukta 2.13 vratam iti karmanama; nivrttikarma varayati iti satah. [1]

The phrase ity apy asya bhavati should be understood to state not the term followed by iti is a synonym of the term previously explained, but rather, that a passage (nigama) using the term followed by iti occurs about the object to which the previously explained term refers. In the only expanded version of the phrase in the Nirukta, Nirukta 4.19

cyavana rsir bhavati, cyavayita stomanam

cyavdnam iti api asya nigamah bhavanti

the genitive of the pronoun idam refers to the person denoted by the name cyavana, not to the name. Yaska informs us that there are passages which refer to that person using the term cyavana also. Durga takes him to be the seer who is the son of Bhrgu and paraphrases the passage, "There are Vedic passages about him, the seer Bhrgu's son, (who is referred to) also thus, "cyavana." (cyavanam iti evam api asya rseh bhrguputrasya chandasi nigamah santi.)

In a slightly different construction in Nirukta 3.11, the genitive of the pronoun idam also refers to the person denoted, in this case Kutsa, not to the name: "The action of slaying also belongs to him because his friend Indra slew Susna" (atha api asya vadhakarma eva bhavati: tatsakhah indrah susnam jaghana iti).

The understanding of these phrases in this manner, with a genitive referring to an object in syntactic connection with a noun which refers to the denoting speech form, is in accordance with the explanation Yaska gives in Nirukta 1.20 that these are the names of this object (etavanty asya sattvasya namadheyani) and with fairly explicit explanations of forms such as in Nirukta 2.15: "This word 'Kastha' is a name of many objects" (kastha ity etad anekasyapi sattvasya nama bhavati).

Notwithstanding the fact that Nirukta and Vyakarana are called separate Vedangas, and without detracting from his excellent characterization of the distinctive traits of Yaska's methodology, its precedents, and subsequent use, it must be noted that Kahrs overstates in chapter 2 the characterization of nirvacana as an independent tradition of interpretation stemming from Yaska. He (p. 13) contrasts the scantily represented tradition of nirvacana stemming from Ygska with the well-represented tradition of Vyakarana stemming from Panini: "In stark contrast to the situation in vyakarana only one basic work of nirvacanasastra has survived." He mentions ten nighantu texts only in a citation to Sayana (p. 29) and excludes the Niruktasamuccaya of Vararuci altogether. Yaska's Nirukta is not a "basic text" of a "Nirvacanasastra" from which a certain tradition of interpretation distinct from Vyakarana develops; it is a commentary initially on the Nighantu texts and, subsequently, on selected Rgvedic passages. Two tradit ions are intertwined in Yaska's work. The Nighantus upon which Yaska comments are the most ancient in a long and full tradition of lexicography described in four hundred pages by Vogel in his contribution to Gonda's history of Indian literature. [2] Yaska's commentary on these stands at the beginning of a rich tradition of commentary upon such texts, including the Ramasrami on the famous Amarakosa, and his subsequent chapters form part of the rich tradition of Vedic commentary. What Kahrs terms nirvacana and seeks to isolate as semantic analysis deriving from Yaska's Nirukta, means literally 'explanation', belongs to all commentary which takes up terminological analysis, and utilizes Yaska, Panini, and others as authorities. Some of the very syntactic formulae used by Yaska occur also in foundational texts of Vyakarana such as the Mahabhasya (as Kahrs mentions on p. 32) and Kasika. Some of the techniques of Vyakarana, such as analysis into roots and affixes, are evident in the Nirukta (as analyzed by Kahrs on pp. 112-18).

It does not seem to me entirely just to peg the difference between Vyakarana and nirvacana on the relative emphasis of speech form versus meaning, as Kahrs does, following Tibetans and Agnihotra (p. 32). The very ground of the analysis of speech forms in nirvacana, and in Vyakarana, as well as in historical and comparative method, is meaning. Semantic conditions are recognized in grammatical texts as the conditions for the introduction of speech forms in the first place, prior to any phonological operations, as Kahrs recognizes in his discussion of the priority (antarangatva) of meaning in his analysis of Patanjali's Mahabhasya on 1.1.49 (pp. 215ff.). The foundational difference between Yaska's Nirukta and Panini's Astadhyayi seems to me to be that Yaska's work is a commentary while Panini's is a systematic treatise. Yaska's work begins from specific speech forms in their own contexts and seeks, and is obliged, to explain them in that context. Panini's work is much more abstract; it seeks to construct a syst ematic analysis of all speech forms, thereby making the system primary. Durga concludes that Nirukta is the explanation of meaning while grammar is principally systematic description (svatantram evedam vidyasthanam arthanirvacanam vyakaranam tu tu laksanapradhanam). [3] The grammarians themselves describe systematic procedure as an essential part of grammar. Katyana characterizes grammar as consisting of both that which is to be described (laksya) and the means of description (laksana). Patanjali explains that that which is to be described is the speech form (sabda) and the means of description is the rule (sutra). [4] Earlier Patanjali emphasizes that only through a systematic procedure can the vast floods of speech forms possibly be described. He states: "A consummate listing is not a suitable procedure for the comprehension of speech. ... How then are the speech forms to be comprehended? A certain means of description (laksana) consisting of general and specific statements is to be undertaken by means of w hich small effort the Vast floods of speech forms can be understood. What is that? General rule and exception." [5]

That citation of Panini's work attains greater prevalence than citation to Yaska's in Sanskrit commentatorial literature, even in lexicography and perhaps even in Rgvedic interpretation, indicates the great authority systematic linguistic analysis held in India. It indicates that the extent to which systematic analysis superseded unsystematic nirvacanas. The prevalence of citation to Yaska's work in the sphere of Rgvedic interpretation stems from its directly commenting on Vedic passages and words, and from its inclusion of forms not analyzed in grammatical texts. This is consistent with Kahrs's analysis of Nirukta 1.15 on pp. 31-32 (tad idam vidyasthanam vyakaranasya kartsnyam svarthasadhakam ca) and his evaluation of Sayana's statement of the purpose of nirvacana on p. 33. It is the complement of Vyakarana in that it explains words not analyzed by Vyakarana and it accomplishes its own purpose in that it serves to give understanding of the portions of Rgvedic text it comments upon. Yaska states this last as the purpose of the Nirukta in 1.20: "Latter sages, tired of teaching, composed this book in order that subsequent Vedic scholars would be able to comprehend certain passages (upadesaya glayanto 'vare bilmaghahanayemam grantham samamnasisuh).

In the case of a nirvacana explanation of words whose analysis is accounted for systematicaly in grammatical treatises, it is difficult to determine whether a mutually consistent analysis indicates that the grammar later systematized the nirvacana or the nirvacana refers to the systematic grammatical analysis and whether an inconsistent analysis indicates that the grammatical analysis supersedes the nirvacana or the nirvacana disagrees with and deliberately ignores the systematic analysis. Nevertheless, a thorough examination of these cases might contribute to determining the relative dates of Yaska and Panini. This issue is complicated by the fact that the treatises of both are full of interpolations.

On the other hand, in cases such as the Saiva nirvacanas of the term bhairava, where, despite being aware of the Paninian analysis of the term, the authors give several competing explanations, it is clear that the purposes of the nirvacana are something quite different from the elucidation of the text. While one with an eye solely to correct derivation might characterize their enterprise as semantically obfuscating rather than "semantically creative" (p. 6), even without ascribing to their linguistic analysis, understanding their method of discourse permits one to comprehend better what they intended to communicate, in this instance, what these Saivites considered the nature of Bhairava to be. Kahrs rightly validates investigation of their method of explanation and himself contributes extensively to the understanding of it in this work.

Kahrs makes several errors in evaluating what he calls "karaka-analysis." For example, he (pp. 82-83) states that Bhairava is given the role of direct object (karman) in the nirvacana analysis of his name bhijanitad ravaj jatah. On the contrary, Bhairava would be the agent (kartr) of the action of being born denoted by the intransitive root jan in relation to which the roar (rava) is the source (apadana). He states that Bhairava is given the karaka-role of agent (akrtr) in the nirvacana analysis bhani naksatrantrayatiti bherah kalas tam vayanti (sosayanti) iti bhervah (kalagrasarasika ye yoginas) tesam svami, where in fact he is the lord in a master-disciple relation relation, a non-karaka relation signified by a genitive (sesasasthi).

Kahrs's bibliograpy is dominated by Western scholars, to the absence of some important relevant works and a few recent titles. Tokunaga's new edition of the Brhaddevata would have been particularly useful in settling questions of citation in his discussion on pp. 20-21. Tokunaga concluded on p. xx that the Brhaddevata is best preserved in the shorter recension and was originally divided into eight chapters of equal length (127-28 verses). Reference to my own work would have spared Kahrs (pp. 29-47) from repeating the scholarly confusion regarding the lucid concepts of meaning and reference in Indian linguistic philosophy. His conclusions, "let it suffice to establish that the term artha is as complex an entity as its English counterpart 'meaning'" (p. 47) and "the distinction between meaning and thing meant is extremely blurred" (p. 45), are totally inadequate, as are hs translations of the technical terms akrti as "generic shape," akara as "shape," pravahanitya as "permanent in the form of a continuous flux, " and svarupa "form of their own." In these contexs meaning is denotation: the causing of cognition (buddhi, pratyaya) of an object (artha). The object is either an abstract generic property (jati, akrti), an individual object (vyakti, dravya), or in certain limited contexts, such as sculptural reproductions (a falcon-shaped altar or a cow made of flour), a shape (akrti). Cognition brought about by a speech form (verbal cognition sabdabodha) is of a certain type (akara) which generally corresponds to an object which exists independently of the cognizer but need not, for instance, in the case of words for imaginary objects, objects which have no essential nature (svarupa), as Nagesa points out in the passage quoted on p. 45. The cognition corresponds to the independently existing object (artha-karah pratyayah) [6] and is located in the intellect (buddhistha). Thus a certain word generates cognition of a single type (eka-karapratyaya) otherwise referred to as an identical cognition (abhinnapratyaya). Since each cognition arising from the utterance of a certain speech form is identical in kind, the cognitions are not considered to be different cognitions created anew but are considered to be numerically identical, the very same single cognition which lasts from time to time (prava-hanitya). This enduring mental cognition brought about by the utterance of the speech form is truly the object bearing the word-meaning relation to the speech form and may be translated quite suitably by the ordinary term "meaning." Since the Indian linguists were not hung up on behaviorism, they did not shun recognizing mental phenomena, were not preoccupied in defining meaning in purely extensional terms, and so did not share the confusion prevalent in contemporary Western philosophy of language about what the term 'meaning' means. There is no need to project that confusion Onto them.

(1.) Similarly Nirukta 2.24, 3.1, 3.8, 3.9, 3.11, 3.13, etc.

(2.) Claus Vogel, Indian Lexicography, A History of Indian Literature, vol. V, fasc. 4 (Wiesbaden: Otto Harrassowitz, 1979).

(3.) H. M. Bhadkamkar, The Nirukta of Yaska (with Nighantu) edited with Durga's Commentary, BSPS 73, vol. 1 (Pune: B.O.R.I., 1985), 117.

(4.) laksyalaksane vyakaranam ... 14 ... laksyam ca laksanam caitat samuditam vyakaranam bhavati. Kim punar laksyam laksanam ca. Sabdo laksyah sutram laksanam. MBh. I.12.15.-17.

(5.) Tasmad anabhyupayah sabdanam pratipattau pratipadapathah. Katham tarhi ime sabdah pratipattavyah kim cit samanyavisesaval laksanam pravartyam yenalpena yatnena mahatomahatah sabdaughan pratipadyeran. Kim punas lat. Utsargapavadau. MBh. I.5.25-I.6.4.

(6.) arthakara is an upamanpurvapada bahuvrthi compound in which there is uttarapadalopa, formed according to 2.2.24 vt. 12 saptamyupamanapurvapadasyottarapadalopasca and is translated literally "that whose type is the type of the object."

BIBLIOGRAPHY

Aklujkar, Ashok. The Theory of nipatas (particles) in Yaska's Nirukta. PGRDS 42, Pandit Shripad Shastri Deodhar Memorial Lectures (sixth series). Pune: BORI, 1999.

Bakshi, Mukund Jha. Niruktam. [Sanskrit] New Delhi: Mehr. Lachh,, 1982. Bombay: Nirnayasagar Press, 1930.

Bhagavaddatta, ed., comm. Niruktasastram with Hindi Comm. Amritsar: Ramlal Kapur Trust, 1964.

Brahmamuni, Swami, comm. Niruktasammarsah with comm. Sailyanusari. [Sanskrit]. Ajmer: Arya Sahitya Mandal.

Chajjurama, Sastri, ed., comm. Niruktapancadhyayi with Sanskrit and Hindi comm. Delhi: Meh. Lach., 1963.

Prasad, Mantrini. Language of the Nirukta. Delhi: D. K. Publishing House, 1975.

Scharf, Peter M. "The Term 'akrti' and the Concept of a Class Property in the Mahabhasya." Wiener Zeitschrift fur die Kunde Sudasiens 36 (1992), supplement (1993): 31-48.

_____. "Assessing Sabara's Arguments' for the Conclusion that a Generic Term Denotes Just a Class Property." Journal of Indian Philosophy 21 (March 1993): 1-10.

_____. The Denotation of Generic Terms in Ancient Indian Philosophy: Grammar, Nyaya, and Mimamsa. Transactions of the American Philosophical Society, vol. 86, part 3. Philadelphia: APS, 1996.

Sharma, Radhe Shyam. Treatment of Vedic Deities in the Nirukta of Yaska. Ph.D. diss., Panjab Univ., 1982.

Sharma, Umashankara, ed., comm. Niruktam with Hindi Comm. Varanasi: Chowkhamba.

Shastri, A. D. Nirukta: Eka adhyayana. [Gujarati] CGV Studies 15. Surat.

Shastri, A. D., and R. I. Nanavati. Nighantu one Nirukta. [Gujarati] CGV Studies 23. Surat.

Thakur, Amareshwar, trans., comm. Nirukta [Bengali] Univ. of Calcutta, Asutosh Granthamala 5 (1963).

Tokunaga Muneo. The Brhaddevata. Kyoto: Rinsen Book Co., 1997.

Varma, Siddheswar. Etymologies of Yaska. V.I. Series 5. Hoshiarpur: V.V.R. Institute, 1953.

Vishveshvar. Nirukta with Hindi Commentary. Varanasi: Jnanamandal, 1966.

Vyasasisya, Kunwarlal. Niruktasaranidarsana. [Hindi.] Delhi: Itihasa Vidya Prakashan, 1978.

Yudhisthira, Mimamsaka, ed. Niruktasamuccayah vararucikrtah. [Sanskrit] New Delhi: Prhcya vidya Pratisthana; Ajmer: Bharatiya Pracya vidya Pratisthan.