1. Vritti Vartika Appayya Dikisita Commentary on the Functions of Words Edwin Gerow Nagraja Rao H.V

THE

VṚTTIVĀRTTIKA

OR

COMMENTARY ON THE FUNCTIONS OF WORDS

OF

APPAYA DĪKṢITA

Edited and Translated by

EDWIN GEROW

in Collaboration with

H. V. NĀGARĀJA RAO

AMERICAN ORIENTAL SOCIETY

NEW HAVEN, CONNECTICUT

2001

UNIVERSITÄT HAMBURG

Institut für Kultur und Geschichte Indiens und Tibets

Neue Rabenstraße 3, 20354 Hamburg

AMERICAN ORIENTAL SERIES

VOLUME 87

EDITOR-IN-CHIEF

PAUL W. KROLL

EDITORS

GARY BECKMAN STEPHANIE JAMISON PAUL E. WALKER

AMERICAN ORIENTAL SOCIETY

NEW HAVEN, CONNECTICUT

2001

32992

4047

sd

84

THE

VRTTIVĀRTTIKA

OR

COMMENTARY

ON

THE

FUNCTIONS

OF

WORDS

APPAYA

DĪKṢITA

Edited

and

Translated

by

EDWIN

GEROW

in

Collaboration

with

H.

V.

NĀGARĀJA

RAO

AMERICAN

ORIENTAL

SOCIETY

NEW

HAVEN,

CONNECTICUT

2001

x

241

UNIVERSITÄT

HAMBURG

Institut

für

Kultur

und

Geschichte

Indiene

und

Tibeta

Neue

Rabenstraße

3,

20354

Hamburg

COPYRIGHT

©

2001

BY

THE

AMERICAN

ORIENTAL

SOCIETY

All

Rights

Reserved

ISBN

0-940490-17-X

Introduction

Appaya Dīkṣita. The little that is known with certainty about Appaya's life is set forth in the standard sources.1 We can do no better than to resume it here. In a colophon to Kuvalayānanda,

amum kuvalayānandam akarot Appadīkṣitaḥ niyogād veṅkaṭapatēr nṛupādhirikrpāṇidheḥ2

Appaya refers to his patron as “Veṅkaṭapati.” Inscriptional evidence reveals two possible Veṅkaṭas to whom this name could refer: a prince of the Vijayanagara dynasty;3 the other,

Veṅkaṭa II of Penukoṇḍa, of whom inscriptions ranging from 1586 to 1613 A.D. are known.4 The latter is doubtless meant, as it would accord more consistently with the other meagre facts known.

By contrast, in a colophon to the unpublished Śivādityamaṇidīpikā, Appaya mentions one Chinnabomma as his patron, “son of Chinnavīra and father of Liṅgama Nāyaka.” This rājā of Velūr (Vellore, in North Arcot Dt.) is known from at least two inscriptions, dated 1549 and 1566 A.D.5 It seems likely, then, that Appaya was “promoted” from the lesser to

the greater chieftain at some point in his career. These dates give, as possible termini of Appaya's literary activity, 1549–1613.

These dates square also with the well-known traditions associating Appaya with Jagannātha (fl. ca. 1620–65), whose patron was, as seems likely, Shāh Jahān, and who received the title paṇḍitarāja from him. These traditions also recount meetings of the two (though, in one, their relative ages seem to have been reversed6) and even attribute to Appaya an attempt to

outcaste Jagannātha for taking a moslem concubine. Jagannātha was, by contrast, unrelenting in his criticism of the (presumably) older man's work.7

1 Chiefly, S.K. De, History of Sanskrit Poetics, 2 vols., 2nd rev. ed. (Calcutta, 1960) and P.V. Kane, History of Sanskrit Poetics, 3rd rev. ed. (Delhi, 1961).

2 P. 203, N.S.P. ed. (1955).

3 De, H.S.P., I: 222 (with references), who hazards for him a date of “ca. 1535 A.D.” This prince, a son of Acyuta Devarāya (d. 1542), ruled for only a part of a year, before losing the struggle for succession with Sadāśivarāya (r. 1542–76), whose regent was the redoubtable Rāma Rājā, the effective ruler (Burton Stein, personal communication).

4 The latter Veṅkaṭa (referred to as “Veṅkaṭa I” by most authorities, who were apparently not aware of his ephemeral predecessor) was one of the successor chieftains of the last important Vijayanagara ruler, Rāma Rājā (d. 1565) (Burton Stein, Vijayanagara, 120). Sewell gives this Veṅkaṭa's dates as 1586–1614 (A Forgotten Empire, 218).

5 De, ibid., quoting Hultzsch.

6 See A. Gangopadhyay, Contribution of Appaya Dīkṣita to Indian Poetics (Calcutta, 1971), 8. Pages 7–10 of this work provide another version of the meagre information available on Appaya's vita.

7 For example, his Citramīmāṃsākhāndana is a point-by-point refutation of Appaya's (incomplete?) work of that name.

i

The Vrttivārttika

In colophons and other references Appaya's name is variously given as Appa, Appaya, and Appayya — the variation often metri causa.9 Such alternation, while unusual, is less problematic in an evidently non-Sanskritic name. The colophon to Kuvalayānanda, cited above, also states that Appaya was of Bharadvāja gotra and the son of Raṅgarājādhvarindra. Sources variously indicate that he was the fifth,8 or elder,9 son (a younger brother was named Apya, or Āccān10). In all probability, then, Appaya was a brahmin of the northern Tamil country, who did not stray too far from home during his lifetime. He is stated to have lived for 72 years, and composed 100 works.11

His Works. Appaya is chiefly famous for his Śaiva, and to some extent theistic, version of late Advaita. His best-known work in this vein is the Siddhānta-leśa-saṅgraha.12 To him are attributed several devotional stotras, not all Śaivite—among them the Varadarājastava,13 a laud of the deity (a form of Viṣṇu) in one of Conjeeveram's oldest temples. It is, however, his three works on Alamkāra that have become commonplaces of the Indian intellectual tradition. His Kuvalayānanda commentary to Jayadeva's Candrāloka has become one of the standard expositions of the figures of speech—and certainly one of the most extensive, treating of 115 major types.14 The Citramīmāṃsā, mentioned already, apparently incomplete,15 treats again of the figures, but from a more sophisticated and critical point of view. It seems to take up where our treatise, the Vṛttivārttika, leaves off—with a discussion of dhvani, 'suggestion', the third vṛtti, or semantic function of the word. But this discussion quickly evolves into a traditional treatment of the figures, via the type of dhvani known as gunībhūtavyaṅgya, or 'suggestion made secondary (to a figure, etc.).

Our present treatise, usually mentioned as Appaya's third in poetics (but which is sometimes taken as an œuvre de jeunesse16), also appears to be incomplete — unless the “dhvani”

8 De, op. cit., 223.

9 Kane, H.S.P., 320.

10 Whose grandson was the well-known poet Nīlakaṇṭha-dīkṣita. Āccān was also probably the name of Appaya's grandfather. Confusing the matter still further, V. Raghavan has identified several other Appaya Dīkṣitas in adjacent generations of the family (Proceedings, A.I.O.C. [Tirupati, 1941], 176-80).

11 In the Śivalilā-rnavakāvya of Nīlakaṇṭha, quoted Kane, op. cit., 319.

12 Various editions and translations. See Potter, E.I.P., I: 262 — who gives Appaya's dates as 1520-92, citing Ananthakrishna Sastri.

13 Ed. A.V. Gopalachariar, Vāṇī Vilās Sanskrit Series, no. 11 (Śrīraṅgam, 1927). Our Vrttivārttika takes many of its examples from this stotra. A tradition current among pundits has it that Appaya became a Śaivite as a result of a personal grudge; in any case, as an Advaitin, he was ecumenical: viṣṇur vā śamkaro vā śrutiśikharagirām astu tātparyabhūmiḥ | nāsmākam tatra vādah prasāratī (Śivārkamaṇidīpikā, Appaya's comm. on Śrīkaṇṭha's Brahma-sūtrabhāṣya: hand copy provided by HVNR).

14 And this without treating of the “verbal” figures at all (śabdālāmkāra)! See De, op. cit., 223 n. 3.

15 And was apparently so for Jagannātha also, whose Khaṇḍana breaks off at roughly the same point (treatment of apahnuti alamkāra). Presently available editions of Citramīmāṃsā carry the discussion through two more figures, utprekṣā and atiśayokti.

16 ...though this would imply that the elaborate Varadarājastava is also a youthful work! Appaya's commentary on the poem is clearly later, for he there cites his Kuvalayānanda.

ii

Introduction

of the Citramīmāṃsā represents its completion—discussing only the first two of the by now standard three semantic functions of the word: denotation (abhidhā) and metonymy (lakṣaṇā). Ambiguous references in the text suggest that it may be based on a still earlier work (of Appaya's?—no author is mentioned), the Kāvyasaraṇī. Another prior work in the genre is the Abhidhāvṛttimātrkā of Mukula (end of ninth century), a terse fifteen-verse account of the same two functions, with autocommentary,17 which appears not to have had successors until Appaya—probably because the essence of its account of lakṣaṇā was incorporated into Mammata's tenth-century Kāvyaprakāśa, and thus became part of the mainstream, malgré lui.18

The virtue of Vṛttivārttika lies not in any novel classification, or even analysis, of the types and sub-types of the functions it treats—these are the standard ones, for the most part—rather, it lies in the author's scharfsinnig discussion of certain related matters, such as the way the second meaning of puns is apprehended, the force of contextual determinants in arriving at the “literal” meaning, the difference between equivalence and similarity in judging metonymies, and the effect of compounding on the logic of an expression. Appaya's many illustrative verses, some of them his own,19 though not poetry of the highest order, are also discussed with great acumen, and sometimes even wit. The ironies of life were not lost on this distinguished representative of medieval scholasticism.

This Edition of Vṛttivārttika. The present text represents a significant improvement over the two previously published texts of Vṛttivārttika, that of the Nirṇaya Sāgara Press [Kāvyamālā 36], ed. Śivadatta and K. P. Parab (Bombay, 1893; 3rd ed., 1940) and that of the Saṃskṛta Viśvavidyālaya, ed. Bhāgīrathaprasāda Tripāṭhī and Vāyunandana Pāṇḍeya (Benares, 1978). The N.S.P. edition makes use of at least three sources, one of which may have been the text printed in the Pāṇḍit 12 (1890), ed. R. S. Tailaṅga.20 The Benares edition makes use of the N.S.P. edition as one of its major sources, but adduces three presumably new manuscripts from the library of the Viśvavidyālaya.21 Despite this, the text as printed is in most respects inferior to the older version, containing many editorial blunders and simple misprints. It is evident that the Benares manuscripts stem from different traditions than those used by the N.S.P. editors, consistently disagreeing with them in those places that can be checked, though the differences are not great.

The novelty of the present edition, aside from correcting the many and often obvious blunders of its predecessors, lies in a careful comparison of the two previous editions, and their collation with still another manuscript, from the Houghton Library at Harvard

17 Published by the Nirnaya Sagar Press, 1916, and by others subsequently.

18 Kane, op. cit., 218.

19 Many also are cited from earlier works, going back to Dhvanyāloka—especially those that have become staples of discussion in certain matters.

20 See De, op. cit., I: 230.

21 Introduction, 2.

iii

The Vrttivarttika

University, Indic no. 1268.22 We might thus claim to have produced an “American” edition

of the Vrttivarttika—except for the fact (now evident) that this Harvard manuscript im-

proves only marginally the readings that are derived from the printed texts. But it does, and

quite often, support the N.S.P. edition as against the more recent Benares text, and to that

extent, confirms what most readers would take for granted anyhow. All of the textual dis-

agreements between the N.S.P. and Benares editions have, further, been discussed with

Vidvān H.V. Nāgarāja Rao, of the Oriental Manuscripts Library, Mysore, himself educated

in Vyākaraṇa and Alamkāra, who has supplied his own corrections and emendations

where necessary. The present edition has thus been brought into line with current tradi-

tions in the exposition of the Alamkārasāstra. We have relied on Rao’s advice in nearly all

cases of disagreement (so noted in the apparatus)—only rarissime preferring another

solution, usually based on the Harvard reading (also indicated, of course). It is my convic-

tion that the present text is about as close to the original as we can get—although I cannot

claim that it presupposes a thorough survey of the manuscript tradition, which, for a text

like this, used frequently in the pāṭhaśālā, is copious.

The age of the text is, comparatively-speaking, recent: substantial variations—that affect

the import and direction of the argument—simply have not developed. Most mss. (if not

all) were copied by students in process of learning the text; the task of “editing” such an

intellectual property has, therefore, as much to do with understanding its tight argumen-

tation as it does with strict adherence to canons of stemmata and variation. The presump-

tion, perhaps too powerful, is that Appaya knew better what he wanted to say than any of

his copyists, and said it at least as well as they did. Baldly so stated, one is made aware of the

one-sidedness of such an assumption—given the acute and utterly comprehensive knowl-

edge of the true pundit, of whom many have intervened between Appaya and ourselves.

But still, in view of the fact that the various versions of the text we know about contain no

major “improvements,” and the fact the manuscripts are usually copied by students, it may

be confidently hoped that the version given below (which has, in addition, been subjected

to the rigors of a careful translation) is quite faithful to Appaya’s ideas.

The Harvard manuscript is itself incomplete, breaking off a few lines before the end of

the second (and last) chapter.23 It shows the hands of two copyists, of whom the second

was considerably less careful.24 The chief benefit of the Harvard manuscript is thus con-

fined to the first half of the text, where several improvements have been noted.

Subject Matter and Scope. The Vrttivarttika is an independent text on a subject that

has been, at least since Mammata, a standard subtopic of surveys of Alamkāra. The two

22 The Houghton Library possesses two other mss. of the Vrttivarttika (nos. 1266 and 2331). These, however, are very

clumsily done, incomplete, and not trustworthy.

23 In the midst of a compound: śabda… vaicitryamātrabhedapraṁtyā.

24 The break appears to coincide with the beginning of a new folio [14], coming also in the middle of a word. It is pos-

sible that the first few akṣaras of the new folio are in the old hand.

"powers" of the word that are discussed here are accepted by all schools of thought: denotation (abhidhā) and metonymy (lakṣaṇā). The standard discussion thus serves as a bridge between extra-poetic concerns, those of logic and grammar, notably, on the one hand, and those of alamkāra, stricto sensu, on the other. The Ālamkārikas— bien entendu, since the wholesale adoption of the theses of the Dhvanyāloka in the ninth century—have been interested in demonstrating the relevance and irreducibility of the third "power" of words: suggestion (dhvani, vyañjanā). It is in this vein that denotation and metonymy have been chiefly discussed by the mainstream writers on poetics. Their treatments have been, for the most part, succinct and limited to commonplaces: it is the dhvani, and its uniqueness, that preoccupies them.

Appaya must have had a somewhat different perspective. Those modern historians who consider the Vrttivārttika incomplete because it fails to discuss the third "power" have perhaps missed the point: the discussion here is focused on the first two powers— not on the dhvani. And it aims not, in any sense, to "establish" the two powers: they have, as indicated, themselves become commonplaces of scholastic discourse. Rather, Appaya appears interested in refining the modes and subtypes of these—powers—for neither is "simple" —and in clarifying the interrelationships among these modes. His discussion, in fact, often enlarges the scope of the primary powers—even at the expense of the dhvani (which he fully recognizes, if he does not discuss it). The locus of discussion is, in general, figuration; that is, certain figures provide the necessary pretext for clarifying the function of the two powers. This is particularly true of Appaya's discussion of "pun" (śleṣa)—a substantial portion of the first chapter—in re the apprehension of the second meaning. In this discussion, a sharp line is drawn between denotation and "suggestion."

It would be equally misleading to see the text as devoted, essentially, to figuration. Figures are discussed only as they present some problem touching the broader question of how meanings are apprehended. Figures are, of course, an ideal ground of such a discussion, bringing together in intimate and essential association the three well-known powers, or functions, of words. This was recognized by the first partisans of the dhvani, who understood that the function they were promoting was essential to the success of many figures, but could not be reduced to an aspect of figuration (or, at least, that was their view). Figuration is not the subject of this treatise in another sense also, for it is not through either denotation or metonymy, primarily, that a figure's force obtains.25 This calls for some clarification, inasmuch as figuration was defended by its early exponents as exploiting the notion of the counter-literal (vakrokti). Metonymy, which presumes, by definition, some failure of the literal, has nevertheless not been considered a figure of speech, doubtless because its primary and typical instances tend towards the petrified and, indeed, have become quasi-literal.

25 We are here discussing only the "figures of sense" (arthālamkāra) that depend on some manipulation of the meaning (as opposed to the form) of an expression. "Figures of sound" (śabdālamkāra) convey whatever charm is proper to them by means other than the word as meaning bearer.

The Vrttivarttika

"Hitting the nail on the head" is one of thousands of such frozen metaphors with which our speech abounds; similarly, the ubiquitous Sanskrit instances, such as "the grandstands are shouting" (mañcāḥ krośanti), are nothing but formulatic and easily deciphered sub-stitutions for literal expressions that, indeed, constitute their "meaning": 'the people in the grandstands are shouting'. As we might "say just the apposite thing." Though a figure may, and often does, build on a metonymy, its expressive force lies not in the mechanical substi-tution of one word for another, but in the unexpected apprehension of further meanings that are thereby provoked—"my luv's like a red, red rose," and so on. A figure is thus not capable of that kind of standardization—even though the same kinds of figures regularly evoke the same kinds of "suggestions."26

A figure is, in the Indian view, not a "function" of the word at all, but, making use of the functions (as does all language), it is an "argument." That is to say, it is a mode of thought, a way of formulating or conveying an idea, which has closer relations with logic than with grammar, or even semantics. Its force rests on a combination of terms whose appropriate-ness (aucitya) is determined more by their sense (or reference) than by their form; though it is, in fine, an assertion (like the sentence), its complexity-is a complexity of meaning, rather than of syntax. The simplest of syntax may express a profoundly resonant figure.

Appaya's chief concern, in this treatise, is to clarify the relation between the two primary "powers" of words and various figures of speech. The Vrttivarttika thus constitutes a kind of foil to the much more extensive literature that seeks to clarify the relation between the figures and the third function, suggestion.

Denotation. The quasi-entirety of the discussion in the first chapter, on denotation, concerns the problem of pun (śleṣa). After a brief delineation of the types and subtypes of denotative meaning (see below), Appaya takes up the question, which for him evidently constitutes the limiting case, of multiple meanings that seemingly derive from the same (or similar) word(s). The discussion has three aspects: (a) pleonasm: why does one repeat words having the same (or similar) denotative meaning; (b) ambiguity: how is the denotative meaning in fact determined when a word is also capable of expressing other meanings; (c) willful use of such double-entendre (pun): whether the second ("non-contextual") meaning is gotten by denotation or some other function, that is, by suggestion (as claimed by the ancients).

Note that our concern here is not with association of different meanings via one or another of the figures of speech (viz., structures of thought), or even dhvani—that is the province of poetics, per se. It is restricted to those cases where several meanings arise gram-matically—in the same word or words (or, mutatis mutandis, where several words repeat

26 Many Sanskrit examples have, by dint of repetitive illustration, become stereotyped: "moon-face," etc. The metonymy here is based on the relation of similitude, rather than a more positive or determinate relation, and so may perhaps nev-er achieve the currency of the "literal". Still, even the figures that do depend on positive relations go beyond the metonymy, into the realm of suggestion—e.g., kāvyahētu, atiśayokti (see Glossary, s.v.).

vī

Introduction

the same meaning). Although poetical or stylistic issues are posed by such usages, they are

not our author's main concern either: he wants only to clarify the denotative function itself

in those problematic cases where it impinges on the domain of poetics—for the poets

have, it seems, been able to convert what seems a defect on the mundane level to (in some

cases) an embellishment on the poetic. Without the poetic usages, of pun, etc., no proper

delineation of the mundane is possible.

But Appaya's interest goes beyond merely clarifying the functional boundaries of deno-

tation: he argues resolutely against resorting to the dhvani in order to explain how we

apprehend the second meanings of puns: here, too—and contradicting the authority of

Mammaṭa—he sees only denotation operating. One suspects that there is more at stake

here than meets the eye, but what it might be is not as clear as is the subtext of the follow-

ing chapter, on metonymy. It does not seem likely that any ritual issues were joined in this

way—puns are not met with in that context—or, if they were, would not be so interpreted.

A fortiori, in a Vedāntic context—unless one takes the hardy position that sentences like

tat tvam asi make more sense as double entendres than as plain assertions of identity. (No

one takes this position, as far as I know.)

The term abhidhā (var., abhidhāna) probably originated (as so much other linguistic

terminology) in the usages of Mīmāṃsā, where it signifies, at first, not 'the literal' per se,

but only a type of the 'literal'. The genus 'literal', for the Mīmāṃsā, is designated by the term

śruti—and this term is chiefly met in explaining the architecture of the sacrifice, that is,

how the various injunctions and other statements of the Vedic text are to be constructed (viz.,

construed) in order to effect an orderly sacrificial procedure. Śruti is the most authoritative

such ordering principle, for it depends on no prior presumption—save mere 'audition'—

to have force: it is the plain text, without “human” interpretation. But śruti, as the author-

ities declare, is of three sorts: injunctive (vidhātri), applicatory (viniyoktri), and denotative

(abhidhātri).27 Though the main interest is focused on the second of these, it seems reason-

ably clear that what is at issue is a distinction between the stem of a word, its inflectional

ending, and the syntax of that ending—all three are 'literal', though in different senses. In

the phrase vrihin prokṣanti 'they sprinkle the rice grains', the three “powers” are illustrated

by the stem vrihi-, the accusative ending -ln, and the verbal syntax of that ending, relating

the 'rice' to the 'sprinkling' as its object. This restrictive sense of abhidhāna reflects the care

the Mīmāṃsaka takes in knowing “whether” he is to sacrifice, “what” he is to sacrifice, and

“how” he is to sacrifice it. In this primitive meaning, abhidhā would seem peculiar to nouns

(that is, noun stems)—and this may well have been the term's original sense—but, as the

authorities again make clear, it applies to any thing, expressed śrutyā—including qualities

and actions—inasmuch as these cannot be admitted to the sacrifice except as “modifying”

a substance of some sort. Thus, even the verbal root prokṣ- has an abhidhā—before it is

27 See Apadeva, Mīmāṃsānyāyaprakāśa, para. 68 (Edgerton ed., pp. 64, 206); also Kumārilā ad P.M.S. 3.3.11; Pārthasārathimiśra,

Śāstradīpikā, 1.3.30-35], etc.

vii

The Vrttivarttika

construed in the sentence. It is not difficult to see how this seemingly peculiar Mimamsaka usage became generalized in the later semantic literature, just as the notion of the sentence was being broadened beyond that of mere injunction. When the Mimamsaka priority on action was replaced by more speculative paradigms, abhidha was literally all that was left of the word.

Metonymy. Appaya's treatment of metonymy is, unfortunately, very brief and often laconic. Here too, after offering a typological summary of metonymy, his treatment seems to focus on three issues: (a) relation: in what sense is similitude a "relation" and can metonymy truly be based on such a "relation"; (b) expansion: should the types of metonymy, which are multivalent, be increased to account for additional possibilities of interfacing; (c) restriction: are two of the standard types, in fact, identical?

Indeed, the typology of metonymy is of a different sort than that of denotation. In the latter case, the subtypes appear to be different in se, and exclusive: a meaning is either "original" (rüdha: 'ascribed'; this category includes compounds whose meanings cannot be derived from an analysis of component elements [— see next]), or "derived" from the meanings of its parts (yukta: 'synthetic' — a type that applies, ipso facto, only to compounds of simpler elements); or "both" — that is, in the case of certain compounds, especially in poetic usage, when either analysis leads to an acceptable meaning (and both are usually intended — in fact, a metonymical "pun"). It is evident that the question does not concern how meanings come to be attached to words — the orthodox tradition considers this relation "innate" (nitya); all meanings are in effect "inherent" — but compounding itself somewhat obscures this original clarity, in that the sense of the compound, in some cases (c, above) differs from and overrides the sense of its component parts. Were it not for compounds (which, apparently, figure as human inventions), there would be but one type of "denotation." For our purposes, however, it is enough to note that the types, as set forth, are exclusive (even though one is a category by combination).

Metonymy, on the other hand, evinces several different modes of typological division — that is, different in principle, and capable of, even demanding simultaneous manifestation: much like the categories of verbal and nominal inflection — where the verb, e.g., must be marked for both person and number, but under each head, cannot be marked but for one person or one number (syntagmatic categories?). Thus, a metonymy is either "pure" or "secondary"28 — a distinction based on the kind of relation involved: either "contiguous" or "similitudinary" (indexical or iconic, to use Peirce's terminology — this distinction is exhaustive); but metonymy is also considered in relation to the speaker's intention (for,

28Mukulabhatta, in his Abhidhavrttimatrika, also speaks of a suddha laksana, but contrasts it with saropa and sadhya-vasana lakṣaṇā—as a metonymy where the two terms retain the sense of their distinction. No trace of this notion is found in Appaya—who follows Mammata in contrasting suddha with gauni, defined now by the kind of relation subtended between the terms. A veiled reference to Mukula may read into Appaya's determined defence of the thesis that gauni is indeed a type of lakṣaṇā.

Introduction

unlike denotation, intention may here determine meaning): it is either “conventional” —

that is, in effect, frozen, deprived of any “motivation”—or “motivated”—used for a pur-

pose apart from what it has come to mean by convention (this division, too, is exhaustive).

This latter category evidently impinges on the domain of dhvani, and many of the exam-

ples of this latter involve “motivated metonymy.” It is clear that these four types intersect—the

two major classes are not exclusive to each other—but each has two exclusive subtypes,

which combine with the exclusive subtypes of the other—in effect, a quadrilateral typolo-

gy based on four features: plus or minus “contiguity”, plus or minus “motivation.”

These two ways of classifying metonymy present few problems, and are standard in the

semantic tradition. It is the third kind of classification that creates difficulty, for a metonymy

(whereby a word accepts another meaning owing to some blockage of its primary sense—

usually by syntactic context) may yet retain some vestige of that primary sense—or, at

least, not be incompatible with that primary sense. This gives us a triple possibility: (a) the

primary meaning is set aside, as incompatible (jahallakṣaṇā); (b) it is not, because still

compatible (ajahallakṣaṇā)29; and (c)—mirabile dictu—it may both be set aside and not

set aside (jahadajahallakṣaṇā).30 Standard examples of the three are: “village on the Ganges”

(where, because of the literally watery environment, the name of the river must be read as

‘bank (of the Ganges)’; “the swords are arriving” (where, though “swords” are, per se, inca-

pable of motion, they do remain a part of the reinterpreted expression, ‘men holding

swords (are arriving)’; “the village is on fire” (by hyperbole, where only a part of the village

in actually burning—the “part” of the village on fire is understood literally; the “part” not

must “give up” its meaning in the total expression.

What is at stake here is, despite appearances, important—for, in the semantics of late

Advaita (Appaya was a theistic Advaitin), the correct interpretation of many important

passages from the upaniṣads depended on the admission of this odd third subtype: tat tvam

asi. Both terms, to be interpreted as an identity, as the purport requires, must “give up” a

part (but not all) of their literal meaning: tat (viz., Brahman), here equated with the embod-

ied soul, is understood as “consciousness”—but not as “omnipotence”; tvam (viz., Śvetaketu),

similarly, is understood as “consciousness,” but not as “dependence,” etc. Appaya does not,

however, really take up the question on this general level—one wishes that he had. Rather,

the more limited question is examined, as to how the third principle of division relates to

the others here propounded: whether, for example, metonymies based on similitude also

admit of this three-fold division (as clearly do those based on relations of contiguity); or

whether jahallakṣaṇā is really the same thing as śuddhādhyavasānalakṣaṇā—that is, a met-

onymy operating by a relation of contiguity that, in effect, simply replaces the literal mean-

ing with the figurative (by identifying the two).

29 These two types are termed lakṣaṇalakṣaṇā and upādānalakṣaṇā by others; cf. Viśvanātha, Sāhityadarpana 2.6-7.

30 Viśvanātha does not recognize this third variety.

ix

The Vrttivārttika

This leads us to the final, and fourth way of classifying metonymy—by the degree of “difference” obtaining between the metonymical meaning and the underlying literal meaning: (a) the former may simply be “superimposed” on the other, but does not necessarily “replace” it, or (b) it may, in fact, be “identified” with the other, in effect, replacing it.31 The terminology, and indeed, the problem is here evidently also Advaitic. But the theological implication is muted, the distinction itself reduced to a question of the grammatical means of expression employed: whether both terms mentioned—or only one. It is, rather, shown, by a kind of a fortiori argument, that the types of metonymy important to the Advaitin are safe from attack, by demonstrating that the various types advanced cannot be further reduced in number. But this seems to us a thoroughly legalistic approach—though not one without persuasiveness in an academic tradition. It does establish beyond a doubt the close relation between theology and semantics in medieval India—a relation that the Indians, at least, took cognizance of, as did few of our theologians.

The exact number of subtypes is not entirely clear from Appaya’s treatment. The four principles of distinction are not completely free to combine with one another—and even when they appear to be free, Appaya sometimes sees obstacles to their combination. He, for instance, denies the relevance of the tripartite distinction having to do with the compatibility of the literal meaning, to metonymies based on similitude. On the other hand, the latter two “Advaitic” principles seem only different emphases of the same notion: both concern the “rhetorical” status of the surviving literal meaning in the resultant “metaphor.” Ultimately, Appaya, after an argument that occupies fully half of his treatment of metonymy, rejects the relevance of the distinction between “superimposed” and “identified” (sāropā/ sādhavasānā), as regards “pure” metonymies—in effect, conflating the “superimposed” metonymy with that of the third type that “gives up” (jahallakṣaṇā) its literal meaning. The distinction itself, then, appears less than probative, and cannot typify any “pure” metonymy. If “superimposed” reduces to jahallakṣaṇā, it would seem that “identified” must also—a fortiori. The distinction does retain its force in the matter of “secondary” (gaunī) metonymy—where, of course, the figures “simile” (upamā) and “metaphor” (rūpaka)32 depend upon it.

Before engaging upon this argument of reduction, Appaya sums up the types as follows: “Thus, pure metonymy being five-fold, and secondary being two-fold, motivated metonymy is seven-fold in all.” This account ignores “conventional” metonymies entirely, and does not consider the other possible intersections among the major types. By restricting the tripartite distinction to “pure” metonymies, and, in effect, the distinction between “superimposed” and “identified” to metonymies of similitude (gaunī), the possible types of motivated

31 This terminology is apparently also owed to Mukulabhaṭṭa, subsequently standardized. But Mukulabhaṭṭa places sāropa and sādhavasānā lakṣaṇā on a par with “pure” lakṣaṇā. Appaya, of course, treats them as varieties of “pure” lakṣaṇā.

32 Or, “metphorical identification,” to be precise. See Glossary, s.v.

X

metonymy reduce to five.33 It would probably not be useful to replicate this pattern among

conventional metonymies.34

However that may be, a consideration of the criteria of distinction applied here to met-

onymy demonstrates the range and exhaustiveness of the analysis: metonymies may be dis-

tinguished by the kind of relationship presumed to link the primary and derived meanings;

by the degree of conventionality acquired by the metonymy (the degree to which it admits

the speaker’s intention as determinative of meaning); and by the manner in which the two

meanings persist in and/or contribute to the total effect of the metonymy. For it is essential

to the function of a metonymy that the bypassed literal meaning yet be present in some

fashion in the “reinterpreted” whole meaning—if only as the sense that a “reinterpretation”

has occurred; otherwise, the function could not be distinguished from denotation—a

point, perhaps, that Appaya makes in allowing for a type of “conventional” metonymy.

The origins of the notion of lakṣaṇā are also to be sought in the Mīmāṃsā system,

where it was adduced by the earliest writers as a “power” necessary to attribute to words if

they were to have any practical application at all. This “strong” view of lakṣaṇā evidently

goes well beyond our later metonymy and was, like its view of abhidhā, peculiar to the

Mīmāṃsā; but it follows, as the night the day, from the peculiar premisses of that system.

Partly in order to guarantee the immemorial and changeless character of the Vedic text, the

Mīmāṃsā viewed its component words as expressive only of genera (jāti)—never of indi-

viduals (vyakti).35 But, in the sacrifice, those same words had to apply to individuals: only

individual rice grains could be “sprinkled”—not the genus “rice”! The particularity of all

action was reflected in and assured by lakṣaṇā—that is, the “ability” or “power” of the uni-

versal to limit itself, in reference, to individuals—who were, yet, not truly “individual” (for

any “brown cow” would do, or any finite set of “rice grains”36), but were simply grasped sub

specie universalis—as representations of the universal. In this sense, all words were metonyms

insofar as they were of this world—insofar as I instantiate their eternal messages.37 Again,

by the twin processes of generalization and shift away from a focus on action, lakṣaṇā

becomes that “inherent” power of any word to “mean” something other than what it plain-

33 Viśvanātha’s classification, probably the most elaborate, allows for the sixteen basic types that are obtained by com-

bining in all possible ways (24 = 16) the four two-fold alternatives (he does not admit jahadajahallakṣaṇā) given by

Appaya, and which seem to be accepted by all authors. He then goes on to add still finer discriminations—to “moti-

vated” metonymy—bringing the overall total of possible types (not all exemplified) to eighty! (comm. ad 2.12ab).

Appaya is evidently arguing against such encyclopaedic thoroughness in this work, which is one of the few to ponder

the question of how or whether these multiple varieties actually coexist.

34 Viśvanātha, however, does, proffering such usages as śveto dhāvati ‘the white (horse) runs’ as exemplifying a conven-

tional (purposeless) version of ajahallakṣaṇā. Here, a purely grammatical feature—the ability of the adjective to stand

alone when its noun is easily supplied by context—is considered “metonymical.”

35 See, i.a., Pārthasārathimiśra, Śāstradīpikā, 1.3.10 [ad P.M.S. 1.3.30-35].

36 The ubiquitous “Devadatta,” though a name, and as close to a designation of the “individual” as one can get (lin-

guistically speaking), is still just an “instance” of the class of those named “Devadatta.”

37 It should be noted that this notion of lakṣaṇā is also “inherent” in the sense that it does not depend on the inten-

tionality of any speaker. Later “poetic” (including Appaya’s) versions, of course, do (or may, see below).

The Vrttivarttika

ly means—but what is common to the older and newer notions is the restrictive factor

provided by context: words cannot mean anything at all at any time, but only what “con-

text,” etc., oblige us to impute to them—in order to save their “practical” sentential force.

In Mīmāṃsā, too, are found the earliest speculations on “poetic” extensions of lakṣaṇā—

those “metaphors” that seem grounded only on similarity, rather than deriving from any

necessity of the process of signification itself: “the trees attend upon the sacrifice”38 —sen-

tences that create great problems of interpretation, for no “nonsense” can be admitted into

an eternal corpus. Here, too, lakṣaṇā is appealed to, but interestingly, involves, as basis of

transfer, a notion of “purpose” (rather than merely one of “application”): by this hyperbolic

utterance, the sacrifice is praised, glorified, made attractive to men! In the later literature,

too, lakṣaṇā continues to be treated as a semantic, rather than a poetic, category—but, ulti-

mately, those powers of the word that do overstep lakṣaṇā and do constitute the stuff of

poets—the figures exploiting similitude, etc.—are proven to evince still another śakti—the

dhvani that is essentially purposive, and yet still necessary to account for the ability of lan-

guage to evoke moods and other non-ideational content.

But Appaya, in this treatise, does not take us that far. The Vrttivarttika is best seen as a

prologue to the study of dhvani, and it is undoubtedly for this reason that it ends

where it does.

38 See P.M.S. 1.1.32 and Śābarabhāṣya thereon.

xii

Table of Contents

Introduction

Appaya Dīkṣita ....................................................... i

His Works ............................................................... ii

This Edition of Vṛttivārttika ....................................... iii

Subject Matter and Scope ............................................ iv

Denotation ............................................................... vi

Metonymy ................................................................ viii

Chapter One: Denotation

Introduction: The Types of Denotation ........................... 3

Pleonasm and Sufficiency of Meaning ............................. 6

Multiple Meanings: Principles Determining the Intended Sense . 11

Pun: Where More Than One Meaning is Intended,

How is the Second Meaning Determined? ........................ 17

Chapter Two: Metonymy

Metonymy and its Types ............................................... 31

Extending the Typology of Secondary Metonymy ............... 37

Reducing or Conflating the Types of Pure Metonymy ........... 38

Pun, Metaphor, and Hyperbole ...................................... 45

Vṛttivārttikam, I ........................................................ 49

Vṛttivārttikam, II ........................................................ 67

Bibliography ............................................................... 79

xiii

On the Functions of Words

Vṛttivārttika

Vṛttivārttikam mayānūdyata Udupiśrikṛṣṇam namaskurvatā

May the goddess Sarasvatī be victorious, honored by the wise,

Who illuminates the world with the three functions of the word:

Metonymy, denotation, and suggestion—as Śiva's embodied form

Illuminates the universe with his three eyes. || 1 ||

Chapter One: Denotation

Introduction: The Types of Denotation

Three functions1 of the word have been propounded

In the poetic tradition2 by the compilers of (treatises on the) figures:

Denotation, metonymy, and suggestion. |2|

Now and again our elders have left certain points unclear;

Hence we have undertaken this Explanation of the Functions

In order to remove these doubts. |3|

Among (those functions), it is denotation3 that conveys a meaning by the (unaided)

capacity4 (of the word itself). (Denotation) is three-fold: by ascription, by synthesis, and

by both ascription and synthesis. “Ascription”5 means that (the word) conveys a single (un-

ambiguous) meaning through its unanalyzed capacity alone. It conveys (this meaning)

simply—either without its constituent parts appearing (to the mind at all), or, if they do

appear, by canceling them. For example,

1vṛtti, or ‘mode’. That is, the various ways the word becomes meaningful. The notion of “function” responds to the fact

that one and the same intentional meaning may be realized by alternative and incommensurate strategies at the expressive

level (of words): all authors agree on at least two, “denotation” and “metonymy” — the subjects of this treatise. See below.

2kāvyasaranau, taken by some (e.g., S.K. De, H.S.P., I: 225) as the title of a work. The context here is ambiguous, refer-

ring to “compilers” (pl.)

3abhidhā, lit., “naming,” or “appellation”: the ‘placing on’ the word of a apposite sense, or ‘on’ the thing of a signifying

word. In contrast to the second vṛtti, metonymy, abhidhā always operates immediately, directly, that is, not reflexively

with reference to itself or another. See ch. 2.

4śakti, lit., ‘power, ability’ — a term, as used here, closely approximating the sense of vṛtti (supra), but with this differ-

ence, that śakti expresses the generic notion of ‘meaningfulness’, inherent in the word as part of a symbol system, where-

as vṛtti designates, stricto sensu, the specific modes or manifestations of that śakti, viz., denotation, etc. The terminol-

ogy is, appropriately enough, Śaivite — as is the frequent use of the two terms as loose synonyms.

5rūdhi, akin to what we mean by “convention” — the word’s meaning derives from usage, purely and simply, as the

word is used in the market-place, without reflection or wit: the ‘literal’. But “convention” is very misleading, for most

Indian theories of usage consider the link between sense and word (or word and thing) “immemorial” — not, in prin-

ciple, subject to human alteration (where such alteration is present, we are not dealing with rūdhi). Thus, rūdhi approx-

imates closely to “natural” meaning. Such notions are, of course, related to the privileged status accorded to Sanskrit,

which, being the “gods’ speech,” is not a part of “history.” Indian theorists — or, at least, the grammarians and the ritual-

ists, who are perhaps the archetypal theorists of language in India — do however insist on the consonance between the

paradigmatic character of language and customary usage. Words, as used in the market-place — caeteris paribus — are

those very words the gods use: hence, such usage may be diagnostic even in Vedic exegesis.

The Vrttivārttika

Since they are fortunate who resort to your lotus-feet but once—

Being thereby liberated—feet beloved of lotus-born Brahma,

It is fitting that the jeweled anklets and pearl-broaches

That adorn your feet eternally, should receive a benefit

Beyond (mere) liberation—offerings of white jasmine flowers!16

Here, the words mani ‘jewel’, nūpura ‘anklet’, etc., (convey their meanings directly) without any awareness of constituent meanings. In the case of the word atimukta, however,

though (we are) aware of its constituents (ati-, mukta), their meanings are set aside in the interest of construing the whole word in the sense ‘jasmine’ (vāsantī). Thus, in both cases,

it is by ascription (that the requisite meaning is conveyed).7

Now, even though (in this example) the meanings of the constituent parts of the word

ati-mukta were expressly intended to determine a further (punned) identity between the

‘beauty of the jasmine bud’ and a ‘success beyond that of the liberated’, still, since those

(constituent) meanings ultimately are set aside as the word is understood in the sense

‘jasmine’, even here ascription alone (must be said to operate). It is for this reason that the

word eka ‘single’ has been introduced into the definition—to exclude the possibility that

such cases not fall within (the category of ascription).8

“Synthesis” means that (the word) conveys a single meaning exclusively by reference to

the capacities of its constituent elements. For example,

Above, o Lord! thy navel-lotus—Brahmā’s abode,

Above the darkness that enshrouds thy single strand of navel-hair,

I see revealed before me the highest state of all—

Thy wide bosom, radiant with masses of pearls.9

Here, (compound-) words, such as viriñci-bhavana ‘abode of Brahmā’, nabhi-padma

‘navel-lotus’, romāvalī ‘hair-strand’, etc., (illustrate denotation by) synthesis. It, too, is

[p. 50]

6From Appaya’s Varadarājastava, vs. 52 (identifications thanks to HVNR). The pun in the last line, which is the point

of the verse, is untranslatable: the same word, atimuktalakṣmī, is understood both as the ‘beauty of jasmine (lit., exceeding [ati-] the whiteness of the pearl [mukta])’ and ‘a fortune beyond (ati-) that of the liberated (mukta)’.

7Appaya’s discussion here does not confront the pun in atimukta. This is taken up next. Here, we are to understand that

rūdhī covers both the cases where no constituent is evident, and those where the constituents, though “evident,” are

overridden by a different meaning. The former would include the large class of unanalyzable morphemes: roots, pre-

fixes, etc., as well as underivable nouns (such as those Appaya mentions); the latter would include (perhaps) that even

larger class of derived forms whose meaning is not (except by grammarians) grasped as a result of analysis (the

province of yoga and yogarūdhī, next), such as most noun stems (bhāva), complex verbs (upa-ne-sya-ti), and what we

8avyāptivāranena: since the result is a single meaning (‘jasmine’) that sublates its constituent elements (‘[white]

beyond the pearl]’), and is not directly derived from them (i.e., such ‘derivation’ as there it is also rūdhyā—by custom),

it might be argued that, had the word ‘single (meaning)’ not been introduced in the definition, the case of ati-mukta

might not fall under rūdhī at all, for the word undeniably has more than one śakti. The point is a fine one, for, prima

facie, such cases appear to exemplify the third of the three types of denotation, ‘both’ (yoga and rūdhī); See below.

9Varadarājastava, vs. 63. The verse praises the jeweled chest of Viṣṇu as the embodied form of the highest condition,

spiritual liberation. There is doubtless also an allusion to the wide night sky (Viṣṇu is dark-blue) and its stars.

4

two-fold, for the meaning conveyed by exclusive reference to its constituent elements (arises) either in the absence of a (countervailing) meaning of the whole, or, if such a meaning does present itself to the mind, when that (whole meaning) cannot properly be construed (in the given context). The former has been illustrated.10 The latter is as follows,

The sun may well be the embodiment of the three Vedas—

Still it is shamed by the splendor of thy necklace's jewels;

Indeed, can anyone who hides the virtuous (stars)

As they approach the end of night, be tolerated by the just?11

[A second, punned, meaning of the last half-verse is:

Indeed, can anyone who insults the virtuous

Who have sought refuge in his home be tolerated by the just?]

Here, even though the word niśānta appears to the mind in the unitary sense12 ‘residence, home’, and even though it is expressly intended in that sense in order to determine a (punned) identity between ‘hiding’ stars as they approach the end of the night, and ‘insulting’ virtuous men who have come calling at one’s home, still, it is the (meaning determined by) synthesis, ‘end of night’ (that alone prevails).13

“Synthesis and Ascription” means that (the word) conveys a single meaning by reference both to the capacities of its constituent elements and to that of its unitary form. For example,

The moon, ever pursuing the lunar vow

Of waxing and waning in the two halves of the month,

Does circumambulation of the gods’ abode, o Lord!

As penance, hoping to win the lustre of thy lotus face.14

Here, the word surālaya (illustrates denotation by both) synthesis and ascription, in that (both meanings) — ‘abode of gods’, by the capacities of its constituent elements, and ‘Golden Summit’,15 by the capacity of its unitary form — (are relevant to the construction of the sentence).

10 That is, no “whole meaning” — apart from that which derives from the composition (yoga) of the elements themselves, is evident in “Brahma’s abode,” “hair-strand,” etc.

11 Varadarājastava, vs. 66.

12 samudāyārtha, not the “compound” sense that derives from the elements taken separately, but the sense of the “whole compound,” taken as a unit. Cf. “new port” and “Newport” (the name of a city—which may or may not [but probably once did] have a ‘new port’). This distinction may seem somewhat superficial, but Indian grammarians consider every nominal stem (prātipadika) to have, ipso facto, a proper sense (artha). Compound nouns, though derived from other nouns or words by deletion of inflectional morphemes, are “stems” in that sense, and have a unitary meaning apart from the meaning of their elements—just as do primary nominal derivatives, apart from the “meaning” of the roots, suffixes, etc., that “compose” them. Cf. P. 1.2.45, etc.

13 The “whole,” or unanalyzed sense, is, of course, derived here by rūdhi, custom. Even though one must be aware of it in order to establish the pun, and hence, “get the point,” it still must be set aside, in the final analysis, as irrelevant to the (astronomical) context of the verse. Thus, the meaning based on the analysis of elements (niśā-anta) alone survives.

14 Varadarājastava, vs. 85.

15 I.e., Mount Meru, the abode, specifically, of Viṣṇu, replicated in the generic temple. Here, too, we see the fine distinction between the “whole” sense and the sense of the parts: The parts, per se, give us the genus; the whole, the species (hence the “name”).

The Vrttivarttika

Pleonasm and Sufficiency of Meaning

It should not be objected that ascription alone, rather than ascription and synthesis together, is needed (to explain the meaning) of this word in the sense ‘Golden Summit’ [i.e., ‘Viṣṇu’s abode’ — Mt. Meru] — alleging that, in the verse exemplified, (the requisite meaning) is realized already through the metaphor of “penance” —inasmuch as an identity with the notion ‘divine abode’ has been already predicated of ‘golden summit’, by the capacities of its constituent elements, as being an object of circumambulation.

This view has no merit, for it is improper to seek a remote function, such as a “predication of identity,” to derive a meaning (‘abode of god’), when the same meaning readily emerges from the constituent parts (of the word itself). In this case, the presence of the synthetic meaning (‘abode of god’) is evident to all.

Nor, this being the case, should (the contrary possibility) be entertained, namely, that the unitary meaning need not be postulated — alleging that, even in the case of ‘golden summit’ [viz., Viṣṇu’s abode], the usage is sufficiently explained in terms of the capacities of the constituent elements alone, which have already been stipulated in the general sense ‘residence of a divinity’.

For we find many references to the unitary meaning of (such compounds as) ratna-śānu ‘jewel-peak’, which both Vedic tradition and the lexica specify as ‘an abode of god’ (surālaya).

Otherwise, we could (always) — as (we did) in the case of devatāgāra— (derive the requisite meaning) by (analytical) specification, and the teaching of (unitary meanings in the dictionaries, etc.) would serve no purpose.

It is for this reason that, when such a word (requiring for its interpretation both synthesis and ascription) is employed, there is a risk of pleonasm if, in addition, a word expressing the unitary meaning (only) is also employed.

For example,

May the blessed Lord be auspicious for you all—

Who, like a sun, illuminates the darkness of devotees’ humility,

The objector asserts that there is no need to appeal to both rūḍhi and yoga here, inasmuch as the rūḍhi meaning [Mt. Meru] can be explained as deriving from the metaphor of the moon “doing penance” by “circumambulating” the ‘golden summit’ [the yoga meaning] — thus determining an extended sense, ‘abode of god (Viṣṇu)’.

The objection not only postulates no double function, yoga and rūḍhi, but (for purposes of argument, evidently) supposes that the yoga meaning can be derived from the rūḍhi, though the former is self-evident in the word itself. Appaya’s reply is a version of Occam’s razor: why introduce a non-evident function to explain such a meaning, ‘abode of god’, when that same meaning emerges from a literal analysis of the word itself: sura-ālaya?

The idea seems to be that all we need to know here to make sense of the verse is that surālaya means, by analysis, ‘god-abode’, which notion alone is implied even in names of specific abodes (such as Mt. Meru). Reference to Viṣṇu, per se, would thus be obviated.

I.e., this meaning, which is over and above the etymological meaning, is universally treated as separate from and supervening the latter, both in dictionaries and in the liturgical (or literary) tradition.

6

Chapter 1

Whose eyes are more lovely than the water-born waterlily

That blooms at the break of day.20

Where, however, a word, though realizing its meaning both by synthesis and ascription,

exhausts its fullest implication by evoking a suggested sense that is based on the capacities

of its constituents' meanings, there is no risk of pleonasm if another word is also employed

in the unitary meaning (alone). For example,

Not even thousand-eyed Indra himself

Would reach satisfaction in gazing at the splendors of thy lower lip—

Soft as the kandala flower that glows like the rising moon,

Sister to the deep-red throat of the night-blooming lotus.

Because the word “thousand-eyed” exhausts its implication in evoking the possibility of a

gaze well beyond the ordinary, through the capacities of its constituents' meanings, another

word meaning 'Indra' [viz., vrṣā, 'bull'] has here been (properly) employed.21 Another example:

By your grace (O Indra!), I could humble the pride even of Śiva

Who bears the bow Pināka in his hand — though my weapons are flowers,

The springtime my only ally;

What other bowmen would I need?22

Here, though the word pinākapāṇi ('who has the [bow] Pināka in his hand') is a name

of Śiva, the word exhausts its fullest implication in evoking a sense of the weapon's potency

through the capacities of its constituents' meanings. Another word meaning 'Śiva' [viz.,

hara, 'destroyer'] has thus been (properly) employed.

It might be objected (on the basis of this argument) that the word kusumāyudha ('whose

weapons are flowers'), also expresses a sense of (Kāmadeva's) weapons' potency through

the capacities of its constituents' meanings, and yet no second word meaning 'Kāma(deva)'

is found (in the cited verse).23 This objection has no merit. According to the author of the

Sāhityacintāmaṇi,24 (there is, in effect, such a second word) because there exists the first-

20 Appaya does not comment on this example, but it must concern the phrase nira-ja nalina 'water-born waterlily'. The

repetition of “water” suggests, in English, the defect intended, but the Sanskrit case does not involve so obvious a

pleonasm: it is simply that nira-ja, which means, by yoga, 'water-born', and, by rūdhi, 'lotus', renders otiose the accom-

panying nalina, also 'lotus'. Nira-ja would have been sufficient.

21 As with most epithets in Sanskrit (cf. nira-ja, above), the qualification can stand alone in the sense of the person or

thing qualified: “thousand-eyed” means “the thousand-eyed one,” viz., Indra. This is its ascriptive, or unitary, meaning;

such usage would normally render otiose another word that means 'Indra' — but here, “thousand-eyed,” in its synthet-

ic sense, has another function, that of suggesting the overwhelming character of Indra's gaze (and a fortiori, his inabil-

ity to reach satisfaction); Appaya opines, therefore, that the second 'Indra' is not, strictly speaking, otiose.

22 Kumārasambhava 3.10. Spoken by Kāmadeva.

23 This kind of niggling counter-argument is typical of śāstraic discourse. But it conceals a deeper purpose, in that it

implies that “pleonasm” in such cases would be the rule — and its absence a defect: the reverse of the conventional treat-

ment.

24 Probably Vīranārāyaṇa (early 15th cent.). See De, H.S.P., I: 225-26, 304-5. The work is known only in extract.

The Vrttivārttika

person pronoun, implied by the first-person verbal ending, which expresses a qualification.25 But, in truth, there is no rule to the effect that, whenever a suggested meaning based on the capacities of (a word's) components' meanings is intended, another word expressing a unitary meaning must also be employed. Rather, it is simply the case that such a type does (properly) exist, as has been illustrated. But the (unitary term) is also not expressed, as in this example:

I think it is the Goddess of Speech herself

Who sings thy praises, o Lord,

By entering the tongues of the great poets.

For she wishes to escape the inevitable censure

(That follows upon such an impossible task)

By placing her blunders on another's brow.26

Here, though the meaning of the term vāgiśvarī ‘Goddess of Speech’ suggested by the capacities of its constituents' meanings is intended as a (further) ornamentation of the (verse's) expressed (meaning) — namely, that Sarasvatī does not tolerate, where knowledge is concerned, any exceptions of blundering, as (do others) elsewhere — nevertheless, there is no second use of a word meaning ‘Sarasvatī.’ (And this is the case) even though (the verse also) suggests a figure of speech, “presumption a fortiori” (arthāpatti), namely: “where even the goddess of speech, herself, does not have sufficient capacity to praise the Lord adequately, how can others even presume to assume that office?”27 Similarly, the author of the Ekāvalī [Vidyādhara, 14th c.] shows that no such restriction applies to the use (of a second term expressing the unitary meaning), when, in defining the figure “accumulation” (parikarālamkāra) as “(the mention of qualifications) furthering the implication of the qualified term,” he uses, in the following example, the qualified term only:28

O Narasimha, man-lion! O lord of earth! Thy fame ‘flows in three courses’ — (a Ganges) purifying the three worlds! Who would not find it praiseworthy!29

25 Skt., kūryām. Aham ‘I’ is, of course, implicit in the verb. The point is that “I” too can be understood as a qualification of Kāmadeva, and so, a second term is employed.

26 Varadarājastava, vs. 3; āviśya rasasrāni, lit., ‘has entered into the tongues of the great poets’ — who now speak through me.

27 This suggestion, technically an alamkāra dhvani, Appaya derives from the verse's implication that Sarasvatī speaks only through the tongues of others, never directly.

28 I.e., the term that normally would be a “qualification” — of the Ganges, used here without the word it qualifies, and hence, by a kind of shorthand, the “qualified” term itself. All texts and mss. agree on viśeṣyam, though we might expect viśeṣaṇam.

29 The logic of Appaya's argument is as follows: The Ganges is here mentioned only via one of its epithets, tripathagā, ‘flowing in three paths [i.e., heaven, earth, underworld]’; but Vidyādhara has offered this as an example of parikarālamkāra, where both term and epithet are expected. Ergo, there is no requirement that both terms always be present, in order that both be understood — in this case, the figure itself requires such an understanding. But Appaya is probably stretching a point here. First of all, a qualification is mentioned, ‘purifying the three worlds’; secondly, he overlooks the pun implicit in tripathagā, which, in the verse, serves as predicate to kirti ‘fame’: your fame (like the Ganges) runs in three courses [i.e., extends to the three worlds]. Thus Vidyādhara's point may be only that kirti (as viśeṣaṇa) is severally qualified by epithets that bring out its “quality” — namely, tripathagā and punāṇā.

8

(Here, instead of the word “Ganges,” the poet uses only the epithet tripathāgā), ‘which flows in three courses,’ a term that furthers our understanding that (the god’s renown, like the Ganges) is capable of purifying the three worlds.

Thus, in those cases where a word, significant both in a unitary meaning and in the meanings of its constituent elements, is used to evoke an understanding that is based on the capacities of those elements, we accept two varieties (of such usage)—depending on whether another word, also expressing the unitary meaning, is used or is not used. No restriction applies therefore to the use (of a second term in the unitary sense).

Sometimes, however, a word that realizes its meaning both by synthesis and by ascription exhausts its fullest implication in expressing its (ascribed) unitary meaning only, because, in the specific context, its constituents’ meanings cannot be construed. For example,

O incorruptible (Viṣṇu)! This garland of budding flowers

Set in your belt like a string of celestial rubies,

Makes more brilliant yet the beauty

Of the lotuses budding in your navel—

From which emerge, one after another,

Creators of numberless eons yet to come.30

Here, the term “lotus” (amburuhā, lit., ‘water-growing’) is understood in its (ascribed) unitary sense only, because the meanings of its constituent elements cannot be construed in the context of a lotus that grows from the Lord’s navel.

Sometimes, on the contrary, (a word that realizes its meaning both by synthesis and by ascription) exhausts its fullest implication in expressing the meanings of its constituent elements only, because its (ascribed) unitary meaning cannot be construed. For example,

O lord! even though water lilies and night-blooming lotuses are ‘mud-born’

Only the day lotus is commonly known as such,

Thus, I think, the creator means to posit its nature as flawed

And affirms, O noble one! that not even it can compare to thy face!

Here, the term paṅkajā, ‘mud-born’ is understood according to the meanings of its constituent elements only, because its (ascribed) unitary meaning (viz., ‘day lotus’) cannot be construed in the context of white lotuses, water lilies, etc. (which do not close at night).31

But the Naiyāyikas (deny this; they) say (that in such cases) one must have recourse to metonymy, rather than (explaining them as involving denotation by) synthesis or ascription. Before terms such as “mud-born,” etc., which are already present to the mind as single words, have been construed (in context) with (other words) such as “navel-lotus” or “white

30 Varadarājastava, vs. 62. The reference is to the lotus that emerges from the navel of Viṣṇu at the beginning of each world-eon. From these lotuses emerges Brahmā, the creator, and from him all that exists.

31 Appaya seems here to recognize an almost botanical classification of the various lotuses and water-lilies, and to assert that some terms, which we often take as epithetically synonymous or generic, may in usage intend one species only—as, here, paṅkaja ‘mud-born’ designates the kamala ‘day lotus’ alone.

The Vrttivarttika

lotus," there is no grammatical necessity (ākān̄ksā), absent such a context,32 of construing (them) either in the sense of the generic term 'lotus', as qualified by 'growing in the mud'; or in the sense 'growing in the mud', as qualifying a 'lotus.'33 Nor is there any understanding consequent on the hearing of an unqualified (term) that would serve to construe it.' Therefore, (as we have said, metonymy, rather than denotation must be resorted to) in order that words such as "mud-born," etc., can be construed as presenting to the mind the independent notions 'lotus' and 'growing in the mud.'34 Nor (continue the Naiyāyikas) should the objection to this view be entertained which denies that metonymy is necessarily resorted to in these cases, because (two levels of interpretation may be distinguished:) first, both (meanings, viz., 'lotus' and 'growing in the mud') are independently apprehended, one by synthesis, the other by ascription, and then only, when we undertake to construe (those meanings) in the context of other terms, such as "navel-lotus," does one of the two meanings become mute, inasmuch as it has been contradicted (in context). It is now possible to construe (the sentence) grammatically with the aid of the (uncontradicted) function that remains, for (the other meaning), that of being a 'lotus', having been independently apprehended, is (still) present.35 (This objection cannot be sustained) because it involves the risk of postulating an unnecessary second reading (anusamdhāna) of the word "lotus" in order for it to be present independently by ascription alone (it having already been apprehended ambiguously, along with "mud-born").36 Now, if one were to accept metonymy (in such cases), this defect would not obtain, for there is no need to require again that which is dependent on the means, when the means themselves are present — (and they are present) in the shape of a metonymy that can supply either the qualifying term ["mud-born"] or the term qualified ["lotus"], (both of which are) present in the determinate concept that (invariably) associates 'being a lotus' with 'being born in the mud'. Nor can it be said (by those

32 Lit., in the context of navel-lotuses or white lotuses (which have not yet been mentioned, and are, therefore, not yet present to the mind).

33 In other words, any understanding of 'lotus', consequent on hearing the term paṅka-ja, is context-driven — whether of "mud-born" things taken as lotuses, or 'lotuses' taken as mud-born. Ākān̄ksā expresses the notion of sentential, or act-based implication, such that one element "implies" another — as the verb in active voice "implies" its subject in the nominative, or a sacrificial offering "implies" its instrument in the form of a oblatory substance, etc.

34 The Naiyāyikas do not admit that a word, pronounced but once, can be understood — denotatively — twice, as the Alamkārika view presented above would imply: namely, that the word paṅka-ja is understood denotatively in two different senses, once by rūḍhi, once by yoga, and that the appropriateness of one or the other or both meanings is then determined. Therefore, the apprehension of the second meaning — the ascribed "unitary" meaning — presumes a different power of the word, viz., lakṣaṇā, 'metonymy' [see ch. 2].

35 The "need" to construe, of course, is present, whether or not the ambiguity is removed; the objector's point is that it cannot be satisfied as long as the ambiguity is present. Once the ambiguity is resolved, the sentence may be construed via the implicatory power termed ākān̄ksā. The point is doubtless related to the Mīmāṁsā view that no action is possible unless it can unambiguously be "construed."

36 According to the Naiyāyika, the process suggested involves an otiose second reading of the word padma, which was first to be read by both yoga and rūḍhi, and then, after the other function became "mute," by rūḍhi alone. It is, of course, only the objector to the Naiyāyika view who considers the twin denotative functions (yoga and rūḍhi) adequate to explain the phenomenon in question; the Naiyāyika does not, but instead, resorts to metonymy.

10

refusing to admit the operation of metonymy) that the presentation of one meaning may be a means to the presentation of another—as is the case with metonymy—according to which they might not have to postulate a second reading of the term "lotus" in order for that it be present (to the mind), independently,37 by ascription alone. Thus say the Naiyāyikas.

Chapter 1

Multiple Meanings: Principles Determining the Intended Sense

[p. 54]

Thus, denotation has been described as threefold. Now, when a word has more than one capacity (to convey a meaning), its (denotation is) determined by means of (the principles known as) “conjunction,” etc.38 For it has been said:

When the sense of a word is indeterminate, the causes of its recollection In a specific sense are: conjunction, disjunction, association, incompatibility, Purpose, context, semantic implication, the presence of another word, Capacity, appropriateness, place, time, gender, intonation, etc.39

Here, “purpose” means ‘(speaker's) aim’; “context” means ‘that which (pre)exists in the minds of the speaker and/or hearer’; by “semantic implication” is meant a property capable of suppressing other meanings of a multiple-meaning word that has been used; “the presence of another word” means ‘the cooccurrence of a (second) word having another meaning that is itself related to one of the meanings of a multiple-meaning word’; “capacity” means ‘causal efficiency’ (kāraṇatva); “appropriateness” means ‘suitability’ (arthatā); “gender” refers to (the grammatical categories) ‘feminine, masculine, neuter’; the remaining (terms) are self-explanatory.

(We) now (exemplify each of these in turn). [a: conjunction] In the phrase, “Hari bears conch and discus,” the denotation of the word hari is determined in the sense ‘Lord’ by its conjunction with “conch” and “discus.”40 [b: disjunction] In “Hari bears no conch or discus,” it (is determined) by its disjunction.41 [c: association] In (the phrase) “Rāma and Lakṣmaṇa,”

37 I.e., as above, not in association with the “synthetic” meaning of the term.

38 Having laid out the types of denotation and commented on their interrelation, Appaya now examines the problem of multiple denotations in the same word—double entendre, pun. The perspective shifts from the principle of denotation to its exemplification.

39 This famous pair of ślokas is first attested, as far as we can now tell, in Vākyapadiya, 2: 315–16 (Abhyankar ed., p. 42, with var. lect.). They have been cited, with few exceptions, in every subsequent work on semantics. The examples given here have also, by and large, become traditional. Earlier works, however, anticipate the concerns addressed here, particularly the Brhaddevatā, which uses some of the same terminology in discussing fixation of the Vedic text's meaning (2.118, and see Raja, I.T.M., 48 ff.).

40 Hari may signify not only ‘Viṣṇu’ but ‘lion, monkey’, etc. Mention of two of Viṣṇu's standard insignia are here sufficient to determine which meaning is intended. Note that it is the cooccurrence of meanings that is here important, not the cooccurrence of words. See next.

41 Clearly, here, “disjunction” of meaning, not of word. It is Viṣṇu. not the lion, that may significantly said to be separated from such insignia. If disjunction of word alone were intended, the phrase might well be taken in the other sense. See discussion, below, in re “presence of another word.”

11

The Vrttivārttika

the word "Rāma" is determined in the sense 'scion of the Raghu clan' by association with Lakṣmaṇa.42 [d: incompatibility] But in "shadow and sunlight," the word "shadow" (is determined) in the sense 'absence of sunlight' because of mutual incompatibility with sunlight.43 (Also,) in "Rāma and Rāvaṇa," the word "Rāma" (is determined) in the sense 'Rāmacandra (son of Daśaratha)' because of the incompatibility of conqueror and conquered, that is, (of Rāma) and Rāvaṇa.44 But we deny that "Rāma and Arjuna" could equally well serve as an example (of "incompatibility"),45 where it might be alleged that the denotation (of "Rāma") is determined in the sense 'Rāma of the axe [Bhārgava],' because of the incompatibility of conqueror and conquered, that is, (of Paraśurāma) and the son of Kṛtavirya.46 For, only when the word "Rāma" has been determined to refer to Paraśurāma, is it possible, by recollecting their mutual hostility, to determine that "Arjuna" refers to Kārtavīrya (and not to the better-known Pāṇḍava); and vice versa, only when the (word "Arjuna" has been determined to refer to Kārtavirya), is it possible, (again) by recollecting their mutual hostility (to determine) that "Rāma" (refers to Paraśurāma). This (example) thus involves (the defect of) circularity. Consequently, only when one or the other of two words (is already understood) in a determinate meaning, is it possible, by recollecting a relation of mutual hostility,47 to determine the denotation of the other multiple-meaning word. Thus, the former example alone is to be adopted (as illustrative of "incompatibility"). The (latter) will be an example of "presence of another word" [(h), below].

[e: purpose] In "For liberation (mokṣa), have recourse to Hari," the denotation of the word "Hari" is determined in the sense 'Lord' by virtue of the purpose (mentioned), namely, liberation (from the cycle of existence). [f: context] In "Rāma entered the circuit of the ascetics' hermitage, dangerous of approach," (the denotation) of the word "Rāma" (is determined) in the sense 'scion of the Raghu clan' because of the context (mentioned). [g:

[p. 55]

42 It is Rāmacandra— not (e.g.) Balarāma or Paraśurāma, other well-known incarnations of Viṣṇu—who alone is associated with a younger brother, viz., Lakṣmaṇa. The fluidity of these categories becomes quite clear from these examples: "association" differs from "conjunction," apparently, by being appropriate to relations among animate individuals, rather than inanimate (property) to animate (individual). Such distinctions are, of course, very important in the Nyāya-Vaiśeṣika ontology, from which the tetrad jāti, guṇa, kriyā, and dravya has generally been co-opted.

43 chāyā may mean both 'shadow' and 'beauty'; ātapa, lit. 'heat', is semantically most often associated with the sun, and is so here translated, better to bring out the contrast.

44 The villain, of course, of the Rāmāyaṇa. Excluded are the other Rāmas. This second example is intended again to focus on animates, where "incompatibility" often means 'hostility'.

45 This was the example offered by Mammaṭa (Kāvyaprakāśa 2.19-20). Appaya's criticism is accepted by Nāgeśa, comm. ad loc.

46 Here, the first of the three Rāmas is, it is argued, specified, because of his well-known enmity toward Kārtavīrya, king of the Haihayas— whose name was also Arjuna (n.b., not the Pāṇḍava). The slaying of Kārtavīrya is mentioned several times in the Mahābhārata, notably, bk. 3, ch. 116 (crit. ed.).

47 smrta- seems otiose in the compound. We have been taking pratisamdhāna as 'recollection', but, more precisely, it means 're-connecting (with)'; hence, the compound may be more literally translated, 'by focusing (the mind again) on (the relation of) hostility between them that has (just) been remembered (or, that has become traditional)'.

12

semantic implication] In “The lord of the earth is angry,” (the term) “earth-lord” (is determined) in the sense ‘king’ by virtue of the semantic implication48 of (the notion) anger, which excludes (the possibility of reference to) a ‘mountain’.49 In the example given for “conjunction,” emphasis was put on the universally acknowledged character (prasiddhi) (of the relationship), since no contradiction (results) from the conjunction of conch, etc., also with another entity, such as Indra; in the example for “semantic implication,” on the other hand, (reference to) another entity is entirely excluded. Such is the difference.

[h: presence of another word] In “Gaze upon the Niṣādha, who bears the world,” the denotation of the word “Niṣādha,” which refers generally to a particular country, is determined in the sense of a particular mountain (of that country), because of the proximity of the term “world-bearer,” which means (here) ‘mountain’; similarly, the term “world-bearer,” which refers generally to a king [see above ex.], (is determined) in the sense ‘mountain’, in virtue of the proximity of the word “Niṣādha,” referring to a particular mountain.50 (Despite appearances to the contrary,) this is not a case of circular reasoning, for here, it is not the case that (prior) realization of a (word’s) meaning is required in order that the associated word be determined in (its) denotation; rather, it is the (fact of) association alone (that determines the meaning). The word must be properly formed (vyutpanna) and its meaning such that its connection (with the other word) has been grasped by virtue of that very meaning.51 Similarly, (we do) not (allege) circularity of reasoning in the case of those relationships where, for example, having recollected one relatum after seeing the other, (we opine that) it is the sight of that other, understood in its relationship (to the first relatum), that alone is required for the first to be recollected. (It is) not (the case) that, immediately

48 Of course, any of these modes of determination involve an “implication,” in the broad sense; linga (token, characteristic mark) is understood here in a manner akin to its Mīmāṃsā usage, where it expresses the second of six rationales for integrating disparate elements of the sacrifice into a coherent whole action. As a manner of indicating hierarchical subordination, it is “weaker” than “direct statement” (śruti) but “stronger” than “sentential interrelation” (vākya). It is thus an implicatory power based on the meaning of a word, rather than on its grammatical form or its place in the sentence. An even weaker indicator of subordination, interestingly for our present discussion, is prakarana—where the implication is not based on the word as such, but on its extra-sentential “context.” See Edgerton, tr., M.N.P., 74 et seq.

49 bhū-bhrt, lit. ‘earth-bearer’, said, somewhat after the manner of Atlas, both of mountains and sovereigns.

50 We might well wonder why Niṣadha may not as well refer to the king of that country, a common form of appellation, whose analogues are still in use today—witness “the Holkar,” or “the Scindia.” In Sanskrit, it must be admitted, we would expect Naiṣādha in this sense

51 Appaya seems to be saying (the syntax is convoluted) that it is not one meaning, au préalable fixed, that determines the other, but that the two meanings mutually determine one another by virtue of their very adjunction; it is the association of two words (fully vyutpanna—i.e., properly realized as words, apart from their syntactical relationships) that suffices to delimit the meaning of both. It seems a fine distinction. The case may be analogous to names such as “New York,” where it is merely the collocation that suggests the city, above and beyond new and york, whose individual meanings are not set aside (but which do not “determine” one another, either). Indeed, the word-meanings must be presumed, in this case. As with other such terminology, śabdāntarasāminnidhi may reflect, in its origins, Mīmāṃsā concerns: Kumārila (e.g.) uses the term padāntarasāminnidhi in clarifying the special force of the third kind of śruti, viniyoktrī, inasmuch as the other two kinds do not involve the audition of any word apart from themselves (either as stem or ending). On this, see the Introduction, above, and Kumārila ad P.M.S. 1.3.11 (A.S.S. ed., vol. 4, p. 213).

The Vrttivarttika

after seeing (the other), an additional recollection of the relationship (itself is required).52 Here too, for the denotation of the words “Niṣadha” and “world-support” to be determined, the mere collocation of the words “Niṣadha” and “world-support” is required, each of which is properly formed, and (in virtue of which formation) the meaning of each is grasped. Since a prior understanding of one or the other meaning (per se)53 is not required, there is no circularity of reasoning.

Now, it might then be objected: If this is the case, it would then be pointless to separate the examples of “conjunction,” etc. (from those of “presence of another word”), because also in (those examples, such as) “Hari bears conch and discus,” the denotation (of the word hari) is determined by virtue of the proximity of other words, such as “conch” and “discus.” We dismiss (this objection), because it is possible to determine the denotation (of hari) by means of “conjunction,” etc., (alone), for (by “conjunction,” the requisite) meaning is conveyed through merely hearing the words “conch” and “discus” in their determinate senses. If the objector replies that this is also the case as regards “presence of another word,” we concede the point. But since the words “conch” and “discus” are capable of determining the denotation of hari, etc., in terms of the immediately applicable relationship of “conjunction”—which has that (determination) as its purpose—they ought not to be taken as examples of “presence of another word,” because (“presence” implies) the express articulation of a word (such as hari, and articulation) constitutes only a mediately applicable relationship (to other words that also must be expressly articulated), which would apply only after the more immediate relation (which is based on meaning, rather than articulation) had opportunity.54 Thus, in cases such as “gaze upon the Niṣadha, who supports the world,” where no clear meaning emerges because both terms are ambiguous, the example should be taken as referring, rather, to the less immediate relation, which is also capable of delimiting (the meanings). Thus, there is no confusion (among the examples).

52 I.e., “Father” reminds one immediately of “son”; “son,” of “father.” We do not insist (in order to explain such recollection) that the relationship itself (the “father-son relationship,” as a third term) be also remembered. The example suggests that what is at issue in the case discussed is that, unlike Viṣṇu and his discus, no “relationship” need be presumed as the basis of clarification: merely the audition of the two words niṣadha and bhūbhrt in association is sufficient.

53 Appaya probably means—in terms of the discussion above—that “an understanding of the relationship between the two meanings is not required” which would itself serve to determine the meaning of the ambiguous term.

54 This reasoning is based on the grammatical distinction antaraṅga/bahi-raṅga, by which, in very uncertain parameters, the applicability of injunctive rules of word formation is ordered. The case often arises, both in the Pāṇinean system and in that of Mīmāṃsā, which are in many ways isomorphic, that two injunctive rules apply simultaneously, given certain conditions. A way must be found, not only to order their application, but to make sure that the “right” result obtains. Several protocols are adduced, antaraṅga/bahi-raṅga being, by all accounts, the vaguest. Nāgeśa appears to intend thereby that the rule whose occasioning factors are “contained in” (antaraṅga) those of another rule (ipso facto, bahiraṅga) has first opportunity—just as, when someone may be charged both with “possessing a dangerous weapon” and with “discharging it at someone,” it is normally the “narrower” charge that is regarded as more apposite. Appaya applies this principle to the present case, deciding that a relation based merely on “understanding” a word is antaraṅga with respect to a relation that is based on the necessity of speaking that same word—even admitting that the “under-standing” could not take place without the “speaking.” (By “speaking,” of course, we refer to the “express production” of the word—whether it be by voice or by print.)

14

Chapter 1

Now, (in the Pradīpa to Kāvyaprakāśā), “presence of another word” is (defined as): ‘grammatical agreement with a word of fixed meaning’. By that definition, such examples as “god hostile to the three cities” are intended—where the word “god” is determined in the sense ‘Śiva’, because the term “hostile to the three cities” (tripurārāti) is fixed in its reference (to Śiva).55 We consider this (an) improper(ly restrictive definition), because, where a cooccurring word is fixed in meaning, the determination (of an ambiguous word’s meaning) lies within the purview of “semantic implication” (līnga), and the like. Otherwise, even the examples put forth under “semantic implication,” etc., would run the risk of (finding their meanings) determined by “presence of another word.”56

Nor should an attempt be made to distinguish “semantic implication” from “presence of another word” by restricting the latter to instances of grammatical agreement and the former to instances of grammatical dependence—of which an example would be “the anger of the bearer of the world,” etc.57 For, in both cases, the determination (of meaning) is effected by “semantic implication,” understood via the words themselves (and not their grammatical relationships); hence, such an attempt to distinguish (the two types) is merely a commentator’s device.58 Further, (the objector’s) definition (of “presence…”) does not cover instances such as, “the beast is resplendent with showers.”59 It is universally accepted that such cases should fall (within the domain of “presence…”), inasmuch as the denotations of the words vyāla ‘beast’ and dāna ‘shower’, both of which are ambiguous, are there determined in the senses ‘wild elephant’ and ‘ichor’ by their mutual cooccurrence, even though they are in a relationship of syntactic dependence (rather than syntactically coordinate). Thus, the protocol that we have described here for applying the (disambiguating) notion “presence of another word” is the one that should be adopted.

[i: capacity] In (the phrase), “the cuckoo is inebriated with springtime/liquor,” the denotation of the word madhu ‘springtime, liquor’ is determined in the sense ‘springtime’ in virtue of the latter’s capacity to effect inebriation in the cuckoo. [j: appropriateness] In

55 See Pradīpa, p. 64. The epithet applies only to Śiva, and so is sufficient to fix deva in that sense. But the point of the commentary’s definition is the grammatical agreement between deva and tripurārāti (noun and adjective), which enables that fixation of sense. Appaya considers this an unnecessary restriction.

56 I.e., the two types would become indistinguishable—a cardinal sin in Indian poetics, which adores classification. It is interesting to note that Appaya frames the argument in terms of examples, rather than in terms of the formal definitions.

57 The difference is formal: in the former instance, an adjective agrees with its noun; in the latter, a genitive (or other oblique case) depends upon a noun governing it (but is not in agreement): “royal anger,” as opposed to “the anger of obliquie.” The point of the example, of course, is that mountains (also bhūbhrt) do not suffer emotion.

58 Paribhāṣāmātra: i.e., has no authority apart from the personal affirmation of the commentator. Such pejorative use of paribhāṣā is unusual; it certainly implies that a paribhāṣā was understood as a “trick” whose main function was to rescue a thesis from defeat—but had no intrinsic “authority.” Such was indeed the role of the Pāṇinean paribhāṣās, though they are accorded great respect.

59 vyālo dānena rājate: vyāla (nom.) may mean both ‘elephant’ and ‘snake’; dāna (inst.) both ‘ichor’ and ‘gift’. The example would appear to be excluded from the opponent’s proposed definition of “presence…,” because the words exhibit case dependence, and yet universal opinion has it that it is an example of that very type—as Appaya explains.

The Vrttivārttika

"let them have resort to Hari who are afflicted with the grief of existence," (the denotation of the word) hari (is determined) in the sense ‘consort of Lakṣmī’ [viz., Viṣṇu] in virtue of the latter's appropriateness for removing the grief of existence. (To further clarify) these distinctions: in the example for “purpose,” the dative case expresses the aim to be accomplished; whereas in the example for “capacity,” the instrumental case expresses the means. But in the example for “appropriateness” though neither (means nor end) is expressed, one understands a relation of cause and effect based on the cooccurrence (of words in the sentence), in virtue of their mere suitability (to one another).60 [k: place] In “Hari reposes on the milk-ocean,” the denotation of the word hari is determined in the sense ‘he who has Garuḍa for his ensign’ [viz., Viṣṇu], in virtue of the place (mentioned); [l: time] and in “Hari survives the universal destruction (of the world),” in virtue of the time (mentioned). [m: gender] In “the sun/friend appears,” (the word mitra ‘sun, friend’ is determined) in the sense ‘sun’, by virtue of its masculine ending—for, in the sense ‘friend’, the word has neuter gender;61 [n: intonation] and in the case of the revealed text: “There are to be three upasads for the Sāhna and twelve for the Ahīna,”62 the word ahīna, by virtue of the acute accent on its middle vowel [viz., ahína], is determined in the sense of a particular ritual called the “twelve day (sacrifice),” in which the soma is pressed over that number of days (viz., twelve), in accordance with the teaching (contained in the vārttika): “the suffix -kha [viz., -īna] follows the stem ahan-, ‘day-’, when (the name of) a sacrifice is intended.”63 If intended were the sense of the negative compound: ‘(it is) not deficient’ [viz., a-hīna], the word ahīna would have had the acute accent on its first vowel [viz., áhīna].64

By the term “et cetera” (in the verse samiyogo...) is to be understood both (reference by) imitative gesture and (by) exclusionary gesture. [o] Imitative gesture (abhinaya) is a configuration of the hands, etc., that mimics the visible shape of the (thing whose) meaning (is) intended; [p] exclusionary gesture (apadeśa) is the designation of the intended meaning by “grasping its horns,” (i.e., by pointing to it—deixis—rather than verbal or illocutionary reference).65 (An example of the former is:)

60 Again, a distinction more grammatical than real. Sāmārthya emphasizes the means (instrumental case); artha, the goal or purpose of an action (dative case); auciti implies perhaps both (or their relationship), but expresses neither. The influence of these distinctions of Mīmāṃsā, with its meticulous dissection of the (sacrificial) action, is also patent.

61 Mitra-s (masc.), though meaning ‘sun’ in classical Sanskrit, is a form continuing to refer to the Vedic god, who was often associated with the sun; cf. Avestan Mithras. The neuter mitra-m, though understood in the classical language in animate reference, probably derives from an older ‘contract, compact’. Though not exactly parallel, compare the pun “sun/son,” which can sometimes be distinguished on the basis of their gender concord.

62 So G. Jha, tr., Tantra Vārttika, 1223. The upasad is a ceremony preceding the main pressing of the soma, which, in this case, goes on for twelve days.

63 Vt. 2 and 3 (conjoined here) ad P. 4.2.43. In this sense, the word is to be analyzed, therefore, as ah-īna, with deletion of the n-final (by 6.4.145). The Kāśikā gives this vārttika sub 4.2.42, as an exception to āhnah, ‘a group of days’, the normal result of the rule, when the name of a sacrifice is intended (viz., ‘a sacrifice lasting for a group of days’).

64 So Sanskrit Wörterbuch, I: 576-77.

65 apadeśa, perhaps short for vyapadeśa; the idea is that, by pointing, the thing is singled out, extracted, from other possible references. The readings are confused here, some having upadeśa, ‘teaching’ — probably a lectio facilior.

16

Two years ago their (dual) first shape became apparent;

Last year they (dual) put to shame the swelling lotus buds;

This year they (dual) are of such size as this!

In years to come, how large might they become, thy breasts?

Here, the word iyat 'of such size' (is determined) in the sense of a specific measurement

by means of a movement of the hands indicating the size of (her) breasts at that time. (And,)

Since the demon (Tarakāsura) received his good fortune from his (person)

He ought not (to receive) therefrom his demise;

He who has cultivated be it even a poison-tree

Ought not to be the one to cut it down.66

Here, by (the word) itaḥ 'from this' — whereby (the speaker) designates himself — the

denotation is delimited to that speaker, namely, the transcendent (Lord, Brahmā).67

Pun: Where More Than One Meaning is Intended,

How is the Second Meaning Determined?

In some cases, however, one must apply68 a principle of delimitation (as described above)

more than once to (determine) the denotation of a word that has more than one capacity

(to convey a meaning), and thereby (derives more than one suitable) meaning. In these

cases, denotation is not delimited to one or the other (meaning).69 For example,

By whom Love was overthrown of old, and Viṣṇu's body made into a weapon;

Who twines necklaces of proud serpents and who bears down the Ganges;

Whom the gods call Hara, and whom they praise for his lunar diadem;

May he protect you himself, destroyer of Andhaka, ever husband to Umā!70

66 Kumārasambhava 2.55. Brahmā is speaking, to the effect that another god than Śiva must lead the battle against the

demons. The verse contains a famous bone of contention among the Vaiyākaraṇas, in that it appears to involve an ana-

coluthon and to defy normal Pāṇinean syntax. But can Kālidāsa write bad Sanskrit? (viśavrkṣaḥ is nominative—evi-

dently the subject of a passive construction, and at the same time the logical object of the infinitive chettum, which

demands an accusative. Cf. Vāmana, Kāvyālamkārasūtravṛtti 5.2.22; Mallinātha ad loc.; Kāśikā and Siddhāntakaumudī ad

anabhihite (P. 2.3.1). Nāgeśa (ibid.) emends: ity asampratam.)

67 This example seems a bit far fetched, illustrating, as it does, a normal use of the demonstrative pronoun. Pronouns

are, by their nature, anekārthavācī, but are not, by that token, usually understood as ambiguous or paronomastic.

68 praviṣṭam; var. lect.: aviśiṣṭam, “where, however, a principle…is indecisive…”

69 Appaya here begins a long and interesting discursus on the pun—which differs from the cases considered above

merely in that both meanings are intended, though one may be more “obvious” than the other. Puns therefore pose the

question of how such multiple meanings are apprehended. The principles of delimitation that we have specified would

seem either not to function as we have described them, or to function simultaneously and therefore ambiguously (a

contradiction in terms!). Are both meanings denoted—ergo, by two or more of the principles of delimitation operat-

ing simultaneously? Or is but one of the meanings denoted, while the other is apprehended by some other capacity of

the word, such as suggestion?

70 Parsed in this way, the verse refers to Śiva. The verse is discussed in Dhvanyāloka 2.21, in re the relationship of śleṣa

'pun' to other figures of speech that may thereby be suggested. Some later anthologists attribute it to Candraka (Ingalls,

Dhv., 293).

The Vrttivarttika

[By whom, free from suffering, the chariot was thrown down, and his own body made into a woman; Who, slaying the proud snake, was absorbed in vibration, and who bears up both mountain and earth; Whom the gods praise, saying, "he stole the head of him who afflicts the moon!" May he protect you himself, giver of lands to the Andhakas, ever benign!]71

Here, inasmuch as (the verse can function as a) laud either of Siva or of Visnu, the context is not decisive (to either meaning). (As praise of Siva, it reads:) "May the consort of Uma (uma-dhavah) ever protect you"; (he) "by whom Love was o'erthrown," that is, 'by whom the mind-churning one was made to disappear' (nāsita-manmathena); (other passages are to be resolved as follows:) "the Bali-conqueror's body," that is, 'Visnu's body', which was "led to the state of being [i.e., became] a weapon, formerly (or, in the cities)";72 "he who accomplished (for himself) garlands and bracelets made of lordly king-cobras, such as Sesa and Vasuki"; "he who bore the Ganges"; he "whose head the gods say is accompanied by the moon, and whose name, they say, is Hara"; and "he who has effected the demise of the demon Andhaka."

(As praise of Visnu, it reads:) "May Madhava himself, all-giving (sarvadaḥ), protect you!"73 (He) by whom, "exempt from (the travails of) existence," that is, 'not engaged in transmigratory existence,' the 'chariot' (anaḍ, Ved.), viz., the 'cart', was overthrown; and by whom the body of Balijit, that is, of himself, became feminine, on the occasion of the original churning of the ambrosia; and who slew (hā) the gaping serpent of time (kāliya). "In vibration" (rave), viz., 'in the absolute apprehended as sound' (śabdabraḥmani), (is his [Visnu's]) residence (layah), that is, he has assumed the form of that (original sound); or, alternatively, his is the "circumference of the spokes" (āra-valaya), that is, the 'wheel' (cakra), that slew (ha-) the prideful snakes — one of (his) steps; who, in his boar incarnation, supported the world (bhūmī), that is, the motionless (agam) Govardhana (mountain) and, as well, the motioned (gām) (earth); whose laud, they say, is "moon-afflicter-head-thief," that is, 'thief of the head of him who has afflicted the moon, namely, Rahu';74 he (Visnu) is the "domain-maker," that is, the 'founder of the residence' of the Andhaka kings.75

71Parsed thus, to Visnu. On the whole, the prima facie readings are on the side of Siva; in order to refer to Visnu, the readings must, on the whole, be tortured unconscionably: e.g., dhvastam ano [= anaḍ, 'cart' (R.V.)] 'bhavena; ... bhu-jamgahā ravalayo 'gam gām ca yo 'dhārayat; śaśi-math-śirah-hara. The only epithet pertaining evidently to Visnu is the last; to refer to Siva, it must be parsed: sarvadā umā-dhavah!

72The purāstrikrtah is glossed in most mss. as pureṣv astratām nītaḥ; or, this could be interpreted as purā iṣv-astratām nītaḥ, 'formerly became a weapon, viz., an arrow'.

73Also a common epithet of Kṛṣṇa.

74Rāhu (along with Ketu), an "invisible" planet, understood as the demon who "swallows" sun and moon during eclipse. According to one purāṇic account, Rāhu, disguised as an āditya, stole a sip of the ambrosia churned from the primeval ocean; Viṣṇu, once the fraud was detected, decapitated the demon—but, since Rāhu had drunk of the nectar of immortality, his severed head became fixed among the stars, and continues to "afflict" them periodically; his body, or "tail," became Ketu, source of comets, meteors, etc.

75Presumably, the same "Andhaka" whom Siva slew—but here praised as one of the progenitors of the Yadu clan, hence of Kṛṣṇa.

18

Chapter 1

In (deciding) these two meanings, denotation is the only function (we need appeal to).

It is for this reason that we here have an instance of the figure paronomasia (śleṣa).76

Well, then (says an interlocutor), let denotation function several times in those cases where context, etc., cannot be specified; but in other cases, even after the context has proven decisive in determining a certain meaning, another, non-contextual, meaning is apprehended. What then is our recourse? An example is:

That monarch yonder, rising by the Eastern mountain, lovely, ruddy-orbed,

Steals away the hearts of men with his soft rays.

[That king yonder, risen to prosperity, handsome, beloved in his domain,

Steals away the hearts of his subjects with low taxes.]77

Here, in the context of a description of the rising moon, a non-contextual (reference to a) king, qualified as "prosperous," etc., is understood.78

This (second reference) the older (schools of poetics explain thus): our only resort in such cases is the function of suggestion based on the (literal) capacity of the words themselves,79 for there is no other avenue (available).

The (second meaning) cannot arise here (through) denotation, because this has (already) been determined by context;80 nor (through) metonymy, for the primary meaning (of the words) has not been set aside.

Nor is it the case, as regards suggestion, that once a denotation is determined by context it cannot (further serve as basis) for bringing forth (a second, suggestive meaning),81 for it is precisely because (the suggestive meaning) cannot be determined (by denotation) that it has been postulated, in the first place, as a category separate (from the denotative).82

It is for this reason also that certain words,

76 Appaya, as will be clear, is here disputing the conclusion of Mammata and others that the second meaning of puns is apprehended via another function than denotation. Theirs was the prima facie view that denotation could function only once in a given word or string—any "other" meanings apprehended must be derived via another function—dhvani, "suggestion," for instance.

77 This second, punned meaning is often taken by the commentators in an ironic sense: the verse would thus serve as an example of "covert blame" (see Gerow, Glossary, 285-86). "Beloved in his domain" (raktamaṇḍalah) could also be read, and doubtless was, as, "By whom domains are beloved"—that is, who confiscates property at will. Appaya's reading avoids the irony.

78 I.e., qualified by the same epithets as have been used to describe the moon: not all need be puns (kāntimān, for example). The objector's point is that the second meaning does not here enjoy equality of opportunity with the first, but is apprehended only after the first is fully in place, in and through some ironic appropriateness then apparent (and it is realized that the epithets can refer to someone or something else).

79 śabdaśaktimūlo vyañjanāvyāpāraḥ: this is a technical term of the dhvani theorists. Suggestion (which by definition cannot be expressed "directly") may variously be grounded on that which has been expressed—either "words" as such, or their "meanings" (arthaśaktimūladhvāni), or even on another suggestion. Cf. Dhv. 2.20 et seq.

80 That is, the first meaning apprehended has exhausted the capacity of denotation.

81 This observation counters the implied objection that if denotation cannot arise a second time, once a primary denotation has been fixed by context (see above), then neither can suggestion arise, for, here too, a denotation has similarly been "fixed."

82 In other words, if this reason were allowed to rule out the possibility of the suggestive meaning, then no suggestion at all could ever take place—both an evident impossibility, and contrary to the very conditions that led to the postulation of dhvani in the first place. The dharmigrāhakamāna asserts that no rule can be invoked in such a way as to render its own status nugatory—the mouth can't bite the hand that feeds it!

The Vrttivarttika

[p. 59] which can give rise to another, indecent, meaning when they find their way into belletristic compositions (kāvya), are to be avoided even though their denotative function has been determined in another meaning appropriate to the context. Otherwise (if suggestion is not accepted here), no indecent second meaning would be possible, for once (a primary meaning, on this hypothesis) is determined by context, just as denotation (is exhausted), so also (is) suggestion. And so the defect (of composition) known as “the unrefined” (aślila) would not have been (recognized), as (it has been in this instance): “...he whose mouth is the birth canal of noble speech.”83 Therefore (concludes the interlocutor), (in cases of punning) where denotation, which has been chained (in place) by “conjunction” and the (other principles of determination), finds (its force) blunted, proper indeed is (recourse to) suggestion, as the only (possible) means of explaining the other meaning!

We, on the other hand, assert that the only function (needed) to explain both the contextual and the non-contextual meanings, in cases such as: “That monarch yonder, rising by the Eastern mountain...,” is denotation itself. For, like the contextual (meaning), the non-contextual is subject to denotative determination by the collocation (in a single sentence) of words like rājā, kara, and maṇḍala84—now expressing (in their second, or “non-contextual” reading) ‘royal personages’, and the ‘moneys’ and (subjects’) ‘properties’, etc., that may be seized (= meanings that here too) suitably combine by mutual implication (into a single utterance). For such a collocation is equally capable of determining the denotation (of its several component words) by appeal to the same principle, “presence of another word”— just as was the “contextual” (meaning, by reference to the collocation mentioned).

Those,85 on the other hand, who explain “presence of another word” differently will have to admit that even their version is capable of determining the (meaning of an ambiguous word or phrase), in view of the universally attested fact that words having several (possible) meanings are understood in (such and such) a meaning, according to the principles of association86 mentioned earlier.

83 yasyānanaṁ yonir udāravācāṁ: the aślila word here is yoni, which literally means ‘source, origin’, but “suggests” (inevitably, it seems, now even in transcription) one of its specialized meanings, ‘vulva/vagina’.

84 ... “king (of the night sky/of the realm),” “moonbeam/tax,” “realm (of stars/of courtiers”); the question is how these words get the meanings they have in the second (puned) reading—‘king,’ ‘tax,’ ‘domain’. Appaya asserts that, here too, their mutual compatibility and syntactic cooccurrence help determine their second meaning—just as their “first” meaning was determined. Therefore, denotation must play a role here too.

85 Appaya is probably referring here to commentators, such as Govinda on Mammaṭa, who take śabdāntarasaṁnidhi to mean niyata-arthaka-śabdāntara-sāmannādhikaranya: collocation with a word of determinate meaning. Thus, in devasya purārateḥ the word deva ‘god’ is taken to refer to Śiva, owing to appositional agreement with ‘enemy of the cities’—a common epithet of Śiva (G. ad Kāvyaprakāśa 19/20, p. 64 [A.S.S. ed.]). The difference, as will be made clear, is that, for Appaya, no word in the collocation need be préalablement fixé in its meaning: it is the “collocation” itself that determines.

86 samabhivyāhāra: or, collocation: construing, in the most general linguistic sense; the putting together of a word (or an idea) with the other words (or ideas) with which it occurs. The śloka of Bhartrhari may be said to formulate the possible parameters of that association.

20

Nor should it be asserted that denotation functions only in the contextual meaning because both “presence of another word” and “context” there (serve to determine that meaning, and thus) provide a superfluity of determination. Whereas, in the case of the non-contextual meaning, “presence of another word,” of the sort described, alone functions.87 For we have already said that the capacity to determine denotation that belongs to “context” (and the other principles of determination) is a power that inheres in each of them severally, once the (various) senses arising from a word of multiple meaning are recollected. (Bhartṛhari has said, in the verse cited above:) “When the sense of a word is indeterminate, the causes of its recollection....” So, if some one additional among these (causes is brought to bear in the fixation of the meaning), what more need be said—(other than that) the devices (for provoking) recollection are abundant!

Sometimes, a particular word with multiple meanings is intended88 even in its non-contextual (meaning), as:

When they praise you (o Kṛṣṇa!) for killing (evil) kings such as Kaṃsa As though they were mere deer,

It is, on the contrary, Hari’s condemnation (that is heard—Hari) who once confronted (wild) elephants: the demon(-army) led by Hiranyākṣa!89

Here, the word hari (is an example). For the hostility to elephants that is understood by construing hari syntactically with the word dvipa, ‘(elephant, who) drinks twice’ (helps) determine (the word) even in its non-contextual sense, ‘lion’.90 But, in the case “…he whose mouth is the birth canal of noble speech,” there is no factor similarly determining (the meaning) in the vulgar (i.e., non-contextual, sense of the word yoni). Just as the percipient faculty (manas), intent on recognizing some object, sometimes sets aside (that object) in (favor of recognizing) a bad smell, so usage also sometimes sets aside, in favor of a vulgar

87 The idea seems to be that two outranks one, and that therefore “one” has no opportunity to reach the threshold of denotation. (This would bolster the position of those who derive the second meaning of puns from some other function than the denotative.)

88 It has been assumed heretofore that the “contextual” meaning is intentional; the “non-contextual” accidental or derivative—though this ill suits even the example above, where the “king” is very likely the object of a veiled satire. In any case, at the level of “intentional” language—quite apart from the speaker’s intention—it is the plain meaning of the pun that first comes to mind (smṛti), and through that meaning, the second meaning arises. The case now being brought up by Appaya seems to conflict with that prima facie case.

89 Māgha, Śiśupālavadha 1.39. Intended are two Vaiṣṇava incarnations: Kṛṣṇa and Narasiṃha, the “man-lion,” who slew the elephant-demon Hiranyakāśipu, brother of Hiranyākṣa. The variant hareh (voc.) would have the verse addressed to Viṣṇu; the reading hareh is more ambiguous—as it should be!—construing either with tava (= Kṛṣṇa) or with ...dviṣah that intends Narasiṃha. The verse is a witticism, postulating a rivalry among incarnations.

90 Hari is a common epithet of Viṣṇu in all his forms, and a noun meaning ‘lion’ (it also can mean ‘monkey’)—see above. Elephants and lions are proverbial enemies. The “contextual” (i.e., prima facie) sense of the word is ‘Viṣṇu’, since the verse is evidently a laud to the god—but the word is “doubly appropriate,” we might say, in that the exploits of the god—who has assumed the form of a lion and who kills the elephant-demon, etc.—also bring to mind the “non-contextual” meaning, ‘lion’, which thus functions not simply as word-play, but reinforces the message of the verse. Daṇḍin distinguished between arthaśleṣa and śabdaśleṣa largely in this way—and, of course, the best puns always are thus “appropriate.”

The Vrttivārttika

[p. 60] (meaning), even the context that has determined the sense of a word—for no other reason than that such topics (force themselves on our attention).91 In (explaining none of) these capacities (of the word does) “suggestion” (play any role, hence it) need not be resorted to.

Nor would even the notion of “suggestion” be easy to comprehend in the stated examples, if we did not adopt such factors determining (also the “non-contextual” meaning).92 For, it cannot be asserted that in all cases the function of suggestion (evokes) the understanding of another meaning by setting aside context and the other (means of determination). In that case, the very attribution of such (a power of) determination to context, etc., would be pointless.93 And that (attribution) is certainly not for some unseen purpose94 but, indeed, for (the very obvious one) of stipulating the conditions under which one brings to mind a specific meaning when he hears a word that has several meanings, of which one is intended—as in the example, “bring the saindhava.”95 If in this case also we were to claim that the (contextually resultant) meaning was understood through the function of suggestion, then how can the (stipulations of condition)96 be justified? One cannot stipulate that a canal dug from a common reservoir to a field where water is wanted serve only the field where it is wanted, if another is also dug to one where-it-is not. Or again, (let us suppose that) the function of suggestion appears only (in those cases) where the aspiring poet intends also the non-contextual meaning (of a pun) in order to delight (the hearer), or where (he does not, but produces) the defect of vulgarity, and not in all cases (of double entendre).

Now, if, in order to justify such a stipulation, he (who argues in this way) is obliged to adopt, now and again, the determinative factors we have adduced, or others, he must admit

91 Lit., that such is the nature of the subject. Appaya seems to accept that prurient interests are so powerful as to override even the normal conventions of usage—an observation that few, familiar with colloquial usage in any language, would deny.

92 The conditional marks this as a contrary-to-fact supposition—a favorite argumentative ploy: it is, in fact, our position (delimitation by denotation) that presupposes yours (dhvani); so, even if yours be adopted, ours follows: yours is, at best, an unnecessary embellishment.

93 No one, says Appaya, sees these means (given proper opportunity) sometimes functioning, sometimes not: a “power of determination” that does not “determine” is a contradiction in terms

94 adṛṣṭārtham: a term the ritualists use to distinguish an evidently purposive act (such as winnowing the grain—to remove the chaff, before cooking the rice cake) from one which, though prescribed, appears to have no such purpose (such as pouring ghee into the fire). It is in fact the codification of these adṛṣṭārtha that constitutes the heart of the ritualists’ science—for the “unseen purpose” thereby revealed is nothing less than dharma, right living. There is a tinge of irony in Appaya's remark here.

95 saindhava, lit., ‘derivative of Sindhu’ (the river Indus, and the land of its lower reaches, Sind). Any “derivative” could be meant: ‘salt,’ but also ‘horse’ (both come from that area of the subcontinent). Context will decide: if spoken at table, one sense will be understood, if in the stable, another. A (bad) English example might be, “bring the frankfurter”—which might mean something quite different in the company of several Germans. (Even worse: Kennedy's infamous “Ich bin ein Berliner.”)

96 All mss. read eṣā (fem.); the antecedent is vyavasthā, ‘stipulation.’ Appaya's point is this: if “suggestion” is appealed to to explain wanted or unwanted second meanings of puns, why should it also not be appealed to to explain the choice among meanings that every ambiguous word requires—even where punning is not an issue? Of course, this argument is a reductio ad absurdum, for “denotation” would ipso facto cease to be a “power” of the word, and the various stipulative factors would cease to be relevant in its determination!

perforce that they alone do (serve there to) determine denotation. In those cases, the effort to postulate suggestion will have been pointless. Thus, even in the view of those who defend “suggestion,” it is difficult to avoid the function of denotation.

It follows that since “denotation” is unquestionably a capacity (of words), then it must function also in (determining) the non-contextual meaning (of puns).97 If, following the route declared in the Kāvyaprakāśa and other works, (we say that denotation is) that function of the word which takes the form of conveying a meaning that is directly empowered (by the word itself), then this is that (function). For there is nothing apart from the word (that could) account for our understanding an empowered non-contextual meaning.98 If, on the other hand, denotation (is viewed as) that which enables one to grasp the capacity (of the word), then why would the non-contextual meaning of a (word whose) capacity was not (yet) grasped99 not (also) be understood by denotation? (Could this be explained) in such a way as to be able to say that (denotation) was not (also there) present?

But (did you, Appaya, not say that)100 denotation is that which conveys an understanding through the capacity (of the word itself)? If you do not specify “(understanding) directly,” then denotation will be impossible to distinguish from metonymy, which takes the form of conveying another meaning that is related to the primary meaning already grasped and mediated by the understanding of that (meaning). And it also would follow that there can be no denotation in such cases either, for there too, a prior understanding, in the form of a problem to be resolved, and mediated by a pre-existing context, etc., quickly presents itself (to the mind)—after which, “context” itself (operates to determine the denotation).101 (This objection is) not cogent. The definition (of denotation as) that (function) which enables one to grasp the capacity (of the word) as referring to such and such an object102 is

[p. 61]

97 The idea is that abhidhā is so fundamental a function that no reason can be adduced that would limit its functioning in those cases where several meanings are understood in the course of an utterance.

98 śakya, for śakyārtha. Mammata’s definition differs from Appaya’s (supra) only in stipulating that abhidhā engenders a meaning that is suitably related to the word, rather adducing the engendering function alone. Since the word alone is allowed as means, the two definitions are operationally equivalent.

99 I.e., in the non-contextual sense. This definition differs from Appaya’s in saying that denotation does not convey the meaning, per se, but enables a “capacity” to grasp the meaning. Appaya asks, why should denotation not “enable” one to grasp also the second meaning? In other words, none of these definitions disempowers Appaya’s contention that the second meaning is grasped through denotation alone.

100 nanu introduces an objection, which concludes with iti cen na. The objector returns to Appaya’s own definition (supra), proposing that sāksāt ‘direct, immediate’ be added to it. Without this qualification, it is argued, the “indirect,” or “mediate” understanding that we call “metonymical” would fall within the domain of the “denotative.”

101 The objector here analogizes the several factors determining denotation to the mediating process whereby the metonymical meaning is arrived at. The “understanding” of context precedes the “understanding” of the denotative meaning; therefore, the latter, too, is “mediated.” The former dala of the objection reduced metonymy to denotation; this one reduces denotation to metonymy. Both attack the heart of Appaya’s position—his notion of abhidhā (as competent to explain even puns). If, says the objector, you cannot even propound “denotation” consistently, what am I to make of the rest of your thesis?

102 tadviśaya is added to Appaya’s original definition to preclude the obvious objection that one does not grasp any “capacity” at all, but the one only that is embedded in the word. When “dog” is uttered, I do not understand ‘cat’.

23

The Vrttivarttika

valid even without (adding) the qualification, “directly,” because no overextension (of the notion to be defined) results.103 For whether the (puned) meaning take shape in terms of the same set of words, or take shape as a consequence of a different division of those words, it belabors (the definition of denotation) unnecessarily to add a requirement that there be no interruption by another meaning—that is, that (the apprehension be) “direct.” In (the verse cited above,) “By whom Love was o’erthrown of old…,” (vena dhvastamanobhavena…), there is an immediate apprehension of meaning on the side of Śiva, owing to the currency of the words and the clear relations (between them); on the side, however, of Viṣṇu, “…by whom, free from suffering, the chariot was thrown down…,” the interrelationship of words is not clear,104 and words are used that are either unheard of or obscure,105 or are simply not current—such as “he stole the head of him who afflicts the moon!” and “(he is) giver of lands to the Andhakas.” (These usages) cause us to understand (the meaning) with a delay. (If the qualification “directly” were added), this could not be taken as an instance of “denotation”—which is contrary to everyone’s understanding (of the matter).

In those cases where a word is read in both contextual and non-contextual meanings, but is more current in the non-contextual, and less current in the contextual, it happens (prevailingly) that the meaning first understood is the non-contextual, because of its currency. (But, on the view being examined, namely, that the “latent” meaning of puns is apprehended by some function other than denotation), it is the contextual meaning that (would have to be apprehended) by suggestion, while the non-contextual (meaning) would be apprehended by denotation! This implication is self-defeating, and directly contrary (to what we all take for granted)—for one recollects first the more current meaning, bypassing even context. It is for this reason that the poetic defect known as “obscured meaning” has been accepted, and defined as the use of a word that is capable of being understood two ways in its less current sense.106 (It is do defined) because the understanding of the contextual meaning is interrupted by an understanding of a non-contextual meaning that is more current.

It should not be asserted that because “obscured meaning” has been accepted as a defect, its use in poetic composition is never permissible—even in those cases where the two meanings are intended. For the ancients have denied that it is a defect where pun and

103 anatipraṡaṅgāt: without the qualification “whose fuel is damp” the domain of the subject (pakṣa) of the proposition, “where there is fire there is smoke,” would extend too far, namely to red-hot iron balls, etc., which, though fiery, give no evidence of “smokiness.”

104 asphuṭa; here Appaya means that it is the word-division that is not clear: no one would read dhvasta-mano-bhavena (the prima facie division) as dhvastam ano ’bhavena (the division required on Viṣṇu’s reading). The words themselves may or may not be “clear”—that is still another possible defect.

105 nihatārtha, see below. A word used in a sense that is canceled, or blocked by, a more current reading of the same word.

106 See previous note.

24

"twinning" are concerned.107 And it is evident in (such universally accepted examples of poetic double entendre as) "(he is) giver of lands to the Andhakas," "that king yonder... steals away (the properties) of his subjects," and "for whom [viz, Śiva] the crescent moon is crest-jewel, and whose feet are honored by kings."108 Therefore, only a positive difference among properties conducive to the understanding of a meaning can motivate a difference of function—never the absence of such difference.109 And though there may be a difference between a "direct" and an "interrupted" apprehension (of meaning), the difference between denotation and metonymy is stated uniquely in terms of a positively realized difference in the relation itself between the function and the meaning derived from it.110 In the absence of such a difference, as in cases of morphological punning that involve more current and less current meanings, or in metonymies based on conventional associations, or in the case of suggestion (itself)111—even though all these may illustrate a difference between "direct" and "interrupted" apprehensions of two meanings—no additional function has been adduced beyond denotation, metonymy, or suggestion (which are required in order to explain the respective apprehensions).112 The qualification "directly" has not been added (to the definition of denotation), as being both non-conducive (to an explanation of the apprehensions that are characteristic of the three functions); and positively misleading (in that it posits differences where there are none). Thus, (our definition is without blemish, namely:) denotation is that (function) which enables one to grasp the capacity (of the word) as referring to such and such an object. Therefore, it is difficult to avoid (reliance on) the function of denotation, when (explaining the apprehension) even of the non-contextual meaning (of puns).

107 "Twinning" (yamaka), which I represented as "cadence" in my Glossary, is the obverse of pun: instead of a single utterance understood twice (or in two ways), yamaka repeats the utterance in coordinated parts of the verse, understanding it one way in one, in another in the other.

108 Here, the first two words, rājā rājārcita... illustrate by way of yamaka the same two meanings of the previous example ('king, moon').

109 I take it that Appaya means, by this remark, "not the mere absence of such a property." He thus stipulates that the difference between "direct" and "mediate" apprehensions reduces to one of 'absence' or 'presence' of an intervening element—not sufficient to motivate a difference of function, being external to the function itself. The next remark, true to śāstraic logic, then admits (for purposes of argument) that it is that difference that motivates the distinction at issue.

110 That is, in the way the function makes its meaning intelligible. Metonymy and denotation are not distinguishable in terms of the presence or absence of some intervening element, but in the fact that a condition for the emergence of the metonymical is some failure of the denotative. The metonymical presumes the denotative, and is thus situated in distinctive and "functional" relationship to it—not because it lacks, or possesses, a distinctive "element".

111 The first is the case we have been discussing; the second (laksitalakṣaṇā) is illustrated by such words as dvirepha, 'bee', which has that "meaning" because a more current word for 'bee', bhramara, has "two r's in it"; the third points out that all cases of dhvani involve both the "suggested" meaning and the literal meaning, which is never set aside, as is the rule with lakṣaṇā. And some involve also a delay in apprehension.

112 The three examples involve, respectively, the three functions; each demonstrates both a "direct" and a "delayed" meaning; in none of the three need one appeal to any function other than the one primarily adduced. The case of denotation is thus not in any way unusual: its "second" meanings need no more an extrinsic explanation than do those associated with metonymy or dhvani.

The Vrttivarttika

The Logicians, however (view these matters differently). If (a word's) currency is well established, one recollects its meanings without reference to context and (the other factors listed); in the absence of such currency, there is no recollection of context and the rest. Therefore, one cannot speak of context and the rest as determining denotation. Denotation (they say) can be determined only through the skills that are based on the continuing refinement of our ability to grasp (a word's semantic) capacity.

113 (Context, and the rest,) however, may help to make known the (the speaker's) intention by a process of implicative (reasoning). Thus, when we hear the phrase, “there's hari,” in the absence of any prior (utterance) to which it might be related, we determine (all) the meanings of hari (simultaneously, namely:) 'Yama', 'wind', 'Indra', 'moon', etc. But without knowing the intention of the speaker, the sense is not thereby determined.

114 If one argues in this manner, it will not even be possible to imagine how context and the rest might determine denotation.

Therefore, denotation alone is the function (to be adduced) when (deriving) the meaning of an utterance that involves both plain and recondite (levels), and which (is intended) in both plain and recondite (senses). Once the meanings of its words are thus provided, (both) meanings of such a sentence (arise) with the help of syntactical connection, etc., Thus, one need not resort to suggestion (to derive) either our understanding of the word meaning or the sentence meaning.

Some of our predecessors, in (explaining the) recondite (sense of certain expressions, have asserted that the) denotative meaning (can arise) out of the suggestive function, itself based on (the literal) capacity (of the component words).

117 This they have done with the intention of affirming, necessarily, the presence there of an understood simile, which is itself the basis of our apprehending the (second, more) recondite meaning. In such expressions as, “…risen to prosperity…,” (this second sense) is understood as qualifying an earthruler, who delights the hearts of his subjects with light levies of money, etc., just as the moon (does) with its soft rays. These (predecessors of ours) did not (adopt this view) with any real intention of asserting the presence there of the function of suggestion. Even they offer this (verse) as an example of suggestion of a figure of speech based on the literal power (of words):

113 I.e., by a process of practical approximation that is experiential and cumulative.

114 I.e., without knowing also the intention of the speaker, no fixation of meaning is possible. And this is known, not by semantic function, but by a process of reasoning (linga) — experientially, in other words. The Naiyayikas are, in this respect, behaviorists and nominalists.

115 Lit., it is not possible to suspect (na…śakyasańkam).

116 Appaya appears to take this as a self-defeating conclusion, not requiring further comment.

117 A dhvani may be of a declaration of fact (vastu), of a figure of speech (alamkāra), or of a rasa. The former two are at issue here, though Appaya here discusses only the second (but see below, n. 123). The idea is that the second meaning, though understood literally, is arrived at through a kind of suggestion, based, as he will make clear, on an unspoken similitude.

Chapter 1

Covering over the wide expanse of sky, mounting high above

The earth, fixed there, peaks wide and lofty;

Who on earth, gazing so upon him, crown shimmering with snow,

would not amazéd be—the lord of mountains, Himālaya!118

This is made clear, as well, by the way the puns on the words rāga, etc.—even in their

recondite (meaning)—are explained by reference only to the function of denotation, when

(Ānandavardhana) illustrates as follows the figure samāsokti, which is occasioned by a com-

monality of epithet achieved through punning: “The face of night, stars (like eyes) trem-

bling, is seized by the moon, ruddy (with passion).”119

Nor can the view be maintained that the two lovers, though themselves not (referred to

in the) plain (reading of the verse), are not entirely unrelated to the context (of the verse),

inasmuch as they can be construed as qualifications of the moon and the night, which do

constitute the plain meaning.120 (This view is not tenable) because the recondite meaning

is understood, in accordance with the definition of the figure samāsokti, by virtue of a sim-

ilarity of qualification (with the plain meaning). Therefore, even before the recondite

meaning is-understood as having a relation to the context (of the verse), those qualifica-

tions, on which depend (our understanding of the recondite meaning) must have found

expression through denotation.121 And also because the relation to context (which is not

yet understood) must, here also, somehow or other, come about through the objects of

comparison,122 which have been plainly expressed. For there simply does not exist, in the

entire range of the poetic imagination, a recondite meaning which has no relationship to

the plain meaning at all.

118 Śiśupālavadha 4.19. Suggestion may be classified (see also below) according to the means of suggestion—the most

basic division being that between ‘sound’ (śabda) and ‘sense’ (artha), i.e., language and reference, denotation and impli-

cation. Śabdaśaktimūladhvani does not refer uniquely to the sounds of language as devoid of meaning—but does

include this element. See Ingalls, Dhv., 794, and refs. Here, nagesa can refer to Śiva—also ‘lord of the mountain (=

Kailāsa)’; and the epithets can equally be read to refer to him: digambaram, ‘naked’, gām ākramya, ‘mounting the bull’,

etc. The verse is not discussed in either Dhvanyāloka or Kāvyaprakāśa. Mahimabhaṭṭa mentions it in discussing defects

of śleṣa (Vyaktiveka, 401, Revaprasād Dvivedī ed.). His analysis resembles that discussed here—the second meaning

is arrived at by a parāmarśa (lingaviddhayā, supra). But it seems unlikely that Mahimā would have proposed “suggestion”

as a solution to any problem! The figure at issue is samāsokti (Glossary, 316 ff.). In samāsokti, the implicit subject of

comparison is never mentioned: it is “suggested” via an understood parallelism of epithet, which may involve śleṣa—

that is, the epithetical descriptions can be read also to refer to the implicit term; thus, the figure always involves a sug-

gested simile. Appaya’s point is thus that the real issue here is the pun, not “suggestion,” and that if there is a suggestion

of a figure of speech, it is possible after the pun is recognized, not before.

119 See Dhv. ad 1.13, where this is taken as an illustration of dhvani subordinated. The verse “suggests” two lovers embrac-

ing through a commonality of epithet, e.g. rāga means both ‘red’ and ‘passion’. Appaya interprets Ānandavardhana’s

classification—the clever pun is primary, the element of suggestion secondary to it—as confirming his view that the

meanings of the pun cannot be derived through suggestion.

120 That is, the “moon” and the “night” are depicted themselves as “lovers.”

121 Lit., “denotation must be expressed in those qualifications.”

122 upamāna: in the example, the moon and the night. The subjects of comparison (upameya)—the two lovers—are

implicit in the figure.

The Vrttivārttika

One might raise the following objection: now even though “covering over the wide expanse of sky...” may indeed be taken as an example of the suggestion of a figure of speech based on the literal power (of words), the (type of suggestion from which it is distinguished, namely) suggestion of a matter-of-fact, which is also based on the (literal power of words),123 would become impossible; for the suggestion of a matter-of-fact is to be illustrated in a recondite meaning that is devoid of any figure of speech—and there you accept denotation (as sufficient to explain the resultant meaning)!124 (This objection is) without merit: we have all along accepted (suggestion of a matter-of-fact) in those cases where a matter-of-fact, not otherwise expressed, is understood on the basis of a recondite meaning that is itself apprehended through denotation alone.125 For example,

Saturn (śani) and the thunderbolt (aśani) destroy him on whom thy displeasure falls, o lord of men...126

[p. 64] Here, another meaning of the word aśani, apart from its plain meaning (viz., ‘thunderbolt’), is apprehended by denotation alone—a meaning arising from the apparent contrariety (of aśani and) the word śani (‘Saturn’), with which it is construed.127 The preeminence (of the king’s) power is conveyed by virtue of the thereto subsidiary implication: “though the (two) be contraries, they cooperate in a single effect, namely, furtherance of obedience to thee.” Thus was suggestion of a matter-of-fact based on the literal power of words illustrated by our predecessors.

Here another objection (is possible). Consider this example:

The rain(-poison) born of cloud-snakes

Afflicts women longing for their lovers

With confusion, anxiety, lassitude,

Faintness, loss of consciousness, and dark death.

Here, the recondite sense of the word vīṣa (viz., ‘poison’) is merely suggested, but at the same time brings to completion the literal metaphor: “cloud-snakes.” Mammatācārya, here illustrating suggestion subordinated (to the literal meaning), yet instrumental in the

123 vastudhvani: another of the various classifications of dhvani (see above), perhaps the simplest—by “object” suggested: a ‘state-of-affairs’ (vastu), a ‘figure of speech’ (alamkāra), or an ‘emotional state’ (rasa). Vastudhvani’s scope is thus wide, including irony, most of what we call double-entendre, and wit of all sorts.

124 In other words, the most elemental dhvani, vastudhvani, would disappear as an instance of dhvani: such cases would reduce to denotation—so claims the objector.

125 vastu is not reducible to aprastuta. Appaya will now defend at length this distinction.

126 Example cited in Kāvyaprakāśa 4 (śl. 60, p. 139, A.S.S. ed.), for vastudhvani. Saturn, of course, is associated with bad luck.

127 It must be understood that the prefix a- is often a sign of the negative (though it is not, in the case of aś-ani, ‘thunderbolt’). The apparent opposition with śani forces a remorphemicization (a-śani), which suffices to express (says Appaya) an aprastutārtha.

accomplishment of the literal meaning,128 says that even in the recondite meaning (viz.,

'poison') here conveyed (śakye 'pī), there is suggestion of a matter-of-fact based on the literal power (of words).129 Indeed, so! (As we said before) this (view), that suggestion (is to be accepted) even in the domain of the recondite meaning, was propounded merely for purposes of argument.130 This was seen as a way of affirming emphatically the necessity

that suggestion (be adduced) both in (explaining) matters-of-fact and figures of speech that were implied by an understanding of the recondite meaning. We need not (waste further) effort on it.

Similarly, it might be objected that, if one adopts suggestion (to account) for those matters-of-fact and figures of speech that cannot be derived from an examination of the plain meaning, then this reduces to suggestion based on meaning (of words).131 This being the case, there would (no longer be any reason to distinguish) suggestion based on meaning from that based on the literal power (of words)! (This objection) is cogent. For, even if

that were the case, there would still be reason to distinguish the two, inasmuch as we do not tolerate any pleonasm of synonymy even with the plain (meaning), in relation to a word that is general to both the plain and recondite meanings.132 In this way, we may also set aside many other minor difficulties attendant upon a refusal to adopt suggestion in (explanation of) the literal meanings of recondite (expressions).

Thus ends the first section of the Vrttivārttika, composed by Appaya Dīkṣita, named “Definitive (treatment) of the primary function (of words).”

128 Kāvyaprakāśa 5 (śl. 127, p. 200). Suggestion is again classified according to its relation to the literal, or denotative meaning: it may be absent (citrakāvya), or present and subordinated (gunibhūta), or present and predominant (dhvani per se). This amounts to a classification by ends or purposes, and is fundamentally evaluative—for, according to the dhvani theorists, the best poetry always aims at suggestion.

129 This would, evidently, constitute a challenge to Appaya's view that suggestion plays no role in the apprehension of recondite meanings of puns. Mammata, in fact, does not say this, but it may be implied by the peculiar structure of the verse in question, where one metaphor (cloud-snake) is spelled out, but the complementary metaphor is effected by one word, which bears both meanings (water-poison). Both meanings are therefore “literal”—by Appaya's argument, and by common acceptation; nevertheless, one of them is here taken to be “suggested”—in order to preserve the obvious parallelism with the other metaphor.

130 praudhavāda: a “display” thesis—one not intended as integral to one's own position, but whose “extreme” form may protect another, more moderate, view. Appaya's view is that, in these cases, suggestion is always secondary to denotation; the dhvani school, however, stressed the element of suggestion in order to further their general thesis—not with the specific purpose of deriving the denotative meaning from the suggested.

131 arthaśaktimūla-, opposed to śabdaśaktimūla- (dhvani). See above.

132 That is, it is still considered a pleonasm to employ a word repeating the “plain” meaning, where another word is used in both “plain” and “recondite” senses. This kind of pleonasm could not be defined unless the implication of duplicity based on the “meaning” (artha) is distinguishable from that based on the “word” (śabda).

Chapter Two: Metonymy

Metonymy and its Types

[p. 67]

Metonymy is our next subject of demonstration. It is (defined as) the capacity of a word

to convey a meaning in virtue of a relation between (that meaning and the word's) primary

meaning. Now, when (another) meaning (is conveyed) in virtue of a similitude, (this has

been called) a "secondary" (mode of usage), but is nothing but a kind of metonymy.1 For

even (when we say that "this is) similar to that," there exists a mediate relation2 among (the

referenced) substrata such that similitude is predicated of (all of) them.3

For we do not accept the restriction that metonymy must be based either on an imme-

diate relation or on one giving rise to the concept of a qualified (entity).4 If such were the

case, then "contact inherence" relations5 such as that between the eye and the blueness of

the pot, which are mediate relations that do not give rise to a determinate concept, and

(yet) are comprehended as immediately proximate (pratyāsatti), (would have to be taken

1 gauni (scil., vṛtti, 'mode') and lakṣaṇā were considered, by the Mīmāṃsakas, two coordinate types of word usage. The

issue that Appaya joins is not their essential difference—on that there is no disagreement, the principle of meaning

shift in the one case being similitude, in the other, any other relation—but rather, their architecture. Appaya wishes to

subordinate the former to the latter, making the latter into the sole principle, and thus reducing the four 'modes' of the

Mīmāṃsā to three. For the Mīmāṃsakas, gauni was not a relation in the same sense as were cause-effect, etc., for it was

indeterminate, in the sense that its realization did not appear to modify anything—it was, in Peirce's sense, not "real."

See my "Language and Symbol in Indian Semiotics," PEW 34: 245-60.

2 paramparāsambandha: that is, a relation through a third term, rather than a sāksāt 'direct' relation (see below)—in

this case, the "common property," or tertium comparationis.

3 In other words, the "common property" (being possessed by both relata) is itself a relation—or the ground of a more

abstract relation.

4 sāksātsambandha, viśiṣṭabuddhiyogya sambandha. The former is to be distinguished from paramparāsambandha—

mediate relation (above): contact (samyoga) is an instance ('the pot on the ground'). Possession illustrates the latter—

for when Devadatta possesses a stick, dandi Devadattah, we can say of him: tadviśiṣṭo Devadattah: D. is "qualified"

(viśiṣṭa) by the stick. Similitude, being a relation involving a third term, is excluded from the former notion, and from

the latter, because, even though "her face is like the moon," no notion that the face is qualified by the moon, or that it

"has" the moon, arises.

5 samyuktasamavāyasambandha: a complex relation allowed by the Nyāya. The eye and blueness, though understood as

immediately related, are logically mediated by a third term, the pot, which is blue. The eye maintains a relation of

samyoga 'contact' with the pot; the color, a relation of samavāya 'inherence' to the pot. This "immediacy" would quali-

fy the relation as supporting a metonymy—but no one uses the "eye" and the pot's "blueness" metonymically!

31

The Vrttivarttika

as supporting a metonymy!).6 Further, metonymy is admitted in cases of “metonymical difference” (vyatirekalakṣaṇā), by virtue of a mediate relation that is presumed among (the referenced) substrata such that a contradiction can be predicated of them.7

This is also sufficient to refute the view that secondary (usage) must be different from metonymy because similitude is not a relation—for only that can be called a “relation” which can subtend the predication of a qualified concept. (The usage) “spear-bearing Devadatta”

(is justified) if actual contact (of D. and his spear) exists; on the other hand, although a similitude may indeed exist, no qualified concept (of the form) “Devadatta possesses a lion,” siṃhavān devadattah (arises from the usage ‘D. is a lion, siṃho devadattah).8 (We reject this contention), having already pointed that metonymy might well be presumed where there is an understanding of immediacy involving nothing but a mediate relation unqualified by any determinate concept.9

Metonymy is thus of two sorts: “secondary” and “pure”—depending on whether (the transfer of meaning) is prompted by a relation of similitude, or by a relation other than that. And each of these may be further subdivided, (which two types we may term:) “conventional metonymy” and “motivated metonymy.” Conventional metonymy (is so-called) because it is equivalent to convention(al usage). Motivated metonymy (as its name implies) has an aim (phala), namely, the indication of a meaning beyond (or, in addition to) the intended meaning. As an example of the former,

I suspect, o Viṣṇu! that you have greater affection For her who is herself an ocean of loveliness (= lakṣmī), than for The daughter of the milk-ocean (= Lakṣmī), In that you bear the former on each and every limb of your body, Whereas the other is borne merely on your breast.10

Here, the word lāvanya, ‘beauty’, expressing (literally) ‘saltiness’ (is an example of) conventional metonymy, used in (the sense of) a particularly exquisite (loveliness), on account

6 In other words, such an apparent restriction of definition would actually overextend the definition of metonymy! These relations are immediate—we see the blue, as well as the pot—and do not involve a “determinate concept”—in the sense that blueness is not predicated of the eye.

7 This type of metonymy, admitted by all, is functionally equivalent to irony: “John is a tower of strength”—which often means, ‘John is not a tower of strength.’ Appaya points out that ‘dissimilitude’ cannot be admitted as metonymical without ipso facto admitting ‘similitude’.

8 Again, by “secondary” (gaunī) usage is meant those transfers of meaning based on similitude, rather than on some other relation. This objection, probably based on the Mīmāṃsā, would restrict metonymy (lakṣaṇā) to relations of the second sort only by making the strong claim that similitude is not a relation at all.

9 Mutatis mutandis! The above argument about blue pots is inverted so that relations based on similitude—neither sākṣāt nor viśiṣṭadhīṣamyukta—are included within the metonymical domain.

10 Varadarājastava, vs. 24. Lakṣmī is conventionally represented as reposing on Viṣṇu’s chest. The verse plays on the two senses of lakṣmī: ‘loveliness’ (in general) and ‘Lakṣmī’ (consort of Viṣṇu). The puzzling lāvanyasāgarabhuvi may also conceal a reference to Bhūdevī—Lakṣmī’s rival. The juxtaposition of sāgara and bhū is certainly odd. In any case, none of this concerns the point at issue, which is restricted to the one word, lāvanya, ‘beauty’—whose literal ‘saltiness’ has become a metonym for any ‘spice’?

32

of the similarity (that both salt and beauty possess, namely) an ability to touch the heart,

(the seat of sensibility). Similarly, conventional metonymy is to be suspected when (we

apply) the word “elephant” to the elephant in a painting. This example (demonstrates a

conventional metonymy) of secondary type.11 The pure (type is as follows):

The mango-creeper, extending an arm-like branch,

That jingles with a dark bracelet of swarming bees,

Embraces the southern wind and asks

His health with cuckoo-calls deep throated.

Here, conventional metonymy (of the pure type) is illustrated by the word dvirepha,

which acquires the (meaning) of the word bhramara (‘bee’, the wandering [insect]) by an

etymological analysis (vyutpattȳa) (of the latter word), namely: ‘(that creature whose name)

contains two r’s’.12 Similarly, pure conventional metonymy is to be noted when words like

tvak, ‘skin’ are used to signify the sense of touch, etc.13

Motivated metonymy is seven-fold: according, firstly, as to whether (the literal meaning

is) [i] given up altogether,14 or [ii] not so given up,15 or [iii] both given up and not given

up;16 then, (whether the metaphorical meaning is) [iv] superimposed on, or [v] identified

with (the literal meaning);17 or finally, (whether, as above, the metonymy is) [vi] pure or

[vii] secondary.18

(The first), metonymy (where the literal meaning is) given up altogether, is as follows:

How great a penance, O lord of the mountain!

Must the wishing trees do on the celestial Ganges,

Clad in bark clothing, wearing their topknots,

In order to win those floral ornaments—

The Vrttivarttika

Fresh shoots lovely as coral19—that might

Vie with your lovely lotus feet?

Here, “on the celestial Ganges”20 is metonymically understood in the sense ‘on the banks

of the (celestial Ganges)’, suitable as a locus for the wishing trees. Now, because (“on the

Ganges”), the literal meaning, cannot be construed as a (suitable) locus for them, (we

have) a metonymy whose (literal meaning) has been given up. The motivation (of this

metonymy) is to indicate the exceeding sanctity of those banks, which embellishes the fanci-

ful attribution21 of asceticism to the wishing trees. The well-known example, “a village on

the Ganges,” may also be suspected (of illustrating this type).22 The motivation (of this

usage) is to indicate the extreme coolness of the banks (of the Ganges) as (a place of) delec-

tation for the cowherd.

Now, “metonymical difference” (vyatirekalakṣaṇā) is nothing but a type of a metonymy

whose (literal meaning) has been given up. For example,

You've done well! What can I say?

Your politeness is legendary.

Doing such things, my friend,

May you rest happy for one hundred autumns!23

Here, the words “done well,” etc., metonymically express the contrary of their proper

meaning, being unsuitably construed with one who has done harm (to the speaker). The

motivation is to indicate (the speaker's) own rectitude: “although you've done me a dis-

service, I speak only pleasant words (in reply).”

(Metonymy where the literal meaning) is not given up is as follows: “The swords are enter-

ing, and the lances!” Here, the words “swords,” etc., metonymically express the ‘men hold-

ing the (swords, etc.)’ — the context of the statement being a rout from fear (of the “swords”).24

(Even though swords are incapable of intentional movement), the motion of entering can

19 pravālapadabhānji: lit., ‘affected with the name pravāla (which means both “coral” and “fresh shoots”);

20 suradirghikāyām: lit. ‘on the gods' pond’ — the Ganges before it fell to earth on Śiva's bow.

21 utprekṣā: a figure with various definitions, but the most common is suitable to this usage. It is essentially a rūpaka

where the object of comparison (upamāna) has been resolved into a verbal predicate. Instead of saying, “the trees are

hermits,” the poet says, “the trees practice penance.” See Glossary, 131 ff.

22 Indeed, the example offered by Appaya makes exactly the same point, mutatis mutandis: the literal sense of the loca-

tive, ‘on’ or ‘in’, when attached to a word designating a liquid environment, cannot be construed, when the term requir-

ing the locative needs dry land for its adhikaraṇa, ‘locus’. Both examples, too, explain the motivation of the metonymy

as a transfer of property from the liquid to the solid — this is the element of “suggestion” that dhvani theorists find

in all serious metonymy — but it would ex hypothesi be absent in rūdhilakṣaṇā.

23 Cited in Kāvyaprakāśa 4 (śl. 25, p. 86), as an example of viparıtalakṣaṇā (same meaning). The verse is taken ironical-

ly, spoken by a lady to her lover, who has made love to the go-between instead of simply listening to her message. A sim-

ilar verse, in Prakrit, is found in Dhvanyāloka implicating the messenger in deceit.

24 bhitipalāyanavākye: this element, though deleted in the N.S.P. ed., appears (though often garbled) in all the mss. con-

sulted. The compound (if an ablative tatpuruṣa) is irregular: one would expect bhitipretupalāyana (madhyamalopatpat-

puruṣa?), or no compound at all. Cf. P. 2.1.37 et seq.

here be construed, because they are held by men (who do move): thus (the literal meaning)

is not given up.25 The motivation is to indicate the fact that the (men) strike mercilessly.

In usages such as “the village is burning” or “the forest is flowering” — when only a part

of the village is on fire—we have (an instance of) metonymy (where the literal meaning)

is both given up and not given up, inasmuch as the word “village” intends only a part of the

village, the rest of its proper meaning having been set aside.26 The motivation is to indicate

the wide extent of the conflagration, etc.

Superimposed (metonymy) is as follows:

O lord (Viṣṇu)! since the Earth is ever satisfied

Sipping (the nectar of) your face-moon without blemish,

How is it that this other moon, stained with spots,

Has become a source of delight, with its novel nectar?27

Inasmuch as the word “moon” has here been superimposed on (the word) “face,” because

of a similarity in attractiveness, (this illustrates) a secondary superimposed metonymy.28

The motivation is an intuition (that the two) have the same form. Thus, even when (the two

terms) are uncompounded, as “that rube is a cow”29 or “(her) face is a moon,” a secondary

superimposed metonymy is to be seen.30

A metonymy (where the metaphorical referent is) identified with (the literal referent),

and (the principle of transfer is also) similarity, is as follows:

The wide sky, o four-armed one, did not emerge from your navel,

For your navel, o scion of the Yadus, is found in it!

When sacred texts proclaim such circular paradoxes

Can we know the truth without pondering well?31

Inasmuch as the word “sky” has here been identified with all that is dependent on the Lord,

in virtue of a similarity of subtleness,32 (this illustrates) a secondary identified metonymy.

25 In other words, the notion “sword” enters into the final construing of the meaning, even though is it not enough, by

itself, to give that meaning. In the former type, the element “Ganges” is completely set aside, as the base to which the

locative is attached (although it does, as we have seen, “motivate” the usage).

26 I.e., the word “village,” in the phrase as given, applies only to that part that is actually on fire; the unburnt part is not

“meant.”

27 Varadarājastava, vs. 86. The phrase navasudhārasagocarah can be read punningly, na vasudhārasagocarah, ‘is not an

object of delight for the earth’.

28 See above. This is gauni, because the principle of substitution is resemblance; it is sāropā, because the two terms are

still mentioned separately (‘face-moon’), although treated as if one.

29 We would probably say “turkey.”

30 I.e., the grammatical representation of the metonymy is not here at issue. What counts is the mention of the two

terms. The figure rūpaka, of course, is such a subordinated attribution: ‘face-moon’ < ‘face is a moon’.

31 The first half of this verse appears in Varadarājastava, vs. 59, with a different second half [hand-copy provided by

HVNR]. Our verse involves a pun on antarikṣa, which, in the second line, is understood as Viṣṇu’s “middle,” madhyam

(so Appaya, in his commentary): presumably a reference to V.’s three “steps.”

32 I.e., ephemerality.

The Vṛttivārttika

The motivation is an intuition (that the two) are completely non-different. Now, by “an intuition (that the two) have the same form” we mean that an understanding (has been had) that (the form of the) term serving as locus is affected33 by the form of the term serving as its referent, but without the difference between them being denied. By “an intuition (that the two) are completely non-different” we mean that an understanding (has been had), by (mention of) the referent alone, that the locus is identical (with it). This is (the extent of) the difference between the motivations of superimposed and identified metonymy. Now, superimposition (involves) the mention of two terms, expressing both locus and referent; identification (involves) mention of the referent alone, metonymically, as though it were the locus. This is (the extent of) the difference between superimposed and identified metonymy.

Here, an objection (might be raised): in the example of superimposed metonymy, there is risk of pleonasm if “face-moon” (be understood) as a transfer of the genus “face” to the moon, for the word “face” used (separately).34 Nor could the metonymy (be understood) as (a transfer of) those qualities, such as loveliness, that pertain to the moon, for usage gives no license to attribute a quality to something that does not have it.35 Nor as (a transfer of) the qualities, such as loveliness, that pertain to the face, for no metonymy can function by (suggesting)36 a quality that is devoid of any relation to the primary subject (of assertion).37 (Therefore, in principle, secondary metonymy cannot be asserted of such cases.)

To this, (we reply): no, for, in these cases, we accept a metonymy that (functions) by (suggesting qualities), such as loveliness, that pertain generally to both face and moon; thus neither of the above-mentioned defects touches (our position). This is what the ancients had in mind when they said, “if the usage (takes place) by association of qualities indicated (by the literal meaning), the (metonymy) intended is ‘secondary’.”38

33 uparaktatā: lit., ‘colored by’: modified by, having the properties of (‘affected’ as used in medieval Aristotelianism). The two terms are here given their logical dress as viṣaya (which we translate ‘locus’, doubtless the upamāna, object of comparison—the moon) and viṣayin (‘referent’, upameya, subject of comparison—the face): see next.

34 I.e., the expression would, in effect, mean ‘face-face’. Theorists have long disputed whether metonymy of similarity (gaunī) is effected by transfer of “genus” (here, mukhatva, candratva) to the other term, or by a transfer of quality (guṇa, hence the two names of the type) pertinent to one or the other. See Kāvyaprakāśa ad 2.11 (pp. 46–47), Tantrāvārttika, loc. cit., 316-17. Cf. Raja, I.T.M., 242–45.

35 pravṛttinimittatvāyogāt. I.e., for attributing the qualities of one thing (moon) to another (face)—ordinary language does not operate that way! The question here is “what justifies such usages?”: pravṛttinimittam kim?

36 omukhena, i.f.c.: ‘à titre de, by means of’. Certainly a confusing usage, given the context—and one that evidently confuses the copyist of H!

37 I.e., the reverse situation: the qualities of the face would not even come to mind, since the verse is about a “moon.” These two pakṣas assume, of course, a “loveliness” that is specific to the moon; another, to the face. The dilemma is thus unreal, says Appaya (next).

38 Tantrāvārttika ad P.M.S. 1.4.23 (p. 313), often quoted, e.g., Kāvyaprakāśa ad 2.12 (p. 47). Jha translates: “when the word signifies something (even though it be in no way connected with the expressed meaning), through the similarity of the qualities indicated (by the expressed meaning of the word), then this signification is called ‘gauna’ or ‘secondary’” (T.V., 1: 440).

36

Chapter 2

A metonymy (where a metaphorical meaning is) superimposed on (a literal meaning), and (the principle of transfer is not similarity, but some other relation—a type we have termed) “pure,” is as follows:

O soul guiding (my soul), O lotus-eyed god!

May thy form that now adorns the elephant mountain39

Ever shine forth within me—thy (form) complete

From topknot to toe, a source of joy

Without limit for those of the lotus-glance.40

Here, the word “joy” has been superimposed metonymically on the limbs of the Lord, in function of their being able to effect joy. The motivation is to indicate that none other is so competent to effect joy. A metonymy of identification (is seen) in the usage, “He is joy!” (applied) to Him who is capable of effecting joy, but with suppression of the locus.41 The motivation is to indicate the unexceptioned permanence of the joy. Thus, pure metonymy being five-fold, and secondary being two-fold, motivated metonymy is seven-fold in all.

Extending the Typology of Secondary Metonymy

Here, an objection (might be raised). Why should secondary metonymy be only two-fold? Here too, the distinctions of “giving up” and “not giving up” (the literal meaning) are possible. For (in the examples) “thy face is a lotus” or “the rube is a cow,” (a metonymy) is evident (wherein the literal meaning) is given up. Similarly (a metonymy wherein the literal meaning) is not given up is observed in such cases as “let the cows be brought here,” where both cows and rubes are possible referents.42 (To this objection) we reply: a metonymy is “secondary” when (it is effected) through a relationship of similarity to the literal meaning. And it is not just in the case (you allege) that a relationship of similarity obtains such that a secondary function could be attributed to the primary as well: for also in cases of metonymy (wherein the literal meaning is) not given up, there is attested a relationship of whole to part between the term qualified and the qualification, inasmuch as the term expressing the qualification, “swords” is (also) understood as referring to the term qualified, (“men holding

39 śritahastiśailam: the reference is to the small hill on which Varadarāja’s temple stands—Hastiśaila.

40 Varadarājastava, vs. 105. Appaya’s commentary to the verse presumes another reading of the second line: ānandavr̥ndalasitam sudṛśām asimam: ‘shining with compacted joy, without limit for those who see well [= those who know; women with lovely eyes (!)].’ The reading here is probably corrupt: “those of the lotus-glance” would seem to refer only to lovely women—or, taking the form as a true genitive, to the “lotus glances” of the god: “limitless source of lotus glances.”

41 Apart, that is, from the pronominal reference. As above, reference to the viṣaya (locus) by mentioning the viṣayin (referent) only counts as sādhyaavasāinī lakṣaṇā.

42 In the former case, there is no “real” lotus; therefore the word ‘lotus’ has “given up” its primary meaning. In the latter, the word ‘cows’ refers generally to both groups, one literally, one figuratively.

37

The Vrttivarttika

the swords”).43 Therefore, in cases such as “let the cows be brought here,” (it is better to say

that) two understandings are had, one by denotation, one by metonymy—as in “there is

water in the Ganges and a village stands (on it).” Thus, the distinctions of “giving up” and

“not giving up” (the primary meaning) are not possible in the case of secondary (met-

onymy). This is, then, the classification of metonymy as set forth by the ancients.

Reducing or Conflating the Types of Pure Metonymy

In this regard, the following observations can be made. Among pure (metonymies)

the delineation of a sub-type involving identification (of the referent with the locus) is im-

proper. For this in no wise differs from a metonymy (where the literal meaning) is given

up, inasmuch as (here too, the literal meaning) offers itself completely for the sake of

another.44 And there is no special charm45 (attaching to this sub-type that would set it

apart), even if one did not make the distinction, as there is in a metonymy involving super-

imposition (a charm that derives) from the grammatical juxtaposition of the (referent and

the) locus. So, unlike that case, there would here be no reason to illustrate (“pure”

metonymical identification) as different from a metonymy (wherein is) given up (the liter-

al meaning).

Well then (says an objector): “Let a distinction be made as follows: the function (in

question, ‘identification,’ operates) in terms of a relation of cause and effect;46 (whereas) a

metonymy (wherein the literal meaning is) given up (operates) in terms of a relation of

contiguity, as between the Ganges and its banks, without reference to the relation of cause

and effect. Thus it is possible to understand (the two types) without confusing (them).” (To

this, we reply), not so! The authoritative works (on our subject) have offered the following

illustrations of both superimposed and identified metonymies: “a royal servant,” addressed

as “king”; “a post placed for the sake of Indra,” termed “Indra”; “one who is not a carpenter,”

addressed as “carpenter”; “the tip of the elephant’s trunk,” termed “trunk.”47 These illus-

trate the relations ‘dependence on,’ ‘benefit for,’ ‘skill of,’ and part to whole, respectively,

[p. 72]

43 Appaya means that, if sādrśya is understood in this way, with both literal and metaphorical senses applying within the

same domain (“cow,” meaning both ‘cow’ and ‘rube’), then the same remark can be made regarding other relationships,

such as the “pure” one of “swords entering”—which would evidently eliminate it as a type, reducing it to gaunī (or

jahadajahal-) lakṣaṇā.

44 Using the referent (viṣayin) in place of the locus (viṣaya) is in no wise different from using the locus so as to express

the meaning of the referent.

45 vaicitrya, lit., ‘the quality of that which is variegated (vicitra)’ ; ‘amazement, wonder, delight’: a cover term for the sub-

jective, though mostly intellectual, response to the work of art. ‘Glanz’ and ‘Pracht’ come to mind as better translations.

The property is of course attributed to the work, not to the reader—hence our translation, ‘charm’.

46 The example given was ānando ‘yam, ‘he is joy,’ meaning, he is the cause of joy.

47 agrahāsto ’yam hastah: this is a bit clearer in Sanskrit, where the trunk is termed the elephant’s “hand”—it is really

the tip that “grasps.”

38

Chapter 2

(that illustrate superimposed or identified metonymies) just as does cause and effect.48 Further, in the Alamkārasudhä nidhi,49 even the relation of locus to referent (is so illustrated). Further, even in the absence of a difference in charm, if a difference of kind be accepted on the basis of a mere difference in relation, it would be impossible to refuse to delineate still other types—taking those differences of relation into account—and thus the restriction to just seven types (which has become traditional in accounts of metonymy) would have to be considered invalid.

[Another attempt is made to salvage the distinction.] Well, is it not the case that in the examples of identified metonymy (given above) there is an (additional) revelation of equivalence?50 For when we say, “He [viz., the Lord] is joy!” or “he is king,”51 our understanding of the motivation (of the expression, namely), that ‘joy is invariably associated (with the Lord)’; or that ‘his commands (like those of the king) are not to be transgressed’; and so on, is conditioned on the revelation of (the Lord’s) equivalence with ‘joy’ or (the minister’s) with the ‘king’. Thus, let there be a distinction made in terms of the (additional) charm (that derives) from this suggested (equivalence).

(To this suggestion, we reply:) no, for even in the example of metonymy (wherein the literal meaning is) given up, our understanding of the motivation (of the expression, namely, that an) exceeding sanctity attaches to those (banks), is conditioned on a (prior) realization of the equivalence52 of the stream and its banks. And, in the absence of an understanding that (the banks) share the nature (of the stream), it would not be possible to understand, as belonging to the banks, the exceeding coolness, etc., that in fact belongs to the stream. We do not accept that there is no revelation of identity (in cases of metonymy where the literal meaning is given up). For the motivation of (such) a metonymy is simply, in respect of the banks, the indication of the coolness, etc., that is inherent in the Ganges and which is evoked by association with it—in as much as that (motivation) can be realized by metonymy alone in virtue of the relation between the bank and the Ganges. Furthermore, it is illicit to posit as the motivation of a metonymy a sense that can be derived from the employment of (words) used literally, given that words such as “Ganges” and “banks” here are meant in their primary senses, and that their usage demands no further explanation.53

48 ...but which, by the above argument, being other than cause and effect, would fall outside the domain of sāropā or sādhya vasā nā lakṣaṇā.

49 A work not extant, perhaps by the great Sāyaṇācārya. See De, H.S.P., 1: 225, nt. 3.

50 That is, an “identity” (of this with that) is understood over and above the literal meaning of the predication; abhivyakti, lit., ‘manifestation’, is a term associated with the third function, dhvani.

51 I.e., when one is speaking of someone who is not the king.

52 abheda, lit., ‘non-difference’; here used with the stronger flavor of the Skt. etymology. The point of the usage is grasped only when the term for ‘stream’ is understood as equivalent to the term ‘bank’. Hence, “equivalence” is also at the heart of jahallakṣaṇā,—not a superadded implication that could be alleged to distinguish it from sādhya vasā yā.

53 svāyatte ca śabdaprayoge. This last observation counters the view that abhedābhivyakti could itself be a motivation of metonymical usage: no, says Appaya, for that “identity” is understood grammatically, without recourse to metaphor. It is thus a precondition, and not a “product.”

39

The Vrttivarttika

And, further, as has been observed in the Kāvyasaraṇi,54 “And he who is desirous of praising more effectively, taking advantage of the means (available) in metonymy to convey the surfeit of beauty (proper) to poetry, employs the word “Ganges” in the sense (‘banks’), in order to indicate the extreme sanctity, etc., that pertains to them (and which can be conveyed) through an understanding of them as sharing the nature of the stream.” Therefore, the realization of equivalence (that is characteristic of an identified metonymy) is indistinguishable from that of a metonymy (wherein the literal meaning) is given up. It has been said in the Kāvyaprakāśika,55 “When its banks are referred to by (use of) words such as “Ganges,” etc., consistency of interpretation, in respect of understanding what is the fact of the matter,56 derives from the purpose one desires to convey.” And the Ratnākara57 has said, “From an understanding that the stream and its banks share the same nature, inasmuch as they are communicated through a single word, the qualities of the stream, such as coolness, sanctity, etc., are understood also in respect of its banks, and thus is the (expression’s) purpose achieved.” And so, whether it be predicated on motivation or on something else, no difference is possible between an identified metonymy and one (wherein the literal meaning) is given up; consequently, illustrating them separately is inappropriate.

Similarly, if it is asserted, in regards to the example given above, that superimposed metonymy (operates) by (suggesting) a property common to the subject and object of comparison (and thus is still distinguishable from a metonymy that gives up is literal meaning, our reply is:), this is not appropriate. If that were the case, the mention by another word of that same property would constitute a redundancy. For example:

O highest of beings! the full moon of thy chest-jewel

Bears its full burden of beauty, for it has just

Risen in the autumn sky of thy breast,

Dark as a tamala bloom, where expanses of lovely

Stars shine among white clouds.58

Here, even qualities such as expansiveness and darkness, which are common both to (Viṣṇu’s) breast and the sky, are explicitly stated. If it be asserted (in response) that this metonymy functions by (suggesting) qualities above and beyond those qualities that are stated (explicitly), we answer: no! It is well attested that connoisseurs (are able to) understand the fully developed sense of the (poetic) composition (cited above), despite the fact

54 Another work no longer extant; the model for Vrttivarttika, according to S.K. De (H.S.P., 1: 225-26).

55 P. 44 (A.S.S. ed.), with minor variations.

56 tattvapratipattau, i.e., understanding what is, in fact, being referred to.

57 The name could refer to two works, neither of which is extant, an Alamkāraratnākara, of Śobhākaramitra, also cited by Jagannātha; and a Rasaratnākara, cited by Mallinātha. See De, H.S.P., 1: 226.

58 Varadarājastava, vs. 70. The pun continues in the last phrase, which also means, “… where rows of white pearls shine against the camphor unguent.”

that no common property different from59 the (already) stated darkness, etc., has manifested itself.60 (In any case, extremely general) common properties, such as “thingness,”

lack any touch of the poetic. Nor, finally, does the poet, as he composes, ponder, in such contexts, qualities above and beyond the qualities (already) stated (explicitly).

Well then (replies the objector), let this (example) not be taken as an instance of metonymy (where the metonymical referent is) superimposed on (the literal referent). To express

that (species of metonymy), let us accept a compound of the type mayūravyamsaka, (whose sense would then be) ‘the autumn sky is indeed his chest.’61 In this case, however, let

there be understood a compound of the subject (with the object) of comparison, (whose sense would be) ‘his chest is like the autumn sky.’62 In that case, the comparative particle

iva, which stands in the full sentential expansion (of the compound) and has been elided, expresses a similitude. The poetic theorists agree that such cases do not involve any

metonymy of the term expressing the object of comparsion taken for the subject of comparison.63 Since they accept, instead, a simile (here) with deletion (of one of the elements),

there is no defect.64

(To all this, we reply:) no! For there can be no compound expressing a similitude when the common property, for example, darkness, is explicitly rendered. For that (type of)

compound has been taught only where the common property is not mentioned, in accordance with the (Pāṇinian rule): “the subject of comparison (as prior member) is compounded with (words) such as ‘tiger’, etc., (expressing the object of comparison), so long as

59 B reads vyatirikta, against both NSP, H, which read vyaktyatirikta (B also notes this reading as a variant). But Appaya’s point is that there is no vyakti—and that this does not damage the verse. This is one of the few loci where common

sense and the best mss. appear to disagree.

60 —which example none would criticize as deficient in poetic charm!

61 The objector proposes that sāropā lakṣaṇā be limited to compounds formed according to P. 2.1.72, a catch-all (nipā-ta-) sūtra covering various (mostly Vedic) ad hoc exceptions to normal compounding rules. Nāgeśa ad Govinda

(Kāvyaprakāśā [A.S.S. ed.], 466), discussing the same point, argues that compounds so formed should be interpreted as rūpaka—which is of course the point being made here, as rūpaka, by definition, involves a “metaphorical identification” of upameya and upamāna. Mayūravyamsaka, to be understood as an exception, must be taken as an inverted

rūpaka: ‘peacock-fool’—fool who behaves like a peacock. (Or: ‘peacock-rogue’?)

62 An upamitāsamāsa, by P. 2.1.56, which interprets compounds of the type “man-tiger” as “tiger-like man”—a simile with the order of the terms reversed and the comparative particle suppressed. Most Alamkārikas would take such compounds as rūpaka (see Glossary, s.v. samāsta rūpaka); for the purposes of this objection, however, it is sufficient to note

that a rūpaka is indeed interpreted as a modified simile —with deletion of certain expressive elements, often by compounding.

63 I.e., these cases do not involve metonymy; the figurative expression can be accounted for in terms of standard figures of speech, such as simile.

64 The objector has been defending the thesis that the charm of the “superimposed” metonymy derives from a suggestion of the common property, and thus can be distinguished from jahallakṣaṇā. To Appaya’s response that the common

property is itself dependent on the metonymy, the objector attempts again to distinguish the two cases by the extreme claim that “superimposed” metonymy need not be understood as a metonymy at all, but is rather a simple simile—

wherein reliance on “suggestion” is no bar.

The Vrttivārttika

no common property is mentioned."65 Also the poetic theorists agree that when the common property is explicitly mentioned, (the figure in question is) "metaphor(ical identification)" (rūpaka) based on a superimposed metonymy, because (some feature) constraining a simile is present66—but not a simile per se. And when referring to a "compound of the subject (with the object) of comparison," they (take it as) illustrating only the (type of) simile (upama) that is "deficient in explicit reference to the common property." And where the common property is employed, they would still accept here a "compound of the subject (with the object) of comparison," taking it as illustration of (a simile) that is "deficient merely in the expressive (particle)."67

Well then (responds the objector), if that is so, when the common property is employed let there be a metonymy which functions (to suggest) that similitude,68 since there is no risk of pleonasm when the (particle) expressive of that (similitude) is not employed.69 And when the (particle) is employed, inasmuch as the term denoting the object of comparison will be dependent, for (realizing) its own meaning, on having become the locus of a similitude, (there must still be a metonymy which functions to suggest that similitude) because a metonymy with the subject of comparison cannot be (cogently) asserted.70

(To this, we reply:) no! for if this were the case, there would be confusion between simile, which obtains when a similitude is stated in so many words, and metaphor(ical identification), which obtains when a superimposed metonymy (is employed); thus the discrimination of the figures (from each other) would be compromised. For simile does not consist merely in the (verbatim) indication of a similitude: in "your face is friend to the

65 P. 2.1.56. This type of compound is made only when the common property is not mentioned—but in our case, several are mentioned. Pāṇini's rule is an extension and restriction of 2.1.55, which allows the object of comparison to be compounded with the common property: 'dark as Kṛṣṇa' (kṛṣṇaśyāma).

66 Reading upamābādhakasattvāt; other readings are upamāvācakasattvāt and upamāvācakaśattvāt—this latter perhaps closest to Appaya's precise intention: "because (the particle) expressive of the simile is not present"—a standard condition of rūpaka.

67 The theory of the simile posits four elements: the subject and object of comparison, the property they share, and a particle expressive of the comparison (E.g. "like"). Many varieties are distinguished merely on the ground that one or another of these elements is implicit, rather than explicit. Many quasi-similes, such as rūpaka, require the suppression (e.g.) of the expressive particle, the dyotaka—but this is not always sufficient to identify a rūpaka. We are left then with the view that the verse in question is not a simile, but a rūpaka, and that metonymy, not suggestion, is the source of its "charm."

68 Previous arguments have focussed on the suggestion of the common property as the device whereby the metonymy is accomplished.

69 The pleonasm would involve the unnecessary repetition of the expressive device in the case of a simile already "implied." The "metonymy" would be the "property" itself taken as the expressor of similitude.

70 ...as explained above. If this is granted then a metonymy will have to be admitted as basis of the simile (that has been alleged in such compounds as "chest-sky"). And, of course, the two types of metonymy at issue can still be distinguished. This position also conflicts with Appaya's notion that the charm of metonymy has nothing to do with the suggestion of any aspect of figuration. We take avācyatvād as a second reason, after prasaṅgāt, both governed by lakṣaṇāstu.

42

lotus,” although the word “friend” is used metonymically, there is a complete realization of a simile.71

However, there is a general ambiguity in the discrimination of simile and metaphor(ical identification) which derives from the theory of compounding: a compound, such as “foot-lotus,” when formed according to the “(man-)tiger” model [P. 2.1.56], is taken as a simile; whereas (the same compound), formed according to the “peacock-fool” model [P. 2.1.72], is taken as a metaphor. Consider:

Let the gentle foot-lotus of Ambikā Assure your victory, anklets sweetly jingling!

Here, the focus of attention is on the (goddess’) foot, for (her foot) is what is capable of wearing anklets, etc. The compound is to be interpreted according to the “(man-)tiger” rule [P. 2.1.56], inasmuch as the prior member [“foot”] is the syntactic head. In consequence, a simile (upamā) (is here understood: “foot like a lotus”). But in

...whose foot-lotus the other gods, Bowing low in unending adoration, Ever take as the ornament of their brows…

the focus of attention is on ornamentation, for that is what is accomplished by (placing) lotuses (on the brow), etc. The compound is to be interpreted according to the “peacock-fool” rule [P. 2.1.72], inasmuch as the posterior member [“lotus”] is the syntactic head. In consequence, a metaphor (rūpaka) (is to be admitted: “lotus which is [also] a foot; foot-lotus”). This stipulation is adopted by everyone (in an effort to disambiguate the ambiguity caused by compounding).72

Nevertheless this stipulation does not hold good for compounds like caranāravindam, ‘foot-lotus’ taken as a metaphor,73 inasmuch as the generic character of (being a) lotus is set aside and the word “lotus” —by (suggesting) either a property, such as redness, common (to both feet and lotuses) or the similitude thereon dependent—thereby serves as an occasion for interpreting (the word) “foot” metonymically (i.e., as an “ornament”). An indeterminacy of equal weight ensues:74 for even though one resorts to the “peacock-fool”

71 These cases are to be interpreted as metonymies of the comparative particle: “friend” = “like.” But, ipso facto, all four elements of simile are “expressed.” Appaya's remark applies to the second of the stipulations above. The objector there has asserted a metonymical simile even when the comparative particle is expressed; this would in effect erase the distinction between simile and metaphorical identification (rūpaka), which has traditionally been stated in terms of the implicititude of that element. (In what follows, “metaphor” should be understood as a shortened form of “metaphorical identification.”)

72 The interpretation, it should be noted, depends on syntactic factors extrinsic to the compound, not on the form of the compound itself, and is in that sense stipulative, or conventional.

73 I.e., with latter member primary, on the “peacock-fool” model. The reference is to the verse immediately preceding, “whose foot-lotus the other gods…”

74virahataulyāpatteḥ: viraha here in its logical sense, abhāva. Neither rationale can predominate for each cancels the other.

The Vrttivarttika

model, and takes as syntactic head the latter member of the compound ("lotus"), as expressing either the quality of redness or the similitude (thereon dependent), it is as if he had (also) resorted to the "(man-)tiger" model, for the prior member, the word "foot," in consequence of being considered fit for adornment, has also in effect been taken as the head of the compound.75 (This result follows also) from the general stipulation that, in a relation of genus to quality, it is the genus that is taken as primary, as in "the blue lotus."76 And even if the latter member ("lotus," as metonymically reinterpreted) were taken as head of the compound, the quality of redness or the similitude (thereon dependent) would not, in the absence of the lotus itself, imply adornment.77 Therefore, in the examples of superimposed metonymy, it is improper to assert that the metonymy (operates) by (suggesting) the common property.

It is also improper to assert that the motivation (of a superimposed metonymy) is the "intuition (that the one) has the form of (the other)."78 For, if one accepts that metonymy (operates) by (suggesting) a common property or a similitude (thereon dependent), it is unnecessary to require a (further) apprehension of formal coincidence. The complete meaning that results (from such a metonymy) is: 'this face is lovely', or 'this face is like the moon'; there is, to that extent, no (additional) apprehension of formal coincidence. Furthermore, (if such were the case, an apprehension of coincidence) would follow upon (literal) usage as well—such as “this face is lovely.”79 Nor should it be asserted (in response) that we have overextended the principle in question, because (in the verse cited), there is indeed a determination of formal coincidence, (effected) by the function of suggestion based uniquely on the capacity of the words, inasmuch as both “face” and “moon” are expressed by one and the same word80—just as was the case in “a village on the Ganges,” where a similar determination of equivalence (was effected) between stream and banks. (This is not cogent; consider the example:)

In the presence of your face-moon, this other (moon) of cool rays appears…

75 It should be borne in mind that Indian feet are typically adorned with red unguent, and, of course, this “redness” has now become the major vehicle of the trope. The ambiguity—a nice one!—is that the “foot” is here both adorned and adornment (and both senses are intended).

76 “Blue” modifies “lotus,” not the other way around.

77 It is the lotus, not its color, that adorns the brow; or, the color “red” would not be understood “ornamentally” apart from the mention of “lotus.”

78 tadrūpyapratītiḥ, henceforth, ‘understanding of formal coincidence’. This was the traditional definition, given above.

79 But this is of course absurd: such an apprehension is in fact the contrary of that which is had in such cases.

80 That is to say, a compound word—considered in Sanskrit grammar as a unitary pada. So, says the opponent, Appaya's application of the principle of “apprehension of coincidence” to literal usage constitutes an overextension: that case may be distinguished because there several words are used, here one.

Chapter 2

Inasmuch as it would (by your argument) now be possible to apprehend here a formal coincidence between the “face” and (its attribute) “lovely,”81 it becomes impossible to construe easily the understanding of formal coincidence that follows naturally82 upon the fact that the moon has (also) been attributed the generic character of a face.

And so, even when the terms are not compounded, as “her face is a lotus,” the claim that superimposed metonymy (results) in all cases where there is grammatical agreement between locus and referent is incorrect.

Pun, Metaphor, and Hyperbole

O swan of the hearts of thy courtiers!

O fiery splendor withering the lotuses of thy enemies!

O dark purple one, seeking Durgā!

O sacrificial fire consuming its fuel!

O clever one, establishing pleasure in truth!

O ferocious one, ever preceding victory!

O best of heroes, may thou have dominion,

Like Brahmā's, resplendent, for one hundred years!83

In this example, where a king is to be described, there is no possibility of metonymy because no important similitude is understood that derives from the conjunction of (the “king” with) the words “swan,” etc.84 The similitude that is expressed here is based on common properties, such as “residence in mānasa,” which are realized as identical (only) against a backdrop provided by punning: the mānasa, ‘lake’ (of the swans), is the mānasa, ‘heart’, of the king's courtiers! There is apart from that no well-known similitude here—such as (that implied in) “face-lotus.” Nor (could we argue) that metonymy is made possible (here) by means of that (verbal) similitude itself, for the pun, on which it depends, is grounded on the figure rūpaka, ‘metaphor(ical identification)’; and until that figure is recognized, the (similitude) has no occasion to come to mind. The poet who wishes to convey a double entendre has no (other) means (at his disposal), because the pun cannot be achieved without the figure “metaphor(ical identification).” In the view preferred by the ancients, namely,

81 Because mukhacandra, even when it means “face (lovely as) the moon,” constitutes a single word. This, of course, on the assumption that the full meaning of the metonymy involves the suggestion of the common property.

82 anubhavikasya, ‘based on experience’. The problem seems to be that once the “moon” is understood metonymically, meaning ‘lovely’, it becomes difficult to take it again as the basis of a further identification with a “face.” In other words, the naive understanding we have that the moon is here being treated as “another” face is hard to construe.

83 See Kāvyaprakāśa ad 10.95 (ex. 425, p. 470, Ā.S.S. ed.), illustrating a rūpaka where the metaphorical identification is based on a pun. The “king” is a “swan” swimming in the “hearts” of his courtiers—when it is realized that mānasa is also the name of a holy lake.

84 He has been called a “swan”: but why? Pratyāsatti, ‘proximity’, here refers to the necessary close association of the two components of a metonymy that is a prerequisite of any transfer of meaning. Despite the pratyāsatti of the words, no property common to “king” and “swan” comes immediately to mind.

The Vrttivarttika

that denotation is determined by context, it is impossible, before the metaphor (has been formulated), to express the other meaning of an imposed metonymy, which, as here, is devoid of any relation to context.85 Furthermore, according to the dictum of Cakravartin,86 who is stipulating the difference between metaphor and pun:

If a previously formulated metaphor

Then communicates a pun, the (figure in question)

Is “metaphor(ical identification)”;

Otherwise, it is “pun.”

In the present example, it is clear that the pun is posterior, for it is described as a feature of the metaphor(ical identification).

The fact of the matter is this: in examples such as “O swan of the hearts of thy courtiers!” no metonymy of the word “swan,” in the sense ‘king’, as subject of the description, need be resorted to. The equivalence can be successfully understood merely by construing the sentence, inasmuch as the “king” is mentioned in grammatical agreement with a well-known type of swan.87

It might be asserted (in response to this) that, even though it is possible to understand an equivalence (of two terms without resorting to metonymy), the (complementary) understanding that one (term) has the form of (the other)88 may not be accomplished without metonymy: such an (understanding) is always had when locus and referent are in grammatical agreement. (We) reject (this view), because there is no authority (for any such distinction). And also, as regards the example, “Above, o Lord! thy navel-lotus, Brahma’s abode...,” even though the locus and the referent are in grammatical agreement, it is clear that (the verse) intends to assert the equivalence of the well-known highest region and (the Lord’s) chest because the words eva sākṣāt ‘itself visibly’ (express emphatically such equivalence).89

(The following objection might also be raised:) the example, “O swan of the hearts of thy courtiers!” should be understood as hyperbole,90 (rather than involving metaphor[ical

85 In other words, the context is here provided only by the figure rūpaka itself—Appaya’s point.

86 Most likely, Śrividya Cakravartin, a commentator on Ruyyaka’s Alamkārasarvasva (cf. De, H.S.P., 1: 187)

87 I.e., the “equivalence” that is implicit in the compound itself is enough to “identify” the two terms. Not all appositions are grounds for metonymy.

88 abhedapratītiḥ; tādṛūpyapratītiḥ: it will be remembered that these were the two “motivations” (phala) of sādhya- hyavasāyā and sāropā lakṣaṇā. Appaya, in his immediately preceding argument, appears to have allowed “equivalence” as the motivation even of the latter lakṣaṇā. This query attempts to redraw the distinction.

89 The point concerns the last line of the verse:

I see revealed before me the highest state of all —

Thy wide bosom, radiant with masses of pearls.

A more literal translation would be “I see the highest state itself visibly (before me)…” Here, although both terms are explicit, an abhedapratīti is conveyed (before me).

90 atiśayokti: usually understood as a “distortion” of the relation between a subject and its properties, or between the terms of a relation (Glossary, 97 ff.)—although the objector seems to view atiśayokti as equivalent to sādhya- hyavasānā lakṣaṇā.

46

identification]). We will stipulate different (domains) for metaphor and hyperbole: the

former occurs when there is grammatical agreement between locus and referent, hyperbole

when only the equivalence of the referent (and the locus is apprehended). When the refer-

ent has the form of the locus, (we have) metaphor; when the referent is equivalent to (the

locus, we have)91 hyperbole.

This view cannot be maintained. Consider the verse

The moonlight arising from the moons

Of your toe-nails assumes a wholly new salvific aspect—

As it awakens the heart-lotuses of the faithful

And dries up the ocean of transmigration!

Even though the referent, “moonlight,” alone is explicitly mentioned—its correspon-

ding locus, the effulgence, etc., of the Lord's toenails, having been suppressed—what is alone

evident here is an understanding that (the one) has the form of (the other), inasmuch as

(the verse emphasizes) the difference (between the “moonlight” of the Lord's toenails) and

the well-known moonlight that is capable (merely) of awakening the bloom of lotuses, etc.

Thus, (the difference envisaged by) the pair of stipulations cannot be accepted.92

Now, if we abandon the stipulation that a metaphor (obtains) when the locus takes on

the form of the referent, but a hyperbole, when the referent (is deemed) equivalent (to the

locus); and, instead, adopt the (alternative) view that a metaphor (obtains) when there is

grammatical agreement of locus and referent, but hyperbole when there is (explicit) men-

tion of the referent alone—then no inconsistency at all results in the case, “O swan of the

hearts of thy courtiers,” even though there is an understanding of equivalence (between the

king and the swan), for a metaphor is still possible.93 If, on the other hand, the former stip-

ulation be adopted, out of concern that the difference between metaphor and hyperbole

has been reduced to a matter of mere verbal brilliance, and that the two will no longer be

(considered) figures of sense,94 then the case, “The moonlight arising from the moons…,”

can be taken as (an instance of) metaphor, even though there is mention of the referent

only; let hyperbole be abandoned, (as was done) in the method of the ancients. Similarly,

in “O swan of the hearts of thy courtiers,” let hyperbole be accepted, even though the locus

and referent are in grammatical agreement. What inconsistency (results)? In such a case,

the understanding of equivalence derives from (the fact that the two terms are) construed

together; where there is no defect, therefore, it is pointless to labor mightily at producing a

91 Probably implied also is that the locus cannot be mentioned explicitly.

92 That is, we cannot restrict hyperbole and metaphor as proposed above.

93 …because both terms are mentioned, and grammatically agree.

94 arthālamkāra: the most general division of the figures—those whose effect depends, not on the form of the words

used (e.g., alliteration), but on the propositional sense expressed (e.g., simile). Metaphor, being a form of simile, is

unquestionably a “figure of sense.”

The Vrttivārttika

metonymy.95 And now, if the opponent should say: “in that case, one can assert an under-

standing of equivalence in all cases where the locus and referent grammatically agree, on

account of their being construed together, and thus ‘superimposed metonymy’ will vanish

(as a category)” — we reply: this is the very point we set out to establish!

Thus ends the second section of the Vrttivārttika, composed by Appaya Dīkṣita, named

“Definitive (treatment) of the metonymical function (of words).”

95 …for the notion of equivalence is otherwise obtained.

Vṛttivārttikam, I

Collation [Harvard ms.] {NSP ed.} (Benares)

[Śrīkṛṣṇāya namaḥ]

viśvaṃ prakāśayanti vyāpārair lakṣaṇābhidhādhvananaiḥ |

nayanair iva haramūrtir vibudhopāsyā sarasvatī jayati || 1

vṛttayaḥ kāvyasaraṇāv alaṃkārāprabandhṛbhiḥ |

abhidhā lakṣaṇā vyaktir iti tisro nirūpitāḥ || 2

tatra kvacit kvacid vṛddhair viśeṣān asphuṭīkṛtān |

niṣṭaṅkayitum asmābhiḥ kriyate vṛttivārttikam || 3

tatra śaktyā pratipādakatvam abhidhā || sā tridhā | rūḍhir yogo yogarūḍhiś ca ||

'akhaṇḍaśaktimātreṇaikārthatpratipādakatvaṃ rūḍhiḥ || tanmātreṇa pratipādakatvaṃ

cāvayavārthāpratibhāsād vā² o tatpratibhāse 'pi tasya pratipādye bādhād vā || yathā

yat te padāmbaruhaṃ amburuhāsanedyaṃ

dhanyāḥ prapadyā sakṛd iśa bhavanti muktāḥ |

nityaṃ tad eva bhajatāṃ atimuktalakṣmir

yuktaiva deva maṇinūpuramauktikānāṃ ||

atra maṇinūpurādiśabdānāṃ nāvayavārthapratibhāsāsaḥ³ | atimuktaśabdasyaāvayavārtha-

pratibhāse 'pi vāsantilakṣaṇe pratipādye tasya bādha ity eṣāṃ⁴ rūḍhiḥ | yady⁵ api

atimuktalakṣmīr vāsantīmukulāśobhaiva muktātiśayini sampad ity abhedādhyava-

sāyārthaṃ atimuktaśabdasyaāvayavārtho 'pi vivakṣitas tathāpi tasya vāsantīgatatvenāprat-

ipadanāt tatra rūḍhir eva || etad avyāptivāraṇena lakṣaṇe 'py ekapadam arthavat ||

[sū.] avayavaśaktimātrasāpekṣaṃ padaśaikārthapratipādakatvaṃ yogāḥ || yathā

ūrdhvaṃ viriñcibhavanāt⁶ tava nābhipadmād

romāvalīpadajuṣas tamasaḥ parastāt |

1tatra, NSP [so HVNR]

2“vā” del., B. Subsidiary punctuation in Harvard ms. so indicated: “o”

3avayavārthāpratibhāsāsaḥ, B

4eṣā, B

5yad, H

6virañcibhavanāt, H [so HVNR]

The Vrttivarttika

muktaughamaṇḍitam urasthalam unmayūkham

paśyāmi deva paramaṁ padam eva sākṣāt ||

atra viriñcibhavananābhipadma’romāvalityādiśabdānāṁ yogāḥ8 | avayavaśaktimātreṇa

pratipādakatvam api samudāyārthāpratibhāsād vā tatpratibhāse ’pi tasyāvayavaśakti-

pratipādye ’nanvayād9 veti dvedhā | tatrādyam udāhṛtam || dvitīyam yathā

astu trayimayatānus tava lambanālī-

ratnais tathāpi paribhūyata eva bhānuh |

soḍhaḥ satāṁ bata niśāntam upāgatānām

evam tiraskṛtikṛd īśvara kaḥ suvṛttataịḥ ||

atra niśāntaśabdasya niketanarūpasamudāyārthapratibhāse ’pi niśāntam triyāmāvasānam

upāgatānām10 satāṁ nakṣatrāṇām tiraskṛtikṛd antarhitikṛd evam graham prāptānām11

satpuruşāṇām parābhavakṛd ity abhedādhyavasāyārthāntarasya12 vivakṣitatve ’pi tri-

yāmāvasāne yoga eva || (sū.) avayavasamudāyobhayaśaktisāpekṣam ekārthapratipādaka-

tvam yogarūdhịḥ yathā

pakṣadvayakraśima13poṣavibhāvyamāna-

cāndrāyaṇavrataniṣevaṇa eva nityam |

kurvan pradakṣiṇam upendra surālayām te

lipsur mukhābharucim eṣa tapasyatinduh ||

atra surālayāśabdasya surāṇām ālaya ity avayavaśaktyā samudāyaśaktyā ca kanakācale

yogarūdhịḥ na cāśya śabdasya kanakācale rūḍhir eva kevalāstu mābhūd yogarūdhịḥ o

14udāhṛtaśloke tapascaryotprekṣayāḥ pradakṣiṇikriyamāṇasya kanakācalasyāvayava-

saktyupāṣṭhyādevālābhedādhyavasāyenaapi nirvāhād iti vācyam o tatraivāvayavārtha-

sambhave ’nyābhedādhyavasāyārūpapāramukhanirīkṣaṇāyogāt {!} tatraiva yaugikārtha-

pratibhāsasya sarvasāᶇkṣikatvāc ca | na caivam sati 15devatāgārasādhārāṇaklptāvayava-

sáktimātreṇa kanakācale ’pi pravṛttyupapatteḥ na tatra samudāyaśaktiṁ kālpāniyety api

7 virañcibhavanam, nābhipadmam, H

8 prayogah, B [so HVNR]

9 ‘[na]nvayād, NSP

10 upagatānām, NSP, B

11 eva grahrapāptānām, B [so HVNR]

12 abhedādhyavasāyārtham tasya, NSP, B

13 kṛśima, NSP

"samudāhṛta" kha, ga, NSP

15 devāgāra-, NSP, B

śańkyam° ratnasānuḥ surālaya ityādyabhidhānakośasmṛtiparāmparayā viśiṣya tasya tatra samudāyaśakter api siddheḥ | anyathā devatāgāra iva tatra viśiṣyānuśāsanānapekṣaṇena16 tadvaiyarthyāpatteḥ | ata evaitadr̥śapade sati punah samudāyārthavācakapadaprayoge punaruktadoṣollāsaḥ || yathā bhadrāya bhavatu bhavatām bhagavān bhajamānadainya17timiraraviḥ | divasārambhavikasvaranirajanālinābhirāmataranayanah

|| yatra yogarūdhimato 'pi padaśyāyavavārthaśaktimūlapratiyamānārthagarbhīkāreṇa18 tatraiva viśrāntis tatra samudāyārthavācakapadāntaraprayoge 'pi na punaruktadoṣollāsaḥ || yathā

udyanmrgāńkarucikandalakomalānām unnidraśoṇanalinodarasodarāṇām | prāptum tavādhararucāṃ avalokanena nālam saḥasranayanaḥ sa vṛṣāpi tṛptim ||

atra sahasranayanaśabdasyaāvayavavārthaśaktyāvalokana19 prakārsasambhavapratītes tatraiva tasya20 viśrāntir iti punar indravācipadaprayogaḥ || yathā vā

tava prasādāt kusumayudho 'pi sahāyam ekam madhum eva labdhyā | kuryām harasyāpi pinākapāner dhairyacyutim ke mama dhanvino 'nye ||

atra pinākapāniśabdasya haranāmno21 'vayavārthaśaktyā tadāyudhasāratvapratītes tatraiva tasya viśrāntir iti punarharapadaprayogaḥ | nanu kusumāyudho 'pīty atrayavavārthaśaktyā tadlāyudhasāratvapratītir22 api kāmacāvipadāntaram nopāttam iti cen na ° tatrāpy uttamapuruskāśiptahāmpadasya viśeṣyavācinah sattvād iti sāhityacintāmaṇikāraḥ || vastutas tu nāyam niyamo yad avayavārthaśaktimūlapratīyamānārthavivakṣāyām samudāyārthavācipadāntaram prayoktavyam iti kim tv ayam api prakāro 'sṭity udāhṛtam || tadaprayoge yathā

manye nijaskhalanadośam23 avarjanīyam anyasya mūrdhni viniveśya bahir bubhūṣuḥ ||

16 -apekṣaṇe, NSP [so HVNR] 17 -daitya-, H, B 18 -garbhakāreṇa, H 19 -saktyālokana-, H 20 "tasya" del., NSP 21 śivanāmno, NSP, B 22 āyudhasāratva-, corr. NSP; sudhāsāratva-, B [so HVNR] 23 "ma" + an avagraha superscript, H [the addition is on line three of the ms.]

The Vrttivārttika

āviśya deva rasanāni mahākavinām

devī girām api tava stavam ātanoti ‖

atra vāgiśvarīpadād24 avayavārthaśaktyā pratīyamānāsya25 vidyāviṣaye skhālityāpavādam itaravat sarasvatī na sahatā ity arthasya vācyasyopaskarakatayā26 vivakṣitatve 'pi vācām adhikavatāpi yatrā bhagavatāḥ śrotṛ na paryāptim27 samārthyam bibharti28 tatra katham anye adhikuryur ity arthāpattayālamkāradhvananasattve 'pi na29 prthak sarasvatīvācakapadaprayogaḥ ekāvalikṛtāpi viśeṣyasābhipāyatvam30 31 parikurālamkāralakṣaṇam abhidhāyā °

narasiṁha mahipāla kirtis32 tripathagā tava |

na kasya bhavati ślāghyā punānā bhuvanatrayam ‖

iti tadudāharaṇe bhuvanatrayapavitrikaraṇopanapattyabhipriyāgarbharmin tripathagāpadam viśeṣyam eva prayuñjānena tatprayogāniyamo33 darśitaḥ | tasmād yogarūdhimataḥ padasya yaugikārthasaktimūlapratiyamānārthagarbhikāre34 samudāyārthavāci-pādaprayogāprayogābhyām dvaividhyam evādṛtam | na punas tatprayoganiyamah35 ‖

kvacid yogarūdhimataḥ padasya viṣayaviśeṣe 'vayavārthānanvayāt samudāyārthamātre viśrāntiḥ ‖ yathā

ullāsayaty udarabandhanibaddhadivya-śoṇāśmaraśmikālikāvalir acutyaiśā | āgāmyanekaśatakalpavidhātṛgarbha-nābhyudgatāmburuhakuḍmalapaṅktiśobhām ‖

atrāmburuhapadasya bhagavannābhipadme 'vayavārthānanvayād rūḍhir eva ‖

24 vāgiśvarīpadāvayavārtha-, B; vāgiśvarapadāvayavārtha-, H

25 -saktipratīyamānāsya, H

26 ity avācyopaskarakatayā. H

27 paryāptim, H

28 bibharti, H

29 "na" del., NSP

30 viśeṣyasābhipāyatvam, H

31 parikurālamkāra-, B

32 kirtis, H

33 tatprayoganiyamo, H [misread as "prayuñjāne na tatprayoganiyamo...?"]

34 garbhāreṇa, NSP [evident missprint for "garbhīkāreṇa"]

35 "-yoge ni-" kha, NSP

Vṛttivārttikam, I

kvacit tu samudāyārthānanvayād avayavārthamātre36 tasyā37 viśrāntiḥ || yathā

38kahlārakairavamukheṣv api paṅkajeṣu

lokeśa yat kamalam eva tathā prasiddham |

manye 'bhijātā bhavadāsyatulasya neti

marmaprakāśanam idam vidhinaiḥva kṛptam ||

atra paṅkajāpadāsya kumudakahlārādiṣu39 samudāyārthānanvayād yoga eva || naiyāyikās

tu ° paṅkajādiśabdarūpaikapadopādānarūpāyantaraṅga40 pratyāsattayā nabhikamala-kumudādyanvayāt prāg eva paṅkajanikartṛtvavaisiṣṭyenopasthitasya41 padmatvasya

42padmāśritatvenopasthitasya kartṛvasyā43 ca nābhikamalakumudādyanvaye nākaṅkṣā na

vāviśiṣṭasya44 tadanvayaṽiśiṇī śābdadhīr iti tadanvayārtham svatantrāpadmatva-paṅkajanikartṛtvopasthitaye paṅkajādipadasya lakṣaṇaivābhyupagantavyā na tu rūḍhir

yogo vā ° na ca prathamam yogarūḍhibhyām svārthopasthāne paścān nābhikamalādy-anvayārtham45 viruddhatvittimūkībhave sati46 taditaraṽṛttyaiva svatantrāpadma-tvādyupasthiter ākāṅkṣādisambhavān na tadartham47 iha lakṣaṇeti vācyam ° kevala-rūḍhyādinā svatantrāpadmatvādyupasthityartham śabdasya punar anusaṁdhanā-kalpanāgauravāpatteḥ ° lakṣaṇāpakṣe tu nāyam doṣaḥ yogarūḍhyā48 paṅkajanikartṛtva-samvalitapadmatvarūpāviśiṣṭopasthiter viśeṣaṇaviśeṣyānyataropasthitilakṣaṇādvāratvena49

d vāre sati dvāriṇaḥ punar anupekṣaṇāt50 ° na ca lakṣaṇāyām iva śakyopasthitīḥ51

śakyopasthityantare 'pi dvāram ° yena rūḍhyādinā svatantrāpadmatvādyupasthitāv api

punar anusaṁdhanam kalpya ahuḥ ||

36 avayavārthamātreṇa, B

37 tatra, NSP

38 kalhāra-, B

39 -kalhāra-, B

40 -rūpāntaraṅga-, NSP, B [*-rūpayā 'ntara-" kha, ga, NSP]

41 upasthitapadmatvasya, H

42 padmāśrayatvena-, NSP [so H HVNR]

43 paṅkajanikartṛtvasya, NSP

44 na cāviśiṣṭasya, NSP; navaśiṣṭasya, H

45 -anvayārthaviruddha-, H

46 sati del., NSP

47 tadartham, NSP [so HVRL]

48 doṣaḥ paṅkajani-, NSP

49 -ānāropasthitilakṣaṇa dvāratvena, H; -lakṣaṇādvāratvena, NSP, B

50 anupekṣaṇāt, B [so HVNR]

51 śakyopasthitir lakṣaṇāyām iva, NSP, B

The Vrttivārttika

ittham abhidhā trividhā52 nirūpitā | sa cānekaśaktikasyā śabdasya samyogādyair niyamyate || tathāhuḥ ||

samyogo viprayogaś ca sāhacryaṁ virodhitā | arthah prakaraṇaṁ liṅgaṁ śabdasyānyasya samnidhiḥ || sāmarthyam aucitī deśaḥ kālo vyaktiḥ svarādayah | śabdārthasyānavacchēde viśeṣasmṛtihetavaḥ || iti 53

arthah prayojanaṁ o prakaraṇaṁ54 vaktrśrotṛbuddhisthatā | liṅgaṁ prayuktanānārthapadavācyāntaravyāvṛtto dharmah | śabdasyānyasya samnidhiḥ55 nānārthapadaikavācyasam̉sargyarthāntavācīpadaśaṁabhyāhārah | sāmarthyaṁ kāraṇatvaṁ o aucitī arhatā o vyaktiḥ stripum̐napuṁsakam | śeṣaṁ prasiddham || tatra śaṅkhacakradharo hariḥ56 ity atra śaṅkhacakrayogena57 hariśabdasya bhagavaty abhidhā niyamyate | o aśaṅ-kacakradharo hariḥ ity atra tadviyogena58 | rāmalakṣmaṇāv ity atra rāmapadasya bhāvavirodhād bhārgavakārtavīryayoḥ63 abhidhā niyamyata ity udāharaṇaṁ64 tan na o rāmapadasya bhārgave 'bhidhāniyamaṅe sati tadvirodhapratiṣaṁdhānenārjunapadasya kārtavīrye65 'bhidhāniyamaṅaṁ o tasmims ca sati tadvirodhapratiṣaṁdhānena rāmapadasyeti parāsparāśrayāpatteḥ66 | tasmād anyatarapadasya vyavasthitārthatva eva smṛtatadvirodhapratiṣaṁdhānān nānārthapadasyābhidhāniyamanam iti prāguktam evodāharaṇam anusartavyam67 || idam tu śabdāntarasam̉nidher udāharaṇam bhaviṣyati || harim mokṣāya bhajaty atra mokṣarūpeṇārthena hariśabdasya bhagavaty abhidhā

52 "trividhā" del., B

53 Some versions have saṁsargo viprayogaś ca... [Vākyapadīya, ed. Abhyankar, p. 42]

54 prakaraṇaṁ ca, NSP

55 NSP sam̉nidhiḥ [folio 7 appears written with a different pen. Cf nn. 138, etc.]

56 hariḥ, NSP

57 -samyogena, NSP

58 tadviprayogena, NSP, B

59 cchāyaśabdasyātapena, NSP [so HVNR] saha sahānavasthāna-, NSP [so HVNR]

60 "saha" del., B [so HVNR]

61 iva ity, NSP, B

62 vadhyaghātakabhāvavirodhāt, NSP, B

63 -kārtavīrya-, H

64 iti, B [so HVNR]

65 kārtavīrye, H

66 parasparāśrayatāpatteḥ, H

67 anusartavyam, H

Vṛttivārttikam, I

niyamyate ° viveśa rāmo durdharṣaṁ tāpasāśramaṇḍalam ity atra rāmapadasya prakaraṇād raghuvīre ° kupito bhūbḥrd ity atra kopaliṅgena parvatavyāvṛttenā bhūbḥrd ity68 asya rājani || saṁyogodāharaṇe prasiddhiprācuryam | śaṅkhāder69 indrādāv arthāntare 'pi saṁyogavirodhāt || liṅgodāharaṇe tu sarvathārthāntaravyāvṛttir iti bhedaḥ || niṣadhaṁ paśya bhūbḥṛtam ity atra janapadaviśeṣasādhāraṇasya70 niṣadhapadasya parvatavācibhūbḥṛtpadassamniḥdhānāt parvataviśeṣe abhidhā71 niyamyate ° bhūbḥṛtpadasya ca rājasādhāraṇasya parvataviśeṣāvaciniṣadhapadassamniḥdhīnā72 parvate | na cānyonyāśrayaḥ ° na hy atra samabhi vyāhṛtāśabdena tadarthapratipādanam73 abhidhāniyamanāyāpekṣyate74 kim tu svārthena grhītasaṃsarge 'rthe vyutpanno yaḥ śabdas tatsamabhi vyāhāramātraṁ75 | tathā ca75 yathā sambandhidarśanāt sambandhyantarasmṛtisthale grhītasambandhasya76 sambandhino darśanamātram77 sambandhyantarasmaraṇāyāpekṣyate78 | na tu taddarśanānantaraṁ tatsambandhasmaraṇam apīti79 nānyonyāśrayaḥ ° tathehāpi niṣadhabhūbḥṛtpadayor abhidhāniyamanāya grhītasvārthavyutpattikabhūbḥṇniṣadhaśabdasamabhi vyāhāramātram apeksyata iti tattadarthapratipādanasyānapekṣaṇān nānyonyāśrayāḥ | nanv evaṁ śaṅkhacakrādharo harir ityādāv api śaṅkhacakrādiśabdāntarasamīkṣinīabhidhāniyamanam bhaviṣyatīti vyartham prthak saṁyogāyudāhāraṇam iti cen maivam | tatra śaṅkhacakrādiśabdānām niyatartthatayā śravaṇamātrad arthapratipadakatvena tatpratipāditārthasamyogādibhir niyatatvāt saṁyogādibhir abhidhāniyamanasambhavāt | śabdāntarasamni dhir api tathāstiti ced80 astu ° tathāpi hariśabdādyabhidhāniyamanasamarthe tadarthasamyogādirūpasamnihitasambandhenāntaraṅge śaṅkhacakrādāv upasthite tasyāpi81 vācakatayā vyavahitasambandhena82 bahiraṅgaśabdāntarasamni dher83 nodāharaṇatvam arhati |84 ato yatra niṣadham paśya

68 "iti" del., B

69 || ṅkhāder, H [second daṇḍa erroneously for śa]

70 janapadaviśeṣasādhāraṇasya, H

71 "abhidhā" del., NSP, H

72 -samniḍhānāt, B [so HVNR]

73 -pādanasamabhi-, H

74 -niyamanāpekṣyate, H

75 "tathā ca" del., B [so HVNR]

76 "sambandhidarśanāt sambandhyantarasmṛtisthale grhītasambandhasya" del., B

77 sambandhidarśanam, B

78 -smaraṇād apeksyate, H

79 apīti yathā, H

80 tatrāsti iti cet, NSP, H [so HVNR: tasyā = śaṅkhacakrādeḥ]

81 tatrāpi, NSP, H [so HVNR: tasyā = śaṅkhacakrādeḥ]

82 bahirantarāṅga-, H

83 -samnidhir, NSP, H [so HVNR]

84 signum del., H

The Vrttivārttika

bhūbhṛtam ityādāv ubhayasya nānārthatāya na kasyāpy arthasya svayam85 upasthitis tatraiva bahiraṅgasyāpi tasyā niyāmakatvenodāharaṇam iti na kaścit saṁkaraḥ | yat tu niyatārthaśabdasāmānādhikaraṇyaṃ śabdāntarasaminidhiḥ̣85 tena ca devasya tripurārāter ity atra devaśabdasya saṁkare niyamanaṃ ° tatra tripurārātiśabdasya niyatārthatvād iti tad ayuktaṃ ° samabhi vyāhṛtaśabdasya niyatārthatve tatpratipāditaliṅgādinai va tanniyamanāt86 | anyathā liṅgādyudāharaṇe 'pi śabdāntarasaminidher eva niyāmakatvāpatteḥ̣ | na ca sāmānādhikaraṇye śabdāntarasaminidhir vaiyadhikaraṇye bhūbhṛtaḥ kopa ityādirūpam87 liṅgādikam iti bhedakathanam88 ° ubhayatra śabda-pratipāditaliṅgāder eva niyāmakatvenāsya bhedakathanasyāparibhāṣāmātratvāt | vyālo dānena rājata89 ity atra vyālādānāpadayor ubhayor apy aniyatārthatve 'pi vaiyadhikaraṇye 'pi parasparasamabhi vyāhāreṇa duṣṭagajamadalajalayor abhidhāniyamanasya sarvānubhavasiddhatvenāvaśyāsamgrāhyatvāt tadasaṃgrāheṇa lakṣaṇakathanasyātyantāyuktatvācca ° tasmād asmaduktai va śabdāntarasaminidhiḥ̣ prakriyādarśanīyā | madhunā mattaḥ kokila ity atra kokilamadalakathanāturūpeṇa sāmarthyena vasante madhuśabdasābhidhā niyamyate ° bhajata hariṃ bhavatāpakhinna¯90 ity atra bhavatāpaharanaucityena haripadasya lakṣmīramaṇe ° arthodāharaṇe caturthyā kāryatvanirdeśaḥ91 ° sāmarthyo-dāharaṇe tadubhayānirdeśe 'pi vogyatāmātrena samabhi vyāhāralabhyakāryakāraṇabhāvāvagāma iti bhedaḥ | kṣīrārṇave hariḥ̣ śete92 ity atra deśena pralaye tiṣṭhati harir ity atra kālena ca garuḍadhvaje haripadasyābhidhā93 niyamyate ° mitro bhātīti pumvyakt yā tapane ° suhṛdvācino napuṁsakatvāt | tisra eva sāhnsyopadosa¯94 dvādaśāhinasya iti śrutāv95 ahinapadasya madhyodāttasvareṇāhargasamādhyasutyāke dvādaśāhākhyakratuviśeṣe96 ° ahnaḥ khaḥ kratāv ity anuśāsanāt97 ° hiṇo na bhavatīti naṁsamāse saty ahinapadasyādyudāttatvāt |

85 prathamam, NSP, B 86 tanniyamāt, B [so HVNR] 87 -rūpe, NSP, B 88 bhedakathanam yuktam, B, HVNR 89 rājate, H (saṁdhi not resolved). 90 bhavatāpayuktāh, H; bhavatāpaharam, B 91 karaṇatvanirdeśah, NSP [so HVNR] 92 śic, H 93 "abhidhā" del., NSP; "padasya" corrected to "padam" [so HVNR] 94 sāyāhnasya, NSP; upasodo, H 95 śrutov, corr. H. The injunction forms the topic of adhikarana 8 of Mīmāṁsāsūtra 3.3 [sūtras 15–16]. It is by Śabara attributed to the Jyotiṣṭoma. See also MK, p. 1953. The VPAK does not otherwise find it in extant śruti texts. 96 dvādaśākhyakratuviśeṣe, H. The Dvādaśāha is considered a transform of the Jyotiṣṭoma. See Kumārila ad MS 8.1.17 [adhikarana 10] and MK, p. 2181. 97 The source of this "teaching" is not known. The dictionaries do not give "kha" as the name or in the sense of a 'rite'. [Both ms. and text agree in reading "kha-", rather than "sva-".]

56

98ādisaabdenābhinayāpadeśau99 grhyete100 o abhinayo101 vivakṣitārthākṛtipradarśako hastādivyāpāraḥ o apadeśo102 vivakṣitārthasya śṛṅgagrāhikayā103 nirdeśaḥ |

parāri104 saṁdarśitasaṁniveśau parut105 parābhūtasarojakośau | imāv iyantau punar aiśamau106 te kucau kiyantau parato bhavetām ||

ity atra tatkalāstanaparimāṇa107pradarśakahastavyāpāreṇa iyatpadāsya parimāṇaviśeṣe |

itah sa daityahṁ prāptaśrīr neta evārhati kṣayam | viṣavṛkṣo 'pi saṁvardhya svayam chettum asāmpratam ||

ity atra ita iti svātmanirdeśena tasyā vaktari parameṣṭhiny abhidhā niyamyate | yatra tv anekaśaktasyā108 śabdasyābhidhāniyāmakam anekatra praviṣṭam109 arthe o tatra nānyatarasminn abhidhā niyamyate || yathā |

yena dhvastamanobhavenā balijitkāyoḥ purāstrīkṛto yaś codvṛttabhujāṅgahārāvalayo gaṅgām ca yo 'dhārayat | yasyāhuḥ śaśimacchiroharā iti stutyam ca nāmāmarāḥ pāyāt sa svayam andhakakṣayakāras tvām sarvadomādhavah || [cf. Dhv. 2.21]

atra śivakeśavayor api110 stutyatayā prakaraṇam aviśiṣṭam | ato dhvastamanobhavenā nāśitamanmathena yena111 balijitkāyo viṣṇor vigrahaḥ pureṣv112 astratām nīto o yaś codvṛttaseṣavāsukiprabhṛtināgārājāsampāditahārāvalayakṛtyo o yaś ca gaṅgām adhārayad o yasyāḥ śiraḥ śaśimad āhur amarāḥ stutyam nāma ca hara ity āhur andhakāsurakṣayakāraḥ

98 ādipadena, B

99 -upadeśāu, NSP, H [so HVNR]

100 vivakṣitau, H

101 upadeśaḥ, H

102 upadeśo, NSP, H [so HVNR]

103 "Grasping the horns": treating of a matter by dealing with (pointing to) instances, rather than generally (logic).

104 pūrāri, H

105 marut, NSP [so HVNR]

106 aiśamau, H

107 -parimāṇaviśeṣa-, B

108 anekaśabdasya: corr. in margin; anekārthaśabdasya, B; anekārthasya śabdasya, HVNR, following Kāvyaprakāśa.

109 aviśiṣṭam, NSP, B

110 dvayor api, B [so HVNR]

111 "yena" del., NSP [so HVNR]

112 purā, B [so HVNR]

The Vrttivārttika

sa umādhavas tvām sarvadā pāyād iti | abhavena niḥsamsāreṇa yena anaḥ śakaṭam

dhvastam o yena ca balijit svasya kāyaḥ purāṃrtamathane strītvaṃ nīto o yaś

codvṛttaṃ bhujangaṃ kāliyam hatavān o rave śabdabrahmāṇi layas tādrūpyāpattir

asyeti ravalayaḥ o udvṛttabhujangān hatavad āravalayaṃ cakram asyety ekam padam

vā o agam govardhanam gām ca varāhāvatāreṇa bhūmim ca yo 'dhārayat o yasya

stutyam nāma śāśimacchronara ity ānṛtya śāśinaṃ mathnāti yo raṇus tācchronara iti o

andhakānāṃ rājñāṃ kṣayakaro nivāsakṛt sa sarvado mādhavas tvām pāyād iti

cārthadvaye 'py abhidhaiva vyāpāraḥ | ata evātra śleṣo 'laṃkāraḥ ||

nanu yatrānekatra prakaraṇādyaviśiṣṭam tatra bhavatv anekatrāpy abhidhāvyāpāraḥ ||

yatraprakaraṇasya kvacid arthe 'vastithatve 'py aprākaraṇikam śakyāntaraṃ pratīyate

tatra kā gatiḥ | yathā |

asāv udayam ārūḍhaḥ kāntimān raktamaṇḍalaḥ |

rājā harati lokasya hṛdayam mṛdubhiḥ karaiḥ ||

iti candrodayavarnanaprastāve | atra hy aprākaraṇiko bhūpatir abhyudayādivisiṣṭaḥ

pratīyate | atra prāñcāḥ | tatra śabdasaṅtimūlo vyañjanāvyāpāra eva śaraṇam o gaty-

antarābhāvāt | na hi tasya tatrabhidhā saṃbhavati o tasyāḥ prakaraṇena niyanitattvāt o

nāpi lakṣaṇā o mukhyārthabādhādyabhāvāt o na cābhidhāyā iva vyañjanasyāpi pra-

karaṇena niyantrane tampi na tannirvāha iti vācyam o tasyā dharmigrāhakamānena

tadaniyamyatayaiva siddheḥ | ata eva kāvyāntarniviṣṭasya prakaraṇikakimiṃcidartha-

113 "ca" del., H

114 purā 'mṛtamathane, H

115 udvṛtta-, H

116 udvṛttān, B

117 hatavaravalayaṃ, B, HVNR

118 "ca" del., NSP

119 varāhāvatāre, B, HVNR

120 avagraha del., H

121 "sa" del., NSP, B

122 avagraha del., H

123 nanu ca, NSP, B

124 vyavasthite, B [so HVNR]

125 sarvasya, B

126 "hi" del., B [so HVNR]

127 candre niyamitattvāt, B, H [so HVNR]

128 vyañjanāvyāparasyāpi, B [so HVNR]

129 na del., H

130 tanniyamyatayaiva?

niyatābhidhāvyāpārasyaāpy asabhyapadārthāntarapratyayahetoḥ padaśya heyatvam ° anyathā tatrābhidhāyā iva131 vyañjanāvyāpārasyāpi prakaraṇena niyantrane 'sabhyārthāntarapratyayāsambhavāt yasyānānāṁ yonir udāravācāṁ ity ādāv aśliladoṣo na syāt | tasmāt samyogadhir abhidhā śṛṅkhalitā yatra kuṇṭhatāṁ ety arthāntarāvagame vyañjanam eva kṣamaṁ tatreti ||

vayaṁ tu brūmaḥ || asāv udayaṁ ārūdha ityādau prakaraṇike 'rthe prakaraṇikavad132 aprākaraṇike 'pi rājakaraṇḍalādiśabdānāṁ parasparānvayoyogyanrpatitadgrāhyadhanadeśādivācakānāṁ133 samabhyāhārārūpam abhidhāniyāmakam astīty artha-dvayasyāpi prakaraṇikatva iva prakaraṇikāprakaraṇīkarūpatve 'py ubhayatrābhidhaiva vyāparāḥ ° yathoktasamabhyāhārasyaāpi śabdāntarasamīkṛtābhidhārūpatvena prakaraṇavad abhidhāniyāmakatvāt | yatra śabdāntarasamīkṛtābhidhir anyathā nirūcyate tair yathoktasamabhivyāhareṇa nānārthaśabdebhyo 'rthapratyayasyāpi tadbhedenāpi niyāmakatvam āstheyaṁ eva | vastuto 'yaṁ eva śabdāntarasamīkṛtābhidhir ity apy uktaṁ ° na cāprākaraṇike uktarūpāśabdāntarasamīkṛtābhidhimātram ° prakaraṇike134 tac ca prakaraṇaṁ ceti niyāmakādhikyāt tatraivābhidhā syād iti vācyaṁ ° prakaraṇādināṁ abhidhāniyāmakatvar hi nānārthaśabdajanyāśakyasmṛtau teṣāṁ135 anyatamasya kāraṇatvaṁ yathoktaṁ ° śabdārthasyānāvacchede viśeṣasmṛtihetava iti ° tathā ca teṣu kasya cid api samavadhāne smṛtisāmagri puṣkalāveti kiṁ tadbāhulyaena | yatrakam eva padamṁ nānārthaṁ aprākaraṇike 'pi tātparyavat ° yathā °

karoti kāmsādimahibhrtāṁ vadhāj jano mrgaṇāṁ iva yat tava stavam136 | harer137 hiraṇyākṣapurahsarāsuradvipadvīṣaḥ pratyuta sā tiraskriyā138 ||

ity atra haripadam | tatra139 samabhyāhārtadvipapadapratitagajavirodhādikam aprākaraṇike siṁhe 'py abhidhāniyāmakam asty eva | yatra tu yasyānānāṁ yonir udāravācāṁ140 ityādāv aślile naitādrśaṁ niyāmakam tatrāpi bubhutsitārthavyāsaktasya manasas tam [artham] ullanghyaśabhy e durgandhādau pravṛttivat prakaraṇaniyan-

The Vrttivarttika

tritasyāpi śabdāsya tad [prakaraṇam] उल्लङ्घya viṣayasvabhāvād aśile pravṛttir iti na kvāpi śaktivisaye vyaktir aṅgīkaraṇīyā | yadi caivamvidhāni niyāmakāni nādriyeraṁs tadodārtasthaleṣu vyan̄janavyāpāro 'pi durgrahaḥ141 || tathā hi | prakaraṇādikam उल्लङ्घya vyan̄janavyāpāreṇāpy arthāntarpratyayo na sarvatra vaktum śakyate 142pra-karaṇādīnām abhidhānīyāmakatvākalpanasyāiva vyarthatvāpatteḥ143 || tad dhi

nādṛṣṭārtham kim tu sāndhavam anayayedāv ekarthamatrāparānārthasabdāsthalearthaviśeṣasmṛtivyavasthārtham tat | yadi tatrāpi vyan̄janavyāpāreṇārthāntarapratyayaḥ syāt tadā katham eṣā sidhayet | na hy abhimatānabhimatakṣetrasādhāraṇasya jalāśaya-syābhimatā144kṣetragāminim ekām kulyām vidhyānabhimatakṣetragāminaḥ kulyāntarasyaṅpi kalpane jalasyābhimatakṣetragāmitvavyavasthā kṛtā bhavati | tathā ca yatra kavayituś camatkārāprākaraṇike 'py arthe tātparyāṁ yatra vāślīladoṣas tatraiva vyan̄janavyāpāroṇmeṣo na sarvatreti tadvayavasthāsiddhaye tatra tatra niyāmakāny145 asmaduktāni anyādrśāni146 || vāśyam ādartyavyāni147 teṣāṁ cābhidhānīyāmakatvam148 eva vaktum śakyam iti vyarthas tatra vyaktivākalpanāpariśramaḥ || durvāraś ca tatra vyaktivādāimate 'py abhidhāvyāpāraḥ149 || tathā hi | yadi śaktir abhidhā nirvivādām asty evāprāka­raṇike 'pi sā | yadi ca150 kāvyaprakāśādyuktatirya151 śakyārthapratipādakatvarūpaḥ śabdavyāpāraḥ151 so 'py asty eva | na hi tatrāprākaraṇikāśakyapratipatteḥ śabdād anyaj janakam | śaktigraham apekṣya tatpratipadakatvam abhidheti cet |152 aprakāraṇiko 'rthas tatrāgrhittaśaktikasyābhidhayā kim na153 pratīyate yena sā154 tatra na syāt | nanu śaktyā sāksād pratipāadakatvam abhidhā | sāksād ity aviseṣaṇe155 mukhyārthaśaktigraham apekṣya tatpratipādanavyavadhānena156 tatsaṁbandhyarthāntarāpratipādanarūpāya lakṣaṇāyā

141 durgaṇaḥ, H (the visarga used here resembles the Tibetan rather than the usual Devanāgarī visarga); durghaṭaḥ, B, HVNR

142 H.'s first syllable is badly smudged, but does not appear to represent a normal "pra."

143 See note 141. The symbol probably indicates a "strong break" [= ||].

144 kṣetraika-, B, HVNR

145 niyam'māka'ny, H, corrected in ms.

146 "vā" del., H

147 ādattavyāni, H

148 -kattvam, H [and so throughout]

149 abhidhā[niyama]kattvam eva vaktum vyāpāraḥ, H [copied erroneously from line above].

150 kāvyaprakāśādy..., NSP, HVNR

151 śabdā'vyāpāraḥ, H, corrected in ms.

152 atra/yatra, NSP; yatra, HVNR; kim, B

153 -śaktikasyāpi pratīyate, B

154 sa, NSP, HVNR [negative counterfactual cont.: "...by which it wouldn't be there too?"]

155 ityādiviśeṣaṇe, H

156 tatpratipādanyavyavadhānena, H

60

apy abhidhātvaprasaṅgāt | tathā ca 157 prāg evānuvartamānēna 158 prakaraṇēna tatpramēyopasthāpanam śighrabhāvīti tadyavadhanēnārthāntarapratipādanam nābhidhēti cēn na 159 | tadviṣayaśaktigraham apēkṣya tatpratipādanam abhidhēti lakṣaṇam vinaiva sākṣāttavīśēṣaṇam sāādhu 160 | anatiprasangāt | tatpadārūpēṇa padabhaṅgakrta- padāntararūpēṇa vārthāntarapratipadanavyavadhanābhavarupasākṣāttavīśēṣaṇasya 161 gurutvāt | tathāviśēṣaṇē yēna dhvastamanobhavēnētyādat dhūrjatiṅpakṣē sphuṭatapadasambandhitayā prasiddhiprācuryēṇa ca jhaṭity arthapratipattiḥ vaikunṭhapakṣē yēna dhvastamanobhavēnētyādāv 162 asphuṭapadasambandhitayā 163 śāśimacchiroharaḥ andhakakṣayakara ity anayor aprayuktanihatārthatvēnāprasiddhatayā ca 164 vilambēna pratipattir iti sarvasiddhatvēna tatrābhidhāviṣayē 'pi tadabhavāpatthēḥ || prakṛtāpra- kṛtobhayaparē yatrāprakṛtē prasiddhataram padam tatra

165 prasiddhiprācuryēṇāprakṛtārthatvētēra eva prathamikatanvēnaprakṛtē 'bhidhā | prakṛtē vyaktir iti vaiparityāpattēś ca | prakaraṇam ullaṅghyāpi 166 prasiddhatarārthasyāiva prathamam smṛtēḥ 167 | ata eva dvayārtham aprasiddhē 'rthē 168 prayuktam nihatārthakam ity uktalakṣaṇasya nihatārthasya 169 kāvyadoṣatvam upagamyatē | pra- siddhatarāprakṛtārthapratitiyā 170 vyavadhanāt | na ca nihatārthasya 171 doṣatvāt tatprayogo 'rthadvayavivakṣāsthālē 'pi kāvyē na sambandhitīti vācyam | ślēṣayamakādiṣu tasyā vṛddhair doṣatvasyāpoditvāt 172 | andhakakṣayakarakahṛ rājā harati lokasya | rājā rājārcitāṅghrēranupacitakālo yasya cūdāmaṇitvam ityādidarśanāc ca | tasmāc chabdasyārthapratipādanōpayogitvapratyāsattirūpadharmābheda eva vṛttibhēda- prayojako na tu tadbhēdābhāvah 173 || pratipādanāgatasākṣāttavyavahitavabhēdē 174 saty

157 "prāg ēvārtham anuvarta-" kha, ga, NSP

158 anuvarttamānēna, H

159 cēt | na, H

160 "sādhu" dēl., H, B [so HVNR]

161 -vyavadhanarahitayārupasākṣāttavīśēṣaṇasya, B

162 "dhūrjatipakṣē...dhvastamanobhavēnētyādāv" dēl., B

163 -samnidhitayā, NSP [corr. sambandhitayā]; samdhitayā, B

164 "ca" dēl., B

165 prasiddha-, H

166 ullaṅghyāprasiddha-, H

167 smṛtē, H

168 'rthaprayuktam, H

169 nihitārthasya, B

170 -pratītivyavadhanāt, H, B

171 nihitārthasya, B

172 doṣatvābhasyopapāditatvāt, NSP, H ["doṣatvasyāpoditatvāt" kha, NSP] [so HVNR]

173 nanu tadbhēdābhāvē..., B [so HVNR]

174 -bhedaḥ, B

The Vrttivarttika

api175 saktiśakyasambandharūpapatitipādyapratyāsattibhedenaivābhidhālakṣaṇayor176 bhedadarśanāt | tadbhedābhāve177 prasiddhāprasiddhārthagocarāśabdaśleṣāsthalelakṣitalakṣaṇāsthalelakṣitalakṣaṇāsthale vyangyāsthalel vyangyāsthalecārthadvayapratītyoḥ sākṣāttavyavahitattvabhede saty apy abhidhālakṣaṇāvyaktaitiriktaṛttyadarśanāc178 cety aprayojakasyāvyavastāpādakasya ca sākṣāttaviśeṣaṇa179syāyogāt tadviṣayaśaktigraham apeksya tatpratipādakatvam evābhidhety aprakṛte 'py arthe 'bhidhāvyāpāro durvārāḥ |

naiyāyikās tu | prasiddhipracurye sati prakaraṇādiśūnyasyāpy arthasya smrteḥ pra-siddhyabhäve prakaraṇādimato 'py asmṛteś ca prakaraṇādikaṁ nābhidhāniyāmakam | śaktigrahasaṁskārapāṭavamātraniyamyatvāt tasyāḥ | kim tu liṅgavidhayā tātparyagrāhakaṁ | ata eva harir astīty anārabhyavādāśravaṇe 'pi hariśabdasyā yamanilendra-candrādayo180 'rthā ity avadhyāryata eva vaktuḥ181 kva tātparyam iti182 nāvadāryata ity āhuḥ |

tadrītyā tu183 kathamcid api184 prakaraṇādinābhidhāniyamanam śakyasaṅkam | tasmāt prastutāprastutobhayapare 'pi prastutāprastutobhayavacyārthe185 'bhidhāiva ṛttịḥ | tadupastiteṣu186 padārtheṣv ākāṅkṣādisahakāriṇivaśād ubhayavidhavākyārthapratyaya ity aprastutaviṣaye na pādārthapratītaye vākyārthapratītaye vyaktir upagantavyā | yat tu prācām aprastute187 188śaktimūlavyajanānavaṛttyabhidhānāṁ tadaprastutārthapratītimūlake yathodayam ārūḍha ityādiśeṣaṇaviśiṣṭaḥ prthivīpatịḥ svalpair grāhyadhanair lokasya hṛdayam rañjayati evam tathābhūtaś candramā mṛdulaiḥ kiranair ityādirūpeṇa pratīyamāne189 upamādyarthālamkāre tadavyāṁbhāvadrdhīkaraṇābhiprāyeṇa | na tu tatrāpi vastuto190 vyanjanavyāpārasitvabhiprāyeṇa |

175 saty api” del., B [so HVNR]

176 -samban[dha]-, -abhi[dha]-, -pratyarsatti- (with -r- stricken), H

177 ca, B, HVNR

178 -vyaktiatirikta-, H

179 -sāmyādhinatvenaprastutasyā prakaraṇasambandhāvāgamat | prāg eva tadviśeṣaṇa- erroneously inserted here, H [= NSP, p. 16, ll. 17-18].

180 -aryamānilenendracandrādayo, NSP, H [so HVNR: ex Amara, Nānārthavarga]

181 vaktuḥ param, B

182 ināva..., H [“ti” del.]. This summary of the Naiyāyika position contains no internal punctuation in H.

183 na, NSP; tu na, B [so HVNR]

184 prakaraṇāprakaraṇādinā, NSP

185 prastutobhayavacyārthe, H; prastutāprastutayor abhidhā, B, HVNR

186 ca, NSP, B

187 prastute, H [given also as the reading of “ka”, NSP; other readings: “prāgvām,” “prāñcām”]

188 śabdaśakti-, B

189 pratīyamāne, H

190 first two syllables of folio 17 illegible, H

ācchāditāyatadigambaram uccakair gām

ākramya ca sthitam udagraviśālaśṛṅgam |

mūrdhni skhalattuhinadidhitikoṭim191 enam

udvīkṣya ko bhuvi na viṣmayate nageśam ‖192

ityādi nā śabdaśaktimūlasyālāṃkāradhvaner eva dāharaṇāvatvena tair api likhitavāt |

upodharāgeṇa vilolātarakam tathā grhītam śasi nā niśāmukham iti śleṣāpādita193-

viśeṣaṇasādhāraṇyanimittamasoktyalāṃkārodāharaṇe194 rāgādiśabdānām aprastute

'py abhidhāvṛtter eva śleṣavyavahāreṇa sphuṭīkaraṇāc ca |

na cātra svatoprastutayor api kāminor prastutaśaśinīśāvacchedakatvenānvayān na tatra195

tayoḥ sarvathaiva prakaraṇāsambhava iti vācyam | viśeṣaṇasāmyād aprastutasyā gamyatve

samāsoktir iti lakṣaṇānurodhenāprastutapratipatter196 viśeṣaṇasāmyādhinatvenāpra-

stutasyā prakaraṇasambandhāvagamāt prāg eva tadviseṣaṇeṣv197 abhidhyā vācyatvāt |

prastutopamānvatvādinā yathākathamcid bhāviprakaraṇasambandhasyāpi sadbhāvāc ca

| sarvathāprastutāsambandhasyāprastutasyā kvāpi198 kavivakṣamaṇagocarātvābhāvāt |

nanv evaṁ śabdaśaktimūlasyālāṃkāradhvaner ācchāditāyatadigambaram199 ityādi-

udāharaṇasattve200 'pi tanmūlo vastudhvanir lupyetā201 | niralāṃkāre 'prastutārtha eva

vastudhvaner202 udāharaṇīyatvāt tatra ca tvayābhidhāṅgikārād203 iti cen na | yatrāprastute

'rthe 'bhidhāyaiva204 pratipādyamāne tanmūlam anabhidheyam vastu pratīyate tatra

tadaṅgikārād iti | yathā |

śanir aśaniś ca tam uccair nihanti kupyasi narendra yasmai tvam | iti

191 didhitakoṭim, B

192 "Śi[śupāla] Va[dha] 4.19", B

193 śleṣair [or śleṛṣai] pādita, H [garbled akṣara, with no sign of danda]

194 "rogeṇa" kha, ga, NSP

195 "-cchedakatvenānvayāt tatra" ka, NSP; id. H; "śaśivadanāpādakatvenānvayān na tatra", B [so HVNR]

196 -pratipattir, B [so HVNR]

197 -sāmya...tadviseṣaṇa- del., H [see n. 124]

198 kavivivakṣaṇāgocara-, H; kavivivakṣāgocara-, B, HVNR

199 -ambaram uccakair gām, B, HVNR

200 ityādya dāhāraṇasattve, B [so HVNR]

201 lupyetate, H

202 vastudhvānor, H

203 -kārā iti, H

204 yatrāprastute 'bhidhaiva, H

The Vrttivārttika

atra hy205 aśaniśabdena prastutād vajrād anyasmin samabhivyāhrtaśaniviruddharūpe 'py arthe 'bhidhayaiva pratipādyamāne viruddhāv api tvadanuvartanārtham ekaṁ kāryam kuruta ity upalokyasyā206 prabhāvātiśayah207 pratīyate | ittham208 eva prāciṇair api śabdaśaktimūlo vastudhavanir udāhṛtaḥ ||

naḥ ||

bramim aratim alasaḥdayatām pralayam mūrcchām tamaḥ śarirasādām | maraṇam ca jaladabhujagajam209 prasahya kurute viṣam210 viyoginīnām ||

ity atra 211viṣapadenāprastutam garalam vyaṅgyam vācyasya jalada eva bhujaga iti rūpakasya siddhikṛd iti vācyasiddhyaṅgam guṇībhūtavyaṅgyam udāharatā212mammatācāryeṇāprastute śakye 'pi śabdaśaktimūlo vastudhvanir uktaḥ | satyam213 | aprastutārthapratītimūlatayā pratīyamānayor vastvalamkārayor vyaktyavaśyam-bhāvadārthyāyāśritam aprastutārthe 'pi vyaktir iti prauḍhivādāmatram avalambya tat prayṛttam iti 214nātrābhiniveśaḥ kāryaḥ |

nanv evam215 prastutārthaparyālocanālabhyāyor216 eva vastvalamkārayor vyaktyupagame tatrārthaśaktimūlaiva vyaktir bhaved iti prthag arthaśaktimūladhvaneḥ śabdaśaktimūlo217 na syād iti cen maivam | tathātve 'pi prastutāprastutobhayasādharāṇaśabdasāpekṣatayā218

205 "hi" del., B [so HVNR]

206 uparalokyasya [apparently], H

207 "prakṛ[ta]bhāva" kha; "prastutabhāva" ga, NSP

208 akṣaras nearly illegible, look like "icchavaca"

209 -bhujamgajaram, H

210 viṣaṅ del., H

211 viṣapadenāprastutam, B [so HVNR]

212 manmathācāryeṇa, H; mammayā mammatācāryeṇa, B. The verse [no. 126] occurs in ch. 5 of the Kāvyaprakāśa. Mammata's words, as reported in the B.O.R.I. edition [p. 206], are: "atra hālāhalam vyaṅgyam bhujagarūpasya vācyasya siddhikṛt." [so HVNR]

213 "tat satyam" ga, NSP

214 na tatra, B, HVNR

215 "evam aprastu-" ga, NSP

216 -lambhyayor, H

217 "śabdamūlo" kha ga, NSP

218 -śabdasāpekṣatayā, B, H [so HVNR]

prastuta219mātraparāparāyāśabdāntara220parivṛttyasahiṣṇutvena tatasya pṛthag-vyavasthiteḥ221| evamanye 'pyaprastutaśakye222vyaktyanupagame223sambhāvitāḥkṣudropadravā nirasanīyāḥ|

ityappadīkṣitaviracite224vṛttivārttike mukhyavṛttinirṇayo nāma prathamahparicchedah||

219-mātrāpara-, NSP, H ["mātrapara," kha, NSP] [so HVNR]

220-śabdānta-. NSP [so HVNR]

221pṛthagvaktavyatvyavavasthiteḥ, B

222-śakye 'pi, B

223vyaktyanupagasarīmbhāvitāḥ, H

224apyadīkṣitaviracite, H [or, appai-]; śrīmadappayyadīkṣitaviracite, B

Vṛttivārttikam, II

Collation [Harvard ms.] { NSP ed.} (Benares)

atha lakṣaṇā nirṇīyate | l sā ca mukhyārthasambandhena1 śabdasya pratipādakatvam | tatsādṛśyena pratipādakatvam2 gauny api lakṣaṇāprabheda eva | tatsadrṛśe 'pi tan-nirūpitasādṛśyādhikaraṇatvaparāmpārāsambandhasattvāt | na hi sākṣātsambandhe viśiṣṭabuddhiyogasambandhe vā saty eva lakṣaṇeti niyamaḥ cakṣurāder gha anailyādiṣu3samyuktasamavāyādiviśiṣṭabuddhiyogasya parampārāsambandhasyāpi pratyāsatti-tvopattet | vyatirekalakṣaṇāsthale tannirūpitavirodhādhikaraṇatvādi4 parampārāsambandhamātreṇa tādṛśe5 lakṣaṇāklpteś ca | etena sādṛśyam na sambandha iti gauṇī lakṣaṇāto bhinnā | viśiṣ adhiyogasyāiva sambandhatvāt | samyoge sati daṇḍī devadatta itivat saty api sādṛśye simhāvan devadatta iti viśiṣṭabuddherviśeṣaṇat | iti nirastam | viśiṣṭabuddhiyanīmakaparamparāsambandhamātrasya pratyāsattitvopapatteś tāvata lakṣaṇāklpteś ca darśitatvāt |

tasmāt sādṛśyagarbhatadṛśasambandhanimittatayā gauṇī śuddhā ceti lakṣaṇāyā eva dvaividhyam | iyam ca dvividhatvāpi lakṣaṇā pratyekam dvividhaḥ | nirūḍhalakṣaṇā6 phalalakṣaṇā ca | rūḍhitulatayā7 nirūḍhalakṣaṇā | vivakṣitārthāntaradyotanaphalā phalalakṣaṇā | ādyā yathā |

lāvaṇyasāgarabhuvi praṇayāṁ viśeṣād dugdhāmbarāśiduhitus tava tarkayāmi | yat tāṁ bibharṣi vapuṣā nikhilaiḥ pratikair8 anyām tu kevalam adhokṣajā vakṣasāiva ||

atra lāvaṇyaśabdasya lāvanatvavācino hrdayamgamatvasādṛśyāt suṣamāviśeṣe rūḍhi-lakṣaṇā9 evam ālekhyaṅājādau gajādiśabdasyāpi10 rūḍhilakṣaṇānusamdheyā | idam gauṇyā udāharaṇam | śuddhā yathā |

1 mukhyārthasambandhe, H

2 "pā" del., H

3 samyuktasamavāyādivad, B

4-virodhīkaraṇatvādi-, B [so HVNR]

5 tādṛśe, H

6 rūḍhilakṣaṇā, H

7 rūḍhitulayatā nirūḍhā, H, B [implying that the NSP text is a haplography]

8 pratikarair, H [une syllabe de trop]

9 nirūḍhalakṣaṇā, B

10 nirūḍhalakṣaṇā, B

The Vrttivarttika

kvanaddvirephāvalinīlakaṅkaṇaṁ11

prasārya śākhābhujam āmravallari12 |

kṛtopagūdhā13 kalakaṅthakūjitair

anāmayaṁ prcchati dakṣiṇānīlam ||

atra dvirephaśabdāsya dvau rephau14 yasyaṁ vyutpattayā bhramaraśabdavrttes tadvacye

rūdhilakṣaṇā15 evaṁ tvagādisaabdānāṁ tvagindriyādiśv api śuddhā rūdhilakṣaṇā

draṣṭavyā |

16jahallakṣaṇājahallakṣaṇā jahadajahallakṣaṇā | sāropā sādhavasānā ca17 | śuddhā ca

gaunī18 cety evaṁ saptavidhā phalalakṣaṇā |

jahallakṣaṇā yathā |

tvatpadapadmasuṣamānukṛtiksamāni

prāptum pravālapadabhānji vibhūṣaṇāni |

valkāvṛtā dhṛtajataḥ suradirghikāyām

kalpadrumāḥ kati giriṣa tapaś caranti ||

atra suradirghikāpadasya kalpadrumādhikaraṇatvayogye19 tattire lakṣaṇā |

mukhyārthasya tadadhikaraṇatvenānvayābhāvāj jahallakṣaṇā | kalpadrumānāṁ

tapaścaryotprekṣopakārakam tīraśyātisayitapāvānvatvadyotanāṁ phalam | evaṁ

gaṅgāyāṁ ghoṣa iti prasiddhodāharaṇam apy anusaṁdheyam | tatra gosvāminaḥ

pritaye tirasyādhikāśaityadyotanāṁ phalam |

vyatirekalakṣaṇāpi jahallakṣaṇāprabheda eva | yathā |

upakṛtaṁ bahu tatrat kim ucyate

sujanatā prathitā bhavatā param |

11 -nīlakaṇṭham, H

12 agravallari, H; āmravallarim, NSP [so HVNR]

13 kṛtopagūdhah, H

14 rephau vidyete, B

15 -lakṣaṇe, H

16 "phalalakṣaṇā yathā" add., H; in parentheses, NSP, with variant readings; "phalalakṣaṇā" add., B [so HVNR]

17 "tatrādyā trividha śuddhaiva | sāropā sādhavasānā ca" add., B [so HVNR]

18 gaunī śuddhā, B; "ca" del., H

19 -yogo, H

Vṛttivārttikam, II

vidadhad īdṛśam20 eva sadā sakhe

sukhitam21 āśva tataḥ śaradām śatam ‖

atrāpakāriṇyanvayāyayoge upakṛtādipadaiḥ svārthaviruddham lakṣyate | tvayāpakāre kriyamāṇe 'pi mayivam priyam22evocyata iti svasādhutvadyotanāṃ phalam23 |

ajahallakṣaṇā yathā24 kuntāḥ praviśanti yaṣṭayaś ceti | atra bhītipalāyanavākye25 kuntādipadasya26 tadvatsu27 puruṣeṣu lakṣaṇā | kuntāder api puruṣasāhityena praveśakriyānvayād ajahallakṣaṇā | teṣāṃ nirdayaprahartṛtvādiyotanaṃ phalam |

grāmikadeśādahādau sati grāmo dagdhavān vanam ityādiprayoge grāmādipadasya svārthakadeśaparityāgena tadekadeśe28 vṛttir29 jahadajahallakṣaṇā | dagdhabhūyastvādiyotanam phalam |

sāropā30 yathā

nātha tvadīyam akalaṅkam imaṃ mukhenduṃ

āpīya tṛpyati sadā vasudhā yatas te |

tenaiva kiṃ31navasudhārasagocaro 'bhūd indhuḥ kalaṅkamalinikṛtamadhyabhāgaḥ ‖

atrenduśabdasya mukhe lāvaṇyādisādṛśyād gauṇasāropalakṣaṇā | tādrūpyapratītiḥ phalam | evaṃ gaur vāhīkaḥ mukhaṃ kamalaṃ ityādivyāse 'pi gauṇasāropalakṣaṇā draṣṭavyā32 |

20 vidadhīdṛśam, H

21 sukhatam, B [so HVNR]

22 "avo-", H

23 "phalam" del., B [so HVNR]

24 "yathā" del., H, B [so HVNR]

25 "bhītipalāyanavākye" del., NSP, HVNR; "atra bhītipalāyanavākye kuntā-" kha, ga, NSP; atra hi palāyamānāvākye, H; atra bhītipalāyamānāvākye, B

26 kuntapadasya, B [so HVNR]

27 tadvastutsu, H

28 tadaikadeśe, H, B [so HVNR]

29 vṛttir, B [so HVNR]

30 sāropalakṣaṇā, B; "gauṇasāropā yatha' is what we require," HVNR

31 "tava" ka, NSP

32 "draṣṭavyā" del., B

The Vrttivarttika

gauṇasādhyavasānalakṣaṇā33 yathā |

nābher abhūt tava caturbhuja nāntarikṣam34

yan nābhir eva yadunetar35 iyam tato 'bhūt |

anyon yasamiśrayaparāhatidūṣitam yad

āha śrutis tad avimrśya kim atra tattvam36 ||

atrāntarikṣapadasya bhagavadavalagne sūkṣmatvasādṛśyād gauṇasādhyavasānālakṣaṇā |

sarvathaivābhedapratipattih phalam | viṣayasya viṣayibhedenāpahnutenā37 tad-rūpoparaktatāpratītis38 tadṛūyapratipattis39 viṣayasya viṣayiṇāvabhedapratītīḥ sarvathāivābhedapratipattir iti sāropasādhyavasānālakṣaṇāphalayor bhedaḥ | viṣayaviṣayi-vācakayor dvayor api nirdeśā āropādhyavasāyayor40 bhedaḥ

nanu sāropalakṣaṇodāharaṇe mukhendur ityādāv indvādipadasya mukhādu na mukhatvādinā lakṣaṇā mukhapadasyāpi41 sattvena paunaruktyāpatteḥ | nāpīndugata-lāvanyādiguṇena[syānyatra pravṛttinimittatvāyogāt l nāpi mukhagata-lāvanyādiguṇena[mukhyārthasambandhāraharitaguṇa]42mukhena lakṣaṇāyogāt | maivam | mukhacandrodbhayānugatālāvanyasāmānyādimukhena lakṣaṇopagame doṣādvayasya apy asamiśparśat | idam evābhisamdhyāyoktam vṛddhair | lakṣyamānaguṇair yogād vṛtter iṣṭā tu gauṇateti |

śuddhasāropalakṣaṇā yathā |

āpādam ācikurabhāram aśeṣam aṅgam

ānandasāram aravindṛśām asiṃam |

antar mama sphuratu samtatam antarātmā

ambhojalocana tava śritahastiśailam43 ||

33 -sādhyavasāyālakṣaṇā, B. These two forms seem in free variation, even in the same text; Further, B rarely agrees with NSP. Henceforth, the variation will be ignored.

34 caturbhuja antarikṣam, H

35 eva dunetar, H

36 arthattattvam, NSP

37 viṣayibhedānāpahnāvena, B, HVNR

38 tadrūpoparaktatāpratītis, H

39 -s del., H [apparent cpd. with next word]

40 āropādhyavasānayor, H, B [this variation noted because the two forms occur in the same sentence].

41 mukhādipadasyāpi, B

42 portion in [brackets] del., H

43 -hastaśailam, H

atra bhagavadangeṣv ānandakāritvenānandapadasya sāropalakṣaṇā | ānandakarane itaravailakṣaṇyadyotanāṃ phalam | ānandakāriṇi viṣayanigaranenānando44 'yam iti prayoge sādhayavasāyalakṣaṇā ānandāvyabhicārādyotanāṃ45 phalam | evaṃ śuddhā pañcavidhā gauṇī dvividheti saptavidhā phalalakṣaṇā |

nanu gaunyāṃ api jahadajahallakṣaṇābhedau sambhavatah | mukhaṃ kamalaṃ gauravāhika ityāдау jahallakṣaṇā tāvat sphuṭaiva46 | ajahallakṣaṇāpi47 govāhikobhayaviṣaye gāva ete samāniyantāṃ ityādau drśyate | ataḥ katham dvividhaiva gauṇī | na ca tatraiva tatsādṛśyaṃ asti yena mukhye49 'pi gauṇī vṛttih syāt | śuddhā tv ajahallakṣaṇā kuntaḥ praviśantītyādau50 viśeṣaṇavācino viśiṣṭe51 sam்bhavati viśiṣṭaviśeṣaṇayor ekadeśyekadeśābhāvarūpasam்bandhasattvāt | ato gāva ete samāniyantāṃ ityādāv abhidhālakṣaṇābhyāṃ evārthadvayapratipādanam | gaṅgāyāṃ jalam் ghoṣaś ca tiṣṭhatīti atreveti52 | na gaunyāṃ53 jahadajahallakṣaṇābheda-sam்bhavah | tasmāt saptavidha lakṣaṇābhedā iti | itthaṃ ayam ciraṃtanair pradarśito lakṣaṇāvibhāga iti |

atredaṃ pratibhāti | śuddhāsu sādhayavasāyalakṣaṇāprabheda54kathanam ayuktam | parārthaṃ svasamarpañasyāviśeṣeṇa55 tasyā jahallakṣaṇānātirekāṭ | na hi tadaviśeṣe 'pi sāropalakṣaṇāyāṃ viṣayavācakasāmānādhikaraṇyakṛta[vaicitryavad ihaiva]56 vaicitryāntaraṃ asti yena tadvad eva jahallakṣaṇātāḥ prthag udāhriyeta57 | nanu kāryakāraṇabhavāsaṃbandhena58 vṛttir ity ataḥ eva tataḥ59 pṛthaktvaṃ astu kāryakāraṇabhāvātiriktena gaṅgātirādi[gaṭasamyogādi]60sam்bandhena vṛttau jahalkāryakāraṇabhāvātiriktasam்bandhena vṛttau jahallakṣaṇā ity ucyate |

The Vrttivārttika

lakṣaṇety asamikāropapatteḥ | maivam | rājaki̱yaḥ puruṣo61 rājā | indrārthā sthūṇā indraḥ | ataksāyam taksā | agrāhasto 'yam hasta iti svasvāmibhavatādarthyatatkarmyāvayavāvayavibhāveṣv api kāryakāraṇabhāva iva sāropādhyavasāyalakṣaṇayor ākara-granthesūdāharaṇād62 āśrayāśrayibhavasambandhe 'py alarikārasudhānidhāv udāharaṇāc ca vaicitryāntarābhāve 'pi sambandhabhedamātreṇa vidhāntaropagame saṁbandhabhedāntaram avalambyāpi vidhāntaropavarnanasyā63 pratyākhyātum aśakyatayā saptavidhatvaniṣamāsamānjasya-prasaṅgāc ca |

nanu sādhavasāyalakṣaṇodāharaṇeṣv abhedābhivyaktir asti | ayam ānandaḥ ayam rājetyādāv ānandarājyābhedābhivyaktipūrvakam evānandāvyabhicāritvānul-laṅkhyāśasanatvādiphalapratīteḥ64 | ato vyaṅgyavaicitryeṇa bhedaḥ 'stv iti cen na | jahallakṣaṇodāharaṇe 'pi tire pravāhābhedābhivyaktipūrvakam eva tatra tadgatātiśayita-śaityapāvanatvādiphalapratīteḥ | tire pravāhatādātmyapratibhave65 tatra pravāha-gatātiśayitaśaityādeḥ66 pratyayāsambhavāt | na hi tire gaṅgāsambandhaprayuktam tadgatasvābhāvikaśaityādiyotanam eva lakṣaṇaphalam yena gaṅgādisambandhi-tiratvena67 lakṣaṇayiva tal labhyata iti nābhedābhivyaktir aṅgikriyate | saty api mukhye gaṅgātirāḍipade svāyatte ca śabdaprayoge mukhyaprayogād68 api labhyasyā lakṣaṇāphalatvāyogāt | kāvyasaraṇau69 lakṣaṇāyām70 kāvyaśobhātiśayādhāyakatayāpy71 adhikaram stotukāmas tasyā pravāhatādātmyapratipatyā tadgatātiśayitapāvanatva-dyotanāya tasmin gaṅgāpadam prayuṅkta iti72 darśanāc ca | ato jahallakṣaṇāyām apy abhedābhivyaktir aviśiṣṭā | uktam ca kāvyaprakāśikāyām73 | taṭādīnām gaṅgādiśabda-pratipādane tattvapratipattau hi pratipipādayiṣitaprayojanasanampratyaya iti | 74ratnākareṇāpy uktam | srotastiroy ekaśabdabodhyatvena75 tādātmyapratīteḥ sroto-dharmāḥ śaityapāvanatvādayas tire pratīyanta iti prayojanasiddhir iti | ataḥ

61 puruṣo 'yam, B

62 udāharaṇad, H

63 "varnanasyāpi" kha, ga, NSP

64 -śāsanatvādi-, B [so HVNR]

65 -abhave 'pi, B

66 "śaityapāvanatva(de)ḥ pra-" kha; "śaityapāvanatvapratya-" ga, NSP

67 -tiratvena; H; gaṅgāsambandhitiratvena, B

68 mukhyayogād, H

69 See S.K. De, H.S.P, 1:225-26.

70 kāvyaśobhā-, H

71 (...ādhāyakatayā) pravāhagatatīśayitaśaityādisvābhāvikam pāvanatvam dyotayato ('py) add., B

72 "lakṣaṇāyām...prayuṅkte" is treated as a quotation, NSP

73 Cf. K.Pr. ad 2.10 (B.O.R.I. ed., p. 46).

74 ratnakāreṇa; see S.K. De, H.S.P, 1:226.

75 ekaśabdabodhyatvena; H; ekaśabdabodhyatve nadī-, B

sādhyavasāyalakṣaṇāyāḥ phalakṛto 'yakṛto vā na jahallakṣaṇāto bhedaḥ76 saṁbhavatīti na tadudāharaṇapārthakyam yuktam77 ‖

tathā sāropalakṣaṇodāharaṇe78upamānopameyānugatasādhāraṇādharmaṃukhena lakṣaṇety ayuktam1 tathā sati yatra79 sādhāraṇadharmasyāpi pāntareṇopādānam tatra80 paunāruktyāpatteḥ1 yathā1

tārābhirāmapariṇāhalasatītābhram் tāpicchamecakam uraḥśaradantarikṣam1 prāyaiva deva tava81 kaustubhapūrṇacandraḥ pūrṇām bibharti puruṣottama kāntirekhām ‖

atra hi vakṣo'ntarikṣayoh sādhāraṇaḥ pariṇāhitvamekatvādidharmo 'pi prthag upātta eva1 upāttadharmātiriktadharmamukhena lakṣaṇāstv iti cen na1 upāttamekakatvādivyatiriktasāmānyadharmāsphuraṇe82 'pi sahṛdayānām paripūrṇavākyārthapratītidarśanād dravyatvādisāmānyadharmasya kavitānańgatvāt83 1 antataḥ kavayitur apy etādṛśasthaleṣūpāttadharmātiriktadharmagarbhīkārābhāvāc84 ca1

nanv iha mā bhūt sāropalakṣaṇā1 ura eva śaradantarikṣam iti mayūravyamsakādi-samāsopagame hi sā bhavati1 kim tu taḥ śaradantarikṣam ivety upamitasamāso 'stu1 tatra ca vigrahavākyāsthā luptā ivakāraḥ sādrśyām pratipādayatīti nopamānāvācakasyopameye85 lakṣaṇety ālamkārikāṇām86 abhyupagamah1 tatra tair luptopamāmśikārād iti na kaścid doṣa iti cen na1 mecakatvādisāmānyasya87 śabdopāttatayopamita-samāsāsambhavāt1 upamitam vyāghrādibhiḥ sāmānyāprayoge iti sāmānyāprayoga eva tatsamāsānusāsaṇāt1 ālamkārikair88 api sāmānyāprayoge saty upamābādhakasattvāt89

76 jahallakṣaṇābhedaḥ, H

77 "pārthakyayuktam" kha, ga, NSP

78 -haraṇepamāna-, H

79 tatra, NSP

80 tatra tena, B

81 bata, H

82 uttamekatvādivyatirikta-, H; upāttamekatvādivyatirikta, NSP

83 kathitānangatvāt

84 -dharmam ańgīkārabhāvāc, H

85 upamānāvācayopameyalakṣaṇā, H

86 ālamkāriṇām, H

87 -sāmānyadharmasya, B [so HVNR]

88 ālamkārikai1 rapi, H

89 upamāvācakasattvāt; H; "...vācakasattvā" kha, ga, NSP [this reading may be correct, if, as seems probable, vācaka = dyotaka]

The Vrttivārttika

sāropalakṣaṇāmūlako rūpakālamkārah | na tūpamālamkārā ity abhyupagamāt | upamitasamāse sā mānyadharmavācakaluptopamālamkārasyai va tair udāharaṇāc ca | sā mānyadharmaprayoge 'py upamitasamāsāngikāre hi kevalāvācakaluptāpi tatrodāhriyeta |

nanv evam sādhāraṇadharmaprayoge sādrśyamukhena lakṣaṇāstu | tadvācakasya yathevāder aprayogena paunaruktyāprasangāt | tatprayoge upamānayācakasya sādrśyapratiyogibhutāsavacyāparatayopameyalakṣaṇāyā avācyatvād iti cen na | tathā sati sādrśsyasya śabdopāttvenopamālamkārasya prāptyā sāropalakṣaṇāyām rūpakādir alamkāro nopamety alamkārāvibhāgāsamānjasya prasangāt | na hi sādrśsyasya vācyatva evopamā kamalasuhṛn mukham ityādau suhrdādipadena lakṣaṇāyām apy upamā sampratipatteh | api ca | pādāmbujām ityādau vyāghrādisamāsena rūpakāś ca sambhavena sarvatranavadharāṇe prāpte yatra

pādāmbujaṁ bhavatu vo vijayāya mañju mañjiraśiñjitamanoharam ambikāyāḥ

ityādau pāde sambhavamañjira yogadikam nibadhyate tatra tadānuguṇyāt pūrvapadārthapradhānena vyāghrādisamāsenopamālamkārāḥ | yatra tu yasyāniśaṁ diviṣadaś caraṇāravindam uttamsayanty amitabhaktibharāvanamrāḥ

ityādāv aravindakāryam uttamsanādikam nibadhyate tatra tadānuguṇyād utara padārthapradhānena mayūravyamsakādisamāsena rūpakālamkārā iti vyavasthā sarvair api pratipādyate | na ceyam vyavasthā caraṇāravindam ityādirūpakasthaleṣv aravindādipadānām aravindatvādyākārām apahāyāruṇimādisādhanadharmamukhena tannibandhanasādrśyamukhena vā caraṇādilakṣakatve yujyate | tathā saty utara-

90 "sāmānya-" del., NSP; B; "sāmānyadharma" ka, NSP

91 yathevāder; H; yathevadar, B (so HVNR)

92 paunaruktyaprasangāt, H

93 "tattatprayoge" kha, ga, NSP

94 "sādrśyaśabda-", H; "sādrśyasa"- ka, NSP

95 "sādrśyavācya-", H; "sādrśyavācyatva" ka, NSP

96 tallakṣaṇāyām, H, B

97 vyāghrādeḥ samāsena, NSP, H; vyāghrādīdisamāsena, B [so HVNR]

98 mayūradisamāsena, B [so HVNR]

99 nanv evam, B [so HVNR]

100 -sthale 'pi, B

74

padārthasyāruṇyasādṛśyaguṇasya101 prādhānyayogena102 mayūravyamsakādisamāsāśrayane 'pi vyāghrādisamāsāśrayaṇa iva pūrvapadārthacaraṇasyaiva prādhānyāpattyottaramsanānuguṇyasya virahatulyāpatteḥ | guṇajātyor jātipādhānyasya nīlotpalādau vyavasthiteḥ | uttarapadārthasya prādhānye 'py aravindajāter ivāruṇyasādṛśyādiguṇasyottaramsanānuguṇatvābhāvāc103 ca | tasmāt sāropalakṣaṇodāharaṇeṣu sādhāraṇadharmādimukhena laksyaṇetyuktam104 |

tathā tadrūpyapratītir atra phalam ity apy ayuktam | sāmānyadharmamukhena sādṛśyamukhēna lakṣaṇopagame tadrūpyābhivyaktyayogāt | tathā hi kāntimān105 mukham candrasadṛśam mukham ityādir arthāḥ sampadyate | na caitavatā tadrūpyābhivyaktiḥ | kāntimān mukham ityādiprayoge 'pi tatprasangāt | na ca gaṅgāyām ghoṣa ityādau srotastīrayor abhedādhyavasāyāvan mukhacandrayor ekāśabdabodhyatayā kevalāśabdaśaktimūlavyāñjanāvyāpāreṇa106 tadrūpyādhyavasāyā iti nātiprasanga iti vācyam | tadā mukhakāntimator eva tadrūpyābhivyaktiḥ sambhavena

vaktrendau107 tava saty ayam yad aparah śitāmiśur ujjṛmbhate ityādāv ānubhavikasya108 mukhatvoparaktacandratādrūpyapratyayasyākliṣṭa-samarthanāsambhavāt | tathā mukham kamalam ityādivyāse 'pi sarvatra viṣayaviṣayipadayoḥ sāmānādhikaraṇye sāropalakṣaṇety apy ayuktam |

vidvanmānasahamsa109 vairikamalāsaṃkocadiptadyute durgāmārganaṇīlaloḥita samitsvikāravaiśvānara110 | satyapritividhānadaḳṣa vijayapragbhāvabhīma prabho sāmrājyam varavīra vatsarasatam vairiñcam111 uccaiḥ kriyāṃḥ ||

101 “sādṛśyadiguṇasya” kha, ga, NSP; so B 102 prādhānyādigyogena, NSP, HVNR; “prādhānyāyogena” kha, ga, NSP 103 -guṇasyauttaram-sa-, H 104 uktam, H 105 kāntman mukham, H 106 -vyañjanāvyāpāreṇa, NSP 107 vakrendau, B [so HVNR] 108 ānubhāvikasya, NSP, H [so HVNR] 109 -hamsaḥ, H [“ha” appears to be stricken] 110 -vaiśvānaraḥ, H 111 vairañcam, H

The Vrttivārttika

ity atra hamsādiśabdānām varnye rājani pratyāsattirūpamukhyasādṛśyāpratītyā112 lakṣaṇāsambhavāt | tathā hi viduṣāṁ mānasam hṛdayam eva mānasaṁ sara ityādi śleṣabhittikābhedādhyavasāyālabhyamānasavāsitvādisādharāṇadharmanibandhanam eva tatra sādṛśyaṁ vācyam | tadanyasya113 mukhakamālāsādat prasiddhasādṛśyasyābhāvāt | na ca tena sādṛśyena lakṣaṇa sambhavati tallabhyakasyā114 śleṣasya rūpakotthāpyatayā rūpakopalakṣaṇadasāśrayāṁ115 tasya buddhinthānārohat | na ca śleṣasyāiva116 prāthamyaṁ kim na syād iti vācyam | rūpakam vinā kaver arthadvayavivakṣāyām iha gamakābhāvena śleṣānirvṛtteḥ117 | abhidhāyāḥ prakaraṇāniyamayatvapakṣe prācinābhimate rūpakāt prāk prakaraṇasambandharahitasāropalakṣaṇārthāntarābhidhānasyāivāsambhavāt118 |

rūpakam purvasiddharin śleṣam taj jñāpayed yadi | tadā rūpakam eva syād anyathā śleṣa iṣyate ||

iti śleṣarūpakayor vibhāgaṁ vyavasthāpitavatā cakravartināpy asminn udāharaṇe rūpakopavarnanenātra śleṣāpaścātyatvasya119 sphuṭīkaraṇāc cete | atredam tattvam | vidvanmānasahamsetyādau hamsādiśabdānāṁ varnye rājani lakṣaṇā nāṅgikartavyā | sāmanādhikaraṇyena varnyasya rājñāḥ prasiddhahamsasajjītyādibhir abhedasya vākyārthavidhayāiva120 pratītyupapatter | na caivam abhedapratyayasyasambhave 'pi tādrūpyapratyayo na lakṣaṇāṁ vinā sidhyet | sa eva ca121 viṣayaviṣayipadasāmanādhikaraṇyasthale sarvatra vivakṣita iti vācyam | tatra mānābhāvāt | ūrdhvam viriñci-bhāvanāt122 tava nābhipadmād123 ityādau viṣayaviṣayipadasāmanādhikaraṇye 'py evakārasākṣātpādābhyāṁ urasthale prasiddhaparaparāpadābhedavivakṣāyā124 eva sphuṭīkaraṇāc ca | na caivaṁ vidvanmānasahamsetyādāv atiśayoktir eva syāt | viṣayasyā viṣayitadrūpye rūpakam viṣayābheda 'tiśayoktir iti vyavasthitel | tathā ca viṣayaviṣayi-sāmanādhikaraṇyasthale rūpakam viṣayābhedamātranirdeśasthale 'tiśayoktir iti vyavasthāvibhāga iti vācyam |

112 "sādṛśyapratītya" kha, ga, NSP

113 tadanyamukha-, NSP

114 tallabhyakāsya, B

115 rūpakāpelakṣaṇādaśāśrayāṁ, H [superlinear stroke for "o" placed over "pa"]

116 śleṣasya, NSP

117 śleṣānirṛtteḥ, NSP, B

118 -abhidhānasyavāsambhavāt, NSP [so HVNR]

119 -paścātyatvasya sphuṭī-, NSP; -paścātyatvasya, B [form not attested, lexica] [so HVNR]

120 "virodhayaiva" ka, NSP

121 "viṣaya-" del., H

122 All the mss. read "virañci-", though most read "viriñci-" above.

123 "tava nābhipadmād" del., B

124 prasiddhaparaparābhedavivakṣāyād ata eva, H; -padmapada-, NSP [so HVNR]

hṛtpankajāni125 bhajatām pratibodhayanti

saṃsārasāgaram api praviśoṣayanti |

jyotsnā tvadaṅghrinakhaçandrasamudgateyam

anyādṛśim abhayadām126 prakṛtim bibhartí ||

ity atra nakhakāntirūpaviśayīṅgarāṇena jyotsneti viṣayipadamātranirdese 'pi pañkaja-

vikāśakatvādinā127 prasiddhajotsnāto vyatirekeṇa tattadrūpyamātrapratīter darśanena

128vyavasthādvayasyāpy aśakyāṅgīkārat129 | tathā ca yadi tadrūpye rūpakam abhède

'tiśayoktir iti vyavasthām parītyaja viṣayaviṣayipadasāmānādhikaraṇye rūpakam

viṣayipadamātranirdese 'tiśayoktir iti vyavasthādriyate tadā vidvanmānasaṃsahetāv

abhedapratītāv satyām api rūpakam sambhavatīti na kācid anupapattịḥ | yadi ca tathā

sati rūpakātiśayoktyoḥ śabda130vacitryamātrebhedaprāptyā parasparam arthālam-

kārārūpatvam na syād ity ādyā vyavasthādriyate tadā jyotsnā tvadaṅghrinakhetyādau

viṣayipadamātranirdese 'pi prācīnamaryādāyām atiśayoktim apalhāya rūpakam evaṅgī-

kartavyam | tadvat vidvanmānasaṃsahetau viṣayaviṣayipadasāmānādhikaraṇye 'py

atiśayoktir aṅgīkriyatām | kim anupapannam | tasmād etadrśasthe samabhi vyāhārād

abhedapratītyupagame na kaścid doṣa iti vyartho lakṣaṇopādānaprayāsāḥ | na caivam

saryatra131 viṣayaviṣayīsamānādhikaraṇyāsthale samabhi vyāhārād abhedapratīter vaktum

śakyatvāt kvacid api sāropalakṣaṇā na syād iti vācyam iṣṭāpatteḥ ||

iti śrī appadīkṣitavīracite132 vṛttivārttike lakṣaṇāvṛttinirṇayo nāma dvitīyaḥ paricchedaḥ |

samāptaś cāyam samupalabhyamāno granthaḥ133 ||

125 -pañkajāti, H

126 abhayada-, B

127 -vikāśaka-, H

128 vyavasthādvayasa, H ["dva" del.]

129 aśakyāṅgīkaratvāt, NSP, B

130 H. breaks off at this point, in the middle of the sixth line of text.

131 viṣayaviṣayipada-, B, HVNR

132 śrimadappayyadīkṣitavīracite, B

133 śrisitārāmābhyām namaḥ, B ["samāptaś ... granthaḥ" del., B]

Bibliography

Sanskrit Works

Ānandavardhana, Dhvanyāloka, tr. D. Ingalls et al., H.O.S., vol. 49 (Cambridge, Mass., 1990)

Apadeva, Mīmāṃsānyāyaprakāśa, ed. F. Edgerton (New Haven, 1929)

Appaya Dīkṣita, Kuvalayānanda, N.S.P. ed. (Bombay, 1955)

, Vrttivārttika, N.S.P. ed. (Bombay, 1940); id., Benares ed. (Benares, 1978)

, Citramīmāṃsā, Kāvyamālā, vol. 38 (Bombay, 1941)

, Varadarājastava, ed. A. V. Gopalachariar, Vāṇī Vilās Skt. Ser., vol. 11 (Śrīraṅgam, 1927)

Bhartrhari, Vākyapadīya, ed. K. V. Abhyankar (Poona, 1965)

Bhaṭṭojidīkṣita, Siddhāntakaumudī (New Delhi, 1985) [rpt. of N.S.P. ed., Bombay, 1899]

Jagannātha Paṇḍita, Citramīmāṃsākhaṇḍana (publ. as appendix to Citramīmāṃsā, supra)

Jaimini, Pūrvamīmāṃsāsūtrāṇi, with commentaries by Śabara and Kumārila, Ānandāśrama Skt. Ser. vol. 17 (Poona, 1953)

Kālidāsa, Kumarasambhava, N.S.P. ed. (Bombay, 1955)

Kumārilabhaṭṭa, Śloka Vārttika, tr. G. Jha (Calcutta, 1900-8) [rpt. ed. Delhi, 1983]

, Tantra Vārttika, tr. G. Jha (Calcutta, 1903-24) [rpt. ed. Delhi, 1983]

Māgha, Śiśupālavadha, N.S.P. ed. (Bombay, 1957)

Mahimabhaṭṭa, Vyaktiviveka, Trivandrum Skt. Ser., vol. 5 (Trivandrum, 1909)

Mammaṭa, Kāvyaprakāśa [with commentaries by Nāgeśa and Govinda], Ānandāśrama Skt. Ser.. vol. 66 (Poona, 1929)

Pāṇini, Aṣṭādhyāyī, with Vāmana's and Jayāditya's Kāśikāvṛtti. Kāśī Skt. Ser., vol. 37 (Benares, 1952)

The Vrttivarttika

Parthasarathimisra, Sastradipika, Krishnadas Skt. Ser., vol. 103 (Varanasi, 1988) [rpt. of N.S.P. ed.]

Ruyyaka, Alamkarasarvasva, Kavyamala, vol. 35 (Bombay, 1939)

Vamana, Kavyalamkarasutrani, N.S.P. ed. (Bombay, 1953)

Visvanatha, Sahityadarpana, N.S.P. ed. (Bombay, 1936)

Secondary Sources

S. K. De, History of Sanskrit Poetics, 2nd rev. ed., 2 vols (Calcutta, 1960)

S. K. De and S. N. Dasgupta, History of Sanskrit Literature (Calcutta, 1962)

A. Gangopadhyay, Contribution of Appaya Diksita to Indian Poetics (Calcutta, 1971)

E. Gerow, Glossary of Indian Figures of Speech (The Hague, 1971)

, “Language and Symbol in Indian Semiotics,” Philosophy East & West 34 (1984)

P. V. Kane, History of Sanskrit Poetics, 3rd rev. ed. (Delhi, 1961)

K. Potter, Encyclopedia of Indian Philosophies, vol. 1 (Princeton, 1970)

V. Raghavan, in Proceedings, A.I.O.C. (Tirupati, 1941) [see De, H.S.P. 1:223]

K. K. Raja, Indian Theories of Meaning, Adyar Library Ser., vol. 91 (Madras, 1963)

R. Sewell, A Forgotten Empire (London, 1900)

B. Stein, Vijayanagara, New Cambridge History of India, vol.1, pt. 2 (Cambridge,1989)

80