1. A Glossary of Indian Figures of Speech Edwin Gerow ( See Sanskrit Poetics)

PUBLICATIONS IN NEAR AND MIDDLE EAST STUDIES COLUMBIA UNIVERSITY

Editorial Board

JOHN S. BADEAU CHARLES P. ISSAWI JACOB C. HUREWITZ KARL MENGES

ALEX WAYMAN EHSAN YAR-SHATER (Chairman)

Series A

XVI

This work was approved and edited for publication in this series in 1966 by the Editorial Board consisting at that time of:

JOHN S. BADEAU CHARLES P. ISSAWI

DOUGLAS M. DUNLOP TIBOR HALASI-KUN

JACOB C. HUREWITZ JOSEPH SCHACHT

EHSAN YAR-SHATER

This series, published under the auspices of the Department of Near and Middle East Languages and of the Near and Middle East Institute of Columbia University, consists of monographs, readers and other studies designed to promote systematic research on the Near and Middle East and to further public understanding of the problems of the area. The opinions expressed are those of the individual authors and do not neces- sarily reflect those of the Department or the Institute.

A GLOSSARY OF INDIAN

FIGURES OF SPEECH

by

EDWIN GEROW UNIVERSITY OF WASHINGTON

1971

MOUTON

THE HAGUE . PARIS

c Copyright 1971 in The Netherlands. Mouton & Co. N.V., Publishers, The Hague. No part of this book may be translated or reproduced in any form, by print, pho!,. print, microfilm, or any other means, without written permission from the publi:her:

LIBRARY OF CONGRESS CATALOG CARD NUMBER: 70-111623

Printed in The Netherlands by Mouton & Co., Printers, The Hague

To the memory of Mme. Nadina Stchoupak, whose intention was to compile such a Glossary, and who would have made a better job of it.

PREFACE

It has been possible to compile the present work only by imposing strin- gent limitations on method and subject matter. We have been concerned with the figures as form and have treated them comparatively, as formula- tions of the manifold poetic idea. The comparative emphasis is inevitable if the figures are to be understood as a poetic. We have omitted a great deal that could be said about the figures: we have not considered their historical development, the vagaries of their individual formulation, or the biases and peculiarities of the authors who discussed them. Such questions have been exhaustively treated elsewhere, and they are only a prolegomena to the study of the figures as a poetic. The method we have used implies a system of figures, varieties of the poetic dicendum. We have tried in the Introduction to distinguish this approach from others which have been suggested. We have studied systematically only the first third of the figurative tradition: the pre-dhvani or early ālamkārika period. We justify this limitation on the ground that it is precisely in this period that the figures were studied as form. We thus emphasize the distinctiveness of the early or formative period as a school of poetic criticism and attempt to establish its presuppositions and achievements not merely in terms of what it anticipated (the dhvani) but in terms of what it was: a serious study of the kāvya, a poetic genre which was realized in classical India and which remains the single most impressive monument of Indian literature. The eight works used as a basis for this collection of figures of speech are, in rough chronological order: Bharata, Nāțyaśāstra: part of the sixteenth adhyāya (16.40-84); Bhāmaha, Kāvyālamkāra: second and third paricchedas of six; Daņdin, Kāvyādarśa: part of the first (52-68), all of the second, and part of the third (1-124) paricchedas; Vāmana, Kāvyālamkāravrtti: the entire fourth adhikaraņa;

PREFACE 6 Udbhața, Kāvyālamkārasarasamgraha: the entire work; Agni Purāņa: adhyāyas 343-345; part of 342 (18-33); Rudraa, Kāvyālamkāra: part of the second (13-32), the third through the fifth and the seventh through tenth adhyāyas; Mammața, Kāvyaprakāśa: the ninth and tenth ullāsas. My thanks must first of all be addressed to my teacher, Professor Johannes A. B. van Buitenen, at whose suggestion this study was undertaken and who provided the encouragement and careful criticism needed to com- plete it, and to Professor Louis Renou, under whose strict guidance the first concrete work was done. To Professor George V. Bobrinskoy, who encouraged me to take up the study of Sanskrit, is reserved that particular affection which must always attach to the ādiguru. I must also express my indebtedness to Professor V. Räghavan, whose advice and bibliographical expertise made my stay in Madras more fruitful; to Sāhityaśiromaņi V. Ranganāthan, who first showed me the beauties of the kāvya; and, with profound sorrow, to the memory of Mahopādhyāya Vidyāsāgara K. L. Vyāsarāya Sāstri, whose passing deprives us of another link with the traditions of Indian grammar and pāņinīyaśikșā. A special indebtedness must be acknowledged to Professor David Hadas; in conversations with him the ideas guiding the writing of the Introduction were refined and, it is hoped, given a more immediate relation to contemporary literary criticism. Of particular importance to me has been the encouragement of Professor Tibor Halasi-Kun, without whose many-sided bienveillance this work would not have been published. A portion of the research on which this book is based was supported by the Foreign Area Training Fellowship Program of the Ford Founda- tion, thanks to which I was able to spend one year in Paris and another in Madras (1959-61) before completing the Glossary in 1962. Needless to say, the responsibility for this work attaches to no one but myself, and, if indulgence is to be asked-as it must in a work which pretends to novelty-then I would ask it particularly for the speculative attempt to reassess the history of Indian poetics which constitutes the Introduction. It is here presented as an Introduction, but in fact it is a summation, a work undertaken on the basis of the Glossary, an attempt to make sense of the alamkdra as a contribution to poetics. Seattle, Washington, 1965 E.G.

CONTENTS

Preface 5

Introduction. 9 (I) The Problem 9 (a) Poetic as Aesthetic Psychology 9 (b) Poetic as Rhetoric . 13 (c) The Notion of Alamkāra . 16 (II) History of the Search for System 22 (a) Arthaśabda 22 (b) The Criteria of Differentiation 25 (c) An Alternative System: The Stylistic Argument 26 (d) Rudrata: The First Systematist 35 (i) Simile 35 (ii) Hyperbole 37

(iii) Pun: śleșa 38

(iv) Svabhāvokti. 42 (e) Poetic as an Intellectual Discipline. 48

(III) The System of Figures. 50

(a) The Question of Infinitude 50

(b) The System. 55

(IV) The Adequacy of the Alamkārika Poetic 70

(a) Kāvya 70

(b) Nāțya 74

(c) The Religious Lyric 78

(V) Conclusion 81

(VI) Scope of the Glossary . 84

(VII) Method of the Glossary 85

CONTENTS 8

Postscript

A Glossary of Indian Figures of Speech. 90 Numbers and Abbreviations Used 93 Appendix 95 333 Bibliography

. 337 Index of Figures

. 340

INTRODUCTION

(I) THE PROBLEM

The Indian poetics, or alamkāraśāstra,1 has not been favored in the West with attention commensurate to its position among the traditional intellectual disciplines of India. The interest that has been shown has tended to concentrate on those aspects of Indian poetics which bear some superficial resemblance to theories current in the West and indeed further our expectations (or prejudices) of poetry itself. The result has generally been a haphazard and distorted view of the nature and aims of Indian poetics, and interpretations of the history of that poetics which imply critical standards alien to it. What is needed is a new approach to the entire problem-one which, while retaining its critical distance, attempts to place Indian poetics in the context of classical Indian poetry and culture, and to perceive in some measure what important poetic problems the poetic was in fact explicating. We must not judge it only insofar as it may explicate problems posed by contemporary poetry.

(a) Poetic as Aesthetic Psychology Starting from the Crocean position, whose own roots go back at least to Longinus, most contemporary critics of classical Indian poetics2 tend 1 Lit .: 'science of the ornaments (or figures of speech)'. On the word alamkāra (lit .: 'a making adequate'), cf. "The Meaning of the Word Alamkāra" in J. Gonda, A Volume of Eastern and Indian Studies in Honour of F. W. Thomas (Bombay, 1939), pp. 97ff. 2 Particularly S. K. De, whose views will retain our attention throughout this Introduc- tion. His recently published lectures Sanskrit Poetics as a Study of Aesthetic (University of California Press, 1963) (henceforth referred to as SPSA) state his critical view with cogency and clarity, but it is represented also in many of De's other works: History of Sanskrit Poetics (2nd ed., Calcutta, 1960) (hereafter cited as HSP); Some Problems of Sanskrit Poetics (Calcutta, 1951), pp. 1 ff., etc. The views which De expresses most vigorously are also held by others as, for example, S. N. Dasgupta (with whom he

INTRODUCTION 10 to situate the nature of poetry in the poetic act, in the unique individualized inspirations of nascent literature. A true poetic is really a psychology of creation, an aesthetic and not mere criticism. It finds its culmination in a respect for, and a theory of, genius. Since the poetic act is in fact a kind of communion of the self with an ineffable sacrum, it follows that the poetic work is, for the aesthetician, only a touchstone, a pretext for flights into the metaphysics of creation. And a poetic like that of classical India which is uniquely concerned with the poetic fact is not only irrelevant to the poetic problem but, lamentably, an obstacle and a limitation to any proper understanding of it. The attitude of De towards this poetic reveals itself in distinctions of inner and outer, of truth and show, of insight and pretense, of genius (art) and scholasticism. Naturally, a formalist poetic appears to him as a misguided attempt to impose limitations on the creative freedom of the artist, whose proper business, the sublime, must surpass all limitations. Croce, speaking of poetic knowledge, remarks: Ancora ... col definire l'arte come intuizione si nega che essa abbia carattere di CONOSCENZA CONCETTUALE. La conoscenza concettuale, nella sua forma pure che è quella filosofica, è sempre realistica, mirando a stabilire la realtà contro l'irrealtà o ad abbassare l'irrealtà, includendola nella realtà come momento subordinato della realtà stessa. Ma intuizione vuol dire, per l'appunto, indis- tinzione di realtà e irrealtà, l'immagine nel suo valore de mere immagine, la pura idealità dell'immagine; e, col contrapporre la conoscenza intuitiva o sensibile alla concettuale o intelligibile, l'estetica alla noetica, si mira a riven- dicare l'autonomia de questa più semplice ed elementare forma di conoscenza, che è stata paragonata al sogno ... della vita teoretica ... la discriminazione del vero e del falso concerne sempre un'affermazione di realtà, ossia un giudizio, ma non può cadere sulla presentazione de un'immagine o sopra un mero soggetto, che non è doggetto di giudizio, mancando di qualifica o di predicato.3 wrote History of Sanskrit Literature (Calcutta, 1962), hereafter cited as HSL. cf. pp. 28ff.); A. B. Keith, History of Sanskrit Literature (Oxford, 1928) (hereafter cited as SL), pp. 344ff .; L. Renou, L'Inde classique, II (Paris, 1953), pp. 110-111. 3 "Again ... we deny that it has the character of CONCEPTUAL KNOWLEDGE. Conceptual knowledge, in its true form, which is the philosophical, is always realistic, aiming at establishing reality against unreality, or at lowering unreality by including it in reality as a subordinate moment of reality itself. But intuition means, precisely, indistinction of reality and unreality, the image with its value as mere image, the pure ideality of the image; and opposing the intuitive or sensible knowledge to the conceptual or intelligible, the aesthetic to the poetic, it aims at claiming the autonomy of this more simple and elementary form of knowledge, which has been compared to the dream ... of the theore- tic life ... the discrimination of true and false always concerns an affirmation of reality, or a judgment, but it cannot fall under the head of an image or of a pure subject which is not the subject of a judgment, since it is without qualification or predicate." B. Croce, Breviario di Estetica (Bari, 1954), pp. 19-20, trans. by D. Ainslie, "Breviary of Aesthetic", p. 236.

THE PROBLEM The aesthetic mode of criticism situates the work of art in an ineffable; 11 the very word 'aesthetic' betrays the analogical origin of its arguments. By reducing the artistic experience to sensation, in fact to one sense, the sense of sight, and considering the whole experience in terms of categories borrowed from vision-imagination-Croce is led, necessarily to determine art in non-denotative terms, or better, in anti-denotative terms. Not only is the visual experience more direct in relation to its objects, less mediated by a distinctive symbolism ('aesthetic'), but the materials of visual art are not themselves immediately intelligible objects, but rather colors, lines, shapes, which do not so much express as combine to produce impressions and identifications. It is otherwise with verbal art. Its materials are symbols, whose nature is intelligibility. The word itself is a convention. Such art exists at the level of propositions, references, judgements. Attempts to provide it with objects as ineffable as those of visual art overlook the very mode of existence of verbal art, for whatever impression is derived from verbal art is a secondary function of its sustained use of words. The objects it imitates, unlike those of sight, are already conceptualized-events, characters, general types, moral ideas. Verbal art does not have access even to the immediate objects of the sense of sound, which have been in principle pre-empted by music or song. Its very being is determined as a mode of reason: mediacy of symbol, ideality of object. It depends directly on memory and experience even to be understood as art; these are not functions superadded to the artistic experience to provide perspective. Currently the Crocean style of criticism has come to be regarded as outmoded in the West and is now largely restricted to book and enter- tainment reviews. In serious criticism, the Freudians have appropriated the psyche for their own purposes, and the study of poetry has largely returned to the Aristotelian concern with the poem, the ergon-a formalist criticism which in its interest often parallels that of the classical Indian poeticians, and certainly provides a more congenial platform from which to study them than does the Crocean doctrine we have been discussing. That 'aesthetic' view of art is of interest here only because it appears to dominate much of the historicizing on Indian poetics and because, most emphatically, it was not shared by the classical Indian writers themselves who conceived no broader problem than the differ- entiability (viśeșana) of poetry itself as a genre.4 They were exclusively 4 Cf. R. Wellek and A. Warren, Theory of Literature (Harcourt Brace and World

INTRODUCTION means oriented; or rather, they were convinced that any discussion of 12 the ultimate purpose of poetry falls into equivocation and overextension unless it states the purpose within the intentional structure proper to poetry, and presents that aim as the nature of the expressive potency which is the very life of poetry. Austin Warren, discussing the Crocean emphasis on the creative process and its tendency to overvalue discovery at the expense of formulation, says: "The painter sees as a painter- the painting is the clarification and completion of his seeing The pot is a maker of poems; but the matter of his poems is the whole of his percipient life. With the artist, in any medium every impression i shaped by his art; he accumulates no inchoate experience."5 From the very beginning the speciality of poetic speech was, in the Indian tradition of criticism, understood most characteristically as the figures of speech (alamkāra). Traditional Indian scholarship has em- phasized the importance of the figures to the point of giving the name alamkārašāstra to the whole study of poetry. Uneasy at seeing the subject so degraded, historians have generally explained this usage as a historical survival. The first, more primitive and less discriminating writers, it is argued, described the alamkāras exclusively or in the main; and despite later progress and development in poetic insight the name, being old and therefore hallowed, stuck. The figures, in this interpretation, though they give their name to the study of poetry, are in fact its least importam part. They are haphazard lists of purely extrinsic embellishments irrele- vant to the poetic problem.6 The present Introduction will suggest. and the following Glossary in part establish, that such a view as a theory of the figures is false, and as history is perverse and hypothetical. Oddly enough, the important place that the figures have occupied in most Western poetics is overlooked by the historians of Indian poerks though indeed contemporary poetics does not on the whole find much use for the figures .? The writers on Indian poetics of whom we speak appear to wish to reimpose this modern de-emphasis on the Indian subject matter, a severe distortion of the classical tradition which reserves for the study of the figures a place of honor. The alamkaras not only [paperback ed.], 1956), p. 175: "These [sensuous particularity and figuration] are both characteristics, differentiae, of literature, in contrast to scientific discourse." Ibid., p. 74. Cf. W. K. Wimsatt, Jr. and C. Brooks, Literary Criticism, A Short History (New York, 1962), p. 513. De, HSP, II, pp. 32-34; Dasgupta and Dey [De], HSL, p. 517. ? Cf. Wimsatt and Brooks, p. 102, pp. 142-143, 233-234; also Wellek and Warren, pp. 187ff.

THE PROBLEM constitute an original problem of Indian poetics, but are a continuing 13

preoccupation within the tradition. With few exceptions, the figures have been the major problem for every poetician from Bhamaha to Jagannatha.8 Even those writers-especially those of the dhvani school' -- who question the primordinacy of figuration appear obliged to establish their alternative vis-à-vis the figures, so authoritative is that point of view. By its creative persistence, this concern with figuration cannot be dismissed as a primitive survival. Rather it is indicative of a radically different view as to the kind of discipline poetics is.

(b) Poetic as Rhetoric The emphasis on figuration in the early Indian poetic has suggested to De and others an analogy with Western rhetoric: "Bhāmaha attempts to classify poetic expression into fixed rhetorical categories and, from this point of view, his work possesses the general appearance of a technical manual, comprising a collection of definitions with illustrations and empirical canons for the benefit of the artist desirous of externalizing his ideas."10 A disdain for figuration as such leads the same writer to the remarkable conclusion that the most important single exponent of the early alamkāra theory, Rudrața, is not a writer on poetics at all. "Indeed, the practical nature and scope of his work, like that of Udbhata's, leave hardly any room for discussion of general principles or of speculative aspects of the questions involved. Rhetoric rather than Poetics appears to be his principal theme, as it is of most writers of this system who con- cern themselves entirely with the elaboration of rhetorical categories in which they suppose the whole charm of poetry lies."11 Jacobi has warned against considering the alamkārasāstra a rhetoric, pointing out that the subject matter of Indian poetics is consistently determined in the best examples of ornate kāvya, rather than in the arena of debate and public persuasion.12 An insistance on separating the dis- cussion of the categories of figuration from their exemplification raises

8 The oldest and most recent significant representatives of the alamkāra tradition. from the 7th and 17th centuries respectively. Dhvani 'tone' or 'suggestion'; a syncretistic school of criticism (9-10th century) whose reappraisal of the older alamkāra tradition is still authoritative today. We will deal with this important transition more fully later; pp. 78ff. 10 De, HSP, II, p. 47. 11 De, HSP, II, p. 66. 19 H. Jacobi, trans., "Anandavardhana's Dhvanyāloka", Zeitschrift der Deutschen Morgenländischen Gesellschaft, LVI, p. 392, n. 1.

14 INTRODUCTION

absolutely unfounded theories on the bare bones of the contens. figures. But the figures are nothing but their exemplifications: the : that they have logical reference and structure does not ipso facto : them poetic-in this De is right-but neither does it make therr poetic, as De seems to think. The exemplification demonstrates the ; applicability of the form. Poetry, indeed, does not represent a t language at all, a service of language to an extrinsic end. Accordin: one view, and we think this is the view of the early Indian poetie :..- poetry can be taken as the exploitation of language for its own sake :" poetics as the investigation of language insofar as it escapes the immed :. limitations of utility and achieves a condition of self-illumination w .: we might call beauty. General as they are, the figures are not specialized in their applicati Comparison is inherently neither poetic nor rhetorical: A simile. to ui the formal terminology of the Indian poetic, is a statement about t terms (upameya, upamāna 'subject and object of comparison') tendi -: to the explicit point that they share a common property (sādhāranadhar" :. 'tertium') and expressed through use of some adverbial particle indicatit. likeness ('like', 'as'). Customary law, for example, is based upon tr: careful examination of similarities in this sense: A's behavior was ve: much like B's (in another case), which was judged to be culpable (thes establishing a principle of similarity: the law), and therefore A is alse guilty. The rhetorical use of the figures is equally irrelevant to the :: pure form, their expressive entity. In part, the pure formalism of the figures explains the possibility of confusion between poetic and rhetork: the figures are forms of assertion and judgement which are common :e all intelligent discourse. This point, which the later (medieval) Indian. critics of the figurative school often emphasized in order to establish that the essence of poetry (beauty) cannot be shown on this level of formalise: (propositions) alone,13 was not unknown to tradition (where even the terminology of poetics is borrowed from other formalistic disciplines. particularly grammar and logic), nor, in all likelihood, to the figuratiensts themselves. They situate poetics quite consciously next to the discipline of logic and grammar, because logic and grammar provide the background and approach to the poetic problem. By our notion of a rhetoric, the figures are among the persuast devices which further the course of an argument and produce conviction The argument of course is supposed to be true, and as such can get along 13 Dhvanikāra and Anandavardhana, Dhvanyāloka, ed. Durgāprasād (Bombu). 1891), 1.14-15, 2.29ff.

THE PROBLEM 15 without additional embellishments; but this is not the rhetorical point: conviction and persuasion are functions of opinion; they are not the same thing as the truth, and, as ends superadded to the truth and often conflicting with it, require different and additional means-viz., the figures. Moreover, the figures tend to be selected and defined on the basis of their ability to further conviction, and since this latter is often con- ceived in contradistinction to the truth, the figures seem to pullulate in precisely those areas where language is a tour de force, where language becomes so grammatically striking that its Pracht is substituted by the audience for the content of the assertion. But there is some doubt whether this framework can be applied in dealing with the figures as poetry. Poetry, although it also conveys a truth, has no end beyond the entire comprehension of its own sense; the figures, far from being extrinsic, constitute the very form of its ex- pression and are the very means through which poetry is distinguished and becomes voll Gesinnung. From the figures of speech derives that special charm which is the innermost mark of the poetic apprehension. Furthermore, the non-existence of Indian rhetoric as a discipline vis-à-vis poetic reinforces our view that the figures were conceived as essential in expression: the crucial subject matter of poetics. Alone among contemporary Indian critics, V. Rāghavan of Madras appreciates the formalist view: "So poetry requires not only fact and feeling but a beautiful form also; it has not only to be useful, but primarily attractive". And "poetry is not mere thought ... It will be easier to dissociate love from its physical aspect than to keep the concept of poetry aloof from its form."14 This "form", in which the notion of poetic beauty is resumed, is the subject of study of the ālamkārikas15 in the widest sense. In attempting to do justice to the notion of figuration, Rāghavan states the sense in which the figures, as forms of poetic expression, are a proper subject matter for poetics: "If we try to arrive at a clear definition of poetry with an objective differentia, certainly the definition will revolve round the concept of Alankāra ... the beautiful in poetry, the beautiful form-saundaryam alankārah."16 The form of poetic beauty is not only inseparable but, in the context of poetic language, it defines what we mean

14 V. Räghavan, chap. entitled "The Use and Abuse of Alañkāra in Sanskrit Liter- ature", in his Studies on Some Concepts of Alankāra Sāstra (Adyar Library, 1942), pp. 48, 50. Cf. W. Empson, Seven Types of Ambiguity (Meridian Books, 1960), pp. 268-69 15 Ālamkarikas, those concerned with alamkara; the figurative school. 18 Rāghavan, p. 50; Rāghavan quotes Vāmana, Kāvyālamkārasūtrāņi, 1.1.2.

INTRODUCTION by beauty. The manifest, finished metaphor is an example of the poet 16 possibility of imagination. Metaphor in general defines the limit and poetic imagination; the figures are the structure of that possibility This has been recognized by Jacobi, who says: "Dadurch beansprucht thatsachlich dieser Teil der Poetik eine wohlausgearbeitete Formenlehre des dichterischen Ausdrucks zu sein."17 But the conclusions drawn therefrom by De are of considerably diminished import. for he says of the alamkarikas: "This formal treatment affords their works the general appearance of technical manuals comprising a collection of definitions illustrations, and empirical canons elaborated for the benefit of the aspir- ing poet. Poetry is regarded, more or less, as a mechanical series of verbal devices. A desirable sense must prevail, diversified by means of various tricks of phrasing ... consisting of the so-called poetic figures."18 Though this is true, the notion is taken in so pejorative a sense that the implications of such a poetic are never taken seriously enough to be examined. (c) The Notion of Alamkāra Having situated the poetic problem in a formalistic context, we must now pose the question (following the Indian tradition): What is the specific characteristic of poetic usage, its viseșana? If poetry shares its forms with other rational discourse, there must be a criterion of application, of usage which restricts the form and demonstrates it as poetry. Indian writers, from the beginning, conceived their problem as that of 'poetic differentia', the visesana, which both establishes the poetic genre as a proper subject matter and which, by normative application, distinguishes the finest examples of poetry. Since the forms of the poetic genre are borrowed from logic and grammar, the question of poetic distinction is of course crucial and in fact is reduced to the problem of application: Is there a recognizably poetic exemplification of the standards of utterance The problem of the expressive characteristics peculiar to poetic speech is not an easy one to deal with. The very first writers realized that they were confronted with infinite modes and indefinite usages.1 In fact 17 Jacobi, Introduction to his translation of the Dhvanyāloka, op. cit., p. 392. 18 De, SPSA, p. 28. De continues his devastating, but misplaced, attack on the figura- tionists throughout chapter 2; he is so intent on belittling expressionism that he appears at times to question the need for expression itself. His is a caricature of the "aesthetic" doctrine. Croce, far subtler, insists on the inseparability of expression and intuition, and in effect makes the same point as Räghavan. Breviario, p. 45. 10 Ānandavardhana, in Dhvanyāloka 3.37 (p. 210), is often quoted: "anantā hi

THE PROBLEM 17 the most striking general, and in a sense canonical, difference is precisely that the language of poetry is by nature and design not literal. Accepting this category of expression, poetics appears as the systematization, the regularization of that vast body of usage whose lower forms are labeled current, figurative, idiomatic and which by and large stands in direct opposition to the refined, precise, scientific uses of language whose principle is consistency and univocality. The expressions of poetry are in the broadest sense characterized only by deviating from that literal purpose, by a negative correlation with the universal positive goal of discourse: the truth. (Of course we do not thereby assert that poetry is "false": the "truths" of poetry are the same as those of any other genre; we are discussing here only the question of means, of realization. A poetic statement is a tour de force, its truth realized despite propositions that are, by the standards of science, false.) Nevertheless, the forms constituting intelligent discourse generally are assumed by the poetic, too, whose burden is to redefine their function. 'Similarity' (sādrśya) is one of seven categories admitted in the mīmāmsā logic;20 'comparison' (upamāna) is allowed by nyāya as one of four criteria of apprehension, along with perception, inference, and verbal authority.21 The problem we face is defining the poetic application of this concept. It is not a simple problem like referring the logical form to a certain content for specific exemplification. Poetry has no more content than logic. In the same way as logic, it refers itself to the entire spectrum of human experience and in the same way is concerned primarily with the structure of that whole. Poetry is not a content, but a context; poetic similarity is equally as formal as logical similarity: capable of any exemplification and limited only by a countervailing notion of falsehood.22 Nevertheless, the exemplification has a general quality which it is the business of the poetic to make precise. The content of the figure is important as a final test of validity (hence the form of the alamkāra texts themselves: definition and exemplification).

vāgvikalpāh" ('endless are the modes of utterance'); we return to the issue in more exhaustive fashion below, p. 50ff. * Śālikanātha, Prakaraņapoñcikā (Benares, 1903-1904), p. 110; discussed in Jhā, Pūrva Mimāmsā, pp. 61-62. As a 'padārtha', sādršya for the mīmāmsakas replaces the 'viseşa' of the nyāyavaiśeşika. 11 Pratyakşa, anumiti, śabda; cf. Annambhatța, Tarkasamgraha, section 36ff. (p. 24) 22 ineffective. Understood not as the logically false, of course, but as the inappropriate, the

INTRODUCTION 18 The exemplification does not in any way abridge the validity of the form, but simply shows that it is capable of poetic representation. . Here is situated the solution to the problem of the distinctiveness of poetw expression. Although the figures provide the referential background and the formal structure of the discipline, it is not the figures as such which are poetic, but rather their characteristic relation to that general content, their "misuse". Poetic expression which is "true", nevertheless deviates from the norms and standards of literal expression, which is the most usual (or most obvious) apparatus for defining what is true; poetic expression is false, yet it does not serve the ends of falsehood. This dilemma has been particularly agonizing for moralist interpreters of literature from Plato onwards. Poetry uses the structures of logic (as modes of thought) not for an extrinsic purpose, but simply to explore the limits of their misuse, to see what freedom we may have of them. It is essential that the form, though falsely applied, express the same truth as the literal application would support; "freedom" is not taken in the abso- lute sense of disregarding the structures of consistency and the like.23 It does have the sense, of crucial importance for poetry, of the undecided, the suggestive, the implied, the ironic. When the poet says, "Oh that I were a glove upon that hand, that I might touch that cheek", he uses the forms of language (here a predication having the logical force of identifica- tion) to express the impossible, the nonsensical-for only a madman would consider himself a glove. But the form-logical identification-is immediately understood as poetry; that is to say, the proper intentional sense is grasped despite, and also because of the misuse of forms: "I am far away and wish to approach closely enough to touch her gentle, soft cheek". Poetry is limited only by the conventions of such non-literal usage. In each case it is similarly impossible to separate the form (intentional logic as expressed grammatically) from the poetry; the poetry consists in the misuse of the form, within narrow limits. The study of poetic expression consists in the identification of those logico-grammatical forms of expression which are capable of such liberal or ironic exemplifica- tion. In the above example, even the law of identity has a poetic counter- part. Poetic comparison likewise contains an element of the impossible: # For instance, Ruyyaka, author of the Alamkārasarvasva, discussing the figure virodha ('contradiction') states that a resolution (samādhăna) of the contradiction is necessary; the apparent inconsistency is simply a matter of stating a consistency more forcefully. Ruyyaka, Alamkārasarvasva (in Kāvyamālā, no. 35) (Bombay, 1939), p. 154. Cf. Empson, Seven Types of Ambiguity, pp. 199ff.

THE PROBLEM 19 "The professor winked at me so hard that his face was like a concertina with a hole in it."24 This nevertheless suggests exactly the property which is so striking in the subject. The unlikelihood of the ascription is the test of the simile: literally, the quality named (hole, with its suggested correlate, the pleats of the concertina) is present in the object of compari- son (concertina) only. The mind of the listener immediately grasps its counterpart in the subject. But the suggestion of the property in the subject must not pass beyond the limits of the absurd or irrelevant; there must be a basis for the comparison in the subject too, however farfetched (the open eye or mouth). The criterion of misapplication is clearly stated in the early definitions of simile. "The [expression of] similitude, in terms of a qualitative aspect, between the subject [of comparison] and an object [of comparison] incompatible with it in place, time, activity, or the like, is called simile."25 Rudrața says: "When a single quality, etc. common to both [the things compared], which is realized (siddha) in one way [in the object], is made real otherwise (anyatra) in the subject, that is simile, which is itself threefold."26 The same necessity of misapplication is present in all the other figures (even those based on sound patterns), but it may be, and usually is, taken for granted. Only rarely is the principle of misapplication itself suggested. Bhāmaha's much discussed mention of vakrokti is the first and most significant such proposition.27 The indefiniteness of this criterion of poetry is widely held to demon- strate the futility of an analytic approach to poetry.28 The manner of posing a problem usually defines the kind of solution required. The definition of indefiniteness is impossible only if the problem is conceived in extenso by assuming that the poetics is an attempt to anticipate every conceivable instance of poetic utterance in its infinite variation and intimate adjustment of context, time, and place. Poetics is, in this view, a machine for writing poetry, a kaviśiksā. It is as though we were to belittle logic because it failed to provide us with arguments. In the first place, the poetic will deal only with the significant genera of poetic utterance analytically, though recognizing that in the poetic work many co-occur in a complex unity. The question of what are the genera of indefinite usage is settled in the only way possible: by reference to

24 Joyce Cary, The Horse's Mouth. 25 Bhāmaha, Kāvyālamkāra (Bombay, 1909), 2.30. 26 Rudrața, Kāvyālamkāra (Bombay, 1887), 8.4. 27 Bhāmaha, 2.85; see below, pp. 42ff. 28 See below, pp. 51ff. Cf. Empson, Seven Types of Ambiguity, p. 285.

INTRODUCTION the several standards of definite usage from which the poetic forms deviate 20 wed winch they assume. It is significant that the major categories of poetic analss art provided by the sister disciplines of logic and grammar Partry is nothing but the general possibility of reformulating standards and commanpleces: of course, not randomly. but knowingly, and in such a wa rhat the sense is not lost. but is preserved through a fanciful trial Infeenced I Aristotle's definition of metaphor, most Western inven- lones of the fgures have been based largely on patterns of word usage: morphemx firuranon. Metaphor for Aristotle is indeed the basic figure. for rt n a word usd in other than its literal sense." Syntactical structures htrene ar distorted (for example, when the predicate precedes the wherti m the common figure chiasmos, a second phrase is added show- Ing thn reversal. Many figures are defined by the variations of letters sithn a word: "c'ar", "ne'er","" ... when his golden hayre / In th'Ocean talloees be hath bathed fayre ... ".ª In the Indian poetic, an important plaor resarved for grammatical categories; but even when allowances er mede for the variations in grammatical standard between Sanskrit and Groet or English, a certain difference of emphasis is plain. There are almot no figures which, properly speaking, are syntactically defined (d vaharethya, krama); only one figure (in one author) is defined a a sord mage (Vamana's vakrokti, which resembles Aristotle's meta- phori Patteres of letters are discussed only in reference to rhyming and m certaun lnds of puns. Indeed, the Indian figures were conceived not at all from the angle of the word and its conventional usage, but rather vert oneoted to the proposition, were fundamentally logical in concep- uon " Thas diference of approach is crucial. It not only belies De's reduc- boa of all frgures to "tricks of phrasing", confusing poetic with a Western mhetonc, but it states boldly the sense in which the system of alamkaras reprocnts an inventory of the poetic imagination-and defines poetic Merance ia its concrete universality.

Armode, Peetics, 1457b, 7f. A modern view, very similar, in Wellek and Warren, Theery of Lherature, pp. 183ff. . The Tedor retoriciaes made figures out of the largely grammatical Greek principles of onphony : cyncope, aphaczesis, crasis, metathesis, and the like. In English, the dimin- iated capecity for such clisions made their often artificial use that much more striking. On the grammetical uederpinning of the poetic, see below, p. 22, pp. 64ff. For the Tudor fgares, see W. Taylor's "Tudor Figures of Rhetoric", unpublished dissertation (Oicaeo, 1937). = Spenner, Prochalansion # Even the figures which are caplicitly grammatical in their reference define patterns, sepetitiom, of onage; nee below, pp. 64ff.

THE PROBLEM But the misapplication of the idea, poetically speaking, is only half 21 the figure; the form itself, logical or grammatical in origin, provides the principle of definition, and it is of course on this differentiable aspect of figuration that the treatises concentrate. Yet the manner in which they are composed plays down the underlying logical or grammatical frame- work, and gives some plausibility to the view that the figures are mere collections or ornaments empirically discovered. The view is evidently related to that which asserts the irrelevancy of the figures to poetics generally Among the early writers, only Rudrata attempts to categorize the figures according to their principles of definition; his treatment, though ingenious, is broad and mentions only four groupings for some sixty figures. Not until the close of the early figurative period, in the works of certain of the encyclopedists, is there any really serious attempt to make the outward presentation of the figures conform to their inner logic. Ruyyaka must be mentioned as the writer who has gone farthest in this direction. Much of our argument is based on his system.33 But the out- ward arrangement of the figures, though it may technically be that of a list, cannot hide the intentional structure which underlies the definition of each figure and, in fact, relates each figure to others in terms of significant variations, implying, if not stating, a universe of figures, a system (at least) of definition. The subsequent history of alamkāra in India is emphatically not a question of groping after aesthetic universals, but of what requirements this implicit system of definition imposes upon the universe of poetic discourse: What relations of subordination and superordination are to be accepted; how one figure can participate in the idea of another (as śleșa in simile); what constitutes a variety of another figure and hence can be dropped or subordinated; whether, in fact, the form of the figure mani- fests any inherent adaptability to misapplication (vakrokti); what further suggestion is based on the misapplication (dhvani); whether, indeed, the

33 Manifesting an insight to me mysterious, Keith concludes: "The division [of figures] even in Ruyyaka is not logical" (Keith, SL, p. 399). The Alamkārasarvasva of Ruyyaka has been translated into German by Jacobi (Zeitschrift der Deutschen Morgenländischen Gesellschaft, LXII, pp. 289-336, 411-458, 597-628) and provided with illuminating notes. Jacobi's longer article, somewhat misleadingly entitled "Ueber Begriff und Wesen der poetischen Figuren in der Indischen Poetik" (= Kgl. Ges. d. Wiss. Göttingen, Nachrichten. Phil-Hist. Klasse, 1908), is primarily an exposition of Ruyyaka's syncretistic views. It is nevertheless one of the rare serious examinations of the ālamkārika position as a poetic. The respect which Jacobi pays to Ruyyaka ought to be generalized to the entire school.

INTRODUCTION notion of poetic deviation is uniquely or even adequately rcalized in th 22 universe of discourse called figuration (whether it may not be bett poetry, as dhvaniy; realized in some other non-figurative function of what the principle of definition is, and whether it suffices for poetry." not, how it relates to other poetic principles. (II) HISTORY OF THE SEARCH FOR SYSTEM (a) Arthaśabda The system of the figures and the manner in which they express or fail to express the inexpressible are topics which condition the entire history of Indian poetics for the antifigurationists as well as the figurationists We have stated that the figure as form must have a poetic application must be employed so as to convey its intent despite some elemental disregard of the sense and limits of the figure itself. But the notion of figuration in its most general form (vakrokti) is not an adequate guide either to the universe of figures or, in fact, to the manner of poetry itself; for the inexpressible aspect, the element of deviation, varies considerably in character from one major category of figuration to another and is itself partly a function of the logical and grammatical genera implicit in the figures. This is why the Indian poeticians proved so uninterested in the universals of poetry-its essence, spirit, and the like-committed as they were to the notion that their propositions were directed to a discriminable subject matter and had always to make explicit the actual variation of intent and understanding in the poetic language. The time-worn division of figures into artha ('meaning') and sabda ('word or grammatical form') is the first and most obvious attempt to characterize deviation concretely. Certain figures, as we have said. are basically misapplications of a logical or propositional form. For example. simile (A is like B) and hyperbole, which is the poetic variety of predics- tion itself (A is B, or A has B), are understood despite the fact that the statement cannot be true: "They have yarns / of a skyscraper so tall they had to put hinges / on the two top stories so to let the moon go by"." viz. "the skyscrapers are tall, very tall". Other figures involve no aspeet of intention at all, but merely reflect variations in the structure of the language itself and of its grammar (conceived, of course, on many levels- phonology, morphology, syntax). Perhaps the most elemental such figure 34 Carl Sandburg.

is alliteration, which is an arrangement of the phonology of the language, 23 "poetic" because the regularities of recurrence and the limitations on the occurrence of certain phonemes (or aspects of phonemes) is not charac- teristic of usual speech, which is far more randomly organized. Most other figures grouped as sabda are similarly regularizations of grammatical features which ordinarily occur ad libitum. Yamaka ('cadence'), certainly the most maligned of all the figures, is basically the Indian correspondent to our rhyme: the repetition of a sequence of syllables at predetermined positions in a metrical pattern, but not restricted to the end of lines as in most Western poetry. Meter itself, though traditionally the subject of another discipline and not treated as poetic, is still capable of the same poetic regularization. Much Indian scientific and religious literature is composed in simple meters, often varieties of the epic śloka; poetic meters in contrast are distinguished by their complexity, and by the absolute regularity of their syllabic quantification (ārya meters apart, but these are non-Sans- kritic in origin and even Jayadeva "poeticizes" them in these terms). Instead of eight-syllable feet with only three syllables fixed as to length (the śloka), poetic meters range in length from eleven to twenty-seven syllables (averaging around twenty) with no variation permitted in the quantification of individual syllables. Kālidāsa's meter, the mandākrāntā, used so effectively in the Meghadūta, has quarters of seventeen syllables arranged in the invariable sequence:

LLLL SSSSSL LSLLSLL35

The authors of the Dhvanyaloka make much of the difference in poetic quality of the figures of artha and sabda, which they rechristened alamkāra and citra kāvya, respectively; clearly, the deviation in the two cases has different force, because the norms in the two cases are fundamentally different. Poetry based on an intentional structure appeals to the intellect and the understanding-the vakrokti has the aspect of true suggestion, of meanings not said as such-whereas the poetry whose deviations are grammatically based is addressed more to the ear and to those ineffable harmonies which may stir the soul but whose sense is difficult to com- prehend. The vakrokti in citrakāvya is not so much a conveying of mean- ing as the imposition of modes of repetition (forms) on what is in principle an inchoate, unstructured, and fundamentally unintentional level of

5 As kaś cit kāntāvirahaguruņā svādhikārāt pramattah (1°).

INTRODUCTION 24 expression-a kind of elemental ordering or creation in the stuff of lan- guage. Despite the wealth of figures in the Indian poetic and despite the ex- actitude and breadth of their exemplification in the treatises, modern critics have not shown much interest in the universe implied by the figures, preferring to extrapolate theories of poetic beauty from the rare and sketchy references to terms like vakrokti, rīti, and śobhā.36 If they are concerned with the figures at all, it is philologically, as an exercise in the history of a text tradition.37 Critics concerned with intellectual history (Ingalls, for example) often adopt the point of view of the dhvani theorists and are puzzled by the persistent emphasis on figuration.38 Verbal beauty is a consideration proper to poetics, but the initial task of the discipline is to describe the expressive apparatus by which this poetic comprehension is achieved. To begin with, the poetic charm of language has to be taken for granted; it is not an object of investigation, but a criterion of identification. The question necessarily posed by the first figurationists was "how", not "what".39 Of course, in time, when the formal apparatus had been more or less successfully delineated, critics, under the impetus of newer poetic genres, began to speculate on the problem of beauty itself and to seek a single principle which underlies poetic language. We will take up this important transition in its place. It would not be accurate to say that the ālamkārikas were insensitive to beauty, but rather that their task was to define a concrete context in which the discussion of beauty would have meaning. Their view of beauty was that it was best revealed in its own structure (much as Euclid must have thought that the intellectual delight of formal contemplation was best served not in pure subjectivity but in a system of postulates and theorems). 38 Typical is Keith, who, with Jovian disregard, adds: "On the classification of figures of speech no serious thought appears to have been expended" (SL, p. 398). The Indian texts, on the contrary, are almost exclusively devoted to questions of concrete definition: the number of figures and related poetic categories. 87 P. V. Kane's History of Sanskrit Poetics. The figures, or rather the definitions of the figures, are meticulously examined for the light they throw on the chronology of the texts. 88 Cf.D. H. H. Ingalls, "Sanskrit Poetry and Sanskrit Poetics", a part of the Introduc- tion (pp. 2-29) to his translation of the Subhāşitaratnakosa (Harvard Oriental Series, 44). De has called for, but no one has yet provided, a study of the "development of the different conceptions of individual poetic figures ... " (HSP, II, p. 70). This is indeed a desideratum, and one which might well precede speculation on the nature and aims of the alamkāra criticism. Contra De, SPSA, p. 2. Even historically, poetry is the first form of exnression according to the folklorists.

HISTORY OF THE SEARCH FOR SYSTEM (b) The Criteria of Differentiation 25

In early discussions, the structure of figuration was overshadowed by a number of problems. The first was the elaboration, definition and collec- tion of the subject matter. Bhamaha, the earliest writer on the figures whose text has survived,40 was still primarily concerned with blocking out the main elements of the figurative universe. His text is a "collection" par excellence; the figures are arranged in arbitrary groups (which may have had a basis in different text traditions) and are almost entirely devoid of subvarieties. Bhamaha and his follower Udbhata appear to this extent somewhat primitive, but it is not possible to treat the entire tradition as though it had not progressed beyond this stage. It is precisely in the direction of greater order, greater systematization and greater understanding of the implications for the system of the categories and the definitions of figuration, that the tradition has evolved. The great advance of Dandin over Bhämaha is in his arrangement, whenever pos- sible, of the figures according to the canon of subordination. Nevertheless, Bhämaha and Dandin use and implicitly recognize much of the definitional apparatus which is not explicitly stated until later with Rudrata and Ruyyaka. Dandin does not even mention the distinction between artha and śabda, though he groups his figures in such a way as to suggest that he is aware of its importance.41 Bhāmaha's definition of kāvya, "śabdārthau sahitau kāvyam",42 which is discussed most frequently as if it meant 'kavya is word and sense joined' -a rather self-evident proposition at best-seems from the context clearly to recognize the distinction not between meaning and grammatical form, but between types of figures.43 Dandin and Bhämaha are both 40 The relative priority of Bhämaha and Dandin is still a point highly disputed. The passages upon which the intricate textual argument is based are given in P. V. Kane, History of Sanskrit Poetics, pp. 102-133. To us, the textual argument is incon- clusive; we accept the priority of Bhamaha on the premise that he is less concerned with system than Dandin. For opposing arguments, cf. A. B. Keith, SL, pp. 375ff. and De, HSP, I, p. 62. 41 The arthālamkāra in Kāvyādarša, chap. 2; the rest in chaps. 1 and 3. 42 Kāvyālamkāra, 1.16. 43 The preceding line from Bhāmaha (1.15) is translated by De (HSP, II, p. 38): "We, however, accept two kinds of ornaments, referring respectively to word and sense". No trace of the import of this distinction is apparent in De's discussion of 1.16: "śabdārthau sahitau kāvyam" (in SPSA, pp. 18-20), where he seems to take it in the truistic sense of the literal words themselves: "But mere sāhitya of sabda and artha is not poetry; it is grammatical fact, common to all speech " D. T. Tatacharya, expressing the transition from 1.15 to 1.16 (in his learned modern commentary to Bhāmaha, the Udyānavrtti) says "nanu śarīram tāvat prathamam kāvyasyābhidhiyatām yasya bhavatā dvividho'lamkāra ișyate / ucyate ... " But Bhāmaha is probably not discussing the sarira ('body') of poetry at all; cf. below, pp. 29ff.

INTRODUCTION aware of the fully articulated formal analysis of the simile into upameya. mamana sadharanadharma. and dyotaka." and use these categories in distinguishing various similes as well as figures based on simile. (c) An Alternative System: The Stylistic Argument The second problem which hindered the elaboration of formal figurative categories was the lingering discussion of the poetic styles (rīti or mārga), which were originally geographical variations5 in the use or avoidance of certain conventional aspects of Sanskrit syntax, such as long com- pounds. involved etymological forms, certain kinds of alliteration, and the like." Soon what was regional style became merely stylized usage: riti. At first two, then three, then five styles were differentiated on the basis of certain typical conjunctions of the syntactic and grammatical characteris- tics, or gunas ('qualities") relating to usage. The analysis of poetry as riti was fundamentally an empirical enterprise and, as such, not really congenial to the Indian thinkers. It was based on the notion that the occasional use of a particular feature was of greater significance than the idea of the feature or its essential character. Poetry as style was defined as a conventional configuration of certain positively or negatively present characteristics; for example, the rīti called vaidarbhī was defined by the presence of the guņas (sāukumārya, prasādatvam, etc.) and the absence of ojas.47 That the characteristic is present or absent is the essential thing, not what it is or what natural purpose it serves. In a sense the guna|rīti theory completes the alamkāra theory by providing a concrete notion of poetic context lacking in the study of figures as theoretic forms. But it is certainly not the case, as some have supposed, that it was intended to supersede or replace the alamkāra theory.48 # The subject and object of comparison, the common or comparable property, the adverbial indicator of comparison ('like'). Raghavan, Studies on Some Concepts of the Alankāra Sāstra, p. 131. * For complete list, see below, p. 32, note 68. 47 There is indeed a general correlation of the gunas with one of the styles, and their opposites with the other (Dandin, Kāvyādarsa, 1.42), but several exceptions, notably ojas (1.80) demonstrate the variability of the gunas vis-à-vis the styles. In some cases the opposite of the guna is (apparently) also a guna; in others, a defect (dosa). See below, p. 33. Dandin in all cases illustrates both the guna and its opposite, and these illustrations may be taken as instances of the stylistic opposition thus made concrete. # "The decline of the alamkāra system was probably synchronous with, and perhaps hastened by, the rise of the rival riti-doctrine" (De, HSP, II, p. 66). What decline? The alamkāra continues to preoccupy long after the riti has been fossilized as allitera- tion by Rudrata. Attempts to elevate the riti into a competing poetic doctrine or school are made tenuous by the certain identification of but a single writer (Vamana) with it.

HISTORY OF THE SEARCH FOR SYSTEM We have indicated the historical background of the guna|riti theory 27

and its empirical preoccupation with differentiating various total styles by calculating grammatical variables. Vamana was the first to have attempted to state the crucial relation between the gunas and the alam- karas, but in so doing turns the matter on its head and makes of the gunas not the predicates of a definition, but 'qualities', that is, 'virtues' constituting poetry itself.49 Taking Dandin's definition of alamkāra- "kāvyaśobhākarān dharmān alankārān pracakșate"50 Vāmana turns it into a definition of guņa-"kāvyašobhāyāh kartāro dharmā guņāh."51 To this conception of the gunas, the alamkāras are subordinated as specific excitants or heighteners of the beauty thus produced: "tadatiśayahetavas tv alamkārāh".52 To establish this distinction, which is crucial to his theo- ry, he appeals to a number of analogies, most importantly that of the soul and the body. Gunas pertain to the soul of poetry and are inherent, like courage, and essential to it; the figures pertain to the body and, like clothes and ornaments, are incidental and optional-at least this is the interpreta- tion of the tradition which is accepted, in general, by De. In fairness it must be admitted that the text is laconic and may not intend the distinc- tion to be made in quite this way: all Vāmana says is "pūrve ... nityāh" ("pūrve guņā nityāh. tair vinā kāvyaśobhānupapatteḥ").53 Elsewhere he mentions the soul in connection with the definition of rīti, implying a distinction which becomes commonplace later.54 The way Vämana intended the analogy of soul/body presents something of a problem. First of all, it is hard to see just how the gunas are any more 'inherent' than the figures in the poetic expression, if they are indeed present or absent as required for the differentiation of Vāmana's three styles. Dandin expresses himself more appositely: The gunas are

Both writers who discuss style also discuss alamkāra in extenso. The riti theory and Vamana are side-issues in the history of Indian poetics. Reasons for De's ahistorical interest in Vāmana will be proposed below. 49 Playing on the same equivocation in Sanskrit guna as in English "quality". 50 "We consider the figures properties producing beauty in poetry", Kāvyādarśa, 2.1. 51 kārasūtrāņi, 3.1.1. "The qualities are properties which are productive of poetic beauty", Kāvyālam-

59 "The figures, on the other hand, are causes of their pre-eminence", ibid., 3.1.2. 53 "The former are inherent" (the former, viz., the qualities, are inherent; for without them there is no realization of poetic beauty), ibid., 3.1.3 and vrtti. 54 Ibid., 1.2.6, "rītir ātmā kāvyasya", 'style is the soul [ātman] of poetry', and the com- mentary specifies the implication that poetry is thereby analogically the 'body' (sarira) whose soul is the riti. Cf. the definition of the dhvani in Dhvanyāloka 1.1: "dhvanir ātmā kāvyasya". The mention of the 'soul' of poetry is given a different interpretation by De, below, p. 29.

INTRODUCTION essential to the style, not to poetry as such.55 Furthermore, it is hard tr 28 and understand what real difference can be found between 'beauty' "heightening of beauty', particularly if the distinction between 'bodily ornaments' and 'spiritual qualities' (courage, etc.) is analogically intended In the poem, the beauty of the gunas, being essential and inherent, should not require heightening; if it does, then it is not in itself adequate and should not be described as nitya, primordinate. That two entirely different kinds of beauty are meant seems improbable on the basis of Vamana's first two sūtras: "kāvyam grāhyam alamkārāt" and "saundaryam alam- kärah".58 He generalizes the notion of 'ornament' (anticipating the dis- tinction between gunas and alamkāras) to mean beauty itself-a term broad enough to include the gunas, too. As an aspect of his general theory of subordination, Vamana was also the first to try to generalize about the figures of speech. In Vamana we find not only attempts to order the aspects of poetry, but, within the one aspect of figuration, to suggest an architectonic. In fact, Vämana attempts to reduce all the arthālamkāras to varieties of simile,57 a project disas- trously one sided, but showing a clear awareness of how basic comparison is to a large part of the figurative universe. Difficulties of interpretation vitiate Vamana's ambitious program of ordering poetry and in part explain why the guna/rīti theory was simply abandoned by later writers or reduced to a place in the discussion of alliteration (Mammata).58 Nevertheless, Vāmana must be accorded an important role in the history of poetics, for he was the first to explore explicitly the problem of arrangement, pattern, and relations of subordin- ation in the different aspects of poetry. His effort to show that all arthālamkāras were types of simile was at least the first attempt to demon- strate coherence in the figures. It is widely held that Vämana represents a step forward in the develop- ment of Indian poetic thought because he considered for the first time 55 It is possible to see Vämana's preference for one of the styles, the vaidarbhi, as an anticipation of this objection. According to Kāvyālamkārasūtrūni, 1.2.14 the vaidurthi demonstrates all the gunas, the other two, pancali and gaudiy, only some, and this is taken to show the relative excellence of the vaidarbhi. But if the collocation of qualities suffices to define the best poetry, then the vaidarbhi is, in a word, poetry, the others, defective poetry. The equivocation in the word "quality" is drawn out to its limit: on this view Vämana is no longer discussing style, but the normative conditions of composition. For reasons we have indicated most of Vamana's colleagues found it more congenial to discuss this problem in terms of the alamkara theory. 50 'Poetry is perceptible through ornament' and 'Ornament is beauty'. 57 Kāvyālamkārasūtrāņi, 4.2.1 ff. 58 Kāvyaprakāša, chap. 8.

HISTORY OF THE SEARCH FOR SYSTEM the arman or 'soul' of poetry. "The enquiry as to what is the 'soul' or 29

essence of poetry is for the first time definitely posed and systematically worked out by Vāmana ."59 This seems a rather heavy load to put upon what was probably a loose analogy. It is at least clear from Vamana's extremely short definition, "rītir ātmā kāvyasya", that ātman is the predi- cate, the viseşana, and rīti the subject and viseșya. De's statement is, on grammatical grounds, incorrect, for he apparently takes ätman to be the subject of investigation, as if it were said: "the soul of poetry is rīti ... " A small difference, but a crucial one, for we are in fact establishing Vāmana's subject matter: Vāmana is not discussing the soul of poetry at all.60 By this slight reinterpretation of Vāmana's intent, De is able to put forward what is in fact a remarkable theory of historical evolution, according to which Vamana's predecessors discussed only the 'body' of poetry-that is, the figures and other formalistic categories: " ... earlier authors like Bhamaha and Dandin propose to confine themselves chiefly to what they call the kāvyasarira or the 'body of poetry', as distinguished from its atman, its 'soul' or animating principle."61 This is demonstrably and textually false and shows how easily analogies of 'soul' and 'body' can get out of hand and eventually replace in discussion the subjects they were originally meant merely to explicate. There are two references to the 'body' of poetry in Dandin and Bhāmaha, and both are embarrassing for De's historical assumptions. Bhāmaha 1.23 concerns a discussion of the hero (nāyaka). In 1.22 the inappropriateness of portraying this hero's death has been asserted and a reason follows: 'But if it is not the intention [of the poet] to make him coextensive with the body of the poem, and not to show him participating in prosperity, then it is pointless to mention him in praise at the beginning' (1.23). Whatever "kāvyasarīra" may mean in such a context, it clearly has nothing to do with the opposi- tion of soul and body. It appears to signify the poem as a work, as a composition having a certain scope and bulk, and to imply that the hero must be present throughout the work in order to be a hero; he cannot be killed off before the end. Dandin likewise refers to a "sarīra" (1.10) in discussing the origins of poetry: "The inspired sages of old, aware that poetry itself was not enough, and intending the mental formation of their descendants, composed treatises in the form of precepts, relating 59 De, HSP, II, p. 90. Cf. also pp. 36 and 103, and SPSA, pp. 30-31. 60 The use of ätman in the sense of 'chief topic' or 'basic principle' is well attested in the śāstraic literature, cf. Vācaspatimiśra's commentary on Sāmkhyakārikā 12 (topic 99) (= Poona Oriental Series, 10). As here, Vāmana's definition could be expressed "kāvyam rityātmakam". His own gloss is "ritir nāmeyam ātmā kavyasya". HSP, II, p. 34.

INTRODUCTION to speech and its various styles" (1.9). In other words, they established the 30 poetics at the same time as the poetry. He continues: "taih sariram a kāvyānām alamkāras ca darsitah" ('the sages propounded both the body and the figures of poems'). "sarīram tāvad iştārthavyavacchinnā padāval ('now, by "body" I mean a string of words distinguished by a desired meaning'). Clearly Dandin is distinguishing between the poetic 'body and the figures of speech. In the succeeding karikas, he treats of that body (1.11-39), discussing such matters as meter, language, and genres of composition (epic poem, drama, etc.) and their typical content, making clear that he intends by that term, reference to the extrinsic, descriptive categories of poetry (which, as language, poetry may share with other kinds of expression), categories of importance but not relating to the proper expressive power of poetry, its viseşana. The alamkāras indeed are precisely that topic62 and signify, if the contrast with śarīra is to be forced at all, the 'soul' of the discussion. I point this out to show that not only is it false to say that Dandin is preoccupied exclusively with the body of poetry (on the level of textual interpretation), but that, if his own use of the word is to be given any credence, he apparently wishes to distinguish from the kāvyaśarīra precisely that category of discussion (the figures) which De claims constituted its essence. I do not think we need to push Daņdin into an espousal of a doctrine of kāvyātman. Even if the alamkāras are distinguished from the body of poetry, there is probably nothing more intended than the distinction between language (those categories, including meter, which pertain to poetry because it is language and does convey meanings) and poetic language (those categories pertaining to the specific capacities of poetry, in one way or another based on a notion of non-literalism). In this sense, śarira would mean something like 'content' or 'corpus'. The easy verbal analogy of 'body' and 'ornament' was too much to resist. Under the influence of the dhvani theory, the gunas were resuscitated, again analogically, and assigned a relation to the rasa, which had become in its turn the 'soul' of poetry ("dhvanir ātmā kāvyasya");63 the gunas are to dhvani (the spirit of poetry) as qualities of character (e.g., courage) are to the human soul.64 This makes considerably more sense than Vāmana's formulation (though it has been confused with it). In a poem whose basic mood is vīra (the heroic), a stylistic quality such as ojas ('vigor', 62 Taken up in Dandin's two remaining chapters. Cf. M. Winternitz, Geschichte der Indischen Literatur (Leipzig, 1920), III, pp. 12-14. 68 Dhvanyāloka, 1.1. 64 Ibid., 2.7 and Comm.

HISTORY OF THE SEARCH FOR SYSTEM 31

use of long compounds) is essential, whereas a simile, let us say, could or could not be used, depending on other considerations. That this is not how the guna|riti theory was originally intended has been adequately shown.65 One of the points this introduction will make is that the dhvani theory should not be used as a basis for interpreting doctrines that pre- ceeded it. Let us put aside for a moment the controversial problem of soul and body and consider the functional relation, implied in the early alamkāra texts, of the gunas to the figures. It is usually held that Dandin, who alone of the figurationists devotes considerable attention to the gunas as a poetic category, conceived the two terms to be fundamentally identical -the gunas being figures put to a specific use, that of distinguishing the two styles of poetry (by being present in one or the other mārga only), the figures as such being common to both styles. This idea rests on a reading of Dandin 2.3 which appears to me to be capable of another interpretation; the topic is "alamkāra": "kāś cin mārgavibhāgārtham uktāḥ prāg apy alamkriyāh / sādhāraņam alamkārajātam adya pradarśyate". De translates: "For the purpose of classifying the mārgas, some alamkāras have been already spoken of (by me in the previous chapter); now are shown those alamkāras which are common (to both the mārgas)."66 Now, are the figures 'already spoken of' as classifying or separating the margas, in fact the gunas? Or are they simply figures which Dandin has incidentally employed in his illustrations of the various gunas (particularly samādhi, which has been defined as the use of metaphorical expressions in 1.93)? The very common figures, alliteration and pun, have, for instance, been illustrated several times in that discussion. Alliteration, in fact, is treated by Dandin only in the section on madhura guna.67 The phrase that Dandin uses ("uktāh prāg api") does not compel us to identify the guņas and the alamkāras; to do so makes for a number of exegetical problems: Dandin treats the two categories quite distinctly, and never confuses the terms. As usual, discussions of the supposed identity of guņa and alamkāra have involved only terminological quibbles and have not included the broader question of the use of the terms as critical categories in relation to poetry. From this angle, it is clear that

One other puzzling aspect of Vämana's theory is shown in this contrast with the dhvani: the gunas are not for Vāmana contextually appropriate or essential at all; in the best poetry, all ten must be present. They are, as he says, "nityāh", 'invariably present'. The notion of a 'virtue', as potentiality manifested in certain circumstances (as courage), is hard to square with such absolutism. HSP, II, p. 83. 87 Kāvyādarša 1.52-61.

Daņdin does not consider guna and alamkāra identical, and never could have. If Dandin's intention were indeed to employ certain figures to differ- entiate the two styles, there would be no need for a new terminological category (guna). But the major argument against identification are the definitions of the gunas themselves, whose context is that of the spoken language. For gunas relate entirely to the stuff of language-to sound or to its capacity to convey impressions68-and never to ideational, logical, or intentional categories, as do the figures. The last half of Dandin 2.3 does not permit a decision between the two interpretations. As understood by De and Tarunavācaspati, the term sādhāranam ('common') in this passage is taken to refer to the two mārgas previously differentiated: "We will now consider the figures which are common to both styles, not peculiar to one or the other". But again, nothing compels this interpretation, in many ways difficult. Sādhāraņa may also be taken as 'universal' ("having the same ādhāraņa 'basis'"), contrasting, according to my interpretation, with the previous occasional and incidental usage of the alamkāras.69 Here the figures are treated in general, that is, without reference to possible contexts of application (the various margas). Thus we would read the passage: "The group of figures will now be demonstrated universally", that is, in principle and in accordance with their basis and form, not occasionally or for purposes of illustrating the styles. In this interpretation, Dandin's theory of the figures does not differ materially from that of Bhamaha, Udbhata, or Rudrata. For these writers,

68 And in this sense, belong to the sarira of poetry. Their inclusion in Dandin's first chapter is thus decidedly appropriate, and may be seen as completing the discussion begun in 1.10. Cf. above, p. 29. The ten gunas are: śleșa, the employment of a heavy proportion of consonants to vowels; samata, the use of consonants with similar phonic properties; mādhurya, alliteration; saukumārya, the avoidance of harsh consonants and clusters; ojas, the predilection for compounding; prasāda, language easily under- stood; arthavyakti, avoidance of double-entendre and other artificial encumbrances on meaning; udāratva, language conveying a wealth of meaning; kānti. language universally agreeable or commonplace, and samadhi, the use of metaphor. It will lv seen that the first five relate to sound, the latter five to meaning. This both explains and refutes Vamana's highly artificial attempt to define each guna in relation both te sound and sense. Cf. Kāvyālamkārasūtrāņi, 3.1.4 and 3.2.1 ff; also Daņdin. Kuvyaters .. 1.41 ff .; Taruņavācaspati on Kāvyādarša, 2.1. 6º A comparable sense attaches to the term sādhārana in the navyanyaya, where it indicates that fallacy whose reason is too general for its application: "there is tin' on the mountain because it is a possible object of knowledge" (instead of the apposite reason: "because there is smoke"). Sādhārana, when referring to a principle. w translate as 'universal': sāmānya. Cf. Annambhatta, Tarkasamgraha, ed. Y. Athute (= Bombay Sanskrit and Prakrit Series, 55), pp. 44-45 (section 53).

HISTORY OF THE SEARCH FOR SYSTEM however, the gunas were not even considered proper to differentiate 33 styles (that is, to permit judgement of the poetic quality of the styles) and so lost their operational significance. Vaidarbha mārga, which, in the total effect of its ten gunas, falls softly on the ear, may, according to Bhämaha (1.34), be too enervated to be poetic; but a poem, even if gaudiya in vigor and complexity, will be properly honored if it is well figured and avoids excesses (1.35). The gunas, categories pertaining to the poetic context, to the manner of poetry and not to its form and essence, are of little value in discriminating poetry. Insofar as the gunas can be made precise and refer to varieties of deviant usage, they will become figures of verbal effect-kinds of alliteration or degrees of compounding (Rudrața 2.3).70 The discussion of the qualities belongs properly to the prolegomena of poetics, in which is treated the history, aims, and rewards of poetry as well as matters pertaining to the content (śarīra) and conditions of its execution.71 The gunas must have originally been grouped (as the very name implies) with the dosas ('defects', 'sins')-expressive factors which run counter to and vitiate the desired poetic effect. The dosas form a topic in most books on poetics, beginning with Bharata and Bhāmaha,72 and their relevance seems clear, if peripheral: despite the peculiarly expressive form of language which defines and constitutes poetry (named vakrokti by Bhāmaha), poetry still remains indissolubly bound to the general forms and properties of language (nouns, verbs, intonations,

70 It is probably in this sense that the laconic and indeed cryptic remark which Dandin appends to his discussion of the alamkāras (2.366, ed., Tatacharya) is to be taken: "yac ca sandhyangavttyangalakșanādy āgamāntare / vyāvarņitam idam ceșțam alan̄kāratayaiva nah" ('And as for that described in the other śāstra [viz. the Nātyaśāstra]-the san- dhyangas, etc .- this we also wish to understand as alamkāra'). Whether this is just a sop to the already well-established dramatic school of criticism or whether he really means it is hard to decide. The remark, in the manner of a closing aside, is not adequately explained; indeed, it is difficult to conceive how the notion of figuration can be extended to cover such concepts as the five necessary moments in the development of a plot, and the like. Bharata, supposed author of the Nātyasāstra, is careful to distinguish alamkāras from lakșanas and other properly dramatic concepts. That Dandin, whose faculty of discrimination is perhaps the most highly developed of any ālamkārika, should wish to annul such distinctions is improbable. The observation may simply be intended in the sense of wishing Bharata well: "We do not wish here to appear to detract from the dramatists' study of their own proper conceptual terminology, which, taking the term in the broad sense of 'ornament', may collectively be called the alamkāras of their discipline." Compare Vāmana's broad dictum "saunda- ryam alamkārah" (1.1.2). 71 And indeed, as we have remarked above, it is in just this context that Dandin discusses the gunas. 72 Bharata, Nātyašāstra, 16.88-95; Bhāmaha, Kāvyālamkāra, 1.47ff; chap. 4.

34 INTRODUCTION syntactical procedures) which, although they do not permit delineation of a distinctive genre, function as necessary conditions of all utterance and a fortiori of poetic utterance. They impose conditions upon the poetic expression which are effective only in extremis: usually when the poem violates some rule of comprehensibility or grammatical intention (so that it no longer means what it was intended to, for example).73 Such are the dosas. It behooves any poet to be aware of the general, as well as the specific, capacities of his medium, just as the sculptor would not use the same tools on granite and wood. The dosas ('defects', with a connotation of 'sin') clearly parallel terminologically the gunas ('qualities', with a connotation of 'virtue') and like them, pertain to the general preconditions of poetic language. Un- like the dosas, however, the gunas are relevant not in omission, but in commission; they appear to characterize language which is functioning properly. In the definitions of Dandin and Vāmana, several of the gunas are said to be literally 'non-dosas'.74 Most of the others imply a corres- ponding defect, as prasāda implies vyutpanna ('language understood with difficulty'). A few, like ojas, appear basically neutral, relative not to "good" or "bad" but simply to variations in effect: certain poetic contexts call for compounding, certain others do not. But if the gunas are basically the avoidance of dosas and not positive qualities pertaining specifically to the poetic in utterance, it does not appear profitable to situate poetry per se in the gunas. Vāmana's attempt to do precisely that met with little approval, as we have seen. This attempt would make no more sense than would, for example, the definition of poetry in terms of the func- tioning of nouns and verbs. Certainly, poetry cannot be thought of without nouns and verbs, and these are essential to it, but they simply do not relate to the level of poetry on which its specific differences are to be found. In one sense the mere avoidance of negative conditions (doșas) does create a positive condition which it is the business of the poet to cultivate; but as such, the gunas are truistic and do not define poetic expression any more than a properly inflected noun can be con- sidered poetry (although, again, no poetry is possible without properly inflected nouns). In substituting these positive conditions of poetry (gunas) for the specitie differentia of poetry (the figures), Vämana may be said to have attempted a tour de force, but the basic universe of discourse of Indian poctics

78 Daņdin, Kāvyādarśa, 1.7 is to be understood in this sense, no doubt. 74 E.g., arthavyakti ('comprehensibility') defined as aneyarthatvam, 'whose meining need not be reasoned about' (Daņdin, Kāvyādarša, 1.73).

HISTORY OF THE SEARCH FOR SYSTEM 35 was against him. The gunas, like the dosas, may help us to distinguish good poetry from bad poetry, but they will never help us to conceive poetry. And it was to this latter end that all Indian criticism was directed. Moreover (and this is a point not to be taken lightly) the guna theory as a poetic is simply not adequate to the poetry of the Indian classical period. An attempt to define two (Dandin) or three (Vämana) or even five styles seems destined from the start to fall short, to mistake the fundamental character of the highly sophisticated and intellectual ex- pressivism of the classical poem. The obvious quality of this latter is to be found in its complex expressionistic structures-not in the impressions it creates on the ear. If Vämana posed for the first time the problem of the internal organiza- tion of the poetic Geisteswelt, it was not his solution which proved sig- nificant for the tradition. Moreover, the complexity and variety of figur- ation (in which the poetic differentia was felt to reside) is better seen in Bhāmaha and Dandin, where figures were defined according to every conceivable expressive peculiarity, not just that of comparison. It remained for Rudrata75 to attempt the first synthesis, maintaining the structural variety of Dandin, yet proposing an explicit system of classifica- tion which provided for more variables than did the monotonic one of Vāmana. Rudrața introduces a four-fold classification of the figures intended to comprehend all the non-grammatical types, or arthālamkāra.

(d) Rudrata: The First Systematist

(i) Simile Rudrata accepts the fundamental character of simile, as do all other Indian writers. Simile is, indeed, the figure par excellence. Reasons for its priority are not hard to find: little or no poetry is possible without extensive use of simile. This is particularly true of the stanzaic poetry of the Indian classical period, where the poetic work was a sequence of static. reposeful images. But theoretically, too, simile is the cornerstone of the figurative world. Indian writers considered the poetic visesana to be a certain manner of usage-of structures of thought and conception (which are fundamental to all reasoned discourse) but usage which. though entirely successful and meaningful, is not understandable in terms of literal conventions. The standard form of reasoned discourse is the proposition: A is B. The most obvious poctic variation on the 75 Kāvyalamkara, chaps. 7-10 (not to be confused with Bhamaha's work of the same name).

36 INTRODUCTION

proposition is the introduction or the suggestion of an irrelevant second subject (an A'), which equally or more emphatically possesses the same property B. "The candles' ... flames looked at me like the eyes of tigers just waking from sleep" (Joyce Cary). The poetic deviation, concretely speaking, on the first level is this forceful irrelevancy (aprākaranikatva), "the eyes of tigers". The sentence is about candles, not about eyes. In fact, in a statement scientifically directed to candles, tigers' eyes would be considered, except for very occasional references, quite inapposite, relevant only in statements regarding jungle hunting without flashlights and the like. Poetic simile adds poetry to statement by creating a second and parallel universe beyond the apparent and immediate one, a universe whose only claim to relevance is appropriateness in view of a predication, or rather, whose claim on the attention of the reader is precisely a func- tion of its being only similar-having the same property-but not posses- sing any of the other characteristics of relevance such as the same cause, context, motive, etc. Simile, then, is in a sense poetry. It is the basic form of poetry-the reasoned use of irrelevancy. Rudrata groups together, without further attempts at explicit sub-classifications (most of which are obvious by definition), those figures which merely add a dimension to a basic simile.76 For example, the force of simile is a comparison:77 the second subject introduces a comparative standard, which ordinarily heightens the percep- tion of the predicate in the real subject. Such comparison is made explicit by the use of certain adverbial particles: 'like', 'as' (English); iva, -vad, yathā (Sanskrit). But if such explicit comparison is suppressed, we have identification (rūpaka: literally, 'characterizer', though often translated 'metaphor', which is misleading):78 not A is like A', qua B; but A is .A'. qua B. It is still a simile (in comprehension), but its expression is con- siderably more forceful, employing forms which suggest an ontological as well as a perceptual indistinguishability. A further variety occurs when A' is not itself mentioned, but is only suggested through predicating to A (the real subject) a quality or mode of behavior appropriate only te A

76 Kāvyālamkāra, chap. 8. 77 Upama. The word is used both for the family of figures based on simile and for the simile itself. See below. 78 "Anena rūpyate | iti rūpakam". One thing is characterized as though it were anether: it assumes the form (rūpa) of another. Predication is essential to the tigure raps .: metaphor in Aristotle's sense of "figurative usage" need involve only a word which bears another meaning in that context. The latter notion is an aspect of severd Indian figures, notably utpreksā, but never of rūpaka: "The flute of morning stiles in noon-/noon the implacable bassoon ... " (e.e. cummings).

HISTORY OF THE SEARCH FOR SYSTEM (utprekşā).79 The simplest language often conceals such metaphors, 37

their conventionality having deprived them of all poetic force-for example, "the darkness falls" (comparing the darkness to a heavy object). With this elemental priority of simile in poetics may be compared the basic function of analogy (upamana) in the philosophical systems, particularly the mīmāmsā. According to Prabhākara, perception ceases to be inchoate as the mind becomes aware of similarities and differences in its content; knowledge, definite perception, is awareness so determined in similitude; similitude is prior even to the possibility of predication.80 Likewise, poetry comes into existence as soon as simile (the peculiar non-literal simile of poetry) adds its unique dimension of irrelevance which determines the literal predicate of otherwise non-poetic assertion.

(ii) Hyperbole According to Rudrata, the proposition A is B may itself be poetic without the addition of a comparable context, provided the proposition itself is knowingly false. As simile describes accurately by contrasting an irrelevant context, so hyperbole81 is an intentionally accurate or under- standable distortion of the proper relation between the predicate and its subject. No third term is assumed. " ... a skyscraper so tall they had to put hinges / on the two top stories so to let the moon go by" (Sandburg). False: the skyscraper does not have hinges, but true: the skyscraper is very, very tall. Under the general rubric of hyperbole, Rudrata groups those assertions which in some way defy the canonical or assumed relation of a predicate or quality to its subject: for instance, the notion that it is the quality which distinguishes the subject. A range of figures explores the quality which is so distinctive that in certain contexts it merges into qualities of the context and so renders the subject indistinguishable: the whiteness of a girl's sarī as she goes to the tryst in the tropical moon- light "hides" her (tadguņa). The relation of cause and effect is not considered separately by Rudrata as it is by later writers (e.g., Ruyyaka); this relationship, different from similitude, still expresses a natural cohesion between two things in terms

70 Lit. a 'disregarding': the object of comparison is "overlooked", is implicit only in the ascription of its mode to the subject: "The yellow fog that rubs its back upon the windowpanes ... " (T. S. Eliot). Emphatically cat-like! 80 Prabhākara's view as discussed in the Prakaranapaňcikā of Šālikanātha, pp. 52-54; cf. Keith, the Karma Mimamsa, p. 23; Gangānātha Jhā, Pūrva Mimāmsa in its Sources (Benares, 1942), p. 96. 81 Atiśaya, or atiśayokti. Like upamā, the word is used both for the genus and for the most characteristic species. Kāvyālamkāra, chap. 9.

INTRODUCTION of a property, namely that of consequence. Unlike similitude, the rela- tion is not reciprocal. When that cause-effect relationship is conceived in an appropriately distorted way, it is considered a variety of atiśayokti 'hyperbole'. The most obvious example is obtained simply by inverting cause and effect (purva);82 another type expresses a result (effect) without its proper cause (vibhāvanā); still another, the properly functioning cause without its usual result (vyāghāta). The varieties are relatively obvious (as well as limited) when the proper genera are understood. The logical framework underlying this family of figures is evident.

(iii) Pun: śleșa Rudrața believed that simile and hyperbole are the two basic criteria of figuration, but by themselves they establish only two of the four possible categories of poetic usage; the remaining two are established by permutation.83 Simile and hyperbole are not always separable; in certain figures, notably pun and its varieties, they are present in intimate union.84 Likewise, admitting the possibility that neither simile nor hyperbole nor pun are present, we obtain a fourth, somewhat ill-defined group of figures which Rudrata calls vāstava 'natural'.85 Rudrata's notion of pun (ślesa), then, is that figure which is essentially hyperbolic and comparative.86 It might be argued that pun is not an arthālamkāra at all, insofar as it rests upon a chance identity or similarity in the phonemic shape of two different morphemic sequences. Rudrata and the other writers, of course, discuss pun under this aspect (chap. 4), wherein the "charm" is never very far from verbal tour de force, and analogous to conundrums, etc. Certainly puns, like "the ranch named Focus: where the sons raise meat" imply little or no comparison, strictly speaking, for the only property held in common by the 'sun' and the 'son' is the identical pronunciation of their respective morphemes. Such similarity does of course have to be recognized as a common property,

82 Examples of all these varieties will be found in the Glossary. 88 Compare the four existential categories of the Samkhya, defined by permutation of the two terms 'producer' (prakrti) and 'produced' (vikrri). Sāmkhyakūrikā. 3. 84 Śleșa, as the figure double-entendre, means 'coalescence' (of the two meanings). Not to be confused, nevertheless, with the guna "slesa" (above, p. 32. note). Rudrata. Kāvyālamkāra, chap. 10. 85 Kāvyālamkāra, chap. 7. 86 In an unpublished paper, "How Does Pun Differ From Simile" (delivered to the AOS, 1962), I have attempted to justify this thoroughly poetic specialization el the lowly play on words. It is too obviously a play on words to suggest a parallel between Rudrața's 'śleşa' and the fourth Nyāya pramāņa: šabda?

HISTORY OF THE SEARCH FOR SYSTEM and several similes have in fact been defined wherein the tertium is nothing 39 but a pun (e.g., upamāsamuccaya). But Sanskrit is much richer in double-entendres than is English. Several factors combine to produce a far greater inventory of homonyms: the wealth of vocabulary, the lack of a thoroughgoing distinction between concrete and abstract applications of a given word, the great variety of contextual variations permissible for each morpheme, the wide range of derivational affixes in use, and the freedom with which descriptive epithets are formed. Further, the relatively free word order of Sanskrit, and the ability to compound stems and thus to leave aside even the grammatical terminations of words, lends even greater opportunity to the facility of punning. Although Dandin treats ślesa in the chapter on arthālamkāras,87 Rudrata first explicitly recognizes the possibility that pun may indeed function on a level beyond that of mere verbal similarity; the two punned senses may be implicitly comparable.88 Rudrata apparently thought that the simultaneity of apprehension of the comparables in the punned simile added an element of hyperbole which could not be expressed in any other way. Our own poetic education is such as to make us wary of any system in which the most perfect poetic category is the pun. But again, our awareness of the pun is conditioned by the impoverished and relatively exact vocabul- ary of modern Western languages; we ought not to extrapolate from a basis of incompetence. Many critics, De among them, have sought to justify the pun: It is true that it demands an intellectual strain disproportionate to the aesthetic pleasure, and becomes tiresome and ineffective in the incredible and incessant torturing of the language found in such lengthy triumphs of misplaced ingenuity as those of Subandhu and Kavirāja; but sparingly and judiciously used, the puns are often delightful in their terse brevity and twofold appropriateness.89

Keith is even more judicious: Moreover, though we may easily find their paronomasias tedious, there is no doubt that they are frequently rightly called models of twofold appropriateness. and the free employment of figures of speech is often superior to the somewhat

87 Kāvyādarśa, 2.310ff. 88 This is especially true when the two punned senses are not limited to single words in the sentence, but, as is the case in the elegant Sanskrit puns, are extended to the entire sentence by several parallel double-entendres: "asāv udayam ārūdhaļ kāntimān raktamandalaḥ | rājā harati lokasya hrdayam 'mrdubhiļ karaih". Daņdin, Kāvyādarša, 2.311. Such facility is simply lacking in English. Dasgupta and Dey [De], HSL, p. 33.

40 INTRODUCTION rhetorical manner which was introduced into Latin poetry by the practice of declamation in the oratorical schools, which Juvenal so forcibly derides.90

'Paronomasia' is footnoted as follows: "English lends itself only to comic effects, but Greek and Latin authors alike use this device with serious efforts at beauty ... " It is questionable whether such justifications enhance our appreciation of the literary role of the pun, or clarify its peculiar place in classical Sanskrit literature. Keith and De are evidently quite ashamed of the pun; their grudging acceptance of it illustrates once again how alien is their critical judgment to their poetic subject matter. Recent literary develop- ments in the West, however, have focussed more attention on the capacity of the pun. It is no longer possible, after Finnegan's Wake, to say that the pun in English lends itself only to comic effects. The critic John Wain, in terms which might apply almost literally to the maligned Subandhu or Kavirājā, says of Joyce:

With Finnegan's Wake Joyce moved the pun in to the center and made it the main instrument of his writing ... He wanted to present human life as an indi- visible simultaneity and to banish the idea of linear time, so that the last sen- tence of Finnegan's Wake comes back to the first, and the language is given an extra dimension to convey that sense of density, that refusal to isolate experi- ences and take them one at a time.91

Commenting on the rational view of time and events which pun con- tradicts, he says further: "A narrative, unilinear view of experience, which underlay the work of both novelist and historian in the 18th and 19th centuries, is an abstraction. It results from standing back and reasoning about experience, sorting out its thick, knotty textures into manageable threads".92 The critical attitude with which one is invested and which seems so devastatingly appropriate and normal, is often in its more fundamental aspects a function of intellectual history. Wain rather romantically attributes the "unilinear view of experience" to the invention of printing and a consequent degradation of the voice and its many- levelled apprehension. "In literature only people from backward oral areas had an [sic] resonance to inject into the language-the Yeatses, the Synges, the Joyces, Faulkners, and Dylan Thomases".93 The language is being remade to serve the purposes of new apprehensions of reality:

90 Keith, SL, p. 351. 91 New Republic, Aug. 7, 1965, p. 20. 92 Ibid. The influence of McLuhan is obvious 93 Ibid., p. 21.

HISTORY OF THE SEARCH FOR SYSTEM multidimensional, simultaneous, at once abstract and concrete like a 41 Picasso painting. The life history of Yuri Zhivago is told in a manner as far removed as possible from the old linear narrative form which progressed from one event to the next against a tidily arranged 'background'. Quite apart from the fact that the story itself lurches from one coincidence to another there is no separable 'foreground' or 'background'. Everything that happens-the death of a man, an idea occur- ring to a philosopher or a line of verse to a poet, a storm, the birth of a child, an outbreak of street fighting, an evening party at which people make speeches- seems to occur on the same level of significance and at the same level of significance and at the same closeness to the camera-eye.94 The notion of time itself is being broken down into a new model which resembles in principle the pun. I do not want to suggest that this apprehension of time, which is really non-time, the irruption of non-history and immediate experience into art, was that which in medieval India favored and selected the pun above all other manners of literary expression. Differences must be admitted. The history of our recent "anti-history" is itself peculiar; for it is a reaction against the extreme rationalistic view of time expressed in the theory of progress in the nineteenth century. Nevertheless, the newer literary expressions, emphasizing inchoate experience, do serve as a more sugges- tive approach to the understanding not only of classical Indian literary modes, but also to the metaphysical and philosophical apprehensions which they imply-their ideals. By all accounts, the notion of linear time was not one of the leitmotifs of Indian civilization. On the contrary, it expressed in myriad ways not the sequential but the total existential implication of each moment: the entire responsibility of the karmic act; the cyclic return of all the worlds to their pre-created state; the eternal transmigration of souls, the de-emphasis of "this moment" and its claim to exclusive metaphysical validity, the doctrine of creation as "play", au fond inexplicable-all these puzzling aspects of Indian intellectual history suggest why the pun is appropriate in Sanskrit literature. The theory of signification which honored the pun was itself a reflection of such a view of reality. If manifestations are themselves simultaneous aspects of a whole, the meanings by which we respond to them will likewise emphasize multi-dimensionality and simultaneity. The theory of imagery we have been discussing is one attempt to deal with language as a total instrument. Still, the realization of the pun in Sanskrit literature is a far cry from its more recent enshrinement in English. For generations,

Ibid.

42 INTRODUCTION

the pun was a comic device. Even now, the language will not support the burden of simultaneous apprehension which writers like Joyce have put upon it: the language itself is deformed, made more like an echo chamber of distorted and malformed words which imperfectly suggest several adjacent ideas. "'Jack the Nipple', said Wolmbs puffing deeply on his wife, 'is not only a vicious murderer but a sex meany of the lowest orgy'." The example is not Joyce, but John Lennon, M.B.E., to whom Mr. Wain is devoting the review quoted. How far from this weak and tawdry language, struggling with the shadow of an idea, is the elegant stability of the Sanskrit ślesa, spinning out its burden of duplicity in neat, precise verse, with never a phoneme out of place! One has the feeling that if Joyce had been able to write in Sanskrit, he could have been himself and Matthew Arnold, too.

(iv) Svabhāvokti The final category of Rudrata's system, vāstava ('natural': literally 'real'; derivative adjective from vastu 'thing', as realis from Latin res), would appear to contradict the idea of figuration itself, which is predicated on the notion of systematic deviation from the norms of real utterance. "Real [is that class of figures] wherein the nature of a thing is described; and this must be pregnant of sense, but not ironical, comparative, hyper- bolic, or punned."95 Rudrața evidently develops his category vāstava within the tradition of the much discussed figure svabhāvokti 'natural description'.96 Bhāmaha, the earliest writer in the figurative tradition proper, is already not quite sure of the credentials of svabhāvokti. Although he gives an example. he feels obliged to add that "some consider svabhāvokti a figure" ; presumably some do not.97 The occasion for his malaise is the obvious opposition in terms between svabhāvokti and vakrokti, which Bhāmaha in another famous passage has declared to be the basic condition of all the figures.9s 95 "vāstavam iti taj jñeyam kriyate vastusvarūpakathanam yat | pușțārtham avipuria"; nirupamam anatiśayam aśleşam |/", Rudrața, Kāvyālamkāra, 7.10. 96 Svabhāva and svarūpa being synonymous. Note the evident relation to the tirst Nyāya pramāņa: pratyakșa. 97 De states the case much too categorically and mistranslates also: "Such svabiw' ... is not acceptable to Bhāmaha who refuses to acknowledge svabhāvokri as a peete figure at all". Kuntaka, Vakroktijīvita, A Treatise on Sanskrit Poeties, ed. De (Caleuta. 1961), p. xx. 98 Kāvyālamkāra, 2.85. "Saişā sarvaiva vakroktir anayārtho vibhāvvate , vatne 'ssa"! kavinā kāryah ko 'lankāro 'nayā vinā", "This latišayokti 'hyperbole'| is notlung but vakrokti; by means of it the sense is displayed. The poet must make an effort in tts regard, for what figure is there which lacks [an element of] it?" But do the pronouns refer to the vakrokti, or to the original subject, atisayokti? Both are feminine.

HISTORY OF THE SEARCH FOR SYSTEM He does, however, in discussing the poetic genre anibaddha (isolated 43

verses not bound together by any continuing theme or story),99 admit the desirability of both vakrokti and svabhāvokti. Dandin, though he shows less hesitation about the figure, which he treats in some detail, does appear to distinguish svabhāvokti more from vakrokti generically.100 But vakrokti represents poetry by synechdoche and thus doubt is cast on the poetic status of "natural description". Kuntaka, drawing the conclusions of this discussion, rules out svabhāvokti as a figure.101 Of modern commentators, V. Räghavan accepts the distinction be- tween the two concepts: "Indeed, there are cases which do not show any determinable and definable deviation, cases which we call 'natural descrip- tion'",102 but he appears to be of the opinion that deviation is a function of word (sabda) alone, "and it is because jāti concerns itself directly with the thing as it is, without any great śabda vaicitrya, that Bhoja counts jāti as an arthalamkāra and that, the first."103 It does not follow that because a figure lacks abdatvam it becomes of necessity an arthā- lamkāra: Rāghavan avoids the issue posed by Bhāmaha. Vakrokti is a function of both word and sense. Svabhāvokti is a problem not because it fails to show the verbal peculiarities of poetic speech, but precisely because it appears not to involve those of sense either: "maviparītam nirupamam anatiśayam aśleşam". De is unable to elucidate the contradiction implied by the figure svabhāvokti. Seeking in every author notions of poetic essence, De, not inconsistently, discovers it in Bhāmaha's and Dandin's vakrokti: "It seems, therefore, that Bhāmaha regards vakrokti not as an alamkāra but as a characteristic mode of expression which underlies all alamkāras and which thus forms an essential element of Poetry itself, whose meaning can be manifested by vakrokti alone."104 Even if this is so, neither Bhã-

" Misread by De (Kuntaka, Vakroktijivita, p. xvii) as referring to all poetry, not simply to anibaddha. Bhāmaha, Kāvyālamkāra, 1.30: "anibaddham punar gāthāśloka- mātrādi, tat punah | yuktam vakrasvabhāvoktyā ... " D. T. Tatacharya comments on this last: "tad anibaddham punah | vakroktyā svabhāvoktyā ca yuktam eva bhavati". 100 Kāvyādarša, 2.362. "Bhinnam dvidhā svabhāvoktir vakrotiś ceti vānmayam", 'the expressive product is divided twofold: svabhavokti and vakrokti'. 101 102 Kuntaka, Vakroktijīvita, 1.11ff. Cf. De, SPSA, p. 24.

p. 92 "History of Svabhavokti in Sanskrit Poetics", Some Problems of Alankāra Śāstra, 103 Ibid., p. 95. Not only Bhoja, but Dandin himself: "jatis cety adya sālamkrtir

svabhāvokti. yathã ... ", 'Jäti is the first figure ... '. N.B. jāti 'genus' is a term also used for the figure

104 Vakroktijivita, ed. De, p. xviii.

44 INTRODUCTION maha's conditional nor Dandin's categorical acceptance of svabhāvokti (to which term vakrokti is several times opposed as the fundamental alternative) is thereby made coherent or comprehensible. De notes that Daņdin (2.8) refers to svabhāvokti as "ādyālamkrtih" ('the first or primary figure'), but does not appear troubled by this non sequitur vis-à-vis his own theory. The nub of the problem is indeed a misunderstanding of the term vakrokti. Bhāmaha means by it not so much an "essence" (if is under- stood by that concept a generative or constituent principle defining the manner of coming into being)105 as a genus-a term which describes in a general way the fundamental characteristic of all modes of poetic dic- tion-their systematic deviation from a literal norm. Utterance which is literal both in intent and form is not properly poetical. But Bhāmaha is nevertheless aware of certain borderline problems-most acutely, those occasioned by language which is deviant (by the above definition) but which in its deviation has already become conventional: idioms, clichés, and the like.108 The figures leśa and sūkșma (2.86) are specifically rejected because the vakrokti involved has been transformed into a calculus of inference whereby from one thing said or done, another thing can be understood. The grammarians refer to such an utterance, a type of signification, as lakșaņā; the word stands as a token secondary marker for the sense intended and has its meaning because the primary sense is in fact unintelligible in that context: "the grandstands are cheering."107 But more appositely, Bhāmaha objects to the use of cause and effect (hetu) as a poetic figure, deeming it overly conventionalized. His own example illustrates a cliché-ridden usage (2.87): "gato'stam arko bhātīndur

105 "The Indian theorists have almost neglected an important part of their task, viz., to find a definition of the nature of the subject of a poem as the product of the poet's mind ... ". De, ed., Kuntaka, Vakroktijīvita, Introduction, p. xix, n. 19. This inter- pretation of Bhämaha derives ultimately from the dhvani theorists, notably Ānanda- vardhana: cf., Dhv. 3.37 and Comm., p. 208, where Bhāmaha 2.85 is cited. 106 The "conventionality" of Sanskrit literature-a topic much discussed: Kvith, SL, p. 343ff. for a typical view. The stereotypy of theme and style of the ornate poetry has struck all modern commentators, and it has usually been opposed, pejoratively, to a more individualistic ideal-presumably more characteristic of Western poetry since the Renaissance. So Keith, ibid., pp. 345-346, and D. D. Kosambi, ed. Subhāșitara- tnakoșa, Introduction (= Harvard Oriental Series, 42), pp. xlv-lxii. The latter states the stylistic opposition in the ludicrous dress of Marxist social categories. The Indian writers of the classical period were no less aware of the stereotype (a point which should be made more often), but of course were far less unanimous in disapproving of it. We note several cases below, and try to show here how the notion of convention itself was understood poetically. 107 Cf. Mammața, Kāvyaprakāśa, śloka 9.

HISTORY OF THE SEARCH FOR SYSTEM 4 yanti vasaya paksinah" ('the sun has set; the moon shines, the birde a gone to rest'); these three statements are "causes" of the knoledg: is evening." 108 Now, most modern critics have said that Bhamaha reject, hety an the other figures because they are too literal, lacking the clemen' vakrokti essential to a poetic figure.109 Yet this contention is plamr ..: contradicted by the examples in the text. Sūksma and lesa are not ex- amples of literal usage by any standard, and hetu, as instanced in the texts, is not either. None of Dandin's examples (he accepts the poetic quality of hetu) satisfy the literal prerequisites of the conclusive cause-effect relation, as defined in the nyãya-the invariable concomitance of the effect with the cause (vyapti), as: "Where there's smoke, there's fire." Dandin's examples,110 as well as the single one of Bhamaha, are thoroughly poetic in the sense that the logical form is misapplied for effect. Minimally. each of the causes illustrated could have other effects, for example. in the earlier quote, the knowledge that the sun had been eclipsed. The words, or even the individual phrases used ("the sun has set") may indeed be literal and exact, but this is beside the point, for the figure is the relation of cause to effect and not the simple statements which analytically express it. Bhamaha objects to hetu not because he fears that nyāya may be enlarging its domain at the expense of poetry, but

108 The same example in Daņdin, Kāvyādarśa, 2.244, with this explanation. See jñā- paka hetu. 109 So De, HSP, II, p. 50. Bhāmaha says (Kāvyālamkāra, 2.86) "hetuś ca sūkșmo leso 'tha nālankāratayā matah | samudāyābhidhānasya vakroktyanabhidhānatah" (Hetu. sūkșma, and leśa are not considered figures, because the expression of the whole (phrase) lacks an expression of vakrokti' i.e., vakrokti is absent in the composite effect. not necessarily in the form). 110 Kāvyādarša, 2.235 ff. 111 The best-known catalogue is in Rājaśekhara, Kāvyamīmāmsā, chaps. 14-16. Gonda, in his painstaking analysis of the simile in Sanskrit literature, deals with this problem in psychological terms: "When a simile wears out and is no longer alive for the poet or author, when it is no longer the only true expression, springing from an inner urge, it develops into a traditional ornament" (Remarks on Similes in Sanskrit Literature, p. 120). This distinction is most unfortunately put, opposing, as it seems to, inspiration and form. In this monograph Gonda has taken for his problem the determination of the contextual variations in the use of similes over a wide range of literary and semi-literary material. The use of the simile in texts which are non- or semi-poetic appears to him more "natural" than that of the poetic texts per se, which is *artificial" (pp.118-119). But this psychological point of view should not be confused with an expressionistic one: what makes the figure conventional is not that it is "a thing to be imposed on the language from a model" (p. 120), but that the model itself has been identified with a specific content, a specitic application. The model is assumed by all figuration, whether natural, "springing from an inner urge", or highly stylized. This, we think, was Bhāmaha's point.

46 INTRODUCTION probably because of the highly stylized and conventional status which relations of concomitance had come to occupy in Sanskrit poetry. We have only to think of the pearls in the hoods of snakes, and the five arrows of the Love God to realize how insipid the exploitation of such conventional associations had become.111 It is often the function of hetu thus to exploit the association, to draw out its conclusions, as shown in these examples from Dandin: "The wind out of the south, touching springs and sandal forests in the southern mountains, is destined to relieve the weary wanderer",112 and, "The forests are sending forth new shoots, the tanks are full of lotuses, the moon is full; but love turns all this to poison in the eyes of the traveller [separated from his beloved]."113 Bhämaha seems to adopt the special point of view that vakrokti in such borderline cases, though formally present, is no longer recognized as such and may therefore be considered nonexistent. The problem is no more acute than that of any poetic idiom which becomes stereotyped and imitative. Dandin is not so willing to dismiss these figures, thinking that Bhamaha's objection is misplaced. It is not the figure which lacks vakrokti but its application; the poet remains free to exploit other associations, which have not yet been stereotyped. Any figure, in the view we are attributing to Dandin, could be denied poetic value on the basis of conventional application (as Shakespeare's sonnet on stereotyped similes descriptive of a woman's form).114 Conventional usage has been a problem from the earliest period of grammatical speculation, but Indian thinkers have generally preferred to consider it a special variety of literal usage and not itself inherently poetical. What it amounts to in this view is the innovation of a term or a form in a specific, but heretofore unattested, use. The desired meaning, which is assumed and fixed, has, as it were, selected another means of expression as precise as the literal one.115 Such is not the case of the usage styled vakrokti, for the meaning and the form are in essentially negative correlation: all that can be said is that the meaning is not conveyed by the

112 Kāvyādarśa, 2.238. 113 Kāvyādarša, 2.242. 114 Sonnet CXXX: "My mistress' eyes are nothing like the sun; / Coral is far more red than her lips' red: / If snow be white, why then her breasts are dun; / If hairs be wires, black wires grow on her head ... ". 115 For example, the cliché "he hit the nail on the head"; its sense is so strictly deter- mined as the substitute for what it means ("his remark was à propos"), that it may be said to have supplanted that more literal expression. Its usage is as strictly regulated as would be the phrase for which it stands; all attributes of the literal expression have been transferred to the cliché, including that of literalism.

LLIC SEARCH FOR SYSTEM close or literal interpretation of the form, and yet it is conveyed; this 47 indeterminateness is the vakrokti. It can be assumed that neither Bhāmaha nor Dandin intended to oppose svabhāvokti to vakrokti so categorically, for to do so would have been to deny poetic status to svabhāvokti, which neither is willing to do. I think the key to the understanding of svabhāvokti lies in our discussion of conventional discourse. Svabhāvokti is not to be taken as synonymous with "literal" or direct discourse, but rather is a cover term for the poetic possibilities implied by conventional language. This would in itself account for Bhämaha's uncertainty regarding it. The point of view might have been as follows: granted that certain poetic usages may by repetition become conventional and lose their claim to be called poetry, nevertheless, what of the inverse case? Cannot the inherent tendency of language to stereotype, in certain cases and under certain conditions, be given a poetical application? Conventionality, as a formal aspect of language, ought also to have a certain, albeit very limited, poetic scope, just as any structure of language may be so exploited.116 The conventional becomes either more or less expressive than its usual function would dictate. The most striking example is the genre of anibaddha poetry called jāti (it is no accident that this is another name for the figure we are discussing), in which a single detached verse encapsulates the characteristic aspect of a certain general type: "The children, bodies grey with dust, are intent at their play; assuming grave miens and voices, choosing one among them to occupy the station of a King."117 The description here is so vivid, the type brought into such striking perspective, that, in the opinion of Dandin (and following him, Rudrata), a special type of vakrokti has to be allowed which is aviparīta (for it is exact in reference), anupama (for there is no secondary reference), anatisaya (for the only exaggeration is that of the graphic quality and not of the image itself), and aślesa (for the words of the expression are not themselves interesting). A type of vakrokti called svabhāvokti is, as it were, vakra only in the secondary sense-that of the manner of its comprehension -- not with reference to its subject or content. It may be opposed on this subordinate level to vakrokti, which by a misapplication of form, distorts the apprehension of its content.118 116 Rhyme, alliteration. 117 Rudrața, Kāvyālamkāra, 7.32. The discussion, it should be remembered, concerns svabhāvokti as an arthālamkāra; the examples will still illustrate many śabdālamkāra, such as metre, alliteration. 118 Dharminder Kumar, in an unpublished thesis of the Panjab University ("A Study of Dandin") has put the matter justly, although from another perspective: "It is

INTRODUCTION 48 What is implicit in Dandin-that svabhavokti is a category of figures which employ conventionality in a sense which can be called vakra-is explicitly stated by Rudrata; those figures which are considered primarily to illustrate that special quality of apprehension are grouped under the term vāstava. Interestingly enough, besides jāti and hetu, Rudrața deals with sūkma and leśa, as well as a host of figures based on syntactic patterns lending emphasis, or exemplifying (without hyperbole) various kinds of suggestion and inference. The ability of the mind to see an unusual relationship is to Rudrata evidence of poetical intent, even when there is no explicit comparison or secondary reference. Rudrata with one stroke encompasses the whole realm of dhvani119 in literal speech. How misleading to consider svabhāvokti and vāstava merely as literal expression and thus to conclude that the writers are dealing with a con- tradiction, viz., literal speech as poetical.

(e) Poetic as an Intellectual Discipline

It is important to recognize that the vakrokti of the early ālamkārikas does not refer to metaphor in the Aristotelian sense. In fact, Vamana is the only writer to accord special mention to such a figure. The remaining writers borrow metaphor from grammar, as an aspect of the theory of signification,120 and go on to situate poetry primarily on the level of propositions. It is not the usage of individual morphemes that makes poetry (as one interpretation of the dictum "śabdārthau sahitau kāvyam" implies). Vakrokti is to be understood as describing the misapplication of modes of thought and judgement. It is only in this way that we can distinguish poetical from logical or scientific comprehension. In the logical upamiti ('comparison'), the formal elements of comparison are present, as is the understanding of a relation between the subject and a similar. The proofs for the existence of God are often similes: as a product requires a maker, so does this world (a product showing intelligence of design) require a maker.121 The form of simile here is not poetic, because incorrect to think that the term vakrokti denotes an element, while the svabhāvokti is merely a figure, for it does not appear to be sound that all the figures excepting svabhävokti were to Dandin the different forms of vakrokti. There are certainly some other figures as well which are devoid of the element of vakrokti and where the element of svabhavokti is conspicuous by its presence" (pp. 212-213). 110 Scil. which does not contribute to figures based on simile or hyperbole, such as samāsokti (q.v.). 120 The concept of laksanā, or secondary denotation, spoken of above. 121 Sāmkhyakārikā 15; Udayanācārya, Nyāyakusumāñjali, 5.1.

HISTORY OF THE SEARCH FOR SYSTEM 49 it is applied within the limits set by its defining conditions: the similitude is, as far as the figure is a proof, exact. The problem of distinguishing the literal from the poetic is not (as De sees it) that of distinguishing the univocal from the equivocal word-failure to observe that the poem may be literal on one level and poetic on another (that of proposition) has vitiated much of De's otherwise perceptive criticism. In fact, the figures of the early ālamkārikas must be conceived as basic conditions to poetic utterance-unavoidable mannerisms of the formulations we call poetic; for the only alternative is to employ modes of thought and judg- ment exactly (at least in intention): the only alternative to poetry is non- poetry. The poet, by virtue of his decision to express himself poetically, adopts the conditions of imagination-the figurative mode. The interdependence of poetry and non-poetry cuts both ways. Just as the principle of similitude is fundamental to certain logical judgements as well as to poetry, so has the doctrine of metaphorical usage been put to decidedly non-poetic tasks: the mīmāmsakas, for example, long before the poeticians began to analyze poetry for its own sake, were using the same basic tools in exegesis of the Veda. The Veda being eternal and authorless obviously could not have had its origin in specific historical events. The names which abound in the Upanisads were explained as referring either to the continuers of the tradition, or, if that were not possible, as concepts, taking the name as its "meaning": pravāhana, 'the carrying onwards', etc.122 Likewise, the statement "Trees sat at the sac- rificial session", which would appear incompatible with the metaphysical assumptions underlying the doctrine of eternality and infinity, is taken figuratively as referring to the importance and universality of the sac- rifice.123 Śankara adopts the mīmāmsaka techniques of metaphorical analysis to render consonant the many Upanisadic passages which do not favor his doctrine of the single, real ätman, that is, those which seem pantheistic in tone.124 These speculations also suggest a historical relation between poetics and ritualistic interpretation. It does not seem far-fetched that the origins of Indian poetics evolved from such considerations. This would be con- sistent both with the Indian traditions themselves (which group the ālam- kārikas with the mīmamsakas and the grammarians) and with the nature of the discipline. The mīmamsaka point of view really differs very little

122 Kumārila, Ślokavārtika on Mimāmsāsūtra, 1.1.27ff. 123 Mimāmsāsūtra 1.1.32 with Šābarabhāșya. Cf. virodhābhāsa. 124 Vedānta sūtra with Śāmkarabhāşya 2.3.16,17; 2.1.14. Cf. nidarśanā.

50 INTRODUCTION from that of the poetician.125 Both are approaching usage from the angle of its expressive technique; both are attempting to understand the tech- nique as it contributes to a fundamental and correct understanding of the text. But the text, though true (indeed, eternal), is shot through with inconsistencies, and expressions which defy reason and immediate comprehension. The analysis of the text will consist in the exhibition of those techniques as forms, with notes as to how they are to be understood (ipso facto, correctly) and how they are to be made to render sense. The mīmāmsakas and the poeticians differ in several respects, nevertheless: in terms of the refinement of the technique of interpretation, in terms of the canonic character of the texts, and in terms of the scope of the discipline. The mīmāmsakas wish to make consonant a small number of passages which conflict with a canonical and literal majority, while the poeticians have separated off a special subject for their unique attention and are interested only by implication in the origins of their study. It is enough that poetry is also true and serves some extrinsic purpose.126

(III) THE SYSTEM OF FIGURES

(a) The Question of Infinitude

The figures with their subdivisions found in the pre-dhvani texts constitute almost all of the figurative inventory. Later writers have added a few new figures (as arthāpatti),127 but the steady growth in figurative elabora- tion (a point made very often) is mainly one of appearance-a function of regrouping, synthesizing divergent accounts, and accepting subdivisions as independent figures. In late texts, figures based on rasa are elaborated and illustrate the complicated categories of medieval logic. Yet the main types and varieties are stated in Dandin, and the fundamental outline of the subject is achieved in Rudrata. Still, critics from Ānandavardhana to the present have attributed theoretical inadequacy to the study of the figures, alleging that poetic utterance is an undefinable subject matter: 125 It should be borne in mind that the mimamsa, although a school of ritual inter- pretation, is very likely the precursor of all the Indian schools of discursive reasoning: what was at first textual exegesis became contextual exegesis. The oldest name of the mimāmsā appears to have been nyāya; the logical framework of the poetie is its most striking feature. See A. B. Keith, Indian Logic and Atomism, pp. 10-11. 126 An ancillary subject treated in the prolegomena of most early poetie texts: ef. Daņdin, Kāvyādarša, 1.1-10; Mammața, Kāvyaprakāśa, śloka 2. 127 See the Appendix for a list of those figures not defined before the medieval nerind (nost_Mammata)

THE SYSTEM OF FIGURES "anantå hi vagvikalpah" 'the variations of speech are unlimited'.129 More 51

recently, De commented: " ... the poetic intuition differs in each poct ... and there are bound, therefore, to be endless kinds of individual and concrete expression which have their own standards and spheres in each case, and which cannot repeat themselves." Yet it is just the task of describing this futility which De sets before the alamkārikas: " ... it [such endless differentiation] appears to have afforded endless scope to the scholastic ingenuity of later theorists who ... finding hardly anything to systematize in respect of the essentials of theory, occupied themselves in elaborating the details."129 In dealing with the so-called infinitude of the figures, two questions are easily confused. One relates to the form, our chief concern in this Introduction, which we contend was the prime object of the studies of the ālamkārikas; the other derives from the application, the concrete manifestation of the form, in other words, the poetry itself. Evidently here, as in logic or any other formal discipline, the exemplification of science is endless, resting as it does on principles which are not pertinent to the science (observation, etc.). The older writers were aware of this problem and carefully restricted their consideration to the form. Dandin, defining the figures as properties conveying the beauty of poetry, says: "te cādyāpi vikalpyante kas tān kārtsnyena vaksyati" 'now they [the figures] are to be discriminated; who will propound them all?' (2.1). Dandin, by not making clear whether he refers to the figures per se or to their exem- plification, appears to offer justification to those who would deny the theoretic urgency of his task: if the figures cannot be described in their entirety, then why describe them at all? Of course, as in many of these short, indeed laconic kārikās, more than one interpretation is possible. To what does "entirety" refer? The intent of the author is often only the reading attributed to Dandin by later commentators, all of whom were writing under the burden of the dhvani theory, which held that the figures were occasional aspects of the poetic work and were therefore arbitrary in principle and definable ad libitum: anantā hi vāgvikalpāh.130 Before attributing this view to the ālamkarikas themselves, it would appear appropriate to investigate possible alternatives. The verb vik!p -. which Dandin uses in the passive of the causative vikalpyante. does indeed mean 'discriminate', 'render into the form of an alternative' (vikalpa "kind'). But it just as strongly conveys the connotation of 'clothing in 128 Dhvanyāloka 3.37 (Commentary), p. 210. 129 De, SPSA, p. 77; HSP, II, p. 73. 130 Cf. De, SPSA, p. 15.

52 INTRODUCTION the stuff of the imagination'. This sense is conveyed by the commentator Vādijaāghāladeva: "yadyāpi kāvyaśobhākarāḥ [sic] dharmā alankārā ityuktam tathāpi naitāvatālankārasādhāraņalakșaņamātrakathanād eva jñātum pāryante / kim tvadyāpi [utprekșya nirūpyante]131 lakşaņodāha- ranapradarśanena katicit" 132 Although he repeats what had by then be- come the standard remark about the anantatva of the figures, he does render clearly the sense of vikalpyante: 'by an act of imagination, a certain number are now to be fashioned with the aid of definitions and examples'. In other words, it is the multiplicity of manifest figures which cannot be described fully.133 The figure as form may be clothed in an infinitude of referential dresses. Of course it is obvious that, qua poetry (qua kāvya), the form alone defines the force of the expression; its referential content is incidental and may be supplied as appropriate. Dandin appears to confirm this intent in the last kārika of chapter 2: 367: panthāḥ sa eșa vivrtah parimāņavrttyā samkşipya vistaram anantam alamkriyāņām vācām atītya vișayam parivartamānān abhyāsa eva vivarītum alam viśeșān

which translates:

The path [of the figures] has been disclosed by the method of circumscription, Summing up the infinite extent of figurative exemplifications;134 When one goes beyond the subject matter of words [i.e., the verbal science, poetics], Practice alone is able to disclose the [existential] differentiations [of individual figures].

Our attempt to rescue Dandin from the charge of anantatvam is rendered problematical by his repeated references to 'varieties' of known figures

131 Restored by the editor, D. T. Tatacharya (or by V. Krsnamācārya?). The gerund utpreksya may also have the connotation 'looking over', 'comparing'. 132 Daņdin, Kāvyādarša (ed. Tatacharya), p. 64. 133 A remark so obvious that when made it would seem to require another inter- pretation; but in Dandin's text it is a passing remark, one which can appear to antic ipate later theoretical discussions. 134 The Sanskrit term is alamkriya, which is taken as synonymous with alamkar 'figure'. But is it accidental that Dandin uses this exceptional term in those two con texts where by our interpretation he is discussing not the figure but the manifest figure The feminine often has the acceptation of concreteness. But contra, the term alamkrip is required by the metre, and it is used once (4.64, Tatacharya ed.) where clearly n distinction is intended. Also "tān" (masc.) in 2.1. Cf. Kāvyādarsa 2.3 and above p. 37

THE SYSTEM OF FIGURES not defined by him, and presumably to be supplied in addition to those 53 defined (2.96; 168; 309; 347). The last three of these references seem clearly directed to the problem of samsrsti-figures appearing as sub- varieties of other figures, several of which have nevertheless been men- tioned in the main body of the treatise, as e.g., śleșa upamā. This also is a problem that does not touch the structure of figuration itself, but rather involves its concrete application in cases where several figures co-occur. Yet it is also true that Dandin is the writer who elaborates sub-classifications to their greatest extent. Often, as in upamā, his criteria for sub-classifications are not formal, but contextual (nindopamā, cațūpamā, etc.); aware of this, Dandin may indeed be unwilling to over- extend himself in subtleties. Still, even on this level, the problem is really one of context and exemplification. Dandin's lead in trying to involve contextual definitions as subvarieties was not followed by his successors, who stayed, by and large, within the limits of pure formalism. If this is all Dandin meant by cautioning (2.1) against "complete" inventories of figuration, then his warning was indeed heeded, and our defense of him corresponds to the reading of Dandin by the tradition itself. The following attempt to sketch a classification of the figures should not be taken as final. It is based on an enumeration of figures judged different by the tradition, up to and including Mammata (but excluding Bhoja and the Agnipurāna).135 A few notes will be appended on the post-dhvani figures, but they will not be dealt with in detail. Although the alamkāras imply a system of classification, that system is capable of several represent- ations, for it is essentially multivalent. A factor selected as basic by one writer, may, to another writer, appear as secondary. Indeed the figures are defined principally in terms of such emphases, and one is always free to "distort" a figure. The primary characteristic of the figurative universe is not its fixity,136 but its selectivity. The figures realize the potentialities implicit in the norms of grammar and logic in no set or predetermined archetectonic (unlike the theorems of Euclid). We have grouped the figures primarily on the basis of ease of definition. Categories considered as genera in a large number of cases are taken as basic, especially when these genera appear to remove the subjacent figures from immediate cross-relevance (vyatireka bears a much more distinctive relationship to rūpaka, let us 185 That is, we make no effort to encompass the Agnipurāna's figures in this sketch of a system; they will be included in the Glossary. 138 Contra De, "fixed rhetorical categories", SPSA, p. 32.

INTRODUCTION 54 say, than it does to suksma, for both rupaka and vyatireka are subsumed under the classification simile). At all points of the system, however cross fertilization is possible. Dīpaka, which begins its career as a śo bdalamkara, more and more falls under the sway of the most characteristic arthālamkara-simile-not because the form is redefined, but because the context of application becomes more and more specialized; the figure is used to imply similitudes. The classification suggested by Ruyyaka (the most extensive and detailed one supplied by the tradition) is based in part on other considerations (on the reference of the figure: division of logical figures into worldly. verbal, and ratiocinative), but often coincides with our own, focussing on the logical basis of similitude (emphasizing the distinctiveness, the identity or the relative difference of the compared terms).137 We will follow Ruyyaka or Rudrata whenever possible, but will often suggest new orderings. The figure is the form and is represented in the definition; it is distin- guished from its exemplification, which, as we have seen, involves other non-poetic factors and is illimitable. Yet one example is necessary for each definition-a characteristic example which will demonstrate in concrete form the point made by the definition. Figurative poetics is not a study of examples, a posteriori: one example per definition is generally sufficient. It is in fact an a priori science, which treats its exemplification more as a justification and, of course, as a means of conveying an idea. The example shows the characteristic misuse of the logical or other form, but also that the form is correctly apprehended despite the misuse. The definition states only the form and the conditions of the misrepresentation involved. When this position is understood, the question of exhaustiveness still remains, though somewhat mitigated: Are all the expressions which are in principle poetic defined here? Since each figure has a place in a nexus of possibilities, it would be difficult to be completely exhaustive; the system is one which contains many more theoretical figures than actual ones. But the principles and criteria of definition are exhaustive; their possible combinations and permutations are difficult to encompass. Still, each new figure can easily be placed in the system of definition. Rarely indeed is there any suggestion of a novel principle of definition.1 187 Ruyyaka, Alamkārasarvasva (Kāvyamālā, 35), pp. 181, 187, 206; p. 31. 138 The figure arthapatti may be one such; a mode of valid knowledge in the mimāmsa philosophy, but rejected by most other systems; in effect, reasoning a fortiori. First defined as a poetic figure by Ruyyaka, Alamkārasarvasva, p. 196.

THE SYSTEM OF FIGURES In part then, the number of figures depends upon usage-but usage in 55 the narrow sense, not of concrete illustration but of characteristic varia- tion. Two examples of simile which differ only in their reference, and not in their modalities of expression, are considered to be the same figure. It is only when the modality can be shown to have been altered that we have a new figure. For example, the late figure vicitra (Ruyyaka, et al.) would appear to be a special type of cause-effect relationship in which the cause is characterized as prayatna 'effort', but it is in effect a special type of vişama (III).

(b) The System

I. SIMILE, comparison (upama). The adjunction of a literally irrelevant object (called the upamana) which possesses in a higher degree a property (samānadharma) also present in the subject (upameya), the comparison thus calling attention forcefully to that distinctive feature by a kind of transfer of emphasis. Figures based on simile, which add a determination to the basic simile but which remain similes in intent, are of two fundamentally different types: (A) Figures which are variations on the form of the simile: (i) an essential element is implicit.

all explicit upameya upamāna particle type of similitude

implicit implicit implicit

upamā aprastuta- samāsokti dharma (property)

praśamsā prativa- anyokti ubhayanyāsa kriyā (action)

stūpamā These figures, where one of the four elements is implicit, should be com- pared with IIIC below, where implicitude itself is a major element, not sub-joined to simile. The later authors, Udbhata, Rudrata and Mammata, use these and similar formal discriminations to subdivide the figure upamā itself. Since our classification deals only with those figures which have been assigned status as major figures by some author, this formal analysis does not appear here, but it may easily be imagined. Much attention has been paid to the grammatical form of the comparative particle and to the manner in which compounding may be employed.

56 INTRODUCTION (ii) an essential element is repeated. viz. the upameya ananvaya A is like A viz. the simile, but upameyopamā A is like B; B like A inverted viz. the upamāna mālā A is like B, C, D, etc. similes themselves upamārūpaka the AB of the CD (see combine to suggest example in Glossary) a larger simile

This last type is a very special case of what is usually treated as a sub- variety of rūpaka: the repeated metaphorical identifications are all aspects of a larger picture (see samastarūpaka and discussion).

(B) Figures which are modes of conceiving the common property. The samānadharmatā of the simile is its functional sine qua non. Although the property (or mode of behavior) is 'common' and must be so in order for the simile to operate, an element of distinctiveness is nevertheless implied, and it is just as important for, without it, no emphatic transfer would accrue to the upameya; the simile would not be poetic, but realistic. This implicit relative distinctiveness may be conceived according to several principles of relation: (i) the differentiability of the common property as a conceptual mode relating the two things compared. (a) emphasis neither on the difference nor on the similarity:

the mode is comparison upamā A is like B recollection smaraņa B is reminiscent of A inversion pūrva B is more important than A

(b) emphasis is on the similarity (lack of difference):

the mode is identification rūpaka A is B illustration drstānta A? B negative illustra- nidarśanā not A? not B tion (II) denial apahnuti not B: A! confusion bhrantimat B is taken for A doubt samdeha A or B? representation139 samāhita B! (addressing A) substitution sāmya A for B

13₽ Repeated, it appears to be the late figure ullekha, detined in Ruyyaka, p. .".

THE SYSTEM OF FIGURES 57 (to which Ruyyaka adds) transformation140 pariņāma B becomes A (c) emphasis is on the difference (lack of similarity): the mode is excellence vyatireka opinion A excells B mata jealousy some say B; I say A pratīpa B is jealous of A revenge pratyanīka B takes revenge on A (ii) the common property is itself limited as to the scope of its own distinctiveness (modally limited).

is verbal only upamāsamuccaya C is a pun is literal only viśeokti (III) C is metaphorical is inessential sāmya (II) C is absolute

(here could also be put)

is a mode of prativastūpamā action only

(iii) a property of the upamana, that is, one not common, is transferred to the subject, thus suggesting not only the identity of the two things. but the common property itself. This is a special case of B(ii) in which the relative distinctiveness and similarity of the two things are character- ized as referentially distinct, and then only those aspects which in fact do distinguish the compared things are employed in the transferral relationship. Formally these figures are non-similes, that is, similes which function by employing non-C instead of C. A typical case is the so-called pathetic fallacy.

transfer of property apahnuti (III) -- with possession stated vidarśanā transfer of mode of action utprekşā This variety could also be called hyperbolic simile from its manner of misrepresenting the existence of the property in the subject. Ruyyaka. rather tortuously it seems to me, tries to exploit this notion to the extent of trying to make hyperbole (atisayokti) itself an inverted uprek,a. and thus a kind of simile.141

140 See vikriyā upamā. But already recognized by Dandin as a sub-variety of upama (Kavyādarša, 2.41).

141 Alamkārasarvasva, p. 83.

58 INTRODUCTION

Since (A) and (B) are distinguished by criteria which are not alternatives vis-à-vis each other, all of the types listed under (A) may be expanded by the entire array of types under (B), just as is the simplest upama. In the Glossary we have often taken advantage of this indeterminacy in our English examples. The more complex varieties given under (B), for example pratyanīka, could rarely illustrate type (A)'s upamā, much more often a prativastūpamā or an ubhayanyāsa. Likewise utprekșā, suggesting identification rather than stating it, can illustrate several varieties of rūpaka (where the form is not in question). II. HYPERBOLE. The reference is not to two things and a common property but to one thing and its property, or its several properties. In other words, hyperbole is a family of figures founded on the notion of predication. Inevitably, as the reference is to more than one property, there will also be, implicitly or otherwise, reference to a second subject (mīlita, tadguņa). To distinguish these cases from the foregoing, we must appeal to the notion of pradhānatva: we must decide where the main intent of the speaker lies-in the comparison or in the relation of properties. There are three major varieties of hyperbole: (A) Simple exaggeration (we include minimization as a special case). The property or an analogous aspect of the subject is portrayed out of all perspective, thus distinguishing the subject in its very being. exaggeration of a property atiśayokti (in the narrow sense) an action utprekşā (IV) an effect utprekşā (V)142 wealth or udātta nobility ego ūrjasvi affection preyas rasa rasavat sense of a uttara (II) (probably to be considered a type question of conundrum) (B) The relationship of the thing to its property is metaplnsial distorted.

the property is larger than its substratum adhika (II) without a proper substratum višesa (1) 142 These aberrant utpreksā (see above B iii) seem not to be considered separate tiguile, but varieties of atiśayokti, in accordance with the familiar classiticatory tenad: uh guņa, kriyā, dravya.

THE SYSTEM OF FIGURES found in several substrata 59

uncertain as to substrata višeşa (II) a quality shared by several per- samiaya (II) sons tulyayogitå (C) The relationship between properties of a thing is distorted. an essential property represented as another apahnuti (II) paradoxical qualities contradictory qualities leśa (II)

deficiency represented as potency virodha

more than the usual number of epithets viśeşokti (I)

quality predominates unexpectedly over parikara

another pihita

quality fuses with another tadguņa quality fails to fuse, despite similarity atadguņa coalescence of qualities leads to confusion mīlita of two things

At this point the categories of definition return us to the realm of simile; but the emphasis remains on the qualities, and that is what keeps mīlita itself from "fusing" with samāsokti or utprekşā. III. FIGURES based on relationships of causality (hetu) or character (lakşaņa). The remaining arthālamkāras involve relationships other than that of similitude between two or more things (ideas), primarily con- comitance. The poetical vocabulary shows its dependence on the forms of logic most clearly here. Indian logic is based on the notion of necessary concomitance (vyāpti), which is shown to be without exception as demonstrated by syllogism. Such concomitance between two things is ordinarily in one direction only (in class logic, one term includes the other): the relationship may be that of cause and effect. Inference based on vyāpti will determine either the cause from the effect or vice versa.14s The concomitance, however, may be only occasional or conventional and not strictly demonstrable, as when we say that the streams are over- flowing because of rain in the hills; other causes can also be imagined. Cause and effect remains the foundation of the inferential relationship. and this aspect is always present in the play of the tigure on the strictly logical form. Even when the figure does not explicitly formulate an inference, the distortion of the relation of concomitance is generally

148 Nyāya sūtra 1.1.5; Sesavat, parvavat; reasonings of the form sumdnvato drsta appear poetically as arthāntaranydsa. Cf. Keith, Indian Logic pp. 88-9.

60 cast in terms of a universal expectation and its disappointment-the strange and marvelous instead of the usual and mundane. The largest number of causal figures involves the efficient cause (A), to the operation of which can always be analogized the material cause (thus satisfying both nyāya and sāmkhya): as, for example, smoke is always a sign of fire, cloth a sign of threads. A small number of figures, however, involve an appeal to a final cause (B), either as a moral or a justification. Finally (C), some figures are based on entirely conventional concomitances;14 the expectation has the form of suggestion: the simple apprehension of the other term in the concomitance, but lacking a specific element of disappointment or distortion.

(A) figures of the efficient cause (kāraka); cause leads to effect; poetic cause hetu The figure as described by Dandin includes many of the following types: poetic cause kāvyalinga (Hetu is often considered too literal or conventional,145 and this figure, with the "poetic" specified, is substituted.) enchainment of causes kāraņamālā146 cause-effect, incongruous vișama (II) cause, effect with incongruous attributes vișama (IV) cause-effect, inverted (in time) pūrva cause-effect, reciprocal anyonya one cause, several effects viśeșa (III) one cause, effects contrary to each other adhika (I) one cause, effect contrary to that intended vişama (III) one cause, obstructed, carries through to its effect in a modified form one cause, effect at a distance, or at an vyāghāta

interval, or in improper substratum effect realized by a coincident cause asamgati

effect realized without its cause samādhi vibhāvanā (B) Figures of the final cause (jñāpaka) benediction motive represented in an inappropriate subject utpreksā (VI) aśiş

144 Operating, that is, through a laksana: a sign indicating that another sense is to be construed. 146 See above, p. 44. 146 Also typable as a variety of mālā, below VI (D).

THE SYSTEM OF FIGURES moral drawn from an event or particular state 61

justification; a second instance to fortify the nidarśanā (I) first, often in the relationship from general to particular and vice versa (to which Ruyyaka adds) a fortiori con- arthāntaranyāsa clusion arthāpatti

(C) Figures of suggestion (often illustrate the dhvani of the later theory) deduction of a paroksa, stated as such lakşana (indicator) dhvani (thing suggested) anumāna

remark response response praśna remark uttara praise blame aprastuta-

blame praśamsā (III) praise negation (of fact) vyājastuti affirmation (of another) ākșepa affirmation alternative or contrary vakrokti (I) (irony) one thing another paryāyokta pretext embarrassment leśa (I) gestures meaning sūkșma literal description context bhāva (II) associations thing associated avasara context mood preyas (II) context vibhāva mood bhāva bhāva (I)

These figures would correspond in general to Rudrata's category vāstava, with the exception of those involving an element of incongruity so great as to associate them with his notion of hyperbole. We have preferred structure to intention in this categorization, and restrict hyperbole to those figures involving patterns of quality and/or substratum only. Nevertheless, as the universe of figures is multivalent, hyperbole will be found associated with many other categories.147 Similarly, the notion of suggestion can be determined in any relation and particularly in that of similitude, where the figures samāsokti, aprastutapraśamsā, etc. manifest it. There, the pattern of definition would lead us to consider 147 As Daņdin says (2.220) "alankārāntarāņām apy ekam āhuh parāyaņam / vāgi šama- hitām uktim imām atiśayāhvayām" ('the form of expression, called hyperbole, celebrated by Vāgīśa, is considered the sole aim [viz. purport] of [many] other figures'). Bhāmaha's remarks on atiśayokti/vakrokti (2.85 ff.) probably bear on the same point.

INTRODUCTION 62 it secondary to the simile, while here it either functions independeuty or in terms of relations not otherwise categorized. IV. FIGURES based on conjunction or collocation. Distinguished from the former category in that nothing more than coincidental coexistence is implied; no notion of necessity or convention relates the two thines or events. As such, presence is assumed; inference is lacking. The figures are "descriptive", and it is only the quality of the whole description which distinguishes them from mere sentences of non-poetry (see svabhāvokti discussion). (A) conjunction of things appropriate conjunction sama inappropriate conjunction vişama (I) conjunction of usually separate things sahokti separation of usually conjoint things vinokti exchange of one thing for another parivrtti

(B) conjunction of qualities expansion of descriptive adjuncts samuccaya capture of the exact genus in an individual; meticulous description svabhāvokti

V. PUNS. As an arthālamkāra, pun (śleșa) participates in the idea of all the preceding categories (as a śabdālamkāra, see below VI). Pun in general is the simultaneous apprehension of two or more meanings, determined in a given phonemic sequence. (The latter qualification distinguishes pun from the figures of suggestion, where one meaning suggests another.) When the grammatical analysis of the sequence is identical, or nearly so, for the two meanings apprehended, those meanings are a fortiori "compared", and we understand the figure as an arthālamkā- ra. The simultaneity of the apprehension lends an element of hyperbole. at least in manner. Pun resembles the figures of suggestion and colloca- tion in that one of the two meanings is generally prākaraņika ('relevant'). the other aprākaranika. The relevant meaning is, however, often not the obvious one. As we have stated above, the pun in many ways constitutes the figure par excellence. The universe of figures can from one point of view be seen as a meticulous analysis of the components of the pun. As a separate figure, arthaśleșa has been most exhaustively treated by Rudrața, who, as we have seen, devotes an entire chapter to it. Śleşa, as Dandin was the first to observe, is also the tigure which combines

THE SYSTEM OF FIGURES most easily with other figures.148 Many double entendres have been 63 illustrated s.v., upamā, rūpaka, samāsokti, etc., where the pun formally occupies the place of the tertium.148 Puns, which Rudrata considers major figures, can be said to have the double-entendre as their principal force, though they also all retain an aspect of simile. Here, unavoidably, an appeal must be made to the intention of the poet and to the notion of a universe of ideas from which he selects for emphasis, but which never- theless imply an order insofar as his utterances are intelligible. In fine, it should be noted that as a śabdaślesa,150 pun has been minutely distin- guished as to the precise grammatical or syntactical element that carries the double meaning. Our point of view, however, in keeping with the character of the pun as primarily conveying an assertion, is that of the referential or ontological character of the meanings played upon.151

(A) The two levels of punned meaning are ontologically the same: both qualifications (of the subject) avayava (see samāsokti) both nouns (subjects) aviśeșa both moods (terms suggestive of moods) vakra

(B) The two levels are the same, but the meanings are opposed. the literal interpretation carries an incompatibility which is resolved by the second virodhābhāsā carries an outright contradiction virodha (III) 148 "Śleşah sarvāsu pușņāti prāyo vakroktişu śriyam", Kāvyādarśa, 2.363. Empson's recent discussion of ambiguity (esp. chaps. 3,4 and 6) brings Dandin's remark up to date. Moreover, his distinctions recall those of the ālamkārikas. 149 Ibid., 2.28, 87, 161, etc. 150 See below VI (F). 151 A number of English words have been used to translate the Sanskrit ślesa; none, however, conveys the intended significance of that word and all have connotations which are unfortunate. 'Pun', which covers much of the concept, is also used, and perhaps primarily understood as word play of the type that Ogden Nash has made famous. But Nash's punning, like Joyce's, is echo and not literal; the words are often malformed to suggest the second sense. In many cases, especially in Nash, there is no second sense, simply malformation for purposes of rhyme or rhythm. Word play in this sense would probably fall under the Indian category prahelikā ('conundrum') or śleşa vakrokti (qg. vv.). Similarly 'double-entendre', often preferable to pun, errs in the other extreme, including a number of figures which involve two meanings, but have no explicit grammatical basis of expression, such as irony. In these cases, the content alone permits the decipherment of the double-entendre; the ślesa must have grammatical reference. Lastly the learned 'paronomasia' includes cases which are not puns, but only adjunctions of words similar in form but different in meaning: "But from her grave in Mary-bone, / They've come and boned your Mary" (Thomas Hood).

64 INTRODUCTION

(C) The levels of the pun are ontologically different. praise and blame vyāja (see vyājastutí) a question and its answer vakrokti (1), adjectives and complimentary nouns type: kāku tattva pudique, risqué ukti

(D) To the basic pun is added the tour de force that the second meaning alone is referred to. parisamkhyā

VI. GRAMMATICAL AND SYNTACTIC FIGURES: repetitions (śabdālamkāra); figures whose deviation from the norms of standard utterance consists not in the misrepresentation of an idea or judgement, but of the grammati- cal basis of speech itself. Grammar imposes a certain order on speech, in that violation of its rules courts incomprehensibility, but within these rules it permits great variation for emphasis, clarity, or style. Figures which assume as a basis of deviation the standard grammatical (and by that term we intend everything from phonics to the paragraph or even beyond, provided the focus is the means of expression, not the idea expressed : śabdālamkāra) form of the language are of course specific to a given language, at least in the details of their functioning. The poetics of classical India presupposes the Sanskrit language. Even though Prākrits are often referred to, especially in exemplification, they supply no varia- tions, because their use is based on the erroneous notion that the Prākrits were in fact completely describable in terms of Sanskrit grammar. In Sanskrit, as in most languages, scope for explicit violation of the rules of grammar is extremely limited. Movements such as Dada or the San Francisco school may from time to time experiment with these limitations. but in general the acceptable verbal figures are of two types: those which violate otherwise conventional, but not obligatory, patterns, that is. they impose a pattern different from the one which, though not required. ts expected; and secondly, those which formalize what is normally a random dimension of the grammatical material and hus create regularities where none were expected. Examples of the former would be the English inver- sion of subject and verb, which is allowed as "poetic", or the many rhetorical figures which rearrange the sentence for emphasis -chiasmos, for example. The verbal figures described in the Indian poetie texts are, with onl rare and doubtful exceptions, of the second type: those which impose a

THE SYSTEM OF FIGURES 65 form where none existed. A number of reasons can be advanced for this surprising onesidedness. (1) The absence of an Indian "rhetoric", and its attendant notion that somehow the outward arrangement of the assertion is different from, or more influential than, the ideal arrangement deriving from the subject itself (the figures arthālamkāra in general). (2) Convic- tion (persuasion) as an intellectual function was never separated from understanding. (3) The kinds of figures which best exemplify the first type of verbal figure mentioned above depend on fixed word order, notably lacking in Sanskrit, which therefore does not provide a basis of expectation upon which to rearrange the sentence for effect. (4) Persuasion is a natural and not exceptional accompaniment of delectation, and both functions can be realized by formalizing language, either ab ovo, as in the second type of śabdālamkāra, or ideally, through the assertive func- tions of the arthālamkāra. (5) Finally, the general bias of the Indian poetic is against identifying or comparing word-function and sense- function, as would be done if a sense or a purpose were to be derived from a peculiar arrangement of words. We have already referred to this basically divergent attitude in distinguishing the Greek sense of metaphor from the Indian. Here, figures based on words, that is "grammar" narrowly defined, are describable in terms of form alone, and although it is assumed that they produce distinctive impressions when heard, no meaning, à la program music, is ever assigned them.152 In the area of śabdālamkāra, particularly, that subtle appreciation of detail is manifested which is both the Marengo and the Waterloo of the Indian mind. The Indian "enumeration" seems always to have been guided by the idea that if a subject is well enough known in its specifics, its form and structure will be self evident. The monuments of this ap- proach are the grammar of Pāņini and related "grammatical" disquisi- tions in other disciplines, including that of our immediate preoccupation, figures of speech. We distinguish the several levels on which language functions as indicator and means of expression, and find figures exhaus- tively defined for each.

(A) Phonemic repetitions

alliteration anuprāsa

(Dandin defines a type of alliteration based not on the phoneme, but on 1 The closest approach to such teleological argumentation would be the attempt of the dhvani school to subordinate the occurrence of certain figures and gunas to the needs of the primordinate rasa. Cf. Dhvanyāloka 3.10ff.

INTRODUCTION 66 the distinctive feature; hence it could be considered a figure of phonetics. Note that metrics [i.e. verse generally] would appear here also, as a regularization of the feature syllabic quantity.) chekānuprāsa metathetic alliteration (repetition interrupted by distinctive vari- ation) yamaka repetition of fixed sequences of phonemes.

(B) Morphemic repetitions of homonyms punaruktābhāsā of the word or the sense āvṛtti of the syntactical category (subject, verb, etc.) dīpaka (This can also be seen as non-repetition, for one element remains unrepeated, uniting the several phrases.) of the word, but in a different sense (accepta- tion) lāțānuprāsa

(C) Phrasal repetitions. There is no figure involving the repetition of phrases only, for it is hard to see how it could be given a poetic twist. But there are a number of figures which, in addition to the phrasal repeti- tion, involve a characteristic fixation of certain words within these phrases, creating a pattern within a pattern. the predicate of sentence one becomes the subject of sentence two, and so on ekāvalī the subject of each sentence constitutes the most characteristic aspect of its predicate, and the subject of sentence one provides the predicate for sentence two (the inverse of the above) the multiple predicates of several sentences sāra

exactly parallel each other in terms of some larger image (could involve subjects, verbs, etc.) It is only in these three figures that we see any interest shown in syntactical yathāsamkhya

arrangement (see above). Yet even here, the emphasis seems to be on the enchainment of grammatical forms rather than on the irregularity

THE SYSTEM OF FIGURES 67

of the reformation itself. In other words, keeping to our earlier distinc- tion, these figures, too, promote an order where there is normally none, viz. in respect of word patterns across sentence or phrase boundaries.

(D) Repetition of figures the figure itself: the figure mālā, superadded to any appropriate figure, as upamāmālā or rūpakamālā different figures: samsrsti; the figures may be inherently or extrinsically related.

It is in terms of this category of complex figuration that the early Indian poetic approached the problem of composition-the actual manifestation of poetry. Its discussion shows an awareness that the analytic tools of figurative interpretation, though they reveal the ideal structure of poetry, nevertheless do not account for the complex interdependence of various kinds of figuration which we find in any actual poem. The figures rarely occur in hermetic isolation. More often, and especially in the most beautiful examples, it is precisely the number and the intimate blending of figurative types which is striking and worthy of wonder and praise. The Shakespearian trope: "Oh, that I might be a glove upon that hand / That I might touch that cheek!" combines the figures rūpaka, āśiș, nidar- śanā, and various alliterations and meter in its deceptive simplicity. Nevertheless, the charm of such expression, the ālamkārikas would hold, is not, in fact, simple unless complexity itself is given a poetic value.153 What it means as poetry is still dependent on the meticulous isolation of its constituent figures. The collocation has no other meaning or guiding principle behind it.

(E) Repetition in larger grammatical frameworks (paragraphs, chapters, etc.)

imagination the figure bhāvika

15a A plausible argument, which incidentally would serve to explain how it is that certain styles and types of poetic speech (such as similes) can be used in non-poetic contexts: " ... Jesse Jones ... in a hectoring speech ... advised the banks to improve their capital position" (Arthur Schlesinger Jr.). Here the simplest form is used in total isolation from other poetic factors. The premeditation of poetry lies not in the stringing together of such isolated poetic atoms, but rather in the fruitful co-occur- rence of the elements in a whole. The comic effect, on the other hand, can be seen as a disproportion among the complex of elements (which includes content), an exaggera- tion or an over emphasis of some at the expense of the complex image.

Any poetic must ultimately account for the contrasting principles which receive emphatic treatment in other theories of criticism. We meet here, near the end of our list of figures, the imagination, a principle which we might say has been emphasized almost to the point of exclusiveness in many modern poetics and which dominates our contemporary attitudes toward poetry. Originality is a quality highly favored; indescribability, the state of having surpassed conventional limitations, is considered to be the quality of the poetic product corresponding to the faculty of imagina- tion and is considered proof of genius.154 In a certain sense, the poet is thus the "culture hero", possessing as he does the virtues of individuality, inimitability, and creative distinctiveness in an absolute degree. Our attitude toward poetry is as stereotyped as that of classical India, where such notions were in fact largely supplanted by their opposites. In the Indian tradition, then, imagination (bhāvika) is generally described as the ability to make the several images of the individual poetic statements coherent in terms demanded by the work as a larger whole. It is manifested in such things as the plot (the story stringing together the individual statements), by the lack of shocking contrast in its development, by the general appropriateness of one image to its neighbors, and the like.155 Modern critics have been worried by the fleeting attention paid to this "crucial" concept; it is above all this short- sightedness which has earned the Indian poetics its reputation of literary irrelevance.156 We will examine possible historical reasons for this difficulty in the following section; here we shall only remark that the point

154 "One of the greatest limitations of Sanskrit poetics which hindered its growth into a proper aesthetic was its almost total disinterest in the poetic personality by which a work of art attains its ... individual character" (De, SPSA, p. 72). " ... we search in vain for a complete definition or clear discussion of the poetic imagination in the whole range of Sanskrit poetics" (ibid.). "A poetic intuition cannot have a piescribed technique of expression ... it is not an intellectual concept ... nor is there any passage to it from the physical fact or the intellectual concept. It stands by itself" (ibid., pp. 76-77). But on the uniqueness and inexplicability of the creative act, cf. Empson: "Things temporarily or permanently inexplicable are not to be thought of as essentially different from things that can be explained in some terms you happen to have at your disposal ... Explanations of literary matters geometry, and, if you cannot think of a construction, that may show that you would are more like Pure than Analytical be wise to use a different set of methods, but cannot show the problem is of a new kind" (p. 285). 165 Dandin, Kāvyādarśa, 2.364ff. 156 "The Indian theorists have almost neglected an important part of their task, v. to find a definition of the nature of the subject of a poem as the product of the poct mind; this problem is the main issue of Western aesthetics. Only svabhūvokti aild bhavika can be adduced as a proof that the Indian theorists were conscious of the problem ... ", De, ed., Kuntaka, Vakroktijivita, Introduction, p. xix.

THE SYSTEM OF FIGURES 69 of view adopted by the Indian writers has kept them from discussing imagination in the modern sense. Nevertheless, recent literature abounds with attempts to turn bhāvika and similar notions into substitutes for Western critical categories, into steps in the progress of the Indian mind to profounder awareness of beauty.157 Bhāvika is far from being a characteristic of all poetry: it is limited as prabandhaviayo gunah-a quality whose scope is the work. It is not met with, according to Dandin, in those many poetic works which are not wholes: the collections, anthologies, the isolated verses which were characteristic of the classical poetic product. In the systems of the ālamkārikas, bhāvika is a notion of secondary, though considerable, importance, which was best accounted, in their precise terminology, a figure, most likely of śabda, whose grammatical reference was that span of utterance longer than that which would ordinarily be defined in the grammar and which had no specifically grammatical means of indica- tion.158 If bhāvika is a śabdālamkāra, then its reference is to the language itself as a means of expression: it explains matters pertaining to arrange- ment and not to expression; given the complexity and length of certain poems, it tells how to state the appropriate "togetherness" of that mul- tiplicity. Certain poetic works, without in any way augmenting or diminishing their status as poetic works, do have a unity of theme or a coherence of purpose which requires special notice, as differing from both colloquial utterance and from the manner of existence of the events referred to. Within this unity, the imagination of the poet could be said to be working as a limiting factor comprehending and relating the discrete content of the work: in the same way, the poet, composing alliterations, can be relied upon not to break off in the middle with some harsh or irrelevant sound. The imagination as a quality of the whole is an alamkāra of repetition in much the same sense: it is (or is testimonial of) an ex- pression of unity not otherwise present, imposed upon an aspect of the linguistic continuum which is not ordinarily so formed. Life is not made up of coherent or dramatic events as a rule. The classical critics were far from thinking of this "quality of the [whole] work" as the psychological

157 "Nevertheless, the aberrations are at the same time attempts to reach the truth; and in the midst of unlifted shadows one does often perceive a running thread of silver lining", De, SPSA, p. 79. 158 "Bhāvikālamkāras tu mahākāvyeșv eva niyamenāvasthitab" (anon. Comm. "Hrda- yangama" to Kāvyādarsa 2.364-366 [p. 215]). 1 would like to think, for the sake of completeness, that bhāvika was not considered an arthālamkāra at all, since it does not formulate the possibility of any concrete poetic expression or idea. Dandin, however, does treat it at the end of his second chapter (on arthalamkāra), after samsrsti.

INTRODUCTION 70 arena in which genius and talent were to manifest themselves. Of far greater relevance here is the composition of individual stanzas, which are poetry in the Indian sense of the term and which require only limited attainments, such as the ability to speak, think, and compose on the manifold levels of non-referential language. Many, if not most, poems have no other unity. (F) Phonemic limitations whose principle is not inherent (sausabdya), but arbitrary, that is, which rest on some notion having nothing to do with phonics. the pattern of repeated phonemes enables the verse to be shown in pictorial form citra the pattern of repeated phonemes enables the verse to be shown in geometrical form duşkara conundrums: the solution to a problem is hidden in the verse expressing it, which needs to be seen differently prahelikā puns: by another morphological analysis, a different meaning is understood in the sequence. Can be seen as a yamaka whose repetition is only implicit śleşa (śabdaślesa)

(IV) THE ADEQUACY OF THE ĀLAMKĀRIKA POETIC

(a) Kāvya Contemporary critics have too long been dependent on the reformulation of the alamkāra theory proposed by the dhvani writers, Ānandavardhana and Abhinavagupta, for whom the ālamkārika viewpoint is deprived of its independent status and reduced to a moment in a more all-embrac- ing view. Histories of Indian poetics have been little more than attempts to justify this ex post facto subordination. The theory should be con- sidered in its own terms and not for what defects may be predicated of it in terms of another aesthetic. We have tried to keep strictly within the bounds of an alamkara poetic in this introduction. The scope of the discussion ought, at this point, to be broadened to a certain extent, for the characteristic adequacy of a poetic will inevitably correspond in some measure to the poetry with which it was intended to deal. The adequacy of a poetic may be determined in three ways (excepting the disputations which one poetic directs against another):

THE ADEQUACY OF THE "ALAMKRIKA" POETIC (a) Is the theoretic system itself complete; are its constitutive principles 71 adequately worked out in their implications? (b) Does that system adequately distinguish poetry as a genre from other kinds of expression? (c) Is the treatment of poetry internally well differentiated? Does the poetic adequately describe the aims, means and varieties of the poetry which constitutes its subject matter? The first two points have been sufficiently discussed in what precedes, as far as the alamkāraśāstra is concerned. Question (c) can be interpreted in two senses which we might call "contextual" and "metaphysical". De, Dasgupta, and Keith, the standard Western or Westernized interpreters of Indian poetics, prefer to see all poetics addressed to an abstract genre "poetry", to the nature of poetry per se,159 and to account for the variety of poetics in terms of varyingly adequate responses to that problem. In this section we will ask whether the poetic may not be better accounted for by relating it to a kind or style of poetry which was actually cultivated at a given time in India. The alamkāra criticism, from the emphasis it places on the mahākāvya160 and from the style of exemplification it adopts, is evidently addressed to the stanzaic poetry of the Indian classical period. This poetic genre has certain definite characteristics, which shape not only the aims and achieve- ments of the poetry, but, ipso facto, limit the criticism focussing upon it. These factors are well known, and we will not do anything here but sum- marize them in outline and show how certain biases of the poetic turn out to be in fact admirably descriptive of the genre. The poetry of classical India was microcosmic poetry. The locus of composition was a minimal unit of expression, the stanza, and this is to be understood in a quite radical way as excluding larger units of com- position such as the chapter or the work itself.161 The latter are in typical cases not even present, as, for example, in the anthological and thematic collections of Amaru and Mayūra. But even those compositions which have a story or a plot, however loose-the mahākāvya par excellence -- retain the stanza as the unit of composition. Much that appears at tirst blameworthy in classical poetry is explicable in terms of the de-emphasis of the story. The story is never central; it is at best a pretext for stringing 150 160 ... poem as the product of the poet's mind" (De, cited above, n. 156). and Dandin (1.14ff.). Also called sargabandha, treated as poetry par excellence by Bhāmaha (1.18ff.) 101 See Renou, "Sur la structure du Kavya", Journal Asiatique, 1959, pp. 1 ff; the genre stanzaic poetry" is defined, and its stylistics are thoroughly described from the point of view of grammatical usage.

INTRODUCTION 72 together admirable verses-really just a narrative theme. The story may at any time be interrupted by long descriptive irrelevancies on the sunrise, the mountains, the moonset, which appear extraneous by stand- ards emphasizing the unity of plot. Moreover, the overdevelopment and the stylistic complexity of the stanza itself testify to a smaller unit of composition. It is the stanzas and not the work which have a life of their own in the historical traditions of Indian literature. The anthologies are collections of such stanzas from disparate sources. Little value is attached to the whole work, to which stanzas are freely added or from which they are excerpted by the tradition. But what are the aims and possibilities of such a stanzaic style? Evi- dently, the structure of the stanza itself is considered the major achieve- ment of the poet. The character of the stanzas is determined in large measure by traits adapted from the peculiarities of the Sanskrit language: a complex metrics based on syllable length, a sonorous phonemic system, a complexly rich consonant system, an enormous vocabulary enabling richness and overtone in expression and great multidimensionality of the image, and a syntactical system tending in the opposite direction- to great compactness and density through inflection and the compounding of entire clauses. The tendency is toward the expression of one bewilder- ingly complex but stringently coherent idea or image within the stanzaic unit. The stanza imposes its form on the poetic content, which is delivered compactly as image, as figure. The formulation of the principles of stanzaic composition is found adequately expressed in the older ālamkārika treatises and reaches its essential statement in the theory of the figures, an attempt to comprehend the ability of language (its forms of grammar and thought) to express and sustain this ideal microcosm and encompass a second level of mean- ing. Language must do so compactly and rigorously within the structures of coherence and relation of the syntax; language must state imaginations. The typical stanza aims at a richness of intelligibility which is at first overwhelming, and it is that intelligibility which is the aim of the poet. The amount encompassed within the strict syllabic limitations of the longer (or poetic) meters, is so great that a complete translation is often two or three times as long as the original flabby and prolix precisely where the original displays a tense e Translations will tend to be and is most striking in its beauty. Such intelligibility, admittedly an mpactness intellectual value, cannot rush immediately and of the poet; theories of intuition are lamentably inept in explaining tle ill blown from the head significance of the kāvya or, par contre, its ability to provoke delight.

THE ADEQUACY OF THE "ĀLAMKĀRIKA" POETIC The arthālamkāras are the forms of that intelligible richness, from the 73

apparently mundane svabhāvokti, or jāti, which eschewing all figuration per se, attempts to encompass the totality of an event, or a movement, or an individual's characteristic moment: The hawk on high circles slowly many times Until he holds himself exactly poised. Then, sighting with his downcast eye a joint of meat cooking in the Chandāla's yard, he cages the extended breadth of his moving wings closely for the sharp descent, and seizes the meat half-cooked right from the household pot.162 to the elegant upamās of Kālidāsa: You will find her voice subdued, my wife and second life While I'm away; a single cakravāka dove, longing for its mate. Her heart so yearns, as these heavy lonesome days go by That she's become, it seems, a wild lotus struck with frost.163

and the pure delight of the complex half-punned imagery of Mayūra: Deep in the blooms of the lotus; upon salient sharp-honed peaks Alike falling; uniform at birth of day and at the evening hour of rest- May the sun's effulgent rays protect you! (Like travellers) arrived in chorus on the courtyards of three worlds, Bestowing torrid merit, born of constant journey's toil.164 The poetry of the classical period was possessed of its own poetic, in which the major practical issues of that genre were explored fully and subtly. It cannot be denied that the great achievement of Sanskrit poetry lies in its word pictures: the meticulously complete vignettes of the stanza addressed to the mind in contemplative repose. What it attempted to do was limited by that form as well as fulfilled by it. In the poetic of the stanza we do not find discussions of issues which are not, or are only peripherally, pertinent to the form, such as are posed by poetic works whose principle of unity is much more broadly defined. Instead, we nnd preoccupation with imagery, verbal and sensible, connotative and denotative; an awareness of the possibilities of imagery constituted the craft of the poet. Other poetics will, of course, emphasize the ability of ine poet to work in grander media; it would be difficult to describe the roman fleuve in the categories of the alamkāraśāstra. The materials 162 Vidyākara, Subhāşitaratnakoșa (trans. Ingalls), 1150. 168 104 Meghadūta 2.20 (my translation). Sūryaśataka 3 (my translation).

INTRODUCTION with which the novelist-cum-poet works are more vague, more disparate 74

not so exclusively oriented to problems of expression. It becomes legitimate to discuss the conception of the total work of art, the manner in which it reflects its time, the leading ideas of the period, and the like, because the poet is to a great extent concerned with representations of these broader issues. The issue of the genius and the imagination of the poet, under- stood not as a minor sort of combinatory facility165 but as the real fountainhead of the unity-now seen as structure-of a very complicated work, may with justification be posed as a category of serious criticism But when we deal with the poetry of classical India, these notions have little force; they are in fact impositions upon the subject matter. The kāvya poetry was complete in the stanza. The ability of a poet like Kāli- dāsa to compose great works was in fact an ability to compose many beautiful stanzas; the multiplication of stanzas does not alter the critical point of view, for it was not a creative multiplicity. The alamkāra poetics, in describing the materials and the technique of stanzaic composition (to which the notion of concrete subject matter was either conventional or irrelevant), adopted the only possible point of view adequate to the poetry.

(b) Nāțya The foregoing does not imply that there were no other points of view expressed even in classical times. Indeed the internal history of Indian poetics gives much evidence for competing doctrines. But it is conclusive for our view that the doctrines were conceived of as competing only when the poetic genres referred to or implied were losing or had lost their independent status.166 The dhvani doctrine testifies to this amalgamation of previously separate traditions. During the early creative period of the alamkāraśāstra, as far as the texts show that tradition (7th-10th centuries). another poetics coexisted alongside the figurative with little evidence of interaction: that poetics which took as its problem the drama, and which elaborated as its decisive concept the rasa. The oldest extant work of Indian "poetics" is devoted to the drama: the Nātyasāstra of Bharata.167 It is a compendium of topies relating fo

165 Above, p. 67 re the figure bhāvika. 166 Viz. the dramatic and the stanzaic genres. See below. 167 Bibliography in De, HSP, I, pp. 44-45; we have used throughout this werk fl edition of the Gaekwad Oriental Series, ed. M. Ramakrishna Kavi. Currently afP ing is the first complete English translation by Manomohan Ghosh, in the Bibliotle Indica, Calcutta, 1959 -.

THE ADEQUACY OF THE "ALAMKĀRIKA" POETIC the drama and not, strictly speaking, poetic, for it deals with matters as 75 diverse as the construction of the theater and the use of make-up. It is certainly allied to the puräna in style and is not the work of a single author despite the eponymic attribution. For criticism, the dramatic problem presented itself in a different light, and Bharata's treatise attests this difference decisively; the drama is not a creation of the word alone, or even of the word primarily. Of course the language spoken on the stage is an essential element in the drama, but the drama is, in addition and more importantly, a visual spectacle; the language is but one element in the technical materials available to the dramatist. If he imitates human actions and events, he first of all rep- resents (or can represent) human actions and events directly by means of characters, gestures, actions; the realm of immediacy is made available through the actor, a crutch upon which the verbal poet is unable to de- pend. The communal character of the dramatic representation also contributed to its distinctiveness. An audience was an essential part of the produc- tion, a material in which an effect or an imprint was to be produced and whose formation in this sense constituted the end of drama. Although the kāvya literature may have been recited, and historically may have been a lineal descendant of bardic, epic poetry,168 in its developed form it was far too complicated and polished to be enjoyed only viva voce; its audience was more abstract: the learned, often an assemblage of other poets.169 The dramatic production was also determined in a different kind of time, due largely to its dependence on a real audience. The kāvya work could be indefinitely long, for its unity was aggregate and ad libitum both for 108 Though few direct links have been discovered which would mediate the enormous differences in style between the developed forms. Portions of the Rāmayana have 169 often been suggested as links. Cf. Dasgupta and Dey [De], HSL, p. 13. The proper audience for poetry is first extensively discussed in the dhvani literature: cf. "tena brümah sahrdayamanahpritaye tatsvarūpam" (Dhvanyāloka 1.1), and the commentary "sahrdayahrdayāhlādisabdārthamayatvam eva kāvyalaksanam" attributed to a pürvapakșin. The audience as the locus of the dramatic experience becomes doubly important to Abhinavagupta, who analogized that experience to religious eestasy: below, p. 77. Again, this concern with the audience on the part of the dhvani theorists testifies to the syncretistic character of their doctrine, for they were faced with the problem of relating two genres whose audiences were in principle distinet. and whese distinction was self evident. Abhinava's definition of the sahrdaya has eften been quoted "yeşām kāvyānušīlanābhyāsavašād višadībhūte manomukure varņanivatan- mayibhavanayogyatā te hrdayasamvādabhajah sahrdayab" (Locana on Dhv. 1.1. p. 11. Cf. De, SPSA, pp. 54ff.). These connoisseurs may, as the tradition asserts, have gathered occasionally to form a semi-permanent kavisabha, in order to pass judgement case. on the poetical works presented to them. The Tamil sangam is the most illustrious

76 the poet and his reader. The dramatist, however, was working with a time span which itself imposed a kind of unity on his creation, in terms of which a determinate effect was produced in a definite audience. In the dramatic mode, it would, by contrast, not be far wrong to say that the work as a whole constituted the unit of composition, into which entered as conditioning factors the audience, the actors, their actions and gestures, the events portrayed, and, finally, the spoken words of the play.174 In contemporary terms the distinction between the written and dramatic is somewhat obscured, owing to the great diffusion of printed texts. Plays are produced, but they have also become an important part of the written literature. One might wonder, for example, whether A Long Day's Journey into Night was ever meant by its author to be performed. As the unit of composition, the work's special existence, apart from its conditional factors, was expressed in terms of the peculiar relation which all of these factors bore to one another in the understanding of the audience, and which constituted the proper consequence or effect of the dramatic production. Bharata determines this special unity of the work as the rasa or 'dominant mood', and this term has come to denote the leading idea and is perhaps the most distinctive feature of Indian poetics. We cannot develop here the many interesting implications of the rasa theory for dramatic criticism. We are concerned with it only insofar as it provides a contrast both in subject matter and in critical point of view with the alamkära theory, which we hold was addressed to the stan- zaic poetry of the Sanskrit kavya. The two criticisms, coexisting for several centuries, were largely kept compartmentalized not only by their orientation to different subject matter, but also by their emphasis on the means of expression. Neither was interested in the analogical question of creativity.

170 Bharata deals with the spoken word as it affects drama in chaps. 15 (on metre) and 16 (on the figures, gunas, dosas, etc.). Metre is important because of the many set verses, in classical kāvya style, which were placed in the dialogue of the drama. and by which were stated its moods and climaxes, in a manner analogous to the arias of an opera. In the chapter on figuration, Bharata mentions only four figures. i ;. rūpaka, dipaka and yamaka; this has usually been taken as a primitive or germittal form of the alamkārasāstra; Bhāmaha, the next writer whose works are evtant. discusses over thirty figures. It may be, and it is just as likely, that the four figur's were not intended as exhaustive, but were, like the metres of the preceding chapter. presented in abridged form, more to suggest the importance of the subjeet to the dram- atist, who would then be expected to turn to the available manuals of metries au figuration for fuller treatment. The alamkara texts, similarly, refer peripherally f topics which are central in the dramatic tradition, nāyaka, rasa, ete.

THE ADEQUACY OF THE "ĀLAMKĀRIKA" POETIC To the question "How is the rasa expressed?", the dramatic theorists 77

starting from the famous rasa sutra of Bharata proposed various and increasingly subtle explanations involving a minute examination of the range of conditions, effective causes, manner of comprehension, and ultimate status of the rasa. Similarly, the ālamkārikas, having situated their discussion in reference to the question "How is the distinctiveness of poetic speech realized, that is, understood?", spoke to issues focussing on the capacity of language (both verbal and expressive) to convey more or a different sense than the strict employment of its forms would permit. The rasa was realized as an affective, not an intellective, unity-a mode of feeling generated by and transcending the discrete conditions and causes manifested on the stage. The language of the drama played no more important a role, inherently, than did, let us say, the gestures, the portrayal of the characters, the overtones which could be expected from the realization by the audience that this character was indeed Rāma. Abhinavagupta, the most brilliant in the long line of Bharata's commen- tators, carries the view of his predecessor, Bhatta Nāyaka, one step farther and boldly makes an analogy between the rasa experience and that of final salvation: both transcend the limited, temporal character of finite existence (here, the play) and realize the one central unifying theme of that existence (the rasa); they differ, of course, in that the finite existence to which the rasa corresponds is itself only a fiction, and the rasa therefore ceases as soon as the play ceases. The older writers of the alamkāra tradition showed a similar dependence on a theoretical framework bor- rowed from another discipline, but instead of theology, it was logic and to some extent grammar. The two partly complimentary poetics were not of course pursued in conditions of total separation; there is evidence for the effect of one upon the other. Nevertheless, the striking fact about late classical criticism is the continued development of two largely independent theories, addressed to significantly divergent problems.171 The dramatic criticism acknowledges the alamkāra doctrines in Bharata's chapter on the figures, testimony to the fact that the set verses of the drama have an important role to play in the generation of the total effect of the play. Later treatises, such as the Daśarūpaka, take for

171 A striking, though incidental, corroboration of this divorce of poetry and drama is provided by an investigation of characteristic metres in use by the two genres. See Subhāşitaratnakoșa, Ingalls' Introduction (= Harvard Oriental Studies, 44), p. 35; the ālamkārikas sometimes refer to the "other sāstra", as Daņdin, Kāvyādarša, 2.366.

78 INTRODUCTION

granted the divorce of drama and alamkāra and concentrate on topics of more immediate dramatic interest: characters, moods, kinds of drama, and the like. The alamkāra tradition similarly recognizes the rasa as the basis for certain figures, particularly the one called simply rasavat ('having a rasa'). This figure has provoked much speculation by the dhvani writers, for it presages their own doctrine of the dhvani in the figure.172 The original, figurative, expressionistic sense of rasavat is doubtless to be seen in the context of the triad of figures of which it is one: preyas, rasavat, ūrjasvi. All three are expressions of certain kinds of subjective excess and are probably to be thought of as special types of hyperbole, distinguished because of their currency, if not for reasons of form. Preyas signifies an expression overburdened with good intentions and friendliness; rasavat, distinguished by one of the eight canonical rasas (love, pity, etc.), and ūrjasvi, an expression of impertinence, an excess of egoism. All three evidently differ in force from ordinary language. The Indian philosophic emphasis on stability of temperament would doubtless sufficiently distinguish language so emotionally loaded. But, as we say, these evidences of interpenetration are superficial and do not affect the basic divergence in point of view of the two poetics. They represent areas of overlap, and are not proof of a single, constantly developing and more profound aesthetic.

(c) The Religious Lyric With the advent of the dhvani theory, perhaps in the ninth century, we do find an elaborate attempt to reconcile and unify the two divergent poetics of kāvya and nātya.173 The dhvani school (dhvani 'tone, suggestion') retains the leading ideas of both anterior poetics: the dhvani is the rasa. the transcendent emotional significance of the work, recast and redefined (in terms of a literary poetic) as the most essential form of vakrokri, that function whereby language conveys (a) a further sense, or (b) a sense not inferable from its component elements, words, logical forms, and the like. The dhvani turns out to be a more general statement of the expressionist 172 Some more figures were elaborated after the dhvani-rasa theory became popular. they are based on categories borrowed from that theory (cf. Ruyyaka, Alank urasdl vasva, pp. 232ff.). Here we discuss only the early evidence for interrelation. in the last six chapters of Kāvyālamkāra, sums up an entire dramatie theory. Rudrața. is clearly an addendum; it occupies only about one quarter of the whole work. But this

References to rasa are found re some śabdālamkāras: cf. Daņdin, 1.52. 178 See the material on poetics in Introduction to Indian Literature, eds. van Buitenen and Dimock (Asia Society, New York, 196?).

THE ADEQUACY OF THE "ĀLAMKĀRIKA" POETIC poetics, but a statement so general that it encompasses the whole dramatic 79 vocabulary including modalities that are non-linguistic. Poetry and drama are reunited in a theory emphasizing their common expressionistic basis, perhaps testifying in part to the collapse of drama as a living art form and its maintenance in primarily literary terms. The internal differentiation of the dhvani shows how syncretistic it is. Its purest form is rasa itself. A literary poetics has now taken over the portion of dramatic theory previously reserved to the ensemble of largely non-linguistic suggestions-gestures, identifications, etc. Language becomes in this view a surrogate of the real, by its nature able to suggest everything immediately-not only ideas but moods, feelings-not as secondary factors of ideas or as a contextual affectation of some figure. It is a hardy theory, and is argued ably by Anandavardhana in the first chapter of Dhvanyāloka. But the dhvani school allows the older forms of ālamkārika vakrokti, too. In this guise, the dhvani appears as the discrete idea or sense suggested by the non-literal modality of the figure, which in contrast to dhvani as rasa, is primary to whatever rasa it may also express. In this way the alamkāra is integrated into the dhvani theory and, at the same time, its association with the rasa is explained. This relationship for the first time allows the problem to be posed of the context in which the figures are used: Anandavardhana espouses the doctrine of aucitya 'appropriateness', whereby the figure is to be employed only as it furthers the predominant mood: the rasa. Otherwise the discrete sense of the figure predominates over the rasa.174 Finally, a type of dhvani is defined which is realized neither as rasa, nor as figure, but directly as a meaning. This variety corresponds to discrete suggestion, which operates through inference or association: when from the statement of one thing, another (often the contrary) is understood, as for example, irony.175 In fact the ālamkārikas considered this last type, as they did the first, special types of alamkāra. We have discussed several based on inference and suggestion under the heading "System of Figures".176 The dhvani theory really does not add anything to

174 175 Dhvanyāloka 3.32-33, 37-38; type defined at 2.25. Cf. supra, n. 153. Ibid., 2.26. The latter two types, in which the rasa occupies the rank of a subor- dinate element (to the sense or the figure expressed) are thereby termed gunibhūta- yangya. Poetry in which the rasa|dhvani is totally absent, though poetry only by con- vention, is admitted into the system as citrakāvya ('pictorial poetry'): ibid., 3.42. 176 This paradox is explained by the tendency of the dhvani theorists to equate figure with the figure par excellence-simile (upama). They were not prepared, for obvious kārika treatises. reasons, to admit the universality of the definition of figuration implied by the ālam-

80 INTRODUCTION

the materials already elaborated by the figurationists, but it does propose an important redefinition of principle whereby the distinction between literary and dramatic poetics is annulled. It is our view that this marriage of two distinct traditions would not have been possible, even as a theoretical exercise, if the genres to which they referred had not in fact largely lost their respective identities. Poetics, in India as anywhere else, follows poetry, despite the attempts of unsym- pathic critics to argue that it was in essence a normative discipline which saw its ultimate justification in the education of the poet-an effort to prescribe rules to aspiring poets.177 There was no genre in early classical times which exactly corresponded to that implied by the dhvani theory. Both kāvya and nātya can quite successfully be reinterpreted in terms of the dhvani-rasa (in fact this has been so successfully done that the rasa is today often considered the timeless standard for all Indian poetry, from Kālidāsa to the present).178 The poetic form properly corresponding to the dhvani theory is a genre which developed slowly in the late classical period and later became the only really viable literary art in India, the renascent lyric (older stotras, as those of Bāņa or Mayūra, are definitely within the classical kāvya style). The lyric devotional poem is best epitomized in the marvel- ous Gītagovinda of Jayadeva, but is more voluminously represented in the several vernacular literatures. This genre is in fact dramatic poetry; poetry with a narrative basis in the divine event and intended to convey the emotional fervor of the bhakta. It is not surprising therefore that we meet overtones of theology in the poetics, as expressed by the third great dhvani theorist, Abhinavagupta. The poetry itself begins and ends in the service of an increasingly religious ideal: devotion to a personal God. The primordinate rasa, śrīgāra 'love', needs only to be redeployed (the subject matter of poetry has always been considered a mere condition- ing factor, never an independent principle in criticism) from its earlier evidently secular emphasis to a less mundane application. The poctics of devotional poetry has never been more oundly explored than in the dhvani school which continues today to dominate Indian thinking on poetics. Even De, so profoundly affected by Western theories of poctie. and still finding many inadequacies in the dhvani theory, tends to think of it as the typical, most characteristic Indian poetic. 170 177 De, SPSA, pp. 3, 76. 178 Cf. "Sanskrit Poetry and Sanskrit Poetics", a part of Ingalls' Introduction to lus translation of the Subhāşitaratnakoşa. 179 In Dasgupta and Dey [De], HSL. D. 581.

CONCLUSION The final, almost absurd consequences of this alliance of poetry and 81 religion were admitted in the sixteenth century by the school of Rūpago- svāmin in Bengal, who chose to express their theology in poetic terms. Bhakti is elevated into the 'regal' rasa (bhaktirasarāt), qne other forms of affection are considered subordinate rasas.

(V) CONCLUSION We have come a long way from the expressionistic poetics of Bhāmaha and Dandin. The transition is harsh and seemingly total-the grammarian has become the devotional mystic-but the civilization has also changed, and the theory reflects the change. Certain fundamental ideas and approaches are held in common: the expressionistic bias, precluding, even in theories so evidently psychological as the dhvani-rasa, an explanation (if any is needed) of the origin of individual poems; the striking parallelism of the theory as universal poetic and the poetry as stylized creation; the explicit borrowing, by the poetician, of his principles from another dis- cipline, and the consequent absence of doctrines emphasizing the unity of all the arts-"art for art's sake". The final cause is never made to predominate over the efficient cause, the mechanism of poetry. Its aim is always an extension of its operation, not an ultimate justification of means viewed as disparate and secondary. The dhvani, for all its culturally and historically imposed limitations, was not an attempt at such a universal poetic, but one which did at least reconcile drama and literary poetry. No school attempted to define a category of art encompassing not only literary, but truly aesthetic subject matter, such as sculpture, music, and painting. The notion of such a universal or analogical aesthetic did not suggest itself to Indian thinkers, as it has to our own since the Renaissance, because the creative act had always been considered a matter of technique and style embodied in a tradition, evolving from its own material, and not a manifestation of the freely intuiting intellect, the genius. This applies equally to the plastic and the verbal arts. India had art of a high order; but analogical interpretations of different art forms were unknown. We have seen how difficult it proved even to reconcile drama and kāvya, similar in many respects. An aesthetic would be impossible to conceive. What are the intellectual foundations of the alamkāraśāstra, and what are its claims to poetic relevance? We cannot discuss here in great detail the interesting question of the dhvani in the context of medieval poetry-

82 INTRODUCTION

to what extent the dhvani clarifies the aspirations of that poetry and the manner of its prosecution vis-à-vis the stylized poetry of the classical era. It does appear that several points can be made in outline from the angle of our main interest, the alamkāraśāstra. The Gītagovinda, to take the supreme example, is a poem in a sense that no classical kāvya is, be it epic or anthologic: it aims at a religious goal which is secondarily a poetic goal, at the single idea of love and its embodiment in the God Krsna and his consort Rādha, an ideal to be evoked in the hearer by sympathy and involvement. The work serves that purpose in a very direct sense. It is not surprising that the dramatic theory of Bharata suggested itself as a more adequate model for interpreting this new poetico-religious form than did the grammatical and abstractly intellectual alamkāra theory which emphasized the understanding. Written poetry had at this time in fact begun to serve purposes which formerly were considered more appropriate to the drama. From classical times, there is only a single poem known to us which by any stretch of the imagination could be called dramatic in these terms: the Meghadūta of Kālidāsa. It is of course a stanzaic poem of very high quality and observes the important canons of that form: ornate meter, stanzaic independence coinciding with a single, many-sided image, and extensive employment of arthālamkāra. Yet as a work the Meghadūta has a kind of dramatic unity, almost a "plot": the separation of lovers, the voyage of the cloud to the beloved, its message and promise of reconcilia- tion. This form, which we owe to the originality of Kālidāsa, was, how- ever, felt to be so peculiar that it excited only imitations. It constitutes a "genre" of its own. The impact of the Meghadūta, whatever its dramatic form, is never- theless very different from that of the medieval devotional poem. It does not convey a rasa; or rather, whatever rasa is understood (śrīgāra in separation) is certainly understood first in the individual stanzas, then in their aggregation in accordance with the figure rasavat. The Meghaduta is a stanzaic poem held together by an emotional thread instead of a theme or a legend. The genre epitomized by the Gitagorindds demands an accounting of the poem as a whole (the rasa is embodicd in whole chapters or cantos), unlike the Meghaduta, wherein the narrative unity of the work (the voyage of a cloud) becomes the pretext for a sel of lovel intaglios.180 The alamkāraśāstra does not sufficiently account for the older (Pre-

180 The figure bhāvika in manifestation.

CONCLUSION classical) Prakritic lyrics, like those of Hala and the Buddhist canon. 83 Also outside the purview of this poetic was the purely narrative literature, partly gnomic and didactic, such as the Pañcatantra or the classic Brhatka- thã. Here the story, without embellishment, carries the interest of the reader, and the linguistic form is of less import-a simile here and there for exposition. But the Indian tradition does not consider these literary works as kāvya. Kāvya is that literary form whose interest is carried by its ideal or grammatical shape-that shape which is so exhaustively examined by the figurationists. The origins of kāvya are perhaps connected with the realization that language need not always be subservient to a utilitarian object-its reference-whether it is something as mundane as a story or something as remote as the eternal truth. Each language has a proper form, which is independent of all subject matter and which, when recognized as such, can be manipulated according to its inherent canons of excellence, thus defining its own beauty. The kāvya represents such a poetic ideal deter- mined within the formal categories and possibilities of the Sanskrit language. The alamkārašāstra represents a very minor step forward-from poetry to the conceptualization of poetry. It is an attempt to state and arrange the forms which freely used constitute poetry according to their implied ideal categories. The field of interest is thus relatively narrow, and histor- ically is quite precise: a certain modality of language, determined as beautiful in itself. Language determined by an extrinsic end may be secondarily beautiful, but its conceptualization does not properly concern the poetician. Specific references to a subject matter are of course crucial to the perception of the figure and heighten its comprehension; but a subject viewed as technically necessary is a far cry from the subject as a "great idea", a conception infusing a work of art and architectonically becoming its central issue. This option lies outside that of the kāvya, and therefore of the alamkāraśāstra. It even fails to interest later poeti- cians, whose notions of religious bhakti are to some extent more subject- oriented than the pure poetry of the kāvya. Even here the rasa communi- cated by the poem is in the last analysis a function, albeit emotional, of its form alone and not a conception imposed upon the poem; it is under- stood in all the poetics as an expressive function of language itself, hence not decisively different from the vakrokti of the ālamkārikas.

84 SCOFE OF THB GLOSSARY (VI) SCOPE OF THE GLOSSARY

The following glomary is based on those poetic texts which were composed during the formative period of Indian poetics: that period during which the definition of the figure constituted the major problem, situated be- twoca a nccemarily vague prehistory, perhaps characterized by specula- tioms on gumeldosa, and the rise in the ninth century of a poetics based on the dn. The glossary attempts to define every figure and sub-figure discawod in those terts, in accordance with methodological principles which are set forth in the following section. It may not be out of place here to indicate the reasons for restricting the scope of the glossary to the manuals of the carly poeticians excluding the many medieval writers wbo postdate the dhveni: Viśvanātha, Hemacandra, Ruyyaka, Jayadeva, Appayya Dikşita, and Jagannātha, to name the most illustrious. (a) in part our rationale can be inferred from the preceding discussion of the appropriateness of the alankdrasāstra to kāvya. To include later writers would necessitate a greatly increased referential apparatus without substantially incrensing the number of figurative categories treated. Ruyyaka, for example, treats only four figures not named by the writers we bave mduded (the rasa figures excepted). Two of them, parināma and allekhe, are in carlier texts, subvarietics of other figures. It is for his comcious cffort to arrange the figures that Ruyyaka is important (see p. 21f.). Hemacandra names none. The later writers, especially Jayadeva, Appayya, and Jagannätha, add certain figures and elaborate others on the teis of secondary dincriminations. For instance, those of Jayadeva, cooopted also by Appayya and Jagannatha, appear to be a quite late and thoroughly syncretistic attempt to rescue the guna theory in the alamkāras (praokti, lalita, praharsana). (For an inventory of the figures not dofined in thus glossary but found in later writers, see Appendix.) (b) A more crucial issue is raised by the character of the post-dhvani poctics itself. It is a thoroughly eclectic poetics, accepting the dhvani or saalogous functioos as the essence of poetic expression, yet attempting to show the dhoni in the entire range of previously elaborated anal- ytic categories-Agures, gunas, and by absence in the doas-as well as in those categories which are immediately pertinent to the dhvani. The teedency of the later poctics was to syncretism and was not, except for a few writers like Ruyyaka, devoted to questions of definition and ana- lysis The fgures in this late poetics are often sloppily defined, their cosceptualizations traditional and inconsistent. Mammata, the first of the post-dlani encyclopedists (and included in the glossary for that

SCOPE OF THE GLOSSARY reason) already shows this tendency to a remarkable degree. The figure 85

sama, first met in Mammata, may, for instance, be based on a mis- reading of an adverb in the Agnipurāna (see s.v.). Since we have been interested here in the system of the figures and problems associated with the definition of its basic categories, it appears reasonable to concentrate on those authors for whom this too was the major problem, as opposed to an encyclopedist's "completeness". (c) The almost total acceptance of the dhvani (or analogous categories) as a single constitutive principle of poetry refocussed the problems and aims of the later poetics. Accepting the argument of the Dhvanikāra that the figures, although manifesting the dhvani essentially, were not necessary to its expression, later writers, of course, do not consider the figures the central issue of poetics; their interest lies more in demonstrat- ing the dhvani in the figure than in showing the figures as a system of categories comprising poetic expression. The elaboration of figures based on rasa and its categories is illustrative of this tendency. The pre-dhvani poetics thus constitutes a discrete problem whose solution is not facilitated by considering the various transformations that the theory underwent at the hands of the dhvani school and the encyclopedists. Too much has already been done along these lines, and the result has often been complete misapprehension of the aims of the early poeticians. (d) By restricting ourselves to a group of writers having not only an ideological bond, but a historical unity, it is hoped that our study gains a certain consistency which will be useful in further studies of the tangled intellectual history of Indian poetics. The basic poetic categories, all elaborated in these writers, are found without too much fatiguing em- broidery; the manner of defining the figure and arranging it in the universe of poetic figures alone retains our attention.

(VII) METHOD OF THE GLOSSARY We are not dealing historically with individual figures, and therefore shall not consider those aspects or that information about each figure which do not serve to distinguish it from other figures. There will be. in other words, no philological account of the minute changes in defini- tion of which the figures are capable, and which has been the chief concern of most of the Western students of the figures.181 The first task has been 181 Johannes Nobel, Beiträge zür älteren Geschichte des Alamkārasāstra (Inaugural- Dissertation, Friedrich-Wilhelms-Universität, Berlin, 1911). Nobel is dealing only

INTRODUCTION 86 to give a definition, in the light of numerous examples, which will specif, the classifying ideas behind the system of figures. Since we are dealine with different authors, each of whom represents a different system or the same system in a different way, it will be important to understand in ahat sense generalization has been judged legitimate. The example, offered by the authors are of prime importance in separating essential from non- essential difference, since it is only through a comparison of each example with many possible counter-examples that the notion underlying the classification itself becomes explicit. The definitions are of interest primarily as tests whereby two examples are judged to vary in respect of some important feature. The alamkārašāstra represents an extremely close analysis of a specific subject matter-poetic utterance-and the attempt to follow out the reasoning involved in differentiating the con- cepts by an examination of the terminology only misses the point. It ignores the question of what makes an alamkāra possible: the relation of a given figurative or deviate idea to the total possible inventory of such ideas. It is only in its distinctiveness that each figure gets its full measure of significance. Let us take several examples of the kinds of problems one meets with in following this method of definition. a) In the simplest case, the usage of a term is constant and occupies an analogous place in each author's pattern of figures. Upamā is an example of this uniformity. All authors agree that simile is a comparison of two unlike things; in distinction to all other figures based upon an adjunction of unlikes, the end, or intention, of the poet is comparison. No example is offered by the writers which involves a comparisen in the sense indicated and which is not called a simile. b) Often the same name is applied to figures whose examples indicate that there is a significant difference between them. In aprastutaprasansa (or aprastutastotra), praise of something apparently irrelevant (wah a view to condemning the actual subject of the utterance) is impleo by Dandin, who takes the name literally,182 while Udbhata gives ar :astta tion which involves no moral judgement, but only a mention of soe? irrelevant (with a view to suggesting an idea which is the actra' of the utterance).183 Though the definitions themseles gne with a few figures chosen at random and has not, therefore, come to sn system of figures. It is characteristic of the subjeet, howevet, thit de !: the figures two by two, implying the possibility of assigning each heer. 182 183 Daņdin, 2.340. Udbhața, Kāvyālamkārasārasamgraha, 2.4.

METHOD OF THE GLOSSARY 87 for making this distinction (samsa means both 'praise' and 'mention'), we are obliged to do so because of a significant variation in the pattern of examples. Moreover Dandin's example contrasts with that offered for the figure vyājastuti (blame of something apparently irrelevant, with a view to praising the actual subject),184 while Udbhata intends a contrast with the figure samasokti, in which certain attributes of the (unmentioned) subject are mentioned, insofar as they are identical with those of the explicitly mentioned object.185 Because of this contrastive variation, the two versions of aprastutapraamsā must be judged different, since otherwise the distinction between the parallel figures, vyājastuti and samāsokti, would be obscured. c) In the third case, different names have been applied to the same figure. Usually this variation is purely nominal, involving no significant contrasts within the system. For example, a reciprocal comparison is termed variably anyonya ('each other'),186 upameyopamā ('comparison of the subject compared'),187 paraspara ('mutual'), 188 and ubhaya ('both').189 These terms are not met with elsewhere in the discussion of simile; all mean approximately the same thing and all are based upon exactly the same contrast within the universe of similes (statement plus inversion), contrasting with the standard upama (statement only) on the one hand, and, if it is mentioned at all, with viparyaya (inversion without statement) on the other. But in other situations this case requires argument, in that the defini- tions suggest a distinction which cannot be substantiated in the examples. A striking instance of such an intersection we find in the definitions of the terms sambhāvyamānārtha (a type of atiśayokti, defined by Vāmana as the exaggeration of an imagined quality),190 and utpādya, a type of upamā defined by Rudrata, in which the object of comparison is hypoth- esized in order to manifest a property of the subject for the purpose of expressing the actual incomparability of the subject.191 But the examples offered are parallel, and both express conditional similes of exactly the type referred to more precisely by Rudrata. It would seem a matter of

184 Daņdin, 2.343. 185 Udbhața, 2.10. 186 Daņdin, 2.18. 107 Bhamaha, 3.36-7. Note that this term will appear as upameyôpama in the Glossary. We have used the carat there to indicate vowel fusions by samdhi. 188 Agni Purāņa, 344.11. 189 Rudrața, Kāvyālamkāra, 8.9. 190 Vāmana, 4.3.10. 191 Rudrața, 8.15.

taste whether the type should be assigned to the category upama or the category atiśayokti, for an element of comparison as well as an element of exaggeration is indeed present. Since the peculiarity of the figure is best described in terms appropriate to upama, it is listed there. It will be seen that this method, as well as assuming a system of figures, defines such a system. The fact that the method works is the best proof that the figures are not haphazard lists of random verbal phenomena. Despite occasional lapses, more pronounced in Mammata and later writers, there is a consistency in the treatment not only of individual figures, but in the kinds of figures deemed important. Each figure was evidently seen to occupy a place, which was inimitable yet finely attuned to adjacent figures, in the universe of poetic discourse. We do not, of course, mean to imply that there is but one possible system of alamkāras. Indeed, this is clearly not the case, and to maintain otherwise would again deny autonomy to the different systems. d) In the most obvious case, two authors may disagree as to the relev- ance of a given figure in any system of figures. Bhamaha will not admit hetu, sūkșma, leśa, and a few others because he thinks they do not involve an element of deviation, essential to any figure.192 Dandin accepts these figures, but objects to Bhāmaha's utprekșāvayava, stating that it is only a special kind of samsrsti.193 Mammața likewise rejects raśanopamā on the grounds that iteration is an improper discrimination to apply to the figurative idea.194 We have allowed all these figures, in the terms required by the authors who accept them, because our purpose is not to criticize the scope of the figurative idea, but rather to show the develop- ment of which it is capable. e) Such disputes among different authors do not involve the definition of the figure, but simply whether that definition can be included within the idea of figuration. Other problems of relating the several systems are not as clear cut. For example, Rudrata enumerates six kinds of simile, calling one utpādya in which the object of comparison is hypoth- esized.195 Dandin, among his forty-odd kinds of simile, diseusses several which involve "hypothesis" (adbhuta, abhūta, asambhava, ete., but none of these requires specifically the hypothesis of the object per se (rathet. a transfer of property to the object, or a generalization of the objcct. or a predication of an incongruous property in the objeet).126 The type 192 Bhāmaha, 2.86. 108 Daņdin, 2.359. 104 Mammata, 134ff. This and later refs. are to serial order of topies. not 105 Rudrața, 8.15. 196 Dandin, 2.24, 38, 39.

METHOD OF THE GLOSSARY 89 noted by Rudrata, though far more general in intent (being one of six), fits very nicely into Dandin's much larger system, and we have been obliged to distinguish it from the adjacent items of that system. A certain degree of equivocation must be allowed in all such distinctions, since the ultimate level of discrimination is also the most complicated, and certain authors, as Rudrata in the present case, do not move on that level of complexity. The apparent equivocation is really, then, a case of under- elaboration, and we resolve it by requiring greater explicitness than the author himself might. f) A similar problem relates to those authors whose definitions (though not their examples) deviate systematically from the standard definitions because of an overall commitment. Vämana, the most noteworthy case, gives all the arthālamkāras as versions of simile, although this results in some spectacular limitations being put upon those figures (aprastuta- praśamsā, dīpaka, yathāsamkhya) which do not involve the terms of the simile in their standard definitions (upameya, etc.).197 In these cases, we have considered the figures to be the same as their standard counter- parts, both because, by the method of examples, they turn out to be indistinguishable, and because that aspect which would result in their being considered different (for Vämana, that they are similes) is not an issue pertaining to any particular figure, but to all figures. All questions regarding the grouping of the figures have been treated in the notes appended to each figure.

197 Vāmana, 4.3.1 ff.

POSTSCRIPT

Within the perspective of the history of Indian poetry, a case can be mate for the adequacy of the figurative poetic. As a matter of fact, eaa the poetics and the poetry together offers more hope of resolving ad- mittedly difficult problems of aesthetic interpretation thar does te arbitrary application to either the poetry or the poetic of foreign literar: and critical values or notions of aesthetic progress. But a prerecusite of this approach is that the alamkara criticism be taken seriously as criticism, as an attempt to understand a kind of poetry in its form and aim. The question of a universal aesthetic is possible only in historical terms. The false universality of the present may lead the more enthusiass critic to propose values and ideals for all time. In fact, even his eritieis: assumes, and is dated by, the literary and cultural ideas of his age. I: appears quite unfruitful, given the enormous differences in the dinenseas of time and taste, to expect Indian criticism to have explored probes whose relevance is a product of our recent past and of a difleren: spr .:. It is in the realm of its assumptions that one civilization er E:2 tradition is most irreducibly different from another. These ass color the same apparent fact, the same apparent probleri. apparent solutions, and give to them wholly novel dimenso". dimensions, if the historian-cum-critic insists on his own ! We do not intend here an extreme form of historier;' " would be as false to the Indian views of time and histen have criticized. The appearance of universality and essential aspect of the historical sequence of forms w Hegelian terms, could be said to be worked out ths: sequence. This appears to offer more hope for under genius of what was certainly one of the most brilhau. traditions.

POSTSCRIPT 91 Let us then resort, with Samuel Butler's "Presbyterian Knight and Independent Squire" to the study of the figures: For Rhetorick he could not ope His mouth, but out there flew a Trope: And when he hapned to break off I'th'middle of his speech, or cough, H'had hard words, ready to shew why, And tell what Rules he did it by.

A GLOSSARY OF INDIAN FIGURES OF SPEECH

NUMBERS AND ABBREVIATIONS USED

1. The definition of the figure. 2. References to the discussions of the figure in the various authors. The numbers in parentheses indicate examples in the text. 3. Sanskrit example, illustrative of the figure, with expository notes and translation. 4. Example from English or American literature, illustrative of the figure. 5. Discussion of the place of the figure in the system of figures and related topics. I, II, etc. are used to distinguish two or more figures with the same name. AP = Agni Purāna (last half of the 9th century). B = Bhāmaha, Kāvyâlamkāra (early 8th century). D = Daņdin, Kāvyâdarsa (first half of the 8th century). M = Mammața, Kāvyaprakāśa (late 11th century). NS = Bharata, Nātya Śāstra (perhaps 7th century). R = Rudrața, Kāvyâlamkāra (middle of the 9th century). U = Udbhața, Kāvyâlamkārasārasamgraha (early 9th century). V = Vāmana, Kāvyâlamkāravrtti (end of the 8th century). In the Glossary, the symbol * is used to indicate vowels fused through samdhi, e.g., upameyôpamā. The translations of the Sanskrit examples are intended to bring out the figure and are not necessarily complete. The glossary is organized by figures (terms named as such), in Sanskrit alphabetical order. All subfigures are treated in alphabetical order under the main figure to which they pertain. The main figures are set off in the text by centered headings.

glossaries, is: The Sanskrit alphabet, in the order traditionally adopted for dictionaries and aāiiuūrīļe o ai au

k kh g gh n̄

c ch j jh ñ th dh ņ

t th d dh n

p ph b bh m

y 1 V ș s h

atadguņa atadguņa, 'not having that thing's attribute': (1) a figure in which two things or states remain distinguishable in spite of the likelihood or the appropriateness of the one's dominant quality imposing itself upon the other. (2) M 205. (3) dhavalo'si jahavi sundara tahavi tue majjha rañjiam hiaam / rāabharie vi hiae suaha nihitto na ratto'si (Mammata: "Though you are pale, lover, my heart is made bright by you; though you have entered my heart full of passion [redness], you are not enamored [red]"). (4) "Cold-blooded, though with red your blood be graced" (Leigh Hunt). (5) This is an expected tadguna which fails to take place. It differs from nānātva atiśayôkti in that there one thing is said to be twofold, while here two things are said to be twofold; only our expectation of unity is multiplied, not the thing itself. Very few figures involve in their definition an element of expecta- tion, though most in some way exploit it.

atiśaya atiśaya, 'excess': (1) one of the four general categories into which arthâlamkāra are grouped. (2) R 7.9, 9.1 (5) See śleșa; cf. vāstava, aupamya.

atiśayôkti atiśayôkti, 'expression involving an exaggeration': (1) the exaggeration of a quality or attribute in a characteristic way, so as to suggest pre-eminence in its subject; hyperbole. (2) B 2.81-85. D) 2.214-20. V 4.3.10, U 2.11, AP 344.26, M 153. (3) mallikāmālabhāriņvah sarvângīņârdracandanāḥ / kşaumavatyo na lakşyante jyorsnāyām abhisārikāh (Dandin; the whiteness of the girls' dresses is exaggerated

98 GLOSSARY to the point of making them invisible in the moonlight: "Wearing garlands of white jasmine and clothes of linen, their limbs moist with sandal paste, the trysting ladies are hidden in the moonlight"). (4) "Pardon, once more; if you are going to load anything more onto that statement, you want to get a couple of lighters and tow the rest, because it's drawing all the water there is in the river already: stick to facts ... "(Mark Twain; the "weight" of the statement is exaggerated to the point of threatening the seaworthiness of the river packet). (5) Many types of upamā are based upon exaggerations of the common property of one sort or another; these distortions are, however, all subservient to the end of comparison: in hyperbole there is no end other than the magnification of the subject itself. Similarly, in utprekșā, an attribute is figuratively associated with a subject, but the distortion lies in that unlikely association, not in the representation of the attribute itself. In hyperbole, the attribute in its literal form should be naturally inherent in the given subject; it is only its unworldly (lokâtikrāntagocara) or preposterous exten- sion that makes it figurative. Different writers have distinguished different characteristic ex- aggerations. The most common (Bhāmaha, Daņdin, Vāmana, Udbhata) is that of two objects in the presence of one another being made indistinguishable by the property which both share (cf. adhyavasāna). Dandin recognizes the exaggeration of size to the point of ultimate smallness (samśaya), as well as ultimate greatness (ādhikya). Udbhata and Mammata allow the inversion of the first type, where the same subject is considered multiple because of differ- ent qualities (cf. nānātva). In addition, three types are based upon a figurative transference of an attribute from one subject to another (cf. sambhāvyamānârtha) and are distinguishable only with great subtlety from other figures variously defined. Lastly, there is ex- aggeration of a quality by attributing to it the nature of a cause in respect of its own cause (kāryakāraņapaurvâparyaviparyaya). adhyavasāna, 'determination': (1) a type of atiśayôkti in which one thing is characterized as another so as to exaggerate a quality which they in some degree share. (2) M 153. (3) kamalam anambhasi kamale ca kuvalaye tāni kanakalatikāyām / sā ca sukumārasubhagēty utpātape- ramparā kêyam (Mammata: "A lotus grows where no water is: on this lotus are two buds; and the lotus with its buds grows on J golden vine: Who can she be, this concatenation of wonders? Call her fortunate and lovely"). (4) "She seemed to belong rightly to 2

GLOSSARY 99 madrigal-to require viewing through rhyme and harmony" (Thomas Hardy). (5) Here the speaker is describing the girl as though she were a lotus, and in Hardy's example, the girl is described as though she were a song. The point of "indistinguishability" seems to cross the subtle boundary of conscious rapprochement and, as such, intrudes upon the domain of samāsôkti, an abbreviated metaphor in which the subject is not mentioned. I think such cases must be taken as examples of Mammata's sloppy encyclopedism. ananyatva, 'identification': (1) a type of atiśayôkti in which two qualities or attributes, though in fact contrary, are considered indistin- guishable. (2) U 2.12. (3) sa dadarśômām ... tapastejahsphuritayā nijalāvaņyasampadā / kršām apy akrśām eva drśyamānām asamśayam (Udbhata; though Umā is emaciated by her fasting, she appears full blown because of the beauty which her penance imparts: "He saw Umā ... wasted away but appearing full blown in the wealth of beauty born of her ascetic power"). (4) "His departure gave Catherine the first experimental conviction that a loss may sometimes be a gain" (Jane Austen). (5) In these examples, two qualities are mentioned; in the Sanskrit example given under atiśayôkti, two objects are "rendered indistinguishable". It would seem that this latter case is most typical of atiśayôkti, given as it is by most of the writers whether they allow subtypes or not (Bhāmaha, Dandin. Vāmana, Udbhața, Mammața). This figure resembles an exaggerated simile (cf. cațu upamā), but it should be noticed that the qualities compared here are contraries (krśatvam-akrśatvam, "loss-gain"). The aspect of similitude is an incidental consequence of a fortiori premises. One may ask how the present examples differ from rūpaka (metaphor). First, there can be no metaphorical identification of qualities; second, metaphor need not be based on the identification of items somehow contrary. ädhikya, 'superabundance': (1) a type of atiśayôkti in which a quality or attribute is quantitatively exaggerated out of all proportion. (2) D 2.219. (3) aho viśālam bhūpāla bhuvanatritayôdaram māti mātim aśakyo'pi yaśorāśir yad atra te (Dandin: "The extent of your fame. itself measureless, comprehends, O King, the prosperity of the three worlds"). (4) "I will not deccive you; he told me such a monstrous lie once that it swelled my left ear up, and spread it so that I was actually not able to sec around it; it remained so for months, and people came miles to sec me fan myself with it" (Mark Twain). (5) Cf. samśaya, where the attribute is minimized out of all propor-

100 GLOSSARY

tion. This figure is not named by Dandin, but by the commentator; however, it evidently pairs with samsaya and is cncompassed by the "ādi" ('etc.') in 2.216. kāryakāraņapaurvâparyaviparyaya, 'inversion of the sequential relation- ship of cause and effect': (1) a type of atiśayôkti in which the cxag- geration of a quality or attribute is accomplished by expressing it as the cause of that which in the order of nature is its cause. (2) U 2.13, M 153. (3) manye ca nipatanty asyāh kațâkșā dikșu prsthatah| prāyeņâgre tu gacchanti smarabāņaparamparāh (Udbhața; usually the girl's love-lorn glances are the cause of Cupid's shooting the bow; here Cupid beats Uma to the punch-thus expressing, accord- ing to the commentary, how quickly Siva took the tumble: "I think that first the arrows of the Love-God were shot, next her sidelong glances were scattered in the four directions"). (4) "Was it for this that I might Myra see / Washing the water with her beauties white?" (Fulke Greville). (5) Pūrva alamkāra differs from the present case in two respects: There the inversion of the sequential relation- ship is not subordinated to any other consideration, such as the exaggeration of a quality, and temporal inversion is expressed generally, not limited to the one case of cause-effect (not everything which precedes is a cause). nānātva, 'variety': (1) a type of atiśayôkti in which a quality or attribute is exaggerated by considering it multiple, though it is in fact one. (2) U 2.12, M 153. (3) acintayac ca bhagavān aho nu ramaņīyatā tapasâsyāḥ krtânyatvam kaumārād yena lakșyate (Udbhața: "The Lord thought: 'Ay, such loveliness comes from her penance, yet how different is the beauty from that of her youth!'"). (4) "Any customer can have a car painted any color that he wants so long as it is black" (Henry Ford; a many-sided blackness). (5) The present instanee differs from vyatireka alamkāra in two respects: The differentiation attaches to a single quality of a single subject, not to a commen quality of two subjects; hence, the element of comparison is lack.ng. Compare adhyavasāna, or ananyatva, where two qualities are co :- sidered uniform. nirņaya, 'conclusion': (1) a type of samsaya atisayokti in which thea ieae doubt is resolved. (2) D 2.218. (3) nirņetum sakyam astiti tava nitambini / anyathânupapattyaîva payodharabharasteDe "One can decide that your waist indeed is there, O lovehs. otherwise could the weight of your breasts be supporte"ì. Nature H-y's Clay was blending, / Uncertain whai her wor.

GLOSSARY end in, / Whether in female or in male, / A Pin dropped in, and 101

turned the Scale" (Anon.). (5) The name is taken from the com- mentary; see ādhikya. saņśaya, 'doubt': (1) a type of atiśayôkti in which a quality or attribute is minimized to the point where doubt can be entertained as to its existence or nature. (2) D 2.216 (217). (3) stanayor jaghanasyápi madhye madhyam priye tava / asti nâstîti sandeho na me'dyápi nivartate (Dandin: "The narrow waist that intervenes between your breasts and buttocks, O lovely, is it there or is it not? My mind cannot decide this doubt"). (4) "They have yarns ... of the runt so teeny-weeny it takes two men and a boy to see him" (Carl Sandburg). (5) This in the inverse of exaggeration properly speaking, but as it repre- sents just as great a deviation from the normal, Dandin systematically includes it here. Cf. ädhikya. The point of the example is the small- ness of the waist, not the doubt, which is only a psychologically ap- propriate adjunct; hence, this figure differs from samśaya alamkāra. sambhavâsambhava, 'possible, impossible': (1) two types of hyperbole. (2) AP 344.26. (3) (4) No examples. (5) Another one of the mysteries of the Agni Purāņa. saņbhāvyamānârtha, 'whose meaning is imagined': (1) same as utpādya upamā. (2) B 2.81 (83), V 4.3.10, U 2.12, M 153. (5) This figure is also called kalpana by Mammața. Vāmana and Bhāmaha give it no name, but their two examples fit clearly into this category and adhyavasāna. The figure is recognized by six writers: The present four consider it a kind of hyperbole, but Dandin and Rudrata discuss it under simile. Inasmuch as we have supposition of the transferability of a quality from one subject to another, there is a cer- tain exaggeration attendant upon such an irregularly proposed quality. However, the end in all cases cited is comparison, and hyperbole is only a means to that end. Though classifications are by no means systematic, the end does generally serve as the genus. An example of a transfer of property which does not serve the end of comparison would be: "To us the hills shall lend / Their firmness and their calm" (Henry Timrod). Bhamaha's example comes closest, but it still seems to be a simile: "apām yadi tvak chithilā cyutā svāt phaņinām iva | tadā śuklâmśukāni syur angeşv ambhasi voșitām" (2.83; the "skin" [i.e., foam] shed by the waters is transferred to the women as clothes: 'If the loose skin of the waters should fall away, like the skin of snakes, then it would serve as white cloth for covering the bodies of the women in the river').

102 GLOSSARY

adbika adhika (I), 'superabundant': (1) a figure wherein two contraries are said to proceed from the same cause. (2) R 9.26 (27). (3) muñcati vāri payodo jvalantam analam ca yat tad āścaryam / udapadyata nīranidher vișam amrtam cêti tac citram (Rudrața; a reference to the creation myth wherein the primeval ocean gave forth both deadly poison and the Gods' sustinence: "It is amazing that the clouds release both blazing fire and water; that both poison and nectar emerge from the watery sea"). (4) "The long, winding intricate sentences, with their vast burden of subtle and complicated qualifications, befogged the mind like clouds, and like clouds, too, dropped thunderbolts" (Lytton Strachey). adhika (II): (1) a figure wherein a thing is said to exceed or surpass in size or grandeur its own basis or container. (2) R 9.28 (29), M 195. (3) aho višālam bhūpāla bhuvanatritayôdaram / māti mātum aakyo'pi yaśorāśir yad atra te (Mammata; the king's glory cannot be con- tained even by the three worlds; the example is also found in Dandin 2.219 for the term ādhikya atiśayôkti, q. v.). (4) " ... warned me my watch was relieved. It could not have lasted more than two hours: many a week has seemed shorter" (Charlotte Bronte; two hours exceed in duration the thing of which two hours is a part). (5) Mammata in his definition allows for the possibility that the con- tainer exceeds the term predicated upon it, but both his examples show only the reverse (the present case).

anuprāsa anuprāsa, 'throwing after': (1) alliteration. (2) B 2.5-8. D 1.52-59. V 4.1.8-10, U 1.3-10, AP 343.1-11, R 2.18-32, M 104-16. (3) kin tayā cintayā kānte nitāntêti (Bhāmaha: "O lovely. why are you afflicted with doubts?"). (5) Anupräsa is treated by all the writers except Bharata, but Dandin considers it an aspect of madhura gata rather than a figure. The varieties of alliteration considered are (a) repetition of phonetic features (Dandin: sec varnavrri, note): (b) repetition of phonemes or phoneme clusters (parusa, upanegerk .. grāmya, madhura, lalita, praudha, bhadra): (e) variation of vowels within similar consonant strings and vice versa (cheka), and (d) repetition of words or morphemes (lata). Alliteration is carefully distinguished from yamaka (cadence), in that the occurrence of the alliterated elements is not predetermined in verse or verse part. It is. in other words, the phonemes or phonetic features that are being

GLOSSARY 103 repeated, and not verses or verse parts. As stated sub voce, the critical case is that of lāta anuprāsa. While the concept anuprāsa itself is subject to little dispute, various writers distinguish different kinds which are obviously designed to produce different effects on the ear: one melodious, one effeminate, one vigorous, and so on. For this reason, the subject of alliteration is closely tied to the dis- cussion of the different styles (rīti, guna), and various writers (Dandin, Rudrata) attempt to specify the stylistic limits of the different alliterations. Mammata attempts to equate style and alliteration (see vrtti). Anandavardhana, of course, wants to view the questions of style and alliteration as attempts, however partial, by earlier writers to come to grips with the problem of mood (rasa) and the subordination of all discrete elements in the composition to it. Despite these extrinsic differences of opinion, the importance of alliteration in poetry was never questioned, provided that its use corresponded to the effect desired. upanāgarika (perhaps a Prakrit dialect): (1) a type of alliteration in which figure prominently clusters of identical stops (kk, tt) and clus- ters of stops with homorganic nasal preceding (ūk, nt). (2) U 1.5, M 108. (3) sāndrâravindavrndôtthamakarandâmbubindubhiḥ / syan- dibhih sundarasyandam nanditêndindirā kvacit (Udbhața: "Some- where a bumblebee is delighted by the flowing drops of liquid honey from thick clusters of white lotuses"). (5) Upanāgarika resembles madhura anuprāsa of the Agni Purāņa and Rudrața. The term may mean "cultured". Cf. grāmya, to which it is opposed. komala, 'soft': (1) same as grāmya anuprāsa. (2) M 110. grāmya, 'common': (1) a type of alliteration characterized by the absence of clusters and the predominance of liquids and nasals. (2) B 2.6, U 1.6, M 110. (3) kelilolâlimālānām kalaiḥ kolāhalaiḥ kvacit / kurvatī kānanârūdhaśrīnūpuraravabhramam (Udbhața: "Sometimes accompanied by the soft humming of the bee swarms, playfully restless, she simulates the maddening sound of the anklets of Śri wandering in the forest"). (5) Bhämaha apparently considers this type defective or vulgar, but the other two authors allow it as one of the five legitimate types. It is probably to be opposed to upanā- garika, which may mean 'cultured or citified', as opposed to 'rustic. villageois'. Compare Bhāmaha's example "kim tayā cintayā kānte nitāntā" with that offered under upanāgarika (which term Bhāmaha does not use). Mammața calls this figure komala. cheka, 'clever': (1) a type of alliteration characterized by metathetie

104 GLOSSARY

variation of consonants and vocalic substitutions. (2) U 1.3, M 106. (3) sa devo divasān ninye tasmiñ šailêndrakandare / garişthagoșthī- prathamaih pramathaih paryupāsitah (Udbhața: metathesis, as of th-m to m-th, and substitutions, as of i and a for e and o: "The God Śiva spends his days in this cave of the high Himālaya, served by his attendants, principal among the great assemblies"). (5) Both authors distinguish this type from alliteration properly speaking, inasmuch as its effect depends upon variation rather than repetition. parușa, 'harsh': (1) a type of alliteration characterized by a predominance of sibilants and of clusters involving sibilants or "r". (2) U 1.4, AP 343.6-11, R 2.26-28, M 109. (3) lipsūn sarvān so'ntarbrahmôdyair brāhmaņair vrtah paśyan | jihrety agarhyabarhihśesaśayah koșaśūnyah san (Rudrata; -ps-, -rbr-, -hm- and many single sibilants, including the gutteral: "Surrounded by Brahmins who have penetrated the Veda, he sits, watching avaricious people. He is ashamed deep within himself, for he has abandoned his wealth; all that remains to him is a bed of blameless feathers"). (5) The four writers who distinguish five types of alliteration agree only on the name of this one, though two other of the five types seem to be comparable (cf. upanāgarika and madhura, grāmya and lalita). Parușa is said to contribute to ojas guna, but this is a late attempt to rationalize two unrelated systems. Dandin treats all alliterations as aspects of mādhurya guņa. pādânuprāsa, 'foot-alliteration': (1) same as lāța anuprāsa. (2) V 4.1.10. praudha, 'proud': (1) a type of alliteration characterized by clusters of "r" followed by "y" or "n" or any stop except cerebrals or nasals. and clusters of "t" with "p" or "k", (2) AP 343.5, R 2.24-25. (3) kāryâkāryam anāryair unmārganirargalair galanmatibhiḥ / nâkarņyate vikarņair yuktôktibhir uktam uktam api (Rudrața; as -ry-, -rg -. -rn -. -kt -: "The lowborn, heedless ones, who unfettered tread the paths ef unrighteousness as though they had lost their minds. do not heed their duties or proscriptions, even though they be spoken by men of sage counsel"). (5) According to Rudrata's commentator. this anuprāsa is called by others ojas, which was originally detined by Dandin as a stylistic quality consisting in the use of long compounds. Praudha has no apparent counterpart in the five alliterations of Udbhața and Mammața. See parusa. bhadra, 'pleasant': (1) a type of alliteration characterized by the pre- dominance of unvoiced gutterals and cerebrals and by the absence of clustering. (2) AP 343.6, R 2.29-31. (3) utkatakarikaratataļa-

GLOSSARY sphuțapāțanasupațukoțibhih kuțilaih / khele'pi na khalu nakharair 105

ullikhati harih kharair ākhum (Rudrața: "Not in play does the lion rip apart the rat with his hard, curved claws whose tips are quite sharp from evident tearing into the tough hide of elephants' jaws"). (5) The definition is inferred from the example; Rudrata says baldly that this style of alliteration employs "what is left"-the consonants and clusters not used in the other four-specifying that whatever clusters are used must be "agreeable to the ear". madhura, 'lovely': (1) a type of alliteration characterized by clusters of stop following homorganic nasal, double "1" and "r" and "n" in light syllables. (2) AP 343.3, R 2.20-23. (3) bhana taruņi ramaņa- mandiram ānandasyandisundarêndumukhi / yadi sallīlôllāpini gacchasi tatkim tvadīyam me | anaņurananmaņimekhalam aviratašiñjānamañ- jumañjīram | parisaraņam aruņacaraņe raņaraņakam akāraņam kurute (Rudrața: "Tell me, gentle lady with face lovely as the moon steeped in joy, if indeed you are going, sweetly murmuring of love, to the home of your lover, then why does your passing here, feet dripping with lac, with necklaces jangling and anklets sounding incessantly sweet, work in my soul this needless desire?"). (5) This figure resembles upanāgarika anuprāsa in its clusters with homorganic nasal. Rudrața gives rules for the proper use of this alliteration, saying that the quality of "loveliness" will be lost if the "l" is used more than two or three times and that the clusters of stops should not exceed five. The scope of this rule is not specified, but it is probably the śloka. Rudrata lays stress on the importance of observing the proprieties in all five types of alliteration (2.32). lalita, 'gay': (1) a type of alliteration characterized by the unclustered letters "dh", "gh", "gh", "r", "s", and "l" in light syllables. (2) AP 343.4-5, R 2.29-30. (3) malayânilalalanôllalamadakalakalakaņ- țhakalakalalalāmaḥ madhuramadhuvidhuramadhupo madhur ayam adhunā dhinoti dharām (Rudrața: "The spring now afflicts the earth; bees are helpless from drinking sweet honey; the southern wind is amorous with the arguments of kokila birds, muted with drink"). (5) As the example shows, the criteria are permissive rather than obligatory: in the first half-śloka, the "I" is principally employed; in the second, the "dh", which is only to say that the letters given may be employed in a context of unclustered, short syllables. The figure resembles grāmya. lața (lāțīya) (a region): (1) the repetition within the same verse of a word or words having the same meaning but, through the context,

GLOSSARY differing in acceptation. (2) B 2.8, U 1.8-10, M 112-16, V 4.1.10. (3) drştim drstisukhām dhehi candraś candramukhôditah (Bhāmaha; candramukhā is apparently a vocative despite the ending, or it represents secondary sandhi: "Let us sec your face, lovely-to-sec; the moon, moon-face, is risen"). (4) "It was the same rounded, pouting, childish prettiness, but with all love and belief in love departed from it-the sadder for its beauty, like that wondrous Medusa face, with the passionate, passionless lips" (George Eliot). (5) Udbhata and Mammata give an elaborate classification of this figure according to whether the word repeated follows immediately (as here) or is placed at the beginning or the end of the half-verse; similarly, they distinguish words free (having a case termination) from words bound (in compound). Mammata and Vāmana (who calls the figure pādânuprāsa) allow the repetition of the entire half-śloka, provided that the words in both halves are the same as: "yasya na savidhe dayitā davadahanas tuhinadīdhitis tasya / yasya ca savidhe dayitā davadahanas tuhinadīdhitis tasya" (Mammața; in the first half, dava- is attributive to tuhina-, in the second hall, just the reverse; "For him whose beloved is absent, the cool-rayed moon is burning fire; for him whose beloved is present, the burning fire [of the sun] is cool-rayed"). In this case, the alliteration has become for all intents and purposes a yamaka, except that the indi- vidual words are taken as the same words in both utterances, instead of splitting the utterances differently. The figure lātânuprāsa thus occupies the mid-position between alliteration and cadence, differing from the former in its concern with words rather than phonemes, and from the latter in its concern with meaning rather than phonemic sequence. Cf. āvrtti. varņânuprāsa, 'letter-alliteration': (1) same as varņâvrtti. (2) V 4.1.9. varņâvrtti, 'letter-repetition': (1) alliteration. (2) D 1.55, V 4.1.9. (3) candre śarannisôttamse kundastavakavibhrame | indranīlanibhant lakşma samdadhāty anilah [sic] śriyam (Dandin; we prefer the alinah of D. T. Tatacharya and most other Indian editors: "Its marks, dark as sapphires, give the beauty of the bee swarm to the ornament of the autumn night-the full moon, lovely as the jasmine bud"). (5) In Dandin, this is anuprasa in the narrow sense, distinguished from a kind of semi-alliteration in which only phonetie features. such as dentality or gutturalness, are repeated: for evample: cw rājā yadā lakşmīm prāptavān brāhmaņapriyah | tatah prabhri dharmasya loke'sminn utsavo'bhavat ("as soon as that king, beloved

GLOSSARY of Brahmins, attained prosperity, there was a festival of righteousness 107

in the world") where the "s" of esa and the "r" of rāja are both cerebrals, the "j" and following "y" are palatals, "d" and "l" are dentals, and so on. Vāmana, however, distinguishes varnânuprāsa from pādânuprāsa, or the repetition of metrical units (feet); in this context, also, it amounts to alliteration in the usual sense: repetition of identical phonemes in adjacent syllables. Anuprāsa, according to Dandin, consists in observing the mean; the effect is lost if the repeated phoneme is too far away (1.58), or if the phrase is too broken by harsh junctures (1.59). The repetitions must be close enough, but not too close, within these two limits. vrtti, 'mode': (1) a word applied to some or all of the kinds of alliteration. (2) R 2.19, M 105-107. (5) The vrtti is an old element of dramatic theory, mentioned in Bharata, which seems to signify the basic context of the play insofar as it determines a style of representation, similar to Shakespeare's "tragical-comical-historical-pastoral". The term thus has little relevance to poetics and the early writers ignore it. Rudrata, however, uses the word in a neutral sense to mean the five kinds of alliteration taken individually (vrtti literally signifies only 'existence' or 'specific mode of existence') as the modes of alliteration. At the same time, the dhvani theorists were examining the older vocabulary in the light of their novel doctrine, and with their general disposition to belittle or collapse such extrinsic distinc- tions, vrtti was lumped together with other stylistic concepts of the older writers such as guna and rīti (see Dhvanyāloka 3.33; Ānandav- ardhana on the Dhvanyāloka, p. 182). Ānandavardhana says that the vrtti of Bharata is a function of the meaning (vācya), while that of other writers is a function of the outward shape of the words (vācaka). By the latter, he apparently signifies the kinds of allitera- tion as distinguished by Udbhata (upanāgarikâdi). Out of this confusion, Mammata, who everywhere attempts to reconcile the views of the dhvani theorists with the older doctrines, propounds the novel view that the three kinds of alliteration involving phonemic repetition (that is, excepting cheka and lāta) are to be called vrtti. and that these three are the equivalents of the three ritis, or styles of diction, proposed by Vämana, which originally referred to the entire context of word and meaning ("ornate", "limpid". "intense", etc.). but very little else can be expected of Mammata, who represents the

yāloka, p. 6. worst of the syncretistic tendency. Cf. Abhinavagupta on the Dhvan-

108 GLOSSARY

anumāna anumana, 'inference': (1) a figure in which an infcrence is explicitly formu- lated. (2) R 7.56-63, M 182. (3) sāvajñam āgamişyan nūnam patito'si pādayos tasyāḥ / katham anyathā lalāțe yāvakarasatilakapanktir iyam? (Rudrata; reference is to the painted toenails of the beloved: "You must have fallen at her feet, having to return so contemptibly: how else would that row of red lac spots appear on your brow!"). (4) "Scylla is toothlesse; yet when she was young, / She had both tooth enough, and too much tongue: / What should I now of tooth lesse Scylla say? / But that her tongue hath worne her teeth away" (Anon.). (5) The cause (sādhaka) may be inferred from the effect (sādhya), or vice versa; it is essential that the term inferred be paroksa-in some way not obvious. In both our examples, the cause is inferred. The following lines from Somerset Maugham show inference of the effect: "As I walked along the winding road ... I mused upon what I should say. Do they not tell us that style is the art of omission? If that is so, I should certainly write a very pretty piece". In such instances, the effect is usually placed in future time. This figure differs from hetu alamkāra as the active differs from the passive: in the latter figure, a relation of cause-effect is described; in the former, it is used to secure intelligence of one or the other term so related. It is curious that Mammata should reject hetu while accepting anumāna, as the ground of exclusion he advances for the one should apply a fortiori to the other: no figurative usage need be present. Rudrata distinguishes several types which are the equivalents of Dandin's three kinds of hetu: dūrakārya, sahaja. and kāryânantaraja. Rudrata's own version of hetu has no subtypes.

anyôkti, 'saying something else': (1) a figure in which the real subject anyôkti

of comparison is suggested by explicit description of the obiect. where, nevertheless, the two compared terms have no commen property, but only a mode action in common. (2) R S. 4 (5) (3) muktvā salīlahamsam vikasitakamalôjjvalam sarah sarasd"; bakalulitajalam palvalam abhilaşasi sakhe na hamso'si (Rudrata: "Abandoning this pleasant lake with its swans and lotus blooms, you long for the forest pool rough from the flight of cranes: tet. friend, you are no swan"). (4) " ... the men and women who i ! hundred different ways were laboring, as William Allen Wiuite s .! t !.

GLOSSARY to give the underdog a better kennel" (Frederick Lewis Allen). (5) 109

The girl and the swan (as the lover and the lake) share no common property (guna) in the eyes of the Indian aestheticians, which is only to say that the basis of the comparison is to be sought in a verb, in an action (kriyā), rather than in a qualification; cf. vākyártha upamā. Similarly, in the English example, the principal analogy is drawn between the two acts of uplift, though the similarity between the dog and the lower classes is perhaps more vivid than that between the lover and the lake.

anyonya anyonya, 'reciprocal': (1) a figure wherein two things are said to be reciprocally cause and effect. (2) R 7.91 (92), M 187. (3) rūpam yauvanalakşmyā yauvanam api rūpasampadas tasyāh | anyonyam alamkaraņam vibhāti śaradindusundaryāh (Rudrața: "Her beauty is ornamented by her youth; her youth is heightened by her beauty; she is as lovely as the autumn moon"). (4) "The Devil, having nothing else to do, / Went off to tempt My Lady Poltagrue. / My Lady, tempted by a private whim, / To his extreme annoyance, tempted him" (Hilaire Belloc). (5) The reciprocity of cause and effect is the same as being mutually conditioned.

apahnuti apahnuti (I), 'denial': (1) a figure in which the object of comparison is affirmed in place of the subject of comparison. (2) B 3.20 (21). V 4.3.5. (3) nêyam virauti bhrngâlī madena mukharā muhuh | ayam ākrşyamāņasya kandarpadhanuşo dhvaniḥ (Bhāmaha: "It is not a swarm of bees, humming incessantly of honey; it is the sound of the Love-hunter's bow being drawn"). (4) "And there is not a whisper on the air / Of any living voice but one so far / That I can hear it only as a bar / Of lost, imperial music, played when fair And angel fingers wove, and unaware, / Dead leaves to garlands where no roses are" (E. A. Robinson; that is not a whisper. that is music). (5) Cf. tattvâpahava rūpaka. apahnuti (II): (1) a figure in which an essential property of the subject is denied and portrayed otherwise; irony of qualification. (2) D 2.304- 309, U 5.3, AP 345.18, M 146. (3) na pañceşuh smaras tasya sahasram patrinām (Dandin: "The God of Love is not possessed of five arrows; indeed he has a thousand"). (4) "Because these wings are no longer wings to fly / But merely vans to beat the air" (T. S. Eliot). (5)

110 GLOSSARY

Apahnuti is a figure found in all the writers after Bhāmaha, but no unanimity as to its acceptation is discernable. It is related on the one hand (by Bhāmaha and Vāmana) to the tattvapahnava rūpaka (which figure appears only in Dandin), and on the other to the mata alamkāra (as here), wherein the interest attaches to the mis- representation of the subject in a certain way. Subtypes are distin- guished as to the intellective basis (opinion, necessity) of that misrepresentation (Dandin) and as to the mode of its affirmation (mere attribution, transformation: Mammata). See vişaya, svarūpa, śābdī, ārthī. apahnuti (III): (1) a figure in which the subject of comparison is portrayed as possessing a quality which in nature belongs to the object of comparison. (2) R 8.57 (58). (3) navabisakisalayakomalasakalâva- yavā vilāsinī saîșā / ānandayati janānām nayanāni sitâmśulekhêva (Rudrata: "A lovely, wanton lady with limbs as soft as new lotus shoots delights the eyes of men just like the cool-rayed crescent"). (4) "Ask not the Cause, why sullen Spring / So long delays her flow'rs to bear; / Why warbling birds forget to sing, / And Winter Storms invert the year? / Chloris is gone: and fate provides / To make it spring, where she resides" (John Dryden). (5) This figure is just the reverse of adbhuta upamā, where a striking property of the subject is transferred to the object. Cf. asambhava upamā, where the quality is transferred from the subject to the object. ārthi, 'implied': (1) a type of apahnuti in which the misrepresentation is expressed via a transformation of the subject in question. (2) M 146C. (3) amuşmiml lāvanyâmṛtasarasi nūnam mrgadrśah smarah šarvaplu- stah prthujaghanabhāge nipatitah | yad angângārāņām praśamapiśunā nābhikuhare śikhādhūmasyêyam pariņamati româvalivapuh (Mamma- ța: "The God of Love, whose body was consumed [in the fire of] Śiva's [wrath], has now taken up his abode between that doe-eyed maiden's broad thighs-veritable streams of beauty's nectar. See how the thin line of hair on her navel has assumed the form of a wisp of smoke: thus the smouldering coals of Love's body are being extinguished"). (4) "Full fathom five thy father lies; / Of his bones are coral made: Those are pearls that were his eyes: / Nothing of him that doth fade. But doth suffer a sea-change / Into something rich and strange" (Shakespeare). (5) In śābdī, the misrepresentation is accomplished by simple denial and affirmation: hence it is called "explicit". vişaya, 'circumstance': (1) a type of apahnuti in which the misrepresentation is stated to depend upon a difference in point of view or condition.

GLOSSARY (2) D 2.306 (305). (3) candanam candrikā mando gandhavāhaś ca 111

atşişt | stya agnimayī sīstir mayi šītā parān prati (Daņdin: "For me, these things the sandal paste, the moonlight, and the softly blowing southern wind-are made of fire; others may think them cool"). (4) "Those who have crossed / With direct eyes, to death's other Kingdom / Remember us-if at all-not as lost / Viol-at soul, but only / As the hollow men / The stuffed men" (T. S. Eliot). (5) In sarüpa, the misrepresentation is a function of the nature of the thing itself; that is, it amounts to a reinterpretation of that thing. e, titeral': (1) a type of apahnuti in which the misrepresentation is a function of denial and contrary affirmation. (2) M 146C. (3) mita prāzalbhyam pariņatarucah śailatanaye kalamko natvāyam vileeti šasāmkasya vapuşi | amuşyėyam manye vigaladamrtasyan- Atitre [sic] iti śrântā šete rajaniramaņī gādham urasi (Mammața: That is no mere spot which has appeared on the moon's full, brilliant form, O Pårvati; rather I think the courtesan of the Night bes exhausted in tight embrace on his broad chest cool from the Sowing stream of nectar"). (4) "Stay, O sweet, and do not rise! / The light that shines comes from thine eyes: / The day breaks not: it is my beart, / Because that you and I must part" (John Donne; in thit cxample both the daylight and daybreak are misrepresented, the former as the light in ber eyes, the latter in the weak pun. Both are literal, the latter almost too literal). (5) Cf. arthi. swi, 'aatural': (1) a type of apahnuti in which the misrepresentation a expressed as a reinterpretation of the nature of the thing itself. (2) D 2.308 (307). (3) amrtasyandikiranaś candramā nāmato matah | aya evåyam arthátmå vişanişyandididhitih (Dandin; the moon is different to the rejected lover: "The moon is generally considered to have rays of flowing nectar; but it has another soul as well, for its brilliance is steeped in poison"). (4) "Death, be not proud, though some have called thee ! Mighty and dreadful, for thou art not so" (John Donne). (5) Cf. vişaya.

dama (I), 'mentioning the irrelevant': (1) a figure in which the real but implicit subject matter is obliquely referred to by means of an explicit, but apparently irrelevant, subject which, however, stands in a specific relationship to the former. (2) B 3.28 (29), U S.8, AP 345.16, M 151. (3) prīņitapraņayi svādu kāle pariņatam

112 GLOSSARY

bahu / vinā purușakāreņa phalam pašyata šākhinām (Bhāmaha; a courtier is referring to the bounteousness of the king: "Regard the fruit of the trees, pleasing to those who seek it, sweet and ripe in its own time, grown heavy without the aid of man"). (4) "O powerful western fallen star! / O shades of night-O moody, tearful night! / O great star disappear'd-O the black murk that hides the star! / O cruel hands that hold me powerless-O helpless soul of me! / O harsh surrounding cloud that will not free my soul" (Walt Whitman; referring to the death of President Lincoln). (5) This figure is usually distinguished from samāsôkti; for a discussion of its relation to that figure, see samāsôkti. It is apparently the same as the figure paryāya of Rudrata, who does not recognize aprastutapraśamsā. It is also called aprastutastotra in Dandin and the Agni Purāņa. adhyāropa, 'figurative attribution': (1) a type of aprastutapraśamsā in which qualities are attributed to the explicit subject which can apply literally only to the implicit subject. (2) M 152C. (3) kas tvam bhoḥ-kathayāmi daivahatakam mām viddhi śākhôțakam / -vairā- gyād iva vakși sādhu viditam kasmād idam kathyate | -vāmenâtra vațas tam adhvagajanaḥ sarvâtmanā sevate / na cchāyâpi parôpakā- rakarane mārgasthitasyâpi me (Mammața; the tree to which the courtier likens himself is literally incapable of speech: "'Who might you be?' 'I will tell you: think of me as a twisted and accursed śākhoța tree!' 'You seem to be speaking in a spirit of indifference!' 'Well said!' 'Why do you describe yourself thus?' 'On the left over there is a banyan tree which travellers resort to with great relief. But I have no shade to serve others with, though I too grow along the road'"). (4) "Of the Folly of Loving when the Season of Love is past: Ye old mule! that think yourself so fair, / Leave off with craft and beauty to repair" (Thomas Wyatt). (5) Mammata divides intimation in two ways: by considering the relation of the two sub- jects, and by the relation of the qualities expressed to their subjects. This is an example of the latter topic. For an example of intimatien in which the qualities are not thus attributed to the expressed subject. see aprastutapraśamsā. kārya, 'effect': (1) a type of aprastutapraśamsa in which the real subject is an effect and is intimated through a description of its cause. () M 152. (3) yātāh kin na milanti sundari punaś cintā tvayā matkrs kāryā nitarām krśâsi kathayaty evamı sabāşpe mayi lajjamanthera :- rakeņa nipatatpītâśruņā cakşuşā | drstvā mām hasitena bhevpe- ņôtsāhas tayā sūcita (Amaru, quoted by Mammața; a lover thus

GLOSSARY describes the cause of his early return from a journey: "Those who have gone to another country, why should they not return? Beloved, 113

you must grieve for me no longer; you have grown so thin! Even while I speak to you in tears, you look at me with eyes downcast with shame and full of pale tears, while your hysterical laughter surely portends approaching death!"). (4) "With how sad steps, O Moon! thou climb'st the skies! / How silently, and with how wan a face! / What! may it be, that even in heavenly place / That busy archer his sharp arrows tries?" (Sir Philip Sydney; the effect of being in love is described through its cause). (5) Cf. nimitta. tulya, 'equal': (1) a type of aprastutapraśamsā in which the relation be- tween the implicit and explicit subjects is one of similitude, real or apparent. (2) M 152. (5) If the similitude is real, we have sādrśya- mātra; if only punned, samāsôkti; if the implicit subject itself is punned, ślesa. See these terms for examples. For the earlier writers, intimation seems only to have been used where a relation of similitude could be seen; it is often described in the same terms as upama ('simile'), the implicit term being the subject of comparison (upameya). But Mammata broadens the figure to include other relations: that of cause-effect, and general-specific. See aprastutapraśamsā. nimitta, 'cause': (1) a type of aprastutapraśamsā in which the real subject is a cause and is intimated through a description of its effect. (2) M 152. (3) rājan rājasutā na pāțhayati mām devyo'pi tūșnīm sthitāh | kubje bhojaya mām kumāra sacivair nâdyâpi kim bhujyate | ittham nātha śukas tavâribhavane mukto'dhvagaiḥ pañjarāt | citrasthān avalokya šūnyavalabhāv ekaîkam ābhāșate (Mammața; describing the fright caused by the news that the king has set out against his enemies: "'O King, the princesses do not address me! Even the Queens remain silent! Hey, humpback! come play with me! Prince! why aren't you with your friends?' Thus does the parrot, who has been freed by passersby from its cage in your enemy's palace, carry on as he wanders about the empty halls looking at the portraits"). (4) "Help me to seek! for I lost it there; / And if that ye have found it. ye that be here, / And seek to convey it secretly, / Handle it soft, and treat it tenderly, / ... It was mine heart! I pray you heartily Help me to seek" (Sir Thomas Wyatt; the poet is in love, which has resulted in the loss of his heart). (5) Cf. kārya. višeșa, 'speciality': (1) a type of aprastutaprašamās in which the real subject is particular and is intimated through mention of an ap-

114 GLOSSARY propriate universal. (2) M 152. (3) suhrdvadhūbāşpajalapramā- rjanam karoti vairapratiyātanena yah / sa eva pūjyah sa pumān :a nītimān sujīvitam tasya sa bhājanam šriyah (Mammața; this is spoken by a minister of the slain Naraka and urges retaliation on Krsna: "The Prince who wipes away the tears of his friends by taking re- venge on his enemies, he alone is honorable, he is a man and a just man, his auspicious life is a vessel of good fortune"). (4) "But at my back I always hear / Time's winged chariot hurrying near" (Andreu Marvell; to his coy mistress). (5) Cf. sāmānya. śleşa, 'double-entendre': (1) a type of tulya aprastutapraśamsā in which the real subject is intimated by puns or double meanings. (2) M 152C. (3) pumstvād api pravicaled yadi yady adho'pi yāyād yadi praņayane na mahān api syāt / abhyuddharet tad api viśvam itid- śîyam kenâpi dik prakațitā purușôttamena (Mammața; flattery of a king; reference is to the forms of Vișnu: "Even if he deviates from masculinity [from heroism], even if he descends to earth [suffers reverses], even if he is not of great size [not powerful], nevertheless. he upholds the earth; in this way has the expanse of this earth been made manifest by the Great Lord [a great lord]"). (4) " ... A dripping Pauper crawls along the way, / The only real willing out-of-doorer. And says, or seems to say, / 'Well, I am poor enough-but here's a pourer!'" (Thomas Hood; the subject intimated is the rainstorm). (5) Cf. samāsôkti and sādrśyamātra. Śleșa differs from avavava ślesa in that the real subject is there explicit and the pun ancillary. samāsôkti, 'concise speech': (1) a type of tulya aprastutapraśamisa in which the real subject is intimated by puns (or double meanings) on the descriptive qualifications of the explicit subject. (2) M 152C. (6) yenâsy abhyuditena candra gamitah klāntim ravau tatra te ptes pratikartum eva na punas tasyaîva pādagrahah / kșīnenaitad anuş ;::::..... yadi tatah kim lajjase nô manāg | asty evam jadadhāmatā tu Ae yad vyomni visphūrjase (Mammața; this is spoken to a poor man w has demeaned himself by asking alms. The sun and mees (" plicit subjects) are not punned upon, but the descriptive quas tions are as pada, 'ray' and 'foot', ksinda, 'new moen' and ge" ertyless', etc. Note that the last pun requires substitars the phonemes /d/ and /l/ in jadadhamata-jaladhama :: "B w rising have you become so pale, O moon? You should in shine [emulate] him and not be eclipsed by his rays [fall .n 3. and if you have done this through being but a tluin cresecn: of your poverty], you should be ashamed indeed! So be :

GLOSSARY 115 mere fact of your shining in the sky, you are a veritable treasure of coolness [of stupidity]"). (4) "Beneath in the Dust, the mouldy old Crust / of Moll Batchelor lately was shoven, / Who was skill'd in the Arts of Pyes, Custards and Tarts, / And every Device of the Oven. / When she'd liv'd long enough, she made her last Puff, / A Puff by her Husband much prais'd; / And here she doth lie, and makes a Dirt Pye, / In Hopes that her Crust may be rais'd" (Anon .; an epitaph. The real subject of Moll's death and resurrection is suggested by puns on her culinary abilities). (5) In śleșa, the real subject itself is effected through a pun on the explicit subject, not entirely on its qualifications. Cf. also sādrśyamātra. sādrśyamātra, 'mere similitude': (1) a type of tulya aprastutapraśamsā in which the real subject is intimated through the force alone of its similitude with the explicit subject. (2) M 152C. (3) ādāya vāri paritah saritām mukhebhyaḥ kin tāvad arjitam anena durarņavena / kşārīkrtam ca vadavādahane hutam ca pātālakukșikuhare viniveśitam ca (Mammata; the picture is that of a wealthy man wasting his resources: "Taking all the water from the mouths of rivers hereabouts, making it salty and throwing it on the submarine fires and losing it into the secret maws of hell: what indeed has this ocean profited?"). (4) "It's but little good you'll do a-watering the last year's crop" (George Eliot). (5) By mere similitude is meant that no puns or double meanings operate to suggest the implicit subject. See śleșa and samāsôkti. The relation is also between particulars, much as if it were a drstânta with the subject implicit. Cf. viśeșa and sāmānya. sāmānya, 'generality': (1) a type of aprastutapraśamsā in which the real subject is universal and is intimated through description of an appropriate particular. (2) M 152. (3) etat tasya mukhāt kiyat kamalinīpatre kaņam vāriņo yan muktāmaņir ity amamsta sa jadah śrņv anyad asmād api | angulyagralaghukriyāpravilayiny ādīyamāne śanai kutrôddīya gato mamêty anudinam nidrāti nântaļ śucā (Bhallața, quoted by Mammata; the universal here is said to be that the property sentiment of fools is apt to be overextended. Punctuation would help in this example: a comma after kiyat, a period after jadah, a comma after śanaih: "How few words [of sense] come from his mouth; he thinks a drop of dew fallen on a lotus petal to be a pearl of high price! And listen to this: slowly lifting the dewdrop until it melts between the tender movements of his fingers. he cries, 'Where has my pearl flown to?' and he cannot sleep for the pain in his soul!"). (4) DA / Dayadhvam: 1 have heard the key Turn in the

116 GLOSSARY

door once and turn once only / We think of the key, each in his prison / Thinking of the key, each confirms a prison" (T. S. Eljot; the explanation is given by Eliot himself in his notes, quoting F. H. Bradley: "My external sensations are no less private to myvelf than are my thoughts or my feelings. In either case my experience falls within my own circle, a circle closed on the outside ... "). (5) Cf. viśeşa. aprastutapraśaņsā (II): (1) an elliptical simile in which the subject of comparison is referred to by a sign or token, usually a double- entendre based on comparable qualities common to the two things. (2) V 4.3.4. (3) lāvaņyasindhur aparaîva hi kêyam atra yatrôtpalāni śaśinā saha samplavante / unmajjati dviradakumbhatațī ca yatra yatrâpare kadalikāndamrnāladandāh (Vāmana; the other river is, of course, a young lady in the river. The lotuses refer to her eyes. the lobes to her breasts, etc. "Who can that second River of Beauty be- where the lotuses are playing with the moon and the submerged elephant shows his great frontal lobes, and where [are seen] other soft stalks like the trunks of banana trees?"). (4) "Lemon tree very pretty, and the lemon flow'r is sweet; but the fruit of the poor lemon is impossible to eat" (traditional folksong; the poet refers to his disappointed love). (5) This figure resembles the usual samāsôkti inasmuch as the emphasis is placed on recognition of the implicit subject through qualifications which can apply to both subject and object. Vämana is concerned only with those aspects of the several figures which display features of the simile; he departs from tradition in many such cases. His figure samāsôkti is defined as total ellipsis of the subject, by which is probably meant reference through similitude only, not (as here) through punned qualifications. Vimaana would have conformed more closely to tradition by reversing the names of the two figures. Cf. adhyavasāna atišavokti, where the point is the confusion of two things. aprastutapraśamsā (III): (1) a figure in which blame of an implicit sabe: is to be understood through praise of an explicit objeet. ()D . (341-42). (3) sukham jīvanti hariņā vanesr aparasevina" ayatnasulabhais trnadarbhânkurâdibhih (Dandin; this is to be onde! stood as a complaint addressed to an illiberal benelactor: " The geae's deer in the forest think only of serving others and live wihe:a hardship on easily obtainable grasses, darbha shoots, and the Eke" (4)" ... the Dean expatiated upon what is perhaps the most my sicttees characteristic of genius, its tendency to appear among membes

GLOSSARY the human race" (E. M. Forster; an apparent encomium of genius, 117 but in reality a remark directed against mankind as such). (5) This type of aprastutapraśamsā is just the opposite of vyājastuti. For Daņdin, the figure has little to do with samāsôkti; he is the only writer who treats both figures for whom this is true (see samāsôkti). abhivyakti abhivyakti, 'manifestation': (1) intimation. (2) AP 345.7-18. (5) This figure may represent a stage in the prehistory of the dhvani theory. It is described by the author of the Agni Purana as twofold: śruti and dhvani, and the former is then described in terms quite similar to the classical analysis of the kinds of meaning (mukhyā, lākșaņikī, and gaunī). The category dhvani (also called āksepa), which may and should be the gaunī of the preceding triad (cf. Dhvanyāloka, chap. 1), is then subdivided into five common alamkāras: ākșepa, aprast- utastotra, samāsôkti, apahnuti, and paryāyôkta. In addition, the term abhivyakti has become a standard gloss for dhvani in the later writings. Abhinavagupta asserts that the nispatti of Bharata's rasasūtra means abhivyakti (quoted in Kāvyapradīpa). According to S. K. De, the Agni Purāna may have been contemporary with the author of the kārikās of the Dhvanyāloka. The matter is made hypothetical by the terseness of the Agni Purāna, which offers no examples for any of the figures defined.

artha artha, 'sense': (1) a cover term for those figures whose poetic effect was thought to depend on the meaning of the expression rather than on verbal patterns or devices. (2) B 1.16, V 1.1.1, D 3.186, U 5.12, AP 344.1, R 7.9, M chap. 10. (5) Although the arthâlamkāra are on the whole formally defined, the nature of the form differs from that of the more obvious śabdâlamkāra. These latter figures repose upon non-referential criteria, such as morpheme type (see śleșa) and metre (see yamaka). The former involve characteristics attribut- able to the subjects of the utterance or to the relation between the subject and a descriptive phrase, such as comparability (simile), exaggeration (hyperbole), non-literalness, or combinations of these. The basic distinction is that between grammatical form and inten- tional reference, but the formal aspect of both should not be under- estimated (see upamā, vyatireka, śleșa).

118 GLOSSARY arthântaranyāsa arthantaranyasa, 'introduction of another matter': (1) a figure in which a proposition or remark is justified or substantiated by the adjunc- tion of a relevant moral or rationale; apodixis. (2) B 2.71-74 D 2.169-79, V 4.3.21, U 2.4, AP 344.24, R 8.79-84, M 165. (3) priyeņa samgrathya vipakşasamnidhāv upâhitām vakșasi pīvarastane srajam na kācid vijahau jalâvilām vasanti hi premņi guņā na vastuni (Bhāravi, quoted by Vāmana: "She clasps to her full bosom the water-faded garland once offered by her lover in the presence of her rivals, for quality resides in the thought, not the thing"). (4) "Hoist up sail while gale doth last, / Tide and wind stay no man's pleasure" (Robert Southwell). (5) This figure differs from drstânta in that the intention of the speaker is to establish his remark, not to clarify it. The particle "for" may be understood to be connecting the proposi- tion and its substantiation and provides one basis for subdividing the figure (Bhāmaha, Udbhata). Rudrata, however, and to some extent Mammata, consider arthantaranyasa a conjunction of remarks general and specific, while drstânta is a relation of two observations. both specific and neither amenable to the intention of the speaker. Cf. drstânta and ubhayanyāsa. Though founded on a similitude. arthântaranyasa may function through antithesis (vaidharmya: Agni Purāņa, Rudrața, Mammața). That its end is not the description of that similitude distinguishes this figure from upamā. Dandin considers eight subtypes, distinguished with reference to the nature and scope of proof itself : universal (viśvavyāpi), particular (viseșastha). apparent (śleșâviddha), paradoxical, i.e., apparently false (viro- dhavat); and by considering the relation of the speaker to his thesis: disapproval (ayuktakāri), approval (yuktâtman), qualified disapproval (yuktâyukta), and qualified approval (viparyaya). Vāmana alone considers the figure indivisible. ayuktakāri, 'doing wrong': (1) a type of arthântaranyasa in which the situation referred to in proposition and substantiation is condemned by the speaker. (2) D 2.170 (176). (3) madhupānakalār Kang :: nirgato'py alinām dhvaniļ | kațur bhavati karņasya kāminām pūpe" īdrsam (Dandin: "The sounding of the bees, though it issues from throats thick with honey, is harsh to the ears of lovers. What d shame this is!"). (4) "The waters of the Mississippi and Missouri unite and form one river. The water of the latter is exceedingly turbid, and the former clear. When they first meet the waters refuse to mingle .... By degrees the clear, bright waters of the one become

united with those of the other, and the clearness is lost forever. GLOSSARY Virtuous and vicious persons can associate for a time, kecping their 119 characters distinct. But if the associations be continued, the virtuous, pure character will become soiled by the vicious. No one can associate freely with the wicked without becoming in some measure like them" (Robert Blackwell). (5) Cf. yuktâtman and yuktáyukta. yuktâtman, 'essentially correct': (1) a type of arthântaranyāsa in which the situation referred to in proposition and substantiation is ap- proved by the speaker. (2) D 2.170 (177). (3) ayam mama dahaty angam ambhojadalasamstarah / hutâśanapratinidhir dāhâtmā nanu yujyate (Dandin: "This expanse of lotus flowers pains my whole being, yet it may be proper that something as brilliant as fire have a soul aflame"). (4) "'No, I just couldn't feel the same about her again.' 'Well, why feel the same? One has to change as one gets older. Why, then years ago I couldn't be interested in anything later than the Sumerian age and I assure you that now I find even the Christian era full of significance'" (Evelyn Waugh). (5) Cf. ayukta- kāri, where the situation is judged ill. yuktâyukta, 'correct and incorrect': (1) a type of arthântaranyāsa in which the situation referred to in proposition and substantiation is generally or conditionally approved by the speaker but for some reason is, in this case, considered irregular. (2) D 2.170 (178). (3) kşiņotu kāmam šītâmśuh kim vasanto dunoti mām | malinâcaritanı karma surabher nanv asāmpratam (Dandin: "Let the moon consume me if it wants to! Why does the springtime make me suffer? Such a black deed is surely unsuited to the sweet season"). (4) "Before she realized it she was absorbed in what had so often been on her mind lately .... His warm playfulness, his affectionate tenderness --- what had become of it? ... Oh no, no! she caught herself, how can I be thinking of such things again! The sweet desires of the flesh are the nets of Satan" (O. E. Rolvaag). (5) Here the inconsistency is in fact condemned (ayuktatva is siddha). Cf. viparyaya, the reverse. viparīta, 'reversed': (1) same as vaidharmya arthântaranyāsa. (2) R 8.82. viparyaya (I), 'reversal': (1) a type of arthântaranyāsa in which the situa- tion referred to in proposition and substantiation is generally or conditionally condemnable, but is here for some reason approved. (2) D 2.170 (179). (3) kumudāny api dāhāya kim ayam kamalâk arah | na hîndugrhyeşûgreşu süryagrhyo mrdur bhavet (Dandin: "So much do the night lotuses afflict me, how much more ought the day lotus

to burn! For one of the sun's retinue will not be mild when the friends of the moon are harsh!"). (4) "Sweet is the rose. but grows upon a brere; 'Swect is the juniper. but sharp his bough: every sweet with sour is temper'd still. That maketh it be coveted ... So

the more: For casy things, that may be got at will. Most sorts of men do set but little store" (Edmund Spenser). (5) The term "reversal" is prohably to be taken as reversal of vuktayukta. where an otherwise appropriate situation is deemed in some respect inappropriate. Here the inconsistency is aocepted. viperyaya (II): (1) a type of arthantaranydsa in which the substantiation. having the form of a general remark. follows the proposition, which Is particular in reference. (2) U 2.4. (3) śiva apaśyac câtikastāni tapyamåndm tapdmsy vmăm asambhavyapaticchanām kanyānām kå pard gatih (Udbhata: Uma had set her mind on having none but Siva: "[Śiva] watched Uma practicing austerities of unbelievable austerity: what other recourse have girls who desire a perfect hus- band?"). (4) "On the College of Wadham at Oxford being insured from Fire. after a Member had been suspected of an unnatural Cnme: Well did the amorous sons of Wadham / Their house secure from future flame; ! They knew their crime, the crime of Sodom, / And judg'd their punishment the same" (Anon.). (5) Udbhata is the first writer to classify apodixis in this way, but he perversely applies the term viparyaya to that type which the earlier writers consider perfectly normal: a particular remark justified by a general remark, as: "Keep in the heart the journal nature keeps; / Mark down the limp nasturtium leaf with frost" (Conrad Aiken). Udbhata's innovation is, of course, that he allows the general remark to precede, as in the example offered under yuktâtman. Rudrata and Mammata both allow for this same distinction, but do not give it a name. irodhavat, 'contradictory': (1) a type of arthântaranyāsa in which a sceming paradox is justified. (2) D 2.170 (175). (3) jagad ānandayaty eşa malino'pi nišākarah | anugrhņāti hi parān sadoșo'pi dvijêśvarah (Dandin: "The orb of the night, though covered with blemishes, delights the whole world; but then, a Brahmin, even if he have faults, confers favors upon others"). (4) "Before you despise Adam as deficient in penetration, pray ask yourself ... if you ever could, without hard head-breaking demonstration, believe evil of the one supremely pretty woman who has bewitched you. No: people who love downy peaches are apt not to think of the stone, and sometimes jar their teeth terribly against it" (George Eliot). (5)

GLOSSARY This figure differs from viparyaya in that the inconsistency is here 121

accepted, not excused. viśeșastha, 'specific': (1) a type of arthântaranyāsa in which the situation and substantiation are of specific import, that is, do not apply to all men. (2) D 2.170 (173). (3) payomucah parītāpam haranty eva śarīriņām / nanv ātmalābho mahatām paraduhkhôpasāntaye (Daņdin: "The great rain clouds relieve the scorching heat of summer for the wandering ascetics, for it is the office of the great to alleviate the suffering of others"). (4) "What he said had a hateful truth in it, and another defect of my character is that I enjoy the company of those, however depraved, who can give me a Roland for my Oliver" (Somerset Maugham). viśvavyāpin, 'universal': (1) a type of arthântaranyāsa in which the situa- tion and substantiation are of universal import, that is, apply to all men. (2) D 2.170 (172). (3) bhagavantau jagannetre sūryācandra- masāv api | paśya gacchata evâstam niyatih kena langhyate (Daņdin: "The blessed eyes of the world, the sun and moon, even they must set: who can escape his fate?"). (4) "The glorious lamp of heaven, the sun. / The higher he's a-getting, / The sooner will his race be run. ! And nearer he's to setting. / .... / Then be not coy, but use your time, / And while ye may, go marry: / For having lost but once your prime, / You may for ever tarry" (Robert Herrick). (5) The figure is contrasted with viśeșastha. vaidharmya. `difference': (1) a type of arthântaranyāsa in which the verbs of the proposition and its substantiation are opposite in sense; substantiation by antithesis. (2) M 165C. (3) hrdayena nirvrtānām bhavati nrnām sarvam eva nirvrtaye | indur api tathāhi manah kheda- yatitarām priyāvirahe (Rudrața: "Everything delights those whose hearts are full of bliss; to lovers in separation even the cool moon afflicts the mind unmercifully"). (4) "You may think I was seeing lions in the path, but it is never safe to reckon on meeting nothing more formidable than a sheep" (Oliver Onions). (5) Vaidharmya is the same as viparita. The figure is known to the Agni Purāna, but is not named (344.24). Cf. vaidharmya drstánta and remarks under arthântaranyāsa. ślesâviddha, 'invested with double-entendre': (1) a type of arthántaranyāsa in which a pun underlies the attempted substantiation. (2) D 2.170 (174). (3) utpādayati lokasya prītim malayamārutah / nanu dākși- nyasampannah sarvasya bhavati priyah (Daņdin; dākșinya means both 'southern', as applied to the wind, and 'polite', as applied to the

122 GLOSSARY

friend: "The wind from the southern mountain arouses joy amongst men; indeed one born in the South [accomplished in piety] is every- body's friend"). (4) "So round his melancholy neck, / A rope he did entwine, / ... And there he hung, till he was dead / As any nail in town- / For though distress had cut him up, / It could not cut him down!" (Thomas Hood). sādharmya, 'similitude': (1) a type of arthântaranyasa in which the verbs of both proposition and substantiation are parallel in sense, that is not antithetical. (2) M 165. (3) (4) See arthântaranyāsa. (5) This subcategory is arthântaranyāsa itself-a category invented by Mammața to balance vaidharmya, q.v.

avayava avayava, 'member': (1) an arthaślesa in which the second meaning, suggested through puns on certain aspects or qualifications of the primary subject, augments or ameliorates the force of the description. (2) R 10.18. (3) bhujayugale balabhadrah sakalajagallanghane tathā balijit | akrūro hrdaye'sau rājâbhūd arjuno yaśasi (Rudrața; the puns are balabhadrah, balijit, and arjunah which apply to the king as descriptive adjectives but are also the names of great heroes: "In his two arms fortunate of strength [Balabhadra], in overwhelm- ing the whole world a conqueror of his enemies [Balijit], straight- forward [Akrūra] in his heart and glorious [Arjuna] in his fame, was this King"). (4) "The scene in water colours thus I paint" (Thomas Hood; the rainy day is described in "watery" words). (5) I think the point is not that the pun is between an adjective and a proper name, since several other types involve this same feature (avisesa. perhaps tattva), and since the name of the type would itself then be inexplicable. It is to be contrasted with aviśesa, where the pun is on the primary subject, not on any qualifications of it.

avasara avasara, 'occasion': (1) a figure in which a sentiment is expressed through a description of a particular fact strongly suggestive of it. (2) R 7.103 (104). (3) tad idam aranyam tasmin dašarathavacunānspala- navyasanī / nivasan bāhusahāyaś cakāra rak şaļk savam rāmaļ (Rudraļa: through an association with Rama, the idea is intimated that this is a holy place: "In this forest Rama lived, faithful to the com- mands of his father, and with his bare hands, he slew all the demoas"). (4) "The moon shines bright :- in such a night as this, When the

GLOSSARY 123 sweet wind did gently kiss the trees, / And they did make no noise, -in such a night / Troilus methinks mounted the Troyan walls, / And sigh'd his soul toward the Grecian tents, / Where Cressid lay that night" (Shakespeare). (5) Rudrata gives two examples, the first expressing nobility (holiness), the second love; the occasion for Shakespeare's remark is clearly amorous. In this and several other figures, Rudrata delineates what appear to be types of dhvani, or suggestion. On the other hand, the figures leśa, sūkșma, paryāya, and avasara can be related to the earlier figure paryāyôkta, not present in Rudrata's classification and never before subdivided. Rudrața, probably a Kāśmīri and a contemporary of Ānandavar- dhana, may indeed have been familiar with the dhvani theory and may have attempted to incorporate it into a standard alamkāra treatise. The Agni Purāna also propounds a curious view of dhvani (see abhivyakti).

aviśeșa aviśeșa, 'lacking qualifications': (1) an arthaślea in which the double- entendre is expressed in and through the noun or subject of the utterance, instead of through qualifications thereof. (2) R 10.3. (3) śaradindusundararucam sukumārām surabhiparimalām aniśam / nidadhāti nâlpapunyah kaņțhe navamalikām kāntām (Rudrața; navamalikā is both 'jasmine' and a girl's name: "No one who does not deserve it puts jasmine around his neck: she who has the beauty of the autumn moon is sweet and always fragrantly perfumed"). (4) "Synthesis, smoking in a corner / Groans, pulls himself together" (Robert Graves). (5) The point here is that the pun is not carried by the adjectival qualifications (the first half śloka or the phrase "pulls himself together") as it is in the canonic ślesa (cf. aviruddha- kriyā, viruddhakarman). Indeed, the qualifications apply equally to either sense, but those senses are expressed by a single word here (the noun), which, syntactically primordinate, carries the double-entendre.

asaņgati asamgati, 'non-concomitance': (1) same as dūrakārya hetu. (2) R 9.48 (49), M 191.

124 inapplicable qualification, when understood as a pun, becomes a descriptive qualification. (2) R 10.16. (3) parihrtabhujamgasangah samanayano na kuruşe vrşam câdhah / nanv anya eva drstas tvam atra paramêśvaro jagati (Rudrața; here a king is likened to Siva, despite the epithets given, which seem to distinguish him from Siva; these epithets are understood as puns which, in fact, describe the king: "You have avoided the company of snakes [suspicious characters], you have an even number of eyes [do not have three eyes] and do not mount a bull [never decrease prosperity]; still you are another visible Lord of Lords [Śiva] in this world"). (4) "'Oh, Daddy dear, what is a basket?'/Said a youthful and mischievous elf : /'All baskets, me boy, are children of joy. / In fact you're a basket yourself'" (Anon .; "basket" does not seem to apply to the boy until it is understood that, in this dialect, "basket" and "bastard" are synonymous). (5) The name "impossibility" refers to the fact that the distinctive qualifica- tion cannot apply to that subject from which the real subject is being distinguished: samanayanah ('equal-eyed') applies to the king, but not to Śiva who has three eyes; nevertheless, the qualification 'uneven-eyed' is so commonly applied to Siva that the resemblance of the two qualifications is enough to suggest Siva. This ślesa amounts to a vyatireka expressed through puns, but it also fits into the canonic pattern of ślesa, where a qualification generally carries the double-entendre (cf. viruddhakarman).

ahetu ahetu, 'absence of cause': (1) same as viśeșôkti II. (2) R 9.54 (55).

ākșepa, 'objection': (1) a figure in which is expressed an objection to or ākșepa denial of some state of affairs, either real or imagined, either past, present, or future; contradiction. (2) B 2.66-70, D 2.120-68, U 2.2-3, AP 345.14-15, R 8.89-91, M 161. (3) aham tvām yadi nêkşeya kşaņam apy utsukā tatah / iyad evâstv ato'nyena kim uktenâpriyeņa te (Bhämaha; the girl is threatening to expire if her lover leaves her; "If I should not see you, even for a moment, my impassioned soul .... Enough of that! Why should I repeat more unpleasantness!"). (4) "Streets that follow like a tedious argument / Of insidious intent / To lead you to an overwhelming question ... / Oh, do not ask, 'What is it?' / Let us go and make our visit" (T. S. Eliot). (5) Bhāmaha and Dandin define this figure as pratisedhôkti ('the

GLOSSARY 125 enunciation of an interdiction'), the most obvious case relating to an event about to take place which the speaker wishes to prevent or avoid (vakşaymāna, bhavișyat). But the notion of interdiction is also applied to past time, in the sense of contradiction, where the event is too unlikely or preposterous or wonderful (ukta, vrtta). Bhāmaha and those following him, Udbhata and Mammata, leave the matter there, but Dandin offers a third variety by applying the interdiction to present time (vartamāna), where the connotation is that of doubt or suspicion as to which of two alternatives is the more likely or desirable. The general notion of objection or denial is thus strongly qualified by the temporal relation of the events and the speaker and presents an interesting study in miniature of the possible kinds of negation. Rudrata views the figure more in these latter terms, i.e., as a question of mode rather than a mere matter of time, and he distinguishes two varieties: an impossible event (that is, one which is negated) is either conventionally accepted (prasiddha: as, the "burning" of moonlight to the lover), or is entirely irrational (viruddha: as, measuring the sea with a dish). The original idea of preventing an imminent and objectionable event is here completely lost sight of, and the negation is made a quality of the event itself. Daņdin goes on to illustrate twenty-one other types of ākșepa, of two basic sorts. One element in a relation is objected to or denied (either cause or effect, or subject or predicate). For example, a denial of the effect would be: "Baby roused its father's ire / By a cold and formal lisp. / So he placed it on the fire / And reduced it to a crisp. / Mother said, 'Oh, stop a bit! / This is overdoing it!'" (Harry Graham). The remaining seventeen varieties reflect differing affective suggestions which can accompany contradiction. Most illustrate the very first and most obvious sort, that of threatening or prevention; a girl says "don't go" to her lover with anger, despair. irony, bitterness, disgust, etc. In the Agni Purāna, ākșepa is also the genus, equated with dhvani (see abhivyakti), of five figures: ākepa, aprastutastotra, samāsôkti. apahnuti, and paryāyôkta. ukta, 'spoken': (1) a kind of aksepa in which the state of affairs denied or questioned has already occurred. (2) B 2.67 (70). D 2.122 (121). U 2.2-3, M 161. (3) anangah pañcabhih puşpair viśvam vyajayatē- şubhih | ity asambhāvyam atha vā vicitrā vastušaktayah (Daņdin: "The God of Love conquered the whole world with five flower-tipped arrows. This is quite impossible; amazing is the power of things!").

126 GLOSSARY (4) "No, no, for my Virginity, / When I lose that, says Rose, I'nl dye: / Behind the Elmes, last Night, ery'd Dick, / Rose, were you not extremely Sick?" (Matthew Prior). (5) Dandin uses the term vrtta ('occurred'). Compare vakşyamāna and vartamāna, where the facts are about to occur or are occurring. prasiddba, 'established', 'well known': (1) a type of dkşepa in which the question takes the form of an objection to a conventional or well- known fact. (2) R 8.89 (90). (3) janayati samtāpam asau candraka- lākomalâpi me citram / athavā kim atra citram dahati himāni hi bhūmiruhah (Rudrata; during the separation of lovers, it is conven- tional to speak of the "burning" moon: "It is marvellous that the soft-rayed moon causes such a fever; yet perhaps it is not so odd: do not the winter snows consume all things that grow on earth!"). (4) "'You a Magistrate chief', his wife tauntingly said, / 'You a Methodist-Teacher! and caught with your Maid! / A delicate Text you've chosen to handle / And fine holding forth, without Daylight or Candle!' / Quoth Gabriel, 'My Dear, as I hope for Salvation, / You make in your Anger a wrong Application; / This evening I taught how frail our Condition; / And the good Maid and I were but at-Repetition'" (Anon.). (5) Both examples offer a rationale which attempts to meet the objection. This rationale is based upon the situation objected to being well known; compare viruddha āksepa. where the question takes the form of exposing an irrational im- possibility. By "well known", Rudrata refers to the character of certain situations which, though unlikely (the moon burning. a Methodist fornicating), are not entirely unexpected when they do occur. Most caricatures depend upon this basic plausibility of the conventionally implausible. Prasiddha resembles ukta aksep. but the emphasis is modal, not temporal. bhavişyat, 'about to be': (1) same as vakşyamāna āksepa. (2) D 2.120 (125). vakșyamāna, 'about to be spoken': (1) a type of äk sepa in which the state of affairs denied or questioned has not yet occurred. (2) B 2.0 (69), D 2.126 (125), U 2.2-3, M 161. (3) satyam bravimi na tr" !! mām drastum vallabha lapsyase | anyacumbanasamk rantalak sārak tena cakșuşā (Dandin; a threat designed to prevent a state of affairs: "I'm telling you the truth. You'll not be able to see me, lover, with eyes red from the lac of others' lips!"). (4) ".You like words like damn and hell now, don't you?' I said I reckoned so. 'Well, I don't,' said Uncle Jack, 'not unless there's extreme provocation connected with 'em .... Scout, you'll get in trouble if you go around

GLOSSARY 127 saying things like that. You want to grow up to be a lady, don't you?'" (Harper Lee). (5) Vakşyamāna is the same as bhavişyat ākșepa of Daņdin. Compare ukta and vartamāna ākșepa. vartamāna, 'being': (1) a type of āksepa in which the state of affairs denied or questioned is now taking place. (2) D 2.124 (123). (3) kutaḥ kuvalayam karņe karoși kalabhāșiņi / kim apāngam aparyāptam asmin karmaņi manyase (Dandin; flattery is suggested through a fanciful alternative: "Why do you fix a lotus at your ear, my soft- voiced one? Do you think your sidelong glance unable to attract me?"). (4) "In a church which is furnish'd with mullion and gable, / With altar and reredos, with gargoyle and groin, / The penitent's dresses are sealskin and sable, / The odour of sanctity's eau-de- Cologne. / But only could Lucifer, flying from Hades, / Gaze down on this crowd with its panniers and paints. / He would say, as he looked at the lords and the ladies, / 'Oh, where is All Sinners', if this is All Saints'?'" (Edmund Yates; blame is suggested through a fanciful alternative). (5) Only Dandin offers this middle term be- tween ukta, 'spoken' and vakşyamāna, 'about to be spoken'. Though all three types of objection can be reduced to what appears to be this mere outward distinction of time, it is interesting to note the changes of mode which parallel and are probably functions of that distinction. An objection referring to past time (ukta) is inevitably ironical and expresses amazement or suspicion; that referring to future time (vakșyamāna) tends to be hortatory or interdictive and need not be founded upon a hyperbole. Similarly, the āksepa of present time (vartamāna), objecting to something that is in the process of completion but not yet accomplished, has in mind an alternative end, preferable or indifferent. Doubtless it was a consideration like this which prompted Rudrata to recast the definition of āksepa in modal terms: the state objected to is either consistent with convention or impossible to contemplate (prasiddha, viruddha). Complete parallelism is, however, not to be looked for. viruddha, 'contradicted': (1) a type of äksepa in which the objection refers to an impossible situation-one not conventionally realizable. (2) R 8.89 (91). (3) tava gaņayāmi guņān aham alam athavâsa- tpralāpinīm dhin mām | kah khalu kumbhair ambho mātum alam jalanidher akhilam (Rudrata: "I am enumerating your qualities! Enough of my thoughtless muttering! Who indeed would try to measure the whole sea with a pot?"). (4) "How is it that this girl

128 GLOSSARY

could cry at having to tell Sam Bannett she could not think of him, and then treat another lover as she treated the Virginian? I cannot tell you, having never (as I said before) been a woman myself" (Owen Wister). (5) Compare prasiddha, where the situation is conventional though in fact just as impossible. Here the impossibility is carried by the fact itself (counting an infinitude, understanding a woman). Of course, poetic license must be allowed in determining just what facts are admitted. vrtta, 'occurred': (1) same as ukta ākșepa. (2) D 2.122 (121).

āvrtti avrtti, 'repetition': (1) the repetition of a word or an idea in the same or a closely related phrase. (2) D 2.116, AP 343.18-20. (5) Dandin distinguishes three types of repetition: the sense but not the word may be repeated (arthâvrtti), the word but not the sense (padâvrtti), or both the word and the sense (ubhayâvrtti). See s.v. for examples. In the Agni Purāņa, āvrtti is discussed in the same terms as lāțânu- prāsa of Udbhata, as to whether the words are bound or free (paratan- tra or svatantra) and as a part of anuprāsa (along with yamaka). arthâvrtti, 'repetition of the sense': (1) a figure in which the same idea is repeated through different words; paraphrase. (2) D 2.116 (117). (3) vikasanti kadambāni sphuțanti kuțajadrumāh / unmīlanti ca kandalyo dalanti kakubhāni ca (Daņdin; all the verbs mean 'bloom': "The cadamba tree is coming out, the kutaja is flowering, the kandalī bush is blooming, the kakubh is breaking out"). (4) "If a man wished to abstract himself from the world-to remove himself from within the reach of temptation-to place himself beyond the pos- sibility of any inducement to look out of the window-we should recommend him by all means go to Lant Street" (Charles Dickens). (5) In this type of repetition, the important thing is that the words themselves be different. Cf. ubhayâvrtti. ubhayâvrtti, 'repetition of both' (sc. the word and sense): (1) a tigure in which the same word is repeated in the same sense. (2) D 2.116 (119). (3) jitvā viśvam bhavān atra viharaty avarodhanaih viherag apsarobhis te ripuvargo divam gatah (Dandin: "Overrunning this world, my Lord sports with the harem; his enemies, gone to heaven. sport with the Nymphs"). (4) "So loveliness reigned and stillness. and together made the shape of loveliness itself. a form from which life had parted" (Virginia Woolf). (5) Cf. artha- and poddiriti. where one or the other, but not both is repeated.

GLOSSARY padâvrtti, 'repetition of the word': (1) a figure in which the same word is 129

repeated each time in a different sense. (2) D 2.116 (118). (3) utkanthayati meghānām mālā vrndam kalāpinām / yūnām côtkantha- yaty eva mānasam makaradhvajah (Daņdin; the verb is taken first in its literal sense, 'raises the neck', i.e., causes to harken, and then in a figurative sense, "causes to be enamored": "The massing clouds make the flocks of peacocks harken; the God of Love puts longing into the minds of youths"). (4) "Old black rooks flapping along the sky and old black taxicabs flapping down the street" (Joyce Cary). (5) Cf. ubhayâvrtti, where the same word is repeated in the same sense. Padâvrtti is the logical opposite of arthâvrtti, q.v.

āśiș āsiș, 'benediction': (1) a figure expressing a wish for prosperity, good fortune, or reconciliation. (2) B 3.55 (56-57), D 2.357. (3) pātu vah paramam jyotir avānmanasagocaram (Daņdin: "May the supreme Light, best seen by the detached spirit, protect you"). (4) "Let endless peace your steadfast hearts accord / And blessed plenty wait upon your board; / And let your bed with pleasures chaste abound, / That fruitful issue may to you afford" (Edmund Spenser). (5) This figure, appropriately enough, occurs at the end of Bhamaha's and Daņdin's lists. Like several earlier figures (preyas, ūrjasvi, rasavat), it was thought too closely allied to the content of its expression (see note on artha alamkāra) and hence was discarded by later writers.

ukti ukti, 'speech': (1) an arthaślesa in which the second meaning is vulgar or risqué. (2) R 10.14. (3) kalāvataḥ sambhṛtamaņdalasya yayā hasantyâiva hrtâśu lakșmīḥ | nrņām apāngena krtaś ca kāmas tasyāh karasthā nanu nālikaśrīh (Rudrața; apparently this is an encomium of a proper young lady, but a second sense may be obtained roughly as follows: "The good fortune of those she cozzens is in the palm of her hand-she who laughingly accepts payment from her well- ensconced paramour and who will make love at the flick of an eye- brow"; "She holds the beauty of the lotus [the fortune of fools] in her hand and by her side-long glance [with playful glance] is passion [love] inspired [made] in [with] men; she laughing stole the beauty [money] of the moon [of her client] full orbed fin the midst of his friends]"). (4) "Let us roll all our strength and all Our sweetness up into one ball, / And tear our pleasures with rough strife Through

130 GLOSSARY the iron gates of life" (Andrew Marvell). (5) As in the other kinds of arthaślesa which Rudrata describes, the second meaning should further in some relevant way the sense of the first or evident meaning. Mere punning for the sake of punning (word play) is strictly relegated to the realm of śabdaślea. So here the vulgar undertone does not serve the end of lewdness, but rather expands and directs the appar- ently innocent intent of the overtone, which is to say that both examples suggest that innocence is but a veil.

uttara uttara (I), 'answer': (1) a figure in which a preceding remark is inferred from the reply given to it. (2) R 7.93 (94), M 188. (3) bhana mānam anya- thā me bhrukutim vidhātum maunam aham asahā / śaknomi tasya pura- taḥ sakhi na khalu parānmukhībhavitum (Rudrața; from this we are to infer that the girl has been receiving instruction in how to simulate anger in the presence of her lover: "Describe anger another way, friend. I am unable to produce a brow-bent silence! I cannot remain with my face averted before him!"). (4) "'When you call me that, smile!' And he looked at Trampas across the table" (Owen Wister: the Virginian has just been called a "son of a ... "). (5) See praśna. uttara (II): (1) a figure wherein is given a series of fanciful answers to one or more questions. (2) R 8.72 (73), M 121-22. (3) kim maranam dāridryam ko vyādhir jīvitam daridrasya | kah svargah sanmitram suka- latram suprabhuh susutah (Rudrața: "What is death? Poverty. What is sickness? The life of the poor. What is heaven? True friends, a good wife, a fine master, devoted children"). (4) "You are old. Father William,' the young man said, / 'And your hair has become very white; / And yet you incessantly stand on your head-, Do you think, at your age, it is right?' / 'In my youth,' Father William replied to his son, / 'I feared it might injure the brain; / But, now that I'm perfectly sure I have none, / Why, I do it again and again. / 'You are old,' said the youth, 'as I mentioned before, / And have grown most uncommonly fat; / Yet you turned a back-somersault in at the door -- / Pray, what is the reason of that" / 'In my youth. said the sage, as he shook his gray locks, / 'I kept all my limbs very supple / By the use of this ointment-one shilling the box me to sell you a couple?'" (Lewis Carroll). (5) Compare this pocti Allon figure with the conundrum praśnôttara, a series of answers to ques- tions involving a pun.

GLOSSARY utprekșā 131

utpreksā, 'ascription': (1) a figure in which a property or mode of behavior is attributed to a subject literally incapable of sustaining that property, whereby an implicit simile is suggested whose subject (upameya) is the subject receiving the attributed property and whose object (upamāna) is the real basis of that property. (2) B 2.91 (92), D 2.221-34 (222, 224, 226), V 4.3.9, U 3.3-4, AP 344.24-25, R 8.32-37 (33, 35, 37), R 9.11-15 (12-13, 15), M 137. (3) kimśukavyapadeśena tarum āruhya sarvatah | dagdhâdagdham aranyānyāh pašyatîva vibhāvasuļ (Bhämaha; here the red flowers are portrayed as fire, consuming the tree and looking for unburnt parts of the forest: "It is as though fire had climbed the tree in the guise of kimsuka flowers and was looking all about the forest for trees yet unburnt"). (4) "The yellow fog that rubs its back upon the windowpanes, / The yellow smoke that rubs its muzzle on the windowpanes / Licked its tongue into the corners of the evening, / Lingered upon the pools that stand in drains, / Let fall upon its back the soot that falls from chimneys, / Slipped by the terrace, made a sudden leap, / And seeing that it was a soft October night, / Curled once about the house, and fell asleep" (T. S. Eliot; the fog is portrayed as a cat). (5) The figure utprekşā probably comes closer than any other to capturing the sense of the vague term metaphor. Although rūpaka is generally translated `metaphor' (a custom we have followed), its use in the Sanskrit anthologies makes clear that a far more precise meaning is to be attached to the term than 'metaphor' will allow. We have, when the context required such precision, used the phrase "metaphorical identification" for rüpaka, in the sense that two ontologically un- related things are treated gramatically as one thing or, in other words, are identified one with another. The relation of identification is of course directly from one term to another and does not require the interposition of properties, although these may implicitly substantiate the identification. Carl Sandburg's "moon mist mourning veils" or the standard cliché "face-moon" illustrate the necessary explicitness of such identifications. The usual technique for constructing rūpakas is the dvandva compound with the object of comparison (upamāna) in the final position (gramatically free). In English we have elaborated another mode of expression, probably because our language does not encourage explicit compounding to such an extent: namely the subjective genitive, as "the orb of her face", where the object of comparison ("orb") is again the syntactically free term. Utprekșā

132 GLOSSARY

differs from rūpaka in that, instead of the subject and object of comparison being identified with one another, a property character- istic of the object is said of the subject in the most general sense (as predicate, or verb, or even as an independent noun phrase introduced by "as though"). This case is more devious and more universal, though both rūpaka and utprekșā do involve the metaphorical (literally 'carrying over') transfer of something onto something else. It might be said that utprekşā was a rūpaka with suppression of the object. The standard technique, both in English and in Sanskrit, of utprekșā is simply a noun, representing the subject of comparison, followed by a verb or predicate which literally must be understood with the objects of comparison, as: "And one blue parasol cries all the way to school" (Thomas Hood). A parasol can't cry, but a parasol with rain dripping off may be likened to a little girl's face, which can. A rūpaka represents a total identification of two things; an utprekșā is only a partial coalescence through the transfer of a characteristic property or function. Other figures are of course very closely related to utpreksā; perhaps the most significant is samāsôkti, where the subject of comparison is entirely implicit in an expression which in fact represents the object of comparison: "A bird in the hand is worth two in the bush", refers to the advisability of choosing a present advantage (whatever it may be) rather than a future and more attractive advantage. Utprekșā does not subsume com- pletely the subject of comparison in this way: it remains explicit, usually as the subject of the sentence. Mixing these closely related figures in any protracted discourse is, of course, quite common, especially in the case of utpreksā and rūpaka (sometimes enumerated as a separate figure, or utprekşâvayava). Virginia Woolf is partic- ularly rich in such complicated metaphors, as: "Suddenly, as if the movement of his hand had released it, the load of her cumulated impressions [rūpaka] of him tilted up, and down poured in a ponder- ous avalanche all she felt about him [utprekșā]." Notice how easily the one figure can be transformed into the other: "Suddenty, as if the movement of his hand had released their load, her cumulated impressions of him tilted up [utpreksā] and down poured the ponder- ous avalanche of all she felt about him [rūpaka]." It is curious that the figure utpreksā, which in importance is perhaps second only to upamā and rüpaka, and which is recorded by all the writers from Bhamaha onwards, should never have been made the subject of an elaborate subdivision or classification so

GLOSSARY typical of the ālamkārika writers. Only Udbhața even suggests the 133

possibility of subdividing utprekșā, and goes only so far as to enumerate two types (bhāvâbhāva). Even more surprising is the unanimity which is evident in the defining of the figure. Rudrata, though offering no classification, does give six separate definitions of the figure, two of which seem to refer to Udbhata's earlier dich- otomy. These six types will be discussed separately, even though they involve no terminology and no important deviation from the general definition, because of the typical astuteness which Rudrata demonstrates in discussing the principles underlying the various aspects of this important figure. His first definition is the same as that already given, and we will not repeat it here. A mode of action appropriate to one thing is attributed to another, in terms of an implicit simile. That simile, so explained, is nothing but a standard comparison (see upamā) involving a subject, an object, and a real property justifying the similitude (such as those attributed to the cat in Eliot's verse, for example). Now, according to Rudrata, the case is not always so straightforward: for instead of the real property or mode of behavior simply, (a) a second, or subordinate simile (that is, an entire subordinate comparison) may be ascribed to the original subject (type 2), invoking the same three terms; (b) a quality may be attributed to, or implied in another thing, not directly through an upamāna, but in virtue of the relation of both upameya and upamāna to the terms of a further simile which is then under- stood as justifying the first attribution; or, (c) the ascription may not be based upon a real similitude at all, but may be entirely con- ventional. These types follow. utprekșā (II): (1) the ascription of a characteristic to a subject, not in terms of an implicit object of comparison simply, but through the relation of that subject and object to a further subject and object which, as a more general simile, justify the first attribution. (2) R 8.34 (35). (3) āpāņdugaņdapālīviracitamrganābhipatrarūpeņa | šašišañka- yêva patitam lāñchanam asyā mukhe sutanoh (Rudrața: "the spot has fallen on her face"; the immediate simile suggested is that of the beauty marks on the girl's face (subject) and the spots on the moon (object); but this simile suggests a further simile of the girl's face as such (subject) and the moon (object), which in fact justifies and explains the first simile and the attribution based upon it: "A beauty mark has fallen on to the face of this slender-bodied girl, thinking it the moon, for her pale cheeks are decorated with lines of musk

134 GLOSSARY

from the navel of the deer"). (4) "Desolate and lone / All night long on the lake / Where fog trails and mist creeps, / The whistle of a boat / Calls and cries unendingly, / Like some lost child / In tears and trouble / Hunting the harbor's breast / And the harbor's eyes" (Carl Sandburg; here all the parts are made explicit to facilitate comprehension: the first utpreksa [the whistle cries] is followed by the object on which it is based [the child], by which is constituted the immediate simile; but that simile is extended : first subject [whistle] to another [harbor] and first object [child] to another [breast], which further simile [here in the form of a rūpaka] justifies the first). (5) The point here does not concern the utpreksā itself, but only the mode of interpreting the ascription which constitutes the utpreksā; that is, relating that ascription to the simile or similes which it assumes. In this case, the immediate simile so understood is, in a way, not adequate unto itself (compare the next type) because it represents a subordinate and limited aspect of a more universal simile, which situation is suggested by mention of any of its aspects. The form is exactly parallel to the rūpaka called samastavastuvișaya, 'referring to the whole thing (as well as its parts)', and illustrates the generality of the formal framework proposed by the Indian aestheticians. utprekșā (III): (1) the ascription of a characteristic or mode of behavior to a subject, not through direct comparison with an implicit object, as in simile, but via a conventional attribute of that object to which the subject bears a certain relation. (2) R 8.36 (37). (3) atighana- kunkumarāgā puraḥ patākêva drśyate samdhyā | udayatațântaritasya prathayaty āsannatām bhānoh (Rudrața; the dawn does not herald the approach of the sun merely in virtue of its natural relation of precedence to the sun, but, according to the Indian mythology, because the sun's chariot carries a banner the color of the dawn: "The dawn is seen like a flag of deep saffron, heralding the approach of the morning sun hidden behind the eastern hills"). (4) "The Baronet stroked his brow, as if he already felt Bully Bottom's garland" (George Meredith; the Baronet is being compared to Bully Bottom, not directly, in virtue of his "being distinguished by woman", but through the garland which Titania has, in signification. placed upon Bully's brow). (5) The irregularity which this variety of utprekșā accounts for is that of the apparent irrelevance of the terms to one another. In a standard simile, the subject is related to the object through a property which, with some plausibility, can be

GLOSSARY

found in both (though in fact it is limited to the object). So with the 135

utprekşa founded on a standard simile (see above), but in this case, instead of a simple object, there are two terms, related conventionally or accidentally, one of which serves as the technical object of comparison to the subject (as the flag), the other of which, though not sharing any similarity with the subject, gives the rationale for the figurative usage or transfer (as the sun). The figurative or metaphorical ascription is meaningless when the subject is related to either object in isolation: the dawn may behave like the flag only of the sun, the Baronet may wear the garland only of Bully Bottom; yet in both cases the metaphor is understood as a relation between the dawn and the sun, between the Baronet and Bottom. The form which this utpreksā embodies seems to be that of paramparita rūpaka (see utprekșā II). Rudrata, having accounted for certain non-standard similes as foundations for the figure utpreksā, discusses the figure again in terms of the thing said to have the property ascribed (the subject); specifically, in terms of the relation of that subject (or upameya) to that ascribed property. Now, again the point of departure is the standard simile, and here the subject of comparison, as we said, may plausibly be thought of as having the property, in the most real and literal sense. The transfer is justified in terms of a real similarity. Rudrața's type four illustrates this. utprekșā (IV): (1) a metaphorical ascription of a property or mode of behavior justified by an inherently plausible similarity between the explicit object and the implicit subject. (2) R 9.11 (12). (3) ghanasa- mayasaliladhaute nabhasi śaraccandrikā visarpantī | atisāndratayėha nrņām gātrāny anulimpatîvêyam (Rudrața; while the moonlight can- not "anoint" anything, the utpreksa and its implied simile are plausible because of the almost substantial quality of the tropical moonbeams, which do something very near to "flowing" over the body: "In a sky cleansed by the streaming rain of the monsoon clouds, the autumn moon wanders and, almost like an unguent. anoints the limbs of men"). (4) "The very touch of that canvas was enough to make my hand sing. I felt the colour flowing on to it as sweet as cream" (Joyce Cary; as in the Sanskrit, we have the "nlow- ing" of color, but the simile is spelled out by adding "cream" as the object of comparison). (5) The point is not that the usage is not figurative, but that the figuration is plausible in terms of an inherent pattern of behavior present in both the object and implied

GLOSSARY 136 subject. Since the relation is inherent and does not involve the context, Rudrata considers it sambhavita ('hypothetical'). In the next variety of utprekșā, the subject does not possess a plausible relation to the object. utprekșā (V): (1) a metaphorical ascription of a property or mode of behavior justified by an accidental but relevant context. (2) R 9.11 (13). (3) pallavitam candrakarair akhilam nīlâśmakutțimôrvīşu | tārāpratimābhir idam puspitam avanīpateh saudham (Rudrața; the moonbeams make the stucco palace appear as if flowering: "Its spacious floors set with sapphires, the entire royal palace seemed covered with buds in the moonlight; its fine stucco walls were forced into flower by reflections of stars"). (4) "Arthur Donnithorne was ... stared at, from a dingy olive-green piece of tapestry, by Pharoah's daughter and her maidens, who ought to have been minding the infant Moses" (George Eliot; the Pharoah's daughter, being present only in a piece of tapestry, would not ordinarily be characterized as "staring" were it not for the handsome young Arthur, who distracts her). (5) This utpreksā can be seen as one in which the ascription is more in view of the grammatical direct object of the assertion than its grammatical subject. (Note that the words subject and object are not used here as "subject of comparison", etc.) The ascription could be called "transitive": the moonbeams do not behave as vivifiers because of some quality which they inherently possess, but only because of their effect on the stones of the palace; similarly, the portrait does not "stare" because the artist has woven her that way, but because of the presence of an object to be stared at, namely Arthur. It might appear that it was in fact the grammatical object (jewels, Arthur) in these examples which subtended the ascribed quality; an interpretation of this sort is lent authority by both ex- amples being in the passive voice. There are two reasons why this view can not be accepted: the passives can always be expressed as actives with no alteration whatever in the argument, and the Sanskrit authors always speak of the ascribed quality as a kriyā, or simply. 'act.' It is this act which, aside from grammatical expression. constitutes the basic element of the utpreksa, and in terms of the act are defined subject (kartr) and object (kāraņa) to use these well-known terms in a non-standard sense. The subject is the only plausible basis for the act, and It would be perverse

tion differ precisely in the reference of that term (the direct object), or in the ascription these two types of aserip- iscription to the third elevi

GLOSSARY 137 Lastly, a sixth type of utpreksā is possible, where the ascription is not a quality or a mode of behavior at all, but a fanciful rationale for a perfectly literal action (see below). utprekșā (VI): (1) the metaphorical ascription of a motive or rationale. (2) R 9.14 (15). (3) sarasi samullasadambhasi kādambaviyogadūya- mānêva / nalinī jalapraveśam cakāra varşâgame sadyah (Rudrața: "Nalinī takes her bath in the laughing river every day at the onset of the rains, as though she were grieving for the departed geese"). (4) "Her great dark eyes with their long eyelashes touch one so strangely, as if an imprisoned frisky sprite looked out of them" (George Eliot). (5) This is one of the most frequently met types of utpreksā, and yet it differs significantly from those so far described. The fancifulness of the ascription is here more explicit and obvious and seems less to concern the structure of the figure as an ontological treason. A motive or rationale is, of course, less inherent than a mode of behavior or a quality, both of which imply an ontological agent (kartr); the ascription of a motive, on the other hand, neces- sarily involves a bystander as well, and an element of indeterminacy is built into the situation. That this is considered to be utprekșā shows that it is ascription and not the thing ascribed which defines the figure. But since a motive is necessarily ascribed, it might seem that any explanation of whatever sort would qualify as an utpreksā. Rudrața obviates this objection by specifying that the motive ascribed must replace another more obvious, natural, or literal motive. In this sense, we say the "fanciful" ascription of a motive: Nalini really enters the water to bathe, the girl looks at him in a manner which has nothing to do with sprites or imprisonment. The second motive thus plays the same role as the second quality or mode of behavior in that it brings in another term or situation which functions as the object of comparison (imprisoned sprite). What appeared at first to be an irregular utprekşā now appears as a double utprekşā: an ascription of a motive which itself bears a relation of ascription to another [literal] motive. This variety of utpreksā fits into the se- quence of the previous two in a perfectly rational way once its form is understood: just as the ascription of type four took place in reference to the subject alone (first person), and that of type live took place also in reference to the direct object (third person). this type demonstrates those ascriptions dependent upon the second person, or observer. This again illustrates the characteristie insight of the Indian writers into the structure of the figures and their ability

to classify exhaustively and rationally the appearances of poetic inspiration. One other problem is discussed by Dandin and repeated by some commentators: the use of the comparative particle (iva) in some utprekşās has led some thinkers to identify utprekşā and upamā. The example discussed by Dandin is limpativa tamah ('the darkness, as it were, anoints'). Now, the English seems to distinguish the two usages where the Sanskrit does not: we use "as it were", "as if", or "as though" for utprekşā, and "like" or "as" for simile. Dandin likewise insists upon the substance of the difference, even though the word may be the same (2.227-34). A simile relates two similars in terms of a shared quality: the iva attaches to the object of com- parison. In the utpreksā, the iva attaches to the verb, which cannot be considered an object of comparison since it possesses no property.

utprekșâvayava utprekșâvayava, 'component parts of the ascription': (1) a type of meta- phorical ascription (utpreksā) in which further subordinate metaphors explicate and expand the principal ascription. (2) B 3.46 (47), V 4.3.31, 33. (3) angulībhir iva keśasamcayam samnigrhya timiram marīcibhiḥ | kudmalīkrtasarojalocanam cumbatîva rajanīmukham šašī (Vāmana; the principal ascription is: "the moon, as it were, kisses the face of the night"; this is extended by drawing a parallel between fingers playing with a maiden's tresses and the rays of the moon glancing through the darkness (as through tree branches, etc.): "Grasping the darkness with its rays as though it were hair and they were fingers, the moon kisses the face of the night, her eyes demurely closed-lotuses on the lake unbloomed!"). (4) "The Moon, like a flower, / In heaven's high bower, / With silent delight / Sits and smiles on the night" (William Blake; almost identical to Vāmana's example. except that the subordinate parallel is between the moon itself and a flower). (5) Utprekşâvayava is an independent figure in Bhāmaha. and one of the two compound figures (samsrsti) allowed by Vámana (see upamārūpaka); it may be considered in the same contest as utprekşā. It differs from simple utprekșā only in being associated with other and subordinate figures in a "mixed" metaphor. Both the English and Sanskrit examples show an upamā and a rūpaka in conjunction with the principal utpreksa. Later writers consider this figure nothing but one of the many kinds of multiple alamtkura (samsrsti), and its early enumeration as a separate figure probably

GLOSSARY involves no other issues than the extreme frequency with which this 139

particular metaphorical complex is encountered. The poet is always tempted to draw out his ascriptions (utpreksās), especially as they can become quite obscure without much effort, and certainly because an enlarged metaphor is more of a poetic object than an abbreviated one. Vāmana's definition of utprekșâvayava is brief in the extreme and quite misleading at first glance: he says simply: "utprekşāhetur utprekşâvayavah" ('utprekâvayava means the cause of an utprekșā [is given]'). This is to be understood in the sense that the subordinate metaphors or similes define the broader context in which the main utprekșā becomes alive.

udātta udātta, 'lofty': (1) a figure in which great accumulation of wealth or greatness of character (viz. self-denial) is described. (2) B 3.12(11, 13), D 2.300-303, U 4.8, M 176-77. (3) muktāḥ kelivisūtrahāragalitāḥ sammārjanībhir hrtāḥ prātah prānganasīmni mantharacaladbālâñ- ghrilāk șâruņāḥ | dūrād dādimabījaśankitadhiyah karșanti kelīśukā yad vidvadbhavaneșu bhojanrpates tat tyāgalīlāyitam (Mammața: "Bestow- ing benefits is so much a game to your Majesty that in the homes of your advisors the pearls, dropped from necklaces broken in the sports of love, are swept up by the charwomen and, scattered about the borders of the yards, are marked by red lac from the feet of slowly strolling maidens; pet parrots drag the pearls away thinking them to be pomegranate seeds"). (4) "The business was a gold mine, as Sigsag had said. The profits on wine and liquor were of course high; we paid no rent; and the best people in America were our customers. There was no overhead; refrigerators, light, office ex- penses, telephone, glasses, ice and waiters were paid for by the hotel: and a thick golden stream of profit ran into what von Kyling ealled 'the General Welfare'. And along with the profits from the immense turnover of champagne-a hundred cases sometimes being used in some single party-there were of course liberal tips on how to treble that profit on the stock-market" (Ludwig Bemelmans). (5) The figure is always described as being two fold: if great wealth can be appropriated, then it can also be relinquished; the greatness of character is founded upon previously acquired material greatness. The common example is Rama leaving his capital for the forest: "udāttam śaktimān rāmo guruvākyânurodhakah / vihāyopanatam rājyam yathā vanam upāgamat" (Bhāmaha: 'Mighty Rãma, faithful

140 GLOSSARY

to the words of his teacher, abandoned his prosperous and devoted kingdom and entered the forest'). Compare: "Till at length / Your ignorance ... / ... deliver you as most / Abated captives to some nation / That won you without blows! Despising, / For you, the city, thus I turn my back: There is a world elsewhere" (Shakespeare; Coriolanus speaks). The element of exaggeration is not necessarily present, but of course this amounts to a kind of hyperbole (see atiśayôkti) .! Udbhata is careful to distinguish this figure from rasavad alamkāra, for here the evocation of the rasa (for example, vira rasa in the two quota- tions given above) is subordinated to other considerations: a descrip- tion of the forest, or the obloquy heaped upon Coriolanus' enemies. The example from Bhamaha does not support the distinction very well; Udbhata's example is a description of the wealth of the Himâ- laya as a background for Pārvatī's birth. The present figure is one of the group of figures which seem to depend more on their subject matter than on form. Compare rasavad, ūrjasvi, preyas. Except for Mammata, these figures are restricted to the earlier writers. Anandavardhana devotes much significant argument to these figures (especially rasavat) in discussing the relation between rasa and alamkāra; they do show that in the earlier literature the tendency was to include the notion of "mood" within that of "figure", and not the reverse, as happened later.

upamā upamā, 'comparison': (1) the comparison of one thing with a substantially different thing in terms of a property, quality, or mode of behavior which they share; simile. (2) NŚ 16.40-52, B 2.30-33, D 2.14-65 (51-56) (discuss upamādoșa), V 4.2.1-21, U 1.15-21, AP 344.6-21, R 8.4-31, M 125-34. (3) ambhoruham ivâtāmram mugdhe karatalanı tava (Dandin: "Like a pale pink lotus, my sweet! your hand ... "). (4) "My Luve's like a red, red rose" (Robert Burns). (5) Cpamā. one of the four original alamkāras, is in all the rhetorics the most im- portant figure. This is due in part to the universality of the simile in works of art, but another and more cogent reason no doubt concerns the place of simile in the system of the alamkaras. Of the approximately one hundred figures enumerated, perhaps litty are reducible to a basic simile or are describable in terms appropriate to the simile. One of the authors, Vämana, even attempts to state all of the figures involving meaning (arthâlamkāra) as similes, but his

GLOSSARY definitions of certain cases are vicious or too restricted (cf. ślesa). 141

The two systematic writers, Dandin and Rudrata, as well as the compiler Mammata, consider simile the characteristic figure and offer elaborate classifications of it. Upamā was recognized as a category of interpretation as far back as Yāska's Nirukta, and it figures in the Nighantu (3.13), but the term signifies generally metaphorical usage and comprehends what are later considered separate figures (rūpaka, samāsôkti). The broader question of the poetic scope and nature of simile would of course be the most interesting to raise at this juncture. The limitations put upon this work forbid it. A sketch, however, may be in order, insofar as the later alamkāra tradition itself develops along lines which bring into prominence just that discussion. In the dhvani theory and in the work on vakrôkti, the notion is propounded that poetry necessarily involves non-literalness: the poetic passage must refer to a greater range of ideas and things than its immediate words literally convey. In this same context, the importance of simile is probably also to be located. The simile is just such a broad- ening of the expression: a second thing, by nature irrelevant, is brought into the context, whereby the first, the relevant, or subject term is illuminated in a peculiarly characteristic way. The simplest from of non-literalness is just this doubling of the subject. Of course, the non-literal is not at all the irrelevant; the ultimate aim of all poetic diction is coherence, unity, and accuracy and is no wise different from science in this respect. But the skill of the poet lies in his ability to create that coherence out of words and phrases that are constantly fleeing into the shady mists of connotation, constantly avoiding their original meaning and scope. Likewise. the poet who proclaims that his love is like a rose says something far more accurate about that young lady as she is than he could hope to express in terms of her eyes, hair, or physical shape. He does this via an irrelevant discursus which takes us for a moment to the field of botany and associated connotations. This irrelevance is, of course, irrelevant only in the realm of the literal and scientific and constitutes the point of departure for those describing the poctie function of comparison. The ultimate relevance of such oblique reference is at the heart of the poetic problem. By singling ont a thing which is so obviously different-a rose or the moon the poet, by a type of Platonic definition, and by placing it against his subject, immediately cancels out in the reader's mind the entire

142 GLOSSARY

range of literally irrelevant and incomparable aspects and connota- tions of each term so juxtaposed taken separately, and presents only those two things as manifesting some common aspect, the tertium, which by the force of this being abstracted and displayed alone, as it were, redounds to the descriptive credit of the original subject. Simile accomplishes this feat of intellective specification by the use of particles such as "like" (iva). The simile is limited in its expressive power only by the ability of the mind to comprehend the two things as common; the appro- priateness of the simile is primarily a question of the comparing object being precisely proportioned to the subject in just that aspect which is contextually relevant, although in rare cases the object itself may have to be considered. All Indian writers agree on analyzing the simile into the four com- ponent aspects we have introduced: (a) the upameya, or 'thing to be compared': the subject of comparison, through which the simile is related to the literal or outward sequence of ideas which constitute the framework of the poem (compare this literal and grammatical freedom of the subject in upamā with its necessary subordination in rūpaka); (b) the upamāna, or 'agent of comparison': the object introduced to concentrate attention on the essentials of aspect or behavior; (c) the sādhāranadharma, 'shared property': the quality so singled out; and (d) the dyotaka, or 'clarifying' element: the comparative adverb "like" (iva), or a similar indicator. This termin- ology goes back at least to Pānini, who uses it in describing certain compounds which express comparisons (3.1.10, 3.4.45, 2.1.55-56, etc.). The same concepts are also used to describe those figures which depend upon a basic simile but do not express an explicit comparison, such as rūpaka, utprekșā, vyatireka, etc. In this work, the words "subject" and "object" are used in the senses given above unless otherwise specified, and some care must be taken not to confuse this usage with the more common philosophical or grammatical subjects or objects. All comparisons necessarily involve an element of non-identity, but of course the terms of some comparisons are far more "realistic" (sc. similar) than others, as: "Featured like him, like him with friends possessed" (Shakespeare). Vamana alone of our authors seems to have perceived this problem and allows a simile called tattvâkhyāna, or 'literal', where the end is not praise or blame, but merely precision. Simile here seems to leave the strictly poetic realm, in the sense

efined above; who indeed would deprive the scientist of the use of uch convenient explanatory devices as "sodium reacts like potassium 143 a many compounds". Perhaps it is for this reason that Bhāmaha, he earliest writer we know about, avers that no figure can be really )oetic unless it also has a touch of exaggeration, or atiśayôkti B 2.81, 85), associated with it (repeated by Dandin 2.220). Bhāmaha, Udbhata, and Mammata all emphasize in defining upamā that the things compared must in fact be substantially different-'contrary by reason of place, time, or mode of action', as Bhāmaha says. Mammata boldly defines simile as 'similarity in difference' (sādhar- myam bhede). But the other writers, though they consider only similes which by any standard would be considered poetic, do not appear to have been aware of the problem of over-extension, or "end", as it were. Daņdin says only that upamā is sādrśyam ('similitude'). The problem is perhaps more academic than real, since the non- poetic similes are just those where the expressive potential of simile is least well exploited, that is, where so little difference is understood between the terms that comparison itself is almost otiose. It might almost be said: "give a simile something to do, and it will be poetic". The comparison of sodium and potassium is not un- poetic because of the subject matter, but just because, for all practical purposes, the two things are in fact indistinguishable, are like Tweedledum and Tweedledee (note that the last simile is highly poetic). Simile is limited on the one side by the indistinguishability or literal replacability of its terms, but it also has the same limit on the other side, for beyond simile lies the realm of metaphor (rūpaka), where, despite differences great enough to permit scope to simile, the terms of comparison are identified with each other-said in such a way that sameness alone is suggested and not similarity-as in the phrase "realm of metaphor." The mode or the modal reality of the comparison changes, but the terms of its description do not; in rūpaka, for instance, the object of comparison (upamāna) is "projected onto" (meta-phor) or, as we say, identified with the subject of comparison (upameya): not "her face is like the moon" but "her face is the moon" (the moon of her face delights the evening crowd). The common property is usually not expressed, since the aim of metaphor is to suppress all difference; the comparative particle of course is necessarily absent (but cf. utpreksā, where it reappears in a new sense).

In this way, the various figures involving a duplication of the context are explained and reduced to similes. We need not charac- terize them more fully here, since at least half of this work is con- cerned with just that problem. However, some accounting of the various classifications proposed for simile itself is necessary. The general tendency is for the discriminations or subtypes more and more to be based on the quadripartite structure we have given. The earlier writers, however, while obviously recognizing that structure in defining the basic figure, tended to classify simile in terms of the end or final cause of the comparison. This is especially interesting in view of the universal preoccupation with structure, even among the early writers, in classifying rūpaka. This difference seems to reflect the character of the problem under consideration. Since comparison is always a matter of degree, it would appear appropriate to consider the usage of the various degrees, which is not a question of structure, but of the kinds of things compared and the reasons for that selection of things; but metaphor, being identification pure and simple, is never a question of degree (except in the sense that the metaphor can be more or less well specified in its parts, or complete), and the only relevant question concerns the scope of the identification, which has little to do with the things themselves, but is entirely a matter of the poet's employing or not employing the ideal metaphorical type (see rūpaka). In classifying upamā, the non-structural, or contextual, tradition may be said to begin with Bharata himself, for he allows similes of praise (praśamsā) and blame (nindā), as well as three similes which differ as to the degree of comparability intended by the poet: sadrsi, or entirely comparable, that is, where the subject and object possess the same property to a great degree; kimcitsadrśi, where the same subject shares comparable qualities with several objects and is therefore partially comparable; and kalpita, where, strictly speaking, no comparability at all is alleged; that is, no property is described as common to both subject and object, but rather different descriptive properties are assigned to both which are, in fact, similar (the similarity is not literal, but analogical). The other, or structural, tradition may claim almost the same antiquity, for Bhamaha, who specifically objects to the classification by praise and blame as irrelevant (2.37), enumerates only three kinds of upamā, depending only on the grammatical device by which the similitude is expressed. We have mentioned previously only the

GLOSSARY 145 adverb "like" in this connection, but Bhamaha allows two others beside yathêvaśabda upamā, namely samāsa, where the simile is expressed by a compound word instead of the adverb (moon-faced), and vati, ('possessing the suffix -vat'), where the upamāna is also in bound form-bound not by the upameya, but rather by an ad- verbial suffix having the same meaning as iva (moon-like face). It is in this context that the argument as to the relative antiquity of Bhamaha and Dandin finds its moment. The two authors appear to be engaged in mutual refutation. Bhãmaha not only rejects a sequence of similes in exactly the order in which Dandin gives them (nindāpraśamsâcikhyāsā), mentioning ācikhyāsā, which term is peculiar to Dandin, but in reply, Dandin appears to belittle the classification by grammatical type in his rather offhand enumeration of approximately fifty words and conventions for expressing simile (2.57-65). Elsewhere, Dandin objects to figures which are peculiar to Bhāmaha (upamārūpaka, utprekșâvayava, ananvaya, sasamdeha) in 2.358-59. Much controversy has been occasioned by this chronology, and we make these comments only insofar as the problem may com- pliment that of the sequence of analytical models proposed for simile. Daņdin accepts Bharata's point of view entirely, but advances the classification to an undreamed of degree of subtlety. His treatment of upamā is probably unequaled in the history of alamkārašāstra for its length, perspecuity, and philosophical interest. The thirty- four types illustrate a variety of intuitional situations which the upama may facilitate. From the old varieties of 'praise' and 'blame', we progress to 'judiciousness' (ācikhyāsā), 'confusion' (moha), 'amazement' (adbhuta), 'flattery' (catu). All of these are specific ways of representing (or misrepresenting) the basic similitude, the singled-out property. An important distinction introduced by Dandin and accepted by later authors is that between simile of quality and simile of mode of action. The former is regarded as the typical comparison, and is that which has been described above; in the latter, the notion of common property is broadened to include modes of action: in effect, adjectival similes are replaced by verbal similes (he is as swift as a horse; he runs like a horse). Because the subject and object are related now through a verb, the latter type of simile is called väk yârtha, or 'referring to the entire phrase' (of noun and verb). The simple simile, or simile of property, does not involve the verb. Vāmana also mentions this distinction.

The contextual point of view loses ground after Dandin; only the Agni Purāna is fully committed to it. Vāmana allows the three- fold distinction "stuti [for praśamsā] nindā tattvâkhyāna", presumably more on the authority of Bharata than Dandin, but at the same time, he suggests another distinction which soon becomes dominant, and which is clearly based on the method of Bhämaha. This distinction is the first which clearly sets forth the four elements of simile as criteria. A simile in which all four elements are explicit is called pūrņa, or 'complete'; if one or more of the elements is implicit only, the simile is called lupta, or 'deficient'. Bhāmaha's samāsôpamā would be an example of a lupta upamã, since neither the tertium nor the particle are expressed. Udbhata, who follows Bhāmaha in most matters, here adapts the pūrna-lupta distinction to his predeces- sor's three types and comes up with fourteen varieties of simile. Calling lupta samksepa, or 'ellipsis', Udbhața defines four varieties depending on which element or elements are not expressed (tertium, particle, both, both plus subject). (See sāmyavācaka, tadvāci- samksepa.) Mammața takes up this problem again and goes to absurd lengths to illustrate certain possible ellipses (cf. upameyadyo- takalupta). These types usually amount to Bhāmaha's samāsôpamā, but some involve other principles. Udbhata also improves upon Bhāmaha's category vati, where, it will be remembered, the object of comparison was bound by a comparative suffix -vat. Admitting this type, Udbhata then finds certain other morphological contexts where the object of comparison in some form or other appears in bound form with verb-, adverb-, or adjective-forming suffixes. The Sanskrit language, in fact, allows any noun to be made into a verb having the sense of "behaves like X" (see ācāra); likewise an adverbial accusative in -am, always distin- guished from the accusative case, may express the idea of similitude when suffixed to the object of comparison (see namul). Lastly, other taddhita suffixes than -vat are comparative in meaning (-kalpa, g.v.). Rudrața in a way represents a summation of the structural tradition. He allows the same three types as Bhāmaha and Udbhata, calling them vākyôpamā (not the same as Daņdin's vāk yârthôpamū) for Bhāmaha's yathêvaśabdopamā, as opposed to samāsôpamā, which name Rudrata keeps, and pratyaya, or suffixed similes, by which term Rudrata apparently intends all those formed by suflixa- tion as described by Udbhata. In reference to samāsa upamā, it might be remarked that the compound so formed is a bahuvrīhi. or

GLOSSARY 147

adjectival compound, thus distinguishing the compound which forms a simile from that which forms a rūpaka and which is a karmadhāraya type of tatpurușa. The terms of that compound, as we remark else- where, are inverted. From the simile "moon-face(d)", we get the metaphor "face-moon", but this inversion is just a transparent way of illustrating the contrast in compound type. As vākyôpamā, that is, similes expressed through free (not bound) nouns and adverbs, Rudrata admits six types, all of which are known from other authors but whose selection here again typifies the author's preoccupation with system. The first type (unnamed) is a canonical pūrņa upamā with all four elements explicit; the second shows ellipsis of the common property (sāmānyâbhāva-the same as Dandin's vastu upamā). This pair illustrates the standard simile in which the only variable element is the tertium. Next come two similes which may be called reciprocal or reflexive, in which the subject is in effect compared to itself: this may be done either by not mentioning an object at all (ananvaya: "her face is like her face"), or, when an object is given, by immediately proposing that object as subject (ubhaya: "the moon is like her face"). This is Dandin's upameyôpamã. Lastly, we may have similes in which either the prop- erty (tertium) or the object are assumed to be hypothetical for the purposes of the poet (kalpita and utpādya, or abhūta). Rudrața also mentions a simile which is evidently patterned on the standard treatment of rūpaka and which shows very well the influence of the structural approach to the definition of these figures: he distinguishes samastaviśaya from ekadeśin, assuming the standard metaphorical whole (cf. rūpaka). atiśaya, 'pre-eminence': (1) an upamā in which the similitude is expressed by minimizing the difference between the things compared to the point where they appear as bare facts without qualitative differ- entiation. (2) D 2.22. (3) tvayy eva tvanmukham drstam drśyate divi candramāḥ / iyaty eva bhidā nânyā (Daņdin: "Your face is seen on you, the moon is seen in the sky; so much are they different and no more"). (4) "For the time being he had lost the primitive faculty that instinctively classifies the various sensory impressions according to their relative values. One afternoon he saw a transport truck run into an automobile. But this bloody accident impressed him no more vividly than the sight, a few minutes later, of a scrap of news- paper fluttering in the wind" (Carson McCullers). (5) For variations on the theme of minimizing the difference, see catu, tattvâkhyāna.

148 GLOSSARY

Here the common property is not shared to different degrees, yet the two similars are not "confused" as in samsaya. In catu, the common property is shared to different degrees, but that difference is over- looked. See also pratīyamāna vyatireka, bhedamātra vyatireka, vyatirekarūpaka. adbhuta, 'marvellous': (1) an upama in which a real property of the sub- ject is predicated of the object; a presumed similitude is thus ex- pressed. (2) D 2.24, AP 344.16. (3) yadi kimcid bhavet padmam udbhru vibhrāntalocanam / tat te mukhaśriyam dhattām (Daņdin: "If there were a lotus with arching brows and roaming glance, then it would have the beauty of your face"). (4)"'It's lovely, lovely, lovely', she said, with diminishing cadence, ending in pensiveness once more. 'Do you see that little bit just there? No, not where the trees are-that bare spot that looks brown and warm in the sun. With a little sage-brush, that spot would look something like a place I know on Bear Creek. Only, of course, you don't get the clear air here'" (Owen Wister). (5) Cf. abhūta, asādhāraņa. ananvaya, 'lack of consequence': (1) self-comparison; an upamā in which the subject doubles as object. (2) R 8.11 (12). (3) iyam iyam iva tava ca tanuh ... (Rudrața: "Your body resembles itself alone"). (4) "It was always a source of great preoccupation with the ladies that no bit of pad should show through the natural hair. Often they put up a tentative hand to feel, even in the midst of the most absorbing conversation; and then their faces wore the expression which is seen only on the faces of women whose fingers investigate the back of their heads" (Vita Sackville-West). (5) This figure is an independent alamkāra for Bhāmaha (3.44), Vāmana (4.3.14), Udbhața (6.4), and Mammața (135). Daņdin (2.358) considers it equivalent to his asādhāraņa upamā. aniyama, 'absence of restriction': (1) an upama in which the similitude is said to extend to any object exhibiting the common property. (2) D 2.20, AP 344.12. (3) padmam tāvat tavânveti mukham anyac ca tādrśam | asti ced astu tatkārī (Dandin: "Your face resembles the lotus, and whatever may be said to be similar to the lotus -- why your face resembles that as well"). (4) "When, dearest 1 but think of thee./ Methinks all things that lovely be / Are present, and my soul delighted" (Sir John Suckling). (5) Cf. niyama. This figure differs from ācikhyāsā in that there the idea is that we are debating the appropriateness of the object of comparison. anyonya, 'mutual': (1) an upama in which the similitude is made reciprocal.

GLOSSARY 149 (2) D 2.18. (3) tavânanam ivâmbhojam ambhojam iva te mukham (Dandin: "Your face is like a lotus; the lotus is like your face"). (4) "She walks in beauty, like the night / Of cloudless climes and starry skies; / And all that's best of dark and bright / Meet in her aspect and her eyes" (Byron). (5) Anyonya is the same as upameyô- pamā of Bhāmaha, Vāmana, Udbhața, and Mammata; as paraspa- rôpamā of the Agni Purāņa; as ubhayôpamā of Rudrața. abhūta, 'not happened': (1) an upamā in which the object of comparison is a hypothetical universal and is, strictly speaking, nonexistent. (2) D 2.38. (3) sarvapadmaprabhāsāraḥ samāhrta iva kvacit / tvadānanam vibhāti (Daņdin: "Like the distilled essence of every lotus' beauty, your face is lovely"). (4) "I am thinking of him- loosely I admit-very much as some political cartoonist might think of a generalized and consolidated figure that turns a deaf ear to the Bolshevist and his sinister whisperings ... " (Oliver Onions). (5) The idea is that the object is never exposed to the pettiness of mere experience, yet it can be expressed as an extrapolation on experience. The subject is thus, a fortiori, elevated above the mundane. In adbhuta, there is a transfer of property from subject to object; here the object is impossible in its own terms. In utpādya, the object is hypothetical, but not generalized. asambhava, 'impossibility': (1) an upamā in which an incongruous property, in fact belonging to the subject, is predicated of the object of comparison. (2) V 4.2.20. (3) cakāsti vadane tasyāḥ smitacchāyā vikāsinah / unnidrasyâravindasya madhye mugdhêva candrikā(Vāmana: "A smile appears on her bright face like pure moonlight among sleepless lotuses"). (4) "Holt was constantly attentive: the Admiral's flag-lieutenant hung over her like a decorated cliff" (Nicholas Monsarrat). (5) Asambhava differs from asambhāvita only in being stated positively, and from adbhuta in being stated as a proposition. asambhāvita, 'impossible': (1) an upamā in which the subject of comparison is, strictly speaking, nonexistent; that is, the common property proposed is incongruous or unlikely. (2) D 2.39. (3) candrabimbad iva vişam candanād iva pāvakah / paruşā vāg ito vaktrāt (Daņdin: "A harsh word from her mouth would be like poison from the moon's disc or fire from sandal-paste"). (4) "'There's that Bessy Cranage she'll be flauntin' i' new finery three weeks after you're gone. U'll be bound: she'll no more go on in her new ways without you. than a dog 'ull stand on its hind-legs when there's nobody looking'" (George Eliot). (5) This is ironical comparison, for two things are

150 GLOSSARY

compared through a property they do not have in order to express the opposite property. The other forms of unreal comparison (adbhuta, abhūta) are exaggerations only. asādhāraņa, 'particular': (1) an upamā in which the absence of proper objects of comparison is described. (2) D 2.37, AP 344.19. (3) candrâravindayoḥ kakșyām atikramya mukham tava | ātmanaîvâbhavat tulyam (Dandin: "Exceeding the style of the moon or lotus flower, your face is indeed like itself alone"). (4) "'She [Nature] is', the secretary continued, 'like an assemblage of blondes and brunettes, whose tresses-''Oh, bother the blondes and brunettes!' 'Well, she is like a picture gallery, where the features-' 'No, no; Nature is like Nature; why introduce similes?'" (Voltaire, quoted by E. M. Forster). (5) Asādhārana is a variation of ananvaya, from which it differs in that the object of comparison, though ultimately rejected, is mentioned. ācāra, 'conduct': (1) a type of upamā in which the object is expressed in a verbal form. (2) U 1.17,19. (3) sa duhsthīyan krtârtho'pi nihśeșaî- śvaryasampadā | nikāmakamanīye'pi narakīyati kānane (Udbhața: "He[Śiva], all goals fulfilled in his infinite power, nevertheless suffered [behaved himself in the manner of one suffering] in that forest lovely yet filled with the tortures of hell [behaving like hell; helling]"). (4) "The hills belly-rumbled with thunder" (McDonald Hastings). (5) This facility of Sanskrit is a function of the denominative con- jugation, whereby any noun or adjective can be transformed into a verb having the sense of "resembling, acting like X" (Pāņini 3.1.10). In the following example, a comparison is first drawn, then the object is cleverly substituted for the grammatical subject in order to produce a similar effect: "And there he would lie all day long on the lawn brooding presumably over his poetry, till he reminded one of a cat watching birds, and then he clapped his paws together when he had found the word ... " (Virginia Woolf). By other writers, this type of comparison is classified differently: see pratyaya, upameyadyotaka- lupta, dharmadyotakalupta. Compare also colloquial English "to rat", and "to dovetail," etc. ācikhyāsā, 'wanting to explain': (1) an upamā in which the propriety or aptness of the comparison is in doubt. (2) B 2.37, D 2.32. (3) candreņa tvanmukham tulyam ity ācikhyāsu me manaḥ | sa guņo vâstu doșo vā (Dandin: "Your face is indeed like the moon, but I can't decide whether this is a virtue or a vice"). (4) "Some moralist or mythological poet / Compares the solitary soul to a swan: / I am

GLOSSARY 151 satisfied with that, / Satisfied if a troubled mirror show it, / Before that brief gleam of its life is gone, / An image of its state" (William Butler Yeats). (5) The figure is not defined by Bhamaha, who con- siders the term otiose. Here the intention of the speaker is neither praise nor blame (cf. praśamsā, nindā, to which ācikhyāsā stands as third in a triad); the speaker is rather unable to decide between

irtti, 'implied': (1) a type of pūrņa upamā in which the similarity of the two.

two things is inferred or indirectly expressed. (2) M 127. (3) dur- lokah sa samare nidāghâmbararatnavat (Mammața: "His aspect is as painful to look upon as is the summer sun"). (4) "The candles' ... flames looked at me like the eyes of tigers just waking from sleep" (Joyce Cary). (5) A subtlety is intended. A comparison, strictly speaking, is the expression of a relation of similitude between two things which have a property in common. But properties are distinguished from modes of action or behavior. A comparison based upon a like action permits only an inference as to the similitude of the things as possessing properties. In the example, candles are compared to eyes in virtue of their having a capacity in common- "looked at me": the flame looked at me just as the eyes of tigers might look at me. We may then presume that the eyes resemble the flame as to color, size, or what have you, but this, the literal comparison (cf. śābdī) is not made explicit. The same distinction is involved in the definition of the vākyârtha upamā. stpadya, 'invented': (1) an upamā in which the object of the comparison exists only in the poet's hypothesis and exhibits a striking quality of the subject. (2) R 8.15 (16). (3) kumudadaladīdhitīnām tvak sambhūya cyaveta yadi tābhyah | idam upamīyeta tayā sutanor asyāḥ stanâvaranam (Rudrata: "If skin could be imagined on the reflec- tions of the lotus petals (in the pond), and if that skin might be touched-there would be something that might be comparable to the gossamer of her breasts") (4) "The new moon behind her head, an old helmet upon it, a diadem of accidental dewdrops round her brow, would have been adjuncts sufficient to strike the note of Artemis, Athena or Hera" (Thomas Hardy). (5) In adbhuta upamā, a property is hypothetically transferred from subject to object; here te object is hypothesized as a substratum for the property. Compare the following rupaka, where the basis of the identification is likewise bypothesized: "Dr. N . MacBride had fixed upon me his full, mastering eye: and it occurred to me that if they had policemen in heaven.

152 GLOSSARY he would be at least a centurion in the force" (Owen Wister). utprekșita, 'opined': (1) an upamā in which the similitude is expressed as a relative and subjective opinion about which of several objects of comparison is most likely or approriate. (2) D 2.23. (3) mayy evâsyā mukhaśrīr ity alam indor vikatthanaih | padme'pi sā yad asty eva (Dandin: "Enough of the moon's boasting 'I alone rival the beauty of her face'; her loveliness is found in the lotus, too"). (4) " ... I give you your choice which was the bluest-the aimlessly fluttering butterflies, the nodding harebells, or her demure and reprehensible eyes" (Oliver Onions). (5) The name of this simile may be taken in the sense of "reflected, considered", in which case the emphasis in the examples should be placed on the judicious meditation of the speaker vis-à-vis the scope of his simile. Ut- prekșita differs from samśayôpamā in that the confusion in the lat- ter is between the object and the subject, and from nirnaya in that the object in that case is distinguished from its own subject, not putatively, as here, from several other objects. upamānadharmadyotakalupta, 'ellipsis of the object, common property. and the particle of comparison': (1) self-explanatory term. (2) M 134. (3) mrganayanā harate muner manah (Mammața: "Gazelle- eyes' steals the ascetic's mind away"). (4) "Perhaps he lurks in yonder woodbine bower / To steal soft kisses from her lips. and catch / Ambrosial odours from her passing sighs" (William White- head). (5) The compound mrganayanā ('gazelle-eyed') and the phrase "ambrosial odours" are alone relevant here. Each is a simile in miniature when interpreted, for example, "whose eyes are like the eyes of a gazelle". Only the subject is explicit: "eyes" and "odours" are mentioned but once and are taken as the subjects of comparison. Cf. upameyadharmadyotakalupta. upamānalupta, 'ellipsis of the object': (1) an upamā in which the ebjeet ef comparison is not made explicit. (2) M 129. (3) sakalakar .: ""- paraviśrāmaśrīvitaraņam na sarasakāvyasya | drsvate tha nisa ;ex vā sadrśam amśâmśamātreņa (Mammața: "There is notluing see" or heard which even in the smallest part resembles mood pectry -nothing at all which provides such joyful relaxation of all the senses"). (4) "Per Hansa stood there in the darkness of the winter night, looking after the disappearing figure .... No, her equal was tel to be found!" (O. E. Rolvaag). (5) Mammata's example can be tkeu in two ways. The obviousness of the first borders on pettiness: true po- etry is like nothing (ellipsis through non-existence). This wouldamount

GLOSSARY 153 to asadhārana upama; for example: "Fair was this meadow, as thought me overall; / With floweres sweet embroidered was it all; Asfor to speak of gum, or herb, or tree, / Comparison may none y-maked be" (Chaucer). A more likely interpretation is that poetry is implicit; compared to yoga through the qualification "providing relaxation of all the senses". We have followed the latter interpretation in giving the English example : the absence of a proper object is mentioned only to suggest that Per Hansa's wife has supra-feminine qualities. "She could be both minister and father confessor, that woman!" upameyadyotakalupta, 'ellipsis of the subject and particle of comparison': (1) self-explanatory term. (2) M 133. (3) krpāņódagradordandah sa [rājā] sahasrâyudhīyati (Mammața: "With a sword held in his outstretched hand, the King resembles a man having a thousand weapons (sahasrâyudha)"). (4) "Then Jesse Jones brought a new note into the self-congratulatory deliberations. In a hectoring speech, he advised the banks to improve their capital position ... " (Arthur Schlesinger, Jr.). (5) The point here turns upon a rather flimsy grammatical exegesis. "In a hectoring speech" must be taken to mean "in a speech in which he behaved himself like Hector", as the Sanskrit is taken to mean "he behaved himself like Sahasrâyudha". "Himself" becomes the subject of the comparison as grammatical object of the verb "behave" and parallel to "Hector". Hence the ellipsis. Many of Mammata's classifications are similarly far-fetched. Cf. ācāra. upameyadharmadyotakalupta, 'ellipsis of the subject, the common property, and the particle of comparison': (1) self-explanatory term. (2) U 1.17. (3) tām šaśicchāyavadanām ... gaurīm prati mano dadhau (Udbhata: "[Śiva] contemplated Gaurī, whose face had the beauty of the moon"). (4) " ... he recognized the pail-of-water-over-the-head experience ... " (Margery Allingham). (5) In the Sanskrit, "[the beauty of whose] face [is like] the beauty of the moon", only the latter beauty is explicit. In the English, the experience which the pail of water, etc. suffices to characterize is not named in so many words. Cf. upamänadharmadyotakalupta. For the problem of distinguishing such similes from metaphors (rūpaka), see lupta upamā. ubhaya, 'both': (1) same as anyonya. (2) R 8.9. ekadeśin, 'having parts, partial': (1) a multiple upamā wherein several corresponding parts of the subject and object are compared without that comparison being extended to the principal terms themselves.

GLOSSARY 154 (2) R 8.29 (31). (3) kamaladalair adharair iva dasanair iva kevaray virājante | alivalayair alakair iva kamalair vadanair iva nalın,ah (Rudrața: "The lotuses are splendid-their petals like lipy, tha .. filaments like teeth, bees like locks of hair swarming about the face-like blooms"). (4) " ... Mr. Moseley came in and knocke o the counter with a half crown. His face was as red as red ink; and he had a complete new colour scheme, all in browns. Brown the colour of old ale. Golden brown tie like lager. Brown bo shining like china beer handles. Guinness socks. And a new brown bowler, the colour of bitter beer, over his left eye" (Joyce Cary). (5) See samastavastu upamā, in which the major terms are men- tioned. ekabahu, 'singular-plural': (1) an upamā in which the subject is plural and the object singular. (2) NS 16.42 (43). (3) śaśânkavat prakāsante jyotīmsi (Bharata: "The stars shine like the rabbit-marked moon"). (4) "These parties of theirs ... were like chain-smoking: each cigarette was lighted in the hope that it might be more satisfactory than the rest" (Vita Sackville-West). (5) By later writers, non-parallelism of number is considered a defect (vacanabheda dosa). Here we have a curious inconsistency in the general tendency to preserve a classi- fication at any cost. The terms of most similes are, of course. paral- lel: "She stood breast-high among the corn, / ... Like the sweet- heart of the sun" (Thomas Hood). kalpapprabhrti, 'having the form of, etc.': (1) an upama which contains such an expression of comparison instead of the comparative particle ("as", "like"). (2) U 1.21. (3) candālakalpe kandarpam pluşpā mayi tirohite / samjātâtulanairāśyā kim sā śokān mrtā bhavet (U'drha- ta: "While I was hidden there in the form of an outcaste man. Love was consumed; and she [Pärvati], in whom an immense despair was born, appeared about to die of sorrow"). (4) " ... drawn with Püreres- que vigor and dash" (Thomas Hardy). (5) For other examples of similes formed with taddhita suffixes, see s.v. and sadrso. ses kalpita, 'artificial': (1) an upama wherein the similitude is stated i werns of comparable properties of the subject and object, but net threeg? one property, common to both. (2) N$ 16.46 (49), AP 344 21. RS 1 (14). (3) mukham āpūrņakapolam mrgamadalik hitardhapatpaks : te / bhāti lasatsakalakalam sphuțalānchanam indubin:bon: : (Ri:a ta: "Your face, full-cheeked and bearing the beauty marks ef mask. resembles the full moon's orb with its argent spots"). (4) "I s .. the professor winking at me so hard that his face was hke a con-

certina with a hole in it" (Joyce Cary). (5) Cf. sadrśavyatireka. kimcitsadrśi, 'somewhat similar': (1) an upamā in which one subject is compared to several objects by means of several properties. (2) NŚ 16.46 (51), AP 344.21. (3) sampūrņacandravadanā nīlótpalada- lêkşaņā / mattamātangagamanā samprāptêyam sakhī mama (Bharata: "My friend has been at last secured whose face is like the full moon, whose eyes resemble the petals of the blue lotus, whose walk is undulant like a rutting elephant"). (4) "Eustacia's dream ... had as many ramifications as the Cretan labyrinth, as many fluctuations as the Northern Lights, as much colour as a parterre in June, and was as crowded with figures as a coronation" (Thomas Hardy). (5) This figure differs from utpreksita in that property is there unique, from ekadeśin in that the similitudes are there subsidiary in a whole, and from samuccaya in that there we have but one object. gamana, 'going': (1) probably the same as raśanā upamā (2) AP 344.20. (5) Since the Agni Purāna gives no examples, the definition of this unique item is subject to caution. The text reads: "upameyam yad anyasya tad anyasyôpamā matā / yady uttarottaram yāti tadâsau gamanôpamā" ('if the upameya of one term is deemed the upamā of another, and the sequence is continued, then this is gamana'). catu, 'flattering words': (1) an upama in which the real difference between subject and object-that of possessing the common property to a lesser and a greater degree, respectively-though recognized, is voluntarily ignored. (2) D 2.35. (3) mrgêkşanânkam te vaktram mrgeņaîvânkitaḥ śaśī | tathâpi sama evâsau nôtkarșī (Daņdin: "Your glance is learned from the gazelle. The moon is marked with the gazelle itself: he is thus quite similar, but in no way superior"). (4) "Ask me no more where Jove bestows, / When June is past. the fading rose; / For in your beauty's orient deep / These flowers. as in their causes, sleep" (Thomas Carew). (5) Catu differs from atisaya in that there the difference between the subject and object. as far as the common property is concerned, is cancelled. namul (fictive for the accusative absolute in -am): (1) an upama in which the force of the comparison is rendered by means of such a construe- tion. (2) U 1.20. (3) sa dagdhavigraheņâpi vīryamātrasthitātmanā sprstah kāmena sāmānyaprāņicintam acintayat (Udbhața: "Touched by Love, though Love's body was consumed and his force consisted of heroism alone, Śiva remained pensive with cares common to all men"). (4) "Another head came into view from behind the wings of a chair, and its owner glared at us with a Harvard accent" (Peter de

156 GLOSSARY

Vries). (5) The English example illustrates the grammatical point only grosso modo, of course. The comparison of Harvard glances with Harvard speech is expressed via an adverbial clause which is functionally similar to the Sanskrit "had the cares which all men have". Mammața (130) calls this a type of ayotakalupta. tattvâkhyāna, 'literal description': (1) an upamā in which the similitude is assumed to lend itself to a confusion, so that one is obliged to identify the subject and object of the comparison. (2) D 2.36. (3) na padmam mukham evêdam na bhrngau caksușī ime (Daņdin: "That is no lotus, that is a face; those are not bees, they are eyes"). (4) "I have seen roses damask'd, red and white, / But no such roses see I in her cheeks" (Shakespeare). (5) This figure differs from nirnaya in that there a state of doubt or suspicion is assumed, here there is merely a possibility of confusion. In hyperbolic exaggeration, tattvâkhyāna is the next step beyond catu. Cf. tattvâpahnava rūpaka. tattvâkhyäna (II): (1) an upama in which the aim is merely to represent- neither to praise or blame. (2) V 4.2.7. (3) tām rohiņīm vijānīhi jyotişām atra maņdale / yas tanvi tārakānyāsah śakațâkāram āśritah (Vāmana; Rohiņī or Taurus, the "red one", is so called from Al- debaran, the main star: "Know that configuration of stars to be Rohinī which in form resembles a cart"). (4) "To my notion all of the early part of Mourning Becomes Electra has the sinewy and homely narrative strength of-let me reach for a comparison which does him neither too little nor too much honor-a novel by Charles Reade" (Alexander Woollcott). (5) For Vāmana, this is the middle term in the triad stuti ... nindā. Cf. Daņdin, where tattvâkhyāna is replaced by ācikhyāsā-doubt as to the appropriateness of praise or blame. taddhita, 'secondary suffix': (1) an upamā utilizing such a suffix to express the comparison. (2) U 1.20, M 127. (3) (4) For examples, see subtypes vati, kalpapprabhrti. (5) The well-known grammatical term. Mammața distinguishes taddhita upamā from similes formed by samāsa (compounding), q.v. tadvācisamkșepa, 'ellipsis of the comparative particle': (1) same as dyotakalupta. (2) U 1.18. tulyayoga, 'conjunction of equals': (1) an upama in which the object is of strikingly exalted station vis-à-vis the subject. (2) D 2.48 (49). (3) divo jāgarti rakșāyai pulomârir bhuvo bhavān | asurās tena hanyante sâvalepās tvayā nrpāh (Dandin: "Indra keeps watch in heaven and you, O Lord, on earth; demons are slain by him and by you, the

GLOSSARY proud and haughty"). (4) "And then the hyena laughed out. Pleased 157 at such an arrangement! Pleased at having her enemy converted into a dean with twelve-hundred a year! Medea, when she describes the customs of her native country ... assures her astonished auditor that in her land captives, when taken, are eaten. 'You pardon them?' says Medea. 'We do indeed,' says the mild Grecian. 'We eat them!' says she of Colchis, with terrific energy. Mrs. Proudie was the Medea of Barchester; she had no idea of not eating Mr. Slope" (Anthony Trollope). (5) Cf. tulyayogitā alamkāra where the same conjunction is expressed literally and not through the use of figurative devices (simile or metaphor). dyotakalupta, 'ellipsis of the comparative particle': (1) self-explanatory term. (2) M 130. (3) tataḥ ... kāminīgaņdapāņdunā / ... candreņa māhêndrī dig alamkrtā (Mammața; the Sanskrit translates literally as "lover's-cheek-pale moon": "The eastern quarter is adorned by the moon pale as a lover's cheek"). (4) "Her forehead ivory white" (Edmund Spenser). (5) Mammata also includes here certain denom- inative constructions: cf. ācāra and dharmadyotakalupta. This variety of ellipsis is also known as tadvāci, pratyaya, vādi, q.v. dharma, 'property': (1) an upam in which the similitude is spelled out by mentioning the comparable property or aspect of the two terms. (2) D 2.15, AP 344.10. (3) ambhoruham ivâtāmram mugdhe karatalam tava (Dandin: "The palm of your hand is like a pale lotus"). (4) "River roughed up with little waves like the flat side of a cheese grater" (Joyce Cary). (5) Cf. vastu, where that property is implicit. The notion of property is here taken in an exact sense, and presum- ably excludes those similes based on mode of action or result; cf. vākyârtha. Vāmana uses the word guna instead of dharma; cf. the commonplace distinction guņa-kriyā. Here is an example of a simile whose common property is a mode of action: "Richard arrested his resumption of speech, and he continued slowly to fizz like an ill-corked effervescence" (George Meredith). dharmadyotakalupta, 'ellipsis of the common property and the particle of comparison': (1) self-explanatory term. (2) M 131. (3) savitā vidhavati vidhur api savitarati tathā dinanti yāminyah | yāminayanti dināni ca sukhaduḥkhavašīkrte manasi (Mammața; in the Sanskrit, all the upamāna are denominative verbs: "The sun resembles the moon and the moon, the sun; the hours of the night are as those of the day and those of the day, the night for one whose mind is afflicted by the round of pleasure and pain"). (4) "No profane hand shall

dare, for me ... to Bowdlerize my Shakespeare ... " (Anon., quoted in Burton Stevenson; here one is enjoined from turning the author's edition into one like Dr. Bowdler's in the matter of expurgation-the common property). (5) The figure is also known as sāmyatadvācivi- cyava. Cf. ācāra. Udbhața's example is tridaśâdhîśaśārdūlah ('Indra- tiger'), an epithet of Siva; compare a term like "moonstone". dharmalupta, 'ellipsis of the common property': (1) self-explanatory term. (2) M 128. (3) rājīvam iva te vaktram netre nīlôtpale iva (Dandin: "Your face is like a lotus; your eyes are like lotus petals"). (4) "My delight and thy delight / Walking, like two angels white, / In the garden of the night" (Robert Bridges). (5) This figure is also known as vastu, sāmyavācakasamkșepa. Cf. dharma upamā. dharmôpamanalupta, 'ellipsis of the common property and the object of comparison': (1) self-explanatory term. (2) M 132. (3) tunțuņāyamāno marişyasi kanțakakalitāni ketakīvanāni / mālatīkusumasadrkșam bhramara bhraman na prāpsyasi (Mammața: "Buzzing about in the thorny ketakī groves, O bee! you will surely die; yet you will not resemble the mālatī flower"). (4) "For her own person, / It beggar'd all description" (Shakespeare). (5) Mammata's example requires such a tortuous interpretation that this commentator blushes to give it. "You will never attain similarity with the mālati flower" is taken to mean: "the mälatī flower is like nothing else in the world insofar as you are concerned". Cf. upamānalupta and the note thereon. nindā (I), 'blame': (1) an upamā whose intention is to depreciate or belittle and whose object of comparison is therefore pejorative. (2) NŚ 16.46 (48), V 4.2.7. (3) ... kalatram / hālāhalam vişam ivâpagunam (Vāmana; the poison was so deadly it threatened to kill all life: "An evil wife is like the poison Śiva swallowed"). (4) "Both of you are good at keeping secrets-like onions on the breath ... " (Joyce Cary). (5) Ninda is the opposite of stuti, 'praise. Cf. ācikhyāsā, tattvâkhyāna. nindā (II): (1) an upamā wherein, by an ironic depreciation of the object, flattery of the subject is intended. (2) B 2.37, D 2.30, AP 344.21. (3) padmam bahurajaś candrah kşayī tābhyām tavānanam ; samānam api sôtsekam (Dandin: "The lotus is spotted with pollen, the moon wanes; your face, though similar, is more proud"). (4) "If when the sun at noon displays / His brighter rays, / Thou but appear, / He then all pale with shame and fear, / Quencheth his light., Hides his dark brow, flys from thy sight, / And grows more dim : Compared to thee than stars to him" (Thomas Carew). (5) See above.

GLOSSARY niyama, 'restriction': (1) an upamā in which the similitude is said to be 159 limited to the object in question. (2) D 2.19, AP 344.12. (3) tvan- mukham kamalenaîva tulyam nânyena kena cit (Dandin: "Your face may be compared to the lotus and to the lotus alone"). (4) "He looked over his paper with that plump, gratified satisfaction at a chance to shine which in the dog world is the peculiarity of the hound" (Margery Allingham). (5) Cf. aniyama. nirņaya, 'deduction': (1) an upamā in which the two comparable things are distinguished from one another through a deduction based upon, but critical of, their excessive similarity. (2) D 2.27. (3) na padma- syêndunigrāhyasyêndulajjākarī dyutih | atas tvanmukham evêdam (Dandin: "That can't be the gleam of a lotus putting the moon to shame, since the lotus is liege to the moon; it must be your face"). (4) "He was as a ghost, all whose power of wandering free through these upper regions ceases at cockcrow; or rather he was the oppo- site of a ghost, for till cockcrow he must again be a serf" (Anthony Trollope; reference is made to the dependence of the Bishop on his wife). (5) In tattvâkhyāna, the same distinction is made, but without the semblance of an argument. niścaya, 'decision': (1) probably the same as nirnaya. (2) AP 344.12. (5) No example is given, but the commentary on D 2.27 equates this term with nirņaya. padârthavrtti, 'whose scope is the meaning of a word': (1) an upama which expresses a relationship between things in terms of a common prop- erty, not between actions in terms of analogy. (2) V 4.2.3. (3) haritanuşu babhrutvagvimukhāsu yāsām | kanakakaņasadharmā mānmatho romabhedaḥ (Vāmana: "On whose golden bodies, now divested of their clothes of reddish bark, was seen the lovely thin line of hair resembling a string of golden beads"). (4) "Her breast like to a bowl of cream uncrudded ... " (Edmund Spenser). (5) Cf. vakyârthavrtti, dharma. paraspara, 'mutual': (1) same as anyonya. (2) AP 344.11. pūrņa, 'full': (1) an upamā in which the four characteristic elements of the comparison are explicitly stated. (2) V 4.2.5, R 8.5 (6), M 126. (3) svapne'pi samareşu tvām vijayaśrīr na muñcati | prabhāvaprabha- vam kāntam svâdhīnapatikā yathā (Mammața: "Even in the sleep between battles, the Goddess of Victory cleaves to you, O King, like a faithful wife to her excellently beautiful lover"). (4) "The moon was coming up ... making . the houses look like fresh cut blocks of coal, glittering green and blue" (Joyce Cary). (5) The four charac-

160 GLOSSARY

teristic elements of every simile, whether explicit or not, are (a) the subject of comparison (upameya: "houses"), (b) the object of comparison (upamāna: "blocks of coal"), (c) the common property (sādhāraņadharma: "glittering ... "), and (d) the comparative particle (dyotaka: "like"). Cf. lupta and the various terms mentioned for examples of ellipsis. See upama for the problem of translation, and the appropriate terms for a discussion of their meaning and context. Pūrņa is described by Rudrata, but not named. prativastu, 'counterpart': (1) example; an upamā in which the object of comparison is introduced as the subject of another situation which manifests the relevant common property and in which the comparative particle is absent. (2) B 2.34-35 (36), D 2.46 (47). (3) naîko'pi tvādrśo'dyâpi jāyamāneșu rājasu / nanu dvitīyo nâsty eva pārijātasya pādapaḥ (Daņdin: "There is not even one who resembles you among the victorious kings; but then, the coral tree has no imitator either"). (4) "Follow a shadow, it still flies you; / Seem to fly it, it will pursue: / So court a mistress, she denies you; / Let her alone, she will court you. / Say, are not women, truly, then, / Styled but the shadows of us men?" (Ben Jonson). (5) Prativastu is considered by others to be a separate figure, perhaps because the comparative particle is necessarily absent. In this it differs from vākyârtha upamā. pratiședha, 'prohibition': (1) an upamā in which certain flaws of the object are said to vitiate the comparison. (2) D 2.34. (3) na jātu śaktir indos te mukhena pratigarjitum / kalankino jadasya (Daņdin: "Indeed the moon cannot rival your face; mark its cold and blemished air"). (4) "Shall I compare thee to a summer's day? / ... Rough winds do shake the darling buds of May, / And summer's lease hath all too short a date" (Shakespeare). (5) This figure differs from nindā in that here the mood is the indicative rather than the "optative". In vyatireka, the virtue of the subject, rather than the vice of the object, is usually alleged as prohibition. pratyaya, 'suffix': (1) an upamā in which the force of the comparison is rendered by a verbal suffix. (2) R 8.23 (24). (5) Pratyaya is considered by Rudrata to be an ellipsis of the comparative particle. Cf. dyotaka- lupta. The same as ācāra. praśamsā, 'praise': (1) un upamā whose intention is to appreciate or praise, that is, whose upamāna is mejorative. (2) NŚ 16.46 (47). B 2.37, D 2.31, AP 344.21. (3) brahmano'py udbhavah padmas candrah śambhuśirodhrtah | tau tulyau tvanmukhena (Dandin: "The lotus is born of Brahman, the moon is fixed on the brow of Siva:

both resemble your face"). (4) "She stood breast-high among the 161 corn, / Clasp'd by the golden light of morn, / Like the sweetheart of the sun, / Who many a glowing kiss had won" (Thomas Hood). (5) This figure illustrates the definition of the upamāna (object of comparison) as "that term in which the property resides to a higher degree"; by drawing a comparison with an exalted object, the subject necessarily participates in its elevation. Cf. ninda, which can also praise the subject through irony. Praśamsā is the same as stuti. bahu, 'many': (1) an upamā in which a number of different objects are mentioned. (2) D 2.40, AP 344.14. (3) candanôdakacandrâmśucan- drakāntâdišītalaḥ / sparśas tava (Daņdin: "Your touch is cool as moonstone, as the fall of moonbeams, as sandal-water"). (4) "As lightning, or a taper's light, / Thine eyes, and not thy noise, waked me" (John Donne). (5) Bahu differs from utprekşita in that here there is no effort to find the right object; from māla II in that there the multitude of objects manifests a multitude of properties, here there is but one property. bahveka, 'plural-singular': (1) an upamā in which the subject is singular and the object plural. (2) NŚ 16.42 (44). (3) śyenabarhiņabhāsānām tulyârthah [sa kaścit] (Bharata: "He is like eagles, peacocks, and hawks"). (4) "Behold a critic, pitched like the castrati" (Theodore Roethke). (5) See the note on ekabahu. An example of comparing plural with plural is: " ... elegant shoppers wrapped like dainty bears" (Edgell Rickword). Bharata, not an accomplished classifier, neglects the possibilities offered by the Sanskrit dual. mālā (I), 'garland': (1) an upamă in which a series of comparisons are given which not only involve the same similitude, but in which a qualification of that similitude becomes the subject of the following simile. (2) D 2.42. (3) pūșņy ātapa ivâhnîva pūşā vyomnîva vāsaraḥ / vikramas tvayy adhāl lakșmīm (Daņdin: "Victory founded its good fortune on you, just as the heat did in the sun, the sun did in the day, and the day did in the sky"). (4) "He moves among men as most men move among things" (Bernard Shaw). (5) This type of mālā differs from the following in that there but one similitude is stated. It differs from raśanā in that the architectonic moves from substratum to manifestation rather than from subject of comparison to object of comparison. Cf. asambhāvita. mālā (II): (1) an upamā in which one subject is compared to several objects through one or several properties. (2) AP 344.15, R 8.25 (26), M 134. (3) śyāmālatêva tanvi candrakalêvâtinirmalā sā me | hamsîva kalâlāpā

162 GLOSSARY

caitanyam harati nidrêva (Rudrața: "Slender as the dark creeper, spot- less as the new-born, waxing moon, soft-throated as the swan. she steals my reason as do dreams"). (4) "What follows should be prefaced with some simile-the simile of a powdermine, a thunder- bolt, an earthquake-for it blew Philip up in the air and flattened him on the ground and swallowed him up in the depths" (E. M Forster). (5) This figure is the same as kimcitsadrsi, except that here the possibility of one property is allowed, at least by Mammata: "My heart is like a singing bird / Whose nest is in a water'd shoot; / My heart is like an apple-tree / Whose boughs are bent with thick- set fruit; / My heart is like a rainbow shell / That paddles in a halcyon sea; / My heart is gladder than all these, / Because my love is come to me" (Christina Georgina Rossetti). Bhāmaha (2.38) mentions the term mālā, but not in a way that would permit precise definition of its significance. moha, 'bewilderment': (1) an upama in which the two terms of comparison are confused with one another. (2) D 2.25, AP 344.17. (3) śaśîty utpreksya tanvangi tvanmukham tvanmukhâśayā / indum apy anu- dhāvāmi (Dandin: "Now I'm running about after the moon, seeking for your face, for I thought that your face was the moon"). (4) "'When I slung my teeth over that,' he remarked, 'I thought I was chewing a hammock'" (Owen Wister). (5) Moha differs from bhrān- timat alamkāra only in that the comparability of the two confused terms is here necessarily paramount. yathêvaśabda, 'the words yathā (as) and iva (like)': (1) an upamā wherein the force of the comparison is borne by one of these words, the usual adverbial particles of comparison. (2) B 2.31, U 1.16. (3) kşananı kāmajvarôtthityai bhūyaḥ samtāpavrddhaye | viyoginām abhūc cāndrī candrikā candanam yathā (Udbhața: "The moonlight of the full moon, like sandal paste, rouses the sudden fever of love in parted lovers and so increases their suffering"). (4) "And there was Hetty. like a bright-cheeked apple hanging over the orchard wall" (Geerge Eliot). (5) Yathêvaśabda is to be distinguished from those similes expressed through compounding (samāsa). See also dyotakalse :. upamā. raśanā, 'rope': (1) a concatenation of upamās in which the subject of comparison of the first simile is the same as the objeet of comparisen of the following. (2) R 8.27 (28), M 134C. (3) nabha iva vimal !!; salilam salilam ivânandakāri šaśibimbam | šaśibimbam iva lasadiru !! taruņīvadanam śarat kurute (Rudrața: "The autumn scason mtakes

GLOSSARY the crystal water clear as sky, the round, refreshing moon limpid as 163 water, the maiden's coquettish mien like the glancing moon"). (4) "If when the sun at noon displays / His brighter rays,/Thou but appear, / He ... / ... / ... grows more dim / Compared to thee than stars to him" (Thomas Carew). lapta, 'ellipsis': (1) an upamā in which at least one of the four characteris- tic elements is not explicitly stated. (2) V 4.2.6, M 126. (3) (4) See the various subtypes grouped under the names of the element drop- ped: upameya, upamāna, dyotaka, dharma. (5) All the writers implic- itly recognize this type, beginning with Bhāmaha who distinguishes similes containing a particle of comparison (yathêvaśabda) from those formed by compounding and therefore without such a particle. Likewise, Dandin distinguishes dharma and vastu upamās on the basis of the former mentioning the common property and the latter not. The distinction between pūrņa and lupta upamā has, however, become such a commonplace in the later poetics that it is usually imposed by commentators whenever possible upon the earlier writers, even though they manifestly had other reasons for arranging their distinctions in the way they did. Mammata shows the way, being the first writer to ignore completely considerations of subject matter and intention in defining simile in favor of elements of construction. This may be called the triumph of the material principle over the final. I have the following simile to add to Mammata's collection, which seems to exhibit ellipsis of both the subject and object of comparison: "Smell of boot polish like a lion cage" (Joyce Cary). Here someone's boots are being compared to those of a lion tamer. vati, 'the suffix -vat ('like'): (1) an upamā wherein the force of the com- parison is borne by such a particle suffixed to the object of com- parison. (2) B 2.33, U 1.20-21, M 127. (3) dvijātivad adhīte 'sau guruvac cânuśāsti nah (Bhāmaha: "Brahmin-like he studies; guru- like he instructs us"). (4) "Lion-like March cometh in" (W. D. Howells). (5) This is the example par excellence of the use of a taddhita suffix in forming similes. In English, the same word may be used in or out of compound, but in Sanskrit, the morphemes are different: -vat only in compound, iva never in compound. In the ordinary uncompounded simile (cf. vāk yârthavrtti and pādârthavțtti). this type is subdivided into those which express a nominal comparison and those which express a verbal comparison. The examples given illustrate the latter subtype, which seems more natural. Compare the

164 GLOSSARY

phrase " ... drawn with Düreresque vigor and dash" (Thomas Hardy), which expresses a purely nominal similitude (vigor like that of Dürer). This type is recognized by Yāska (3.17), who asserts that it expresses a perfect or total similitude (siddha), as opposed to a partial or presumed similitude expressed by iva. vastu, 'the real thing': (1) same as dharmalupta upamā. (2) D 2.16, AP 344.10. (5) The name implies, according to the commentary, that the emphasis is to be placed on the things compared, rather than on the common property. See the note on lupta. vākya, 'phrase': (1) an upamā in which the comparison is expressed in the form of a phrase, that is, a relation of independent words. (2) R 8.5-16, M 127. (3) svapne'pi samaresu tvām vijayaśrīr na muñcati / prabhāvaprabhavam kāntam svâdhīnapatikā yathā (Mammața: see pūrņa for the translation). (4) "Let us go then, you and I / When the evening is spread out against the sky / Like a patient etherized upon a table" (T. S. Eliot). (5) In this sense, the term is distinguished from similes formed by compounding and those which are telescoped into verbs (cf. samāsa, taddhita, pratyaya). As an instance of the baroque complexity which these classifications can attain, take the two terms pūrņa upamā and vākya upamā. For Mammața, vākya is the first subtype of pūrņa; for Rudrața, pūrņa is the first subtype of vākya. Although both authors define the term vākya in approxim- ately the same way, the system of classification in which the term figures obliges us to modify that meaning slightly and consider its two occurences to be of different scope. For Rudrata, a dharmalupta upamā is a type of vākya on the same level as a pūrņa; for Mammața, it is simply a non-pūrņa and may or may not be a vākya. vākyârthavrtti, 'whose scope is the phrase': (1) an upamā whose two terms extend each to an entire phrase or sentence. (2) D 2.43 (44- 45), V 4.2.3, AP 344.19. (3) tvadānanam adhīrâkșam āvirdašanadīdhiti bhramadbhrngam ivâlakşyakesaram bhāti paūkajam (Dandin: "your face of gently roaming glance and lustrous smile gleams like a lotus with its darting bees and filaments so fine"). (4) "The readers of the Boston Evening Transcript / Sway in the wind like a field of ripe corn" (T. S. Eliot). (5) These similes extend to the entire phrase in the sense that the similitude involves, and in fact is basically a func- tion of, the verb. On the other hand, a "simple" simile expresses a direct relationship between two nouns through a common property (cf. pādârthavrtti) and does not involve the sentence itself, that is, the grammatical association of noun plus verb. A simile extending

GLOSSARY 165 to the verb is thus coextensive with the phrase and is sometimes thought of as a simile of action or mode of behavior. A good test for discriminating such a phrasal simile is this: the same verb is either repeated, as: "The daylight struck down with a pallid glare upon the tatters of soot draping the flue as sea-weed drapes a rocky fissure" (Thomas Hardy), or must be supplied in the other of the two phrases (as in the example from Eliot) to make sense. "My Luve's like a red, red rose" (Robert Burns) requires no such suppletion. vâdilopa, 'ellipsis of vā, etc.': (1) same as dyotakalupta. (2) M 130. (5) vā is a term standing here for the ensemble of comparative particles. Though its usual meaning is 'or', it can be taken in the sense of iva according to Böhtlingk and Roth. vikriyā, 'transformation': (1) an upamā in which the subject of comparison is expressed as a transformation or modification of the object. (2) D 2.41, AP 344.15. (3) candrabimbād ivôtkīrņam padmagarbhād ivôddhrtam / tava tanvangi vadanam (Daņdin: "O slender-limbed, your face seems carved from the moon's circle or raised from the lotus' bud"). (4) "Lowood shook loose its tresses; it became all green ... and it made a strange ground-sunshine out of the wealth of its wild primrose plants" (Charlotte Bronte; here the object is expressed as a transformation of the subject). (5) In the post-dhvani or encyclopaedic writers on figuration, this variety of simile is raised to the status of a separate figure, called pariņāma (transformation). Cf. Ruyyaka, Alamkārasarvasva (KM edition, p. 51). viparīta, 'reversed': (1) probably the same as viparyāsa. (2) AP 344.11-12. viparyāsa, 'transposition': (1) an upamā in which that term which in the order of nature is the subject of comparison is cast in the form of the object, and, similarly, the object term is cast as the subject. (2) D 2.17. (3) tvadānanam ivônnidram aravindam abhūt (Daņdin: "The full-blown lotus was like your sleepless face"). (4) "The fowers did smile, like those upon her face" (William Drummond). (5) By "order of nature", we refer to the definitions of the subject and object as those terms in which the common property resides to a lesser and to a greater degree, respectively. In this type of simile, each of the two terms is expressed in the formal position naturally appropriate to the other, thus exaggerating the prominence of the in fact inferior subject. In catu, there is merely a cancellation of this difference, not an inversion. virodha, 'opposition': (1) an upamā in which the similitude is so expressed as to imply rivalry on the part of the things compared. (2) D 2.33.

166 GLOSSARY (3) śatapattram śaraccandras tvadānanam iti trayam | parasparaviro- dhi (Dandin: "The hundred-petaled lotus, the autumn moon, your face-these three are warring"). (4) "'Speak,' she said, 'thou fairest; Beauty thou impairest ... '" (Henry Constable; here Venus addresses Adonis). (5) The idea seems to be that nothing breeds incompatibility like similitude. In atiśaya, the difference between the terms of com- parison is minimized; here that minimum is dialectically turned into its opposite: mutual contradiction. vyatireka, 'distinction': (1) an upamā of the Agni Purāna whose meaning is unclear. There are no parallels. (2) AP 344.14. (5) The text reads: "yad ucyate'tiriktatvam vyatirekôpamā tu sā" ("where pre- eminence is expressed, that is called vyatirekôpamā"). This figure is probably the same as atiśaya upamā. śrauti, 'audible': (1) an upamā in which the force of the comparison is made explicit. (2) M 127. (3) svapne'pi samareşu tvām vijayaśrīr na muñcati / prabhāvaprabhavam kāntam svâdhīnapatikā yathā (Mam- mața; see pūrņa for the translation). (4) "However, I kept myself safe yet, though I began, like my Lord Rochester's mistress, that loved his company, but would not admit him farther, to have the scandal of a whore, without the joy" (Daniel Defoe). (5) A subtlety is intended: cf. ārthī and vākya upamā. ślesa, 'double-entendre': (1) an upamā in which the common property is replaced by a pun. (2) D 2.28. (3) śiśirâmśupratispardhi śrīmat surabhigandhi ca | ambhojam iva te vaktram (Daņdin; "rival"means "enemy" when applied to the lotus, "similar to" when applied to the face, and Śrī refers to the goddess when applied to the lotus. to "beauty" when applied to the face: "Like the lotus is your face: moon's rival, abode of Śri, perfumed"). (4) "'Now it's time I was up at the office to get my vay-bill and see the coach loaded: for coaches ... is like guns-they requires to be loaded with wery great care, afore they go off'" (Charles Dickens; the venerable Mr. Weller, Sr. speaking). (5) Here we have an example of the ubiquity of ślesa alamkāra; Daņdin regularly expresses interrelationships of figures by considering one a subtype of another. saņśaya, 'doubť': (1) an upamā in which doubt is evpressed as to which of the two things being compared is which. (2) D 220. AP 344.18. (3) kim padmam antarbhrāntali kim te loleksare": " :: a": mama dolāyate cittam (Dandin: "My mind doth pender well: is it a lotus bud with captive bees or a stoe-eyed maiden's face?"). (4) "I observe: 'Our sentimental friend the moon! Or possibly

GLOSSARY (fantastic, I confess) / It may be Prester John's balloon / Or an old 167

battered lantern hung aloft /To light poor travellers to their distress'" (T. S. Eliot). (5) If such a doubt is subjected to reasoning, we have nirnaya; if related to other people's opinion, mata. See also samdeha alamkāra and subtypes. samkşepa, 'ellipsis': (1) same as lupta. (2) U 1.17. (5) Only four types are given by Udbhata: ellipsis of the property, of the particle, of both, and of both plus the subject. See sāmyavācaka, tadvāci. Mammața gives nineteen types of lupta in all. sadṛśapada, "the word 'resembling'": (1) an upamā wherein a word such as sadrśa expresses the force of the comparison. (2) U 1.16. (3) prabodhād dhavalam rātrau kiñjalkâlīnasatpadam | pūrņendubimbena samam āsīt kumudakānanam (Udbhața: "The lotuses were quite similar to the orb of the full moon-freshly white from blooming and drawing the night bees to their pollen cups"). (4) " ... and their other North Oxford acquaintances of the same kidney" (Michael Innes). (5) Udbhata probably intends by this term that large and vague category of words capable of expressing the idea of resem- blance. He thinks of the two most common (yathā, iva) as different, probably in the sense that they set up the norm to which the others approximate. sadrśi, 'similar': (1) an upamā in which two things are represented as fully comparable. (2) NS 16.50, AP 344.21. (3) yat tvayâdya krtam karma paracittânurodhinā | sadrśam na tathaîva syād atimānușa- karmanah (Bharata: "What you did today out of compassion for another could be compared only to the deed of a superhuman soul"). (4) "T. S. Eliot resembles one of those mighty castles in Bavaria which are remarkably visible, famed for their unsightliness, and too expensive to tear down" (Karl Shapiro). (5) Sadrśī is distinguished on the one hand from kimcitsadrsi, where one thing is compared to several others through its aspects (partial similitudes), and on the other from kalpita, in that the similitude is here actually present in both terms, and the common properties apply literally to both subjects; the similitude is not just an analogy of qualities which they severally possess. samastavişaya, 'the whole matter': (1) an upamā in which two things and their several corresponding parts are systematically compared. (2) R 8.29 (30). (3) alivalayair alakair iva kusumastabakaih stanair iva vasante / bhānti latā lalanā iva pānibhir iva kisalayaih sapadi (Rudrața: "The climbing vines resemble maidens, their clouds of bees like

GLOSSARY 168 tresses, their clusters of blossoms like bosoms, their tendrils like clasping arms"). (4) "She summed her life up every day; / Modest as morn, as mid-day bright, / Gentle as evening, cool as night" (Andrew Marvell). (5) See also ekadeśin. These two terms are but tardy extensions of a commonplace distinction usually applied to rūpaka alamkāra. samāna, 'uniform': (1) an upamā in which the common property is replaced by a play on words. (2) D 2.29. (3) bālêvôdyānalātêyam sālakānanasobhinī (Daņdin: "The young girl is like a forest creeper- of beautiful tresses [alaka] and aspect [ānana]" or "beautifying the forest [kānana] of šāl trees [sāla]"). (4) "Why is a lady like a hinge? Because she is a thing to adore" (M. E. W. Sherwood, quoted by Russell Lynes). (5) A play on words differs from a pun in that the latter plays upon a legitimate duplicity of meaning (double-entendre): a word can in context be taken in either of two senses (cf. ślesa upamā). But here there are no words at the base of the play, only the appearance of words (hence the name 'uniform') which must be differently construed to obtain the two desired senses. Only as the construction of the sentence is decided are the words themselves determined. This is, as it were, a syntactical pun. The Sanskrit example is clearer because the component words of the two senses don't even have a common phonemic basis; they are functions of a different analysis of the long compound word sālakānana as sa- alaka-ānana and sāla-kānana. samāsa (1), 'compound': (1) an upamā in which the object of comparison occupies the first position in a compound word. (2) B 2.32, AP 344.8- 9, R 8.17-22. (3) [sā] kamalapattrâkșī śaśânkavadanā (Bhāmaha: "Lotus petal-eyed, moon-faced, she ... "). (4) "Dawn broke in Lon- don, clear and sweet, dove grey and honey" (Evelyn Waugh). (5) Several subtypes are recognized, depending on what element of the simile completes the compound : the common property (as in Waugh's example), the subject of comparison (as in the Sanskrit: compare "pot-belly"), and the Agni Purãna seems to include here compounds of type indusamam ('moon-like'), in which the comparative particle takes second place. It is important to remark that all such com- pounds are adjectival, but that none involve the object of comparison in second position (see rūpaka). samāsa (II): (1) an upamā in which the object of comparison is in an oblique case and is compounded with, that is, followed by, the comparative particle. (2) M 127. (3) atyāyatair niyamakāribhir

GLOSSARY uddhatānām divyaih prabhābhir anapāyamayair upāyaih / šaurir 169

bhujair iva caturbhir adah sadā yo lakşmīvilāsabhuvanair bhuvanam babhära (Mammata: in the Sanskrit, the two terms of comparison, 'arms' and 'powers', are in the instrumental case: "Like Śrīkrsņa, who supports the world with his four arms where Laksmi finds delight, [this King supports the world] with the four royal powers, wide extending, punishers of the haughty, divinely glorious, and eternal"). (4) "There was a great clock ticking, and every time it ticked the tears all fell together with a noise like broken glass tinkling in a plate" (Joyce Cary). (5) In addition to the commonplace ex- ample mentioned in connection with the Agni Purāna in the previous entry, Mammata includes in the present category of simile this bizarre and unparalleled instance whereby we are given to under- stand that the comparative particle (iva, 'like'), when preceded by its object of comparison in an oblique case, is considered to form a compound with it. In the grammatical literature, such compounds are admitted. samuccaya, 'accumulation': (1) an upama in which a second common property cumulates the effect of the first common property. (2) D 2.21, AP 344.13. (3) na kāntyaîva mukham tava / hlādanâkhyena cânveti karmaņêndum (Daņdin: "Not only in beauty is your face likened to the moon, but in its gladdening charm"). (4) "The sun's beams seemed to hit the white road with a directed energy and bounce back like a rubber ball" (Somerset Maugham). (5) This figure differs from utprekșita upamā in that here the properties alone are relevant " ... to hit ... and bounce back ... "); no question is intended as to the adequacy of the object to represent the comparison. Samuccaya differs from mālā upamā in iterating only the property, but not the object. It has no relation whatever to the upamāsamuccaya alamkāra of Rudrața. samanyâbhava, 'ellipsis of the common property': (1) same as dharmalupta upamā. (2) R 8.7 (8). samyatadvacisamksepa, 'ellipsis of the common property and the com- parative particle': (1) same as dharmadyotakalupta. (2) U 1.18. samyavācakasamkșepa, 'ellipsis of the common property': (1) same as dharmalupta. (2) U 1.17. samyôpameyatadvācisamkșepa, 'ellipsis of the common property, the particle of comparison, and the subject of comparison': (1) same as upameyadharmadyotakalupta. (2) U 1.17. stuti, 'praise': (1) same as praśamsā upamā. (2) V 4.2.7.

170 GLOSSARY hetu, 'cause': (1) an upama in which the common property is expressed as the cause of the similitude. (2) D 2.50. (3) kāntyā candramasam dhāmnā sūryam dhairyeņa cârņavam | rājann anukaroși (Dandin: "O King, you rival the moon with your beauty, the sun with your glory, the sea with your steadfastness"). (4) "At the edge of this box there lies a great wooden doll, which, so far as mutilation is concerned, bears a strong resemblance to the finest Greek sculpture, and especially in the total loss of its nose" (George Eliot). (5) Specifically intended is that the common property be expressed grammatically as a cause would be expressed; for example, with the instrumental or, in the English, with "in."

upamārūpaka upamārūpaka (I), 'simile-metaphor': (1) a figure consisting of a rūpaka to which is subordinated, in completion of the image, an upamā ('simile'). (2) B 3.34 (35). (3) samagragaganâyāmamānadaņdo rathânginah / pādo jayati siddhastrīmukhêndunavadarpaņah (Bhāma- ha; according to the commentator, D. T. Tatacharya, the figure concerns only the final attributive compound: "mukham indur iva mukhênduḥ / tasyâbhūtapūrvo darpaņa ivêti"-literally, foot-mirror [rūpaka] for the moonlike faces [upamā]: "May Visnu's foot be victorious, which is the measuring stick of the entire heaven and a new mirror for the moon-like faces of the celestial maidens"). (4) "Thou [West Wind] on whose stream, 'mid the steep sky's com- motion, / Loose clouds like earth's decaying leaves are shed" (Shelley). (5) The definition which Bhämaha gives is clearly different from that for the figure upamārūpaka of Vāmana (see paramparita rūpaka), but his example is inconclusive. Mukhêndu ('face-moon') would by later writers be considered not an upamā, but another rūpaka (see samasta rūpaka); the figure would show then a rūpaka subordinated to another rūpaka, and indeed illustrate a paramparita rūpaka. Our English example appears to illustrate Bhamaha's intention better than his own example: a completely articulated simile (clouds like leaves) is subjoined to the main metaphor (wind-stream) in order to give added force to the identification of property or aspect which that metaphor suggests. Likewise, this independent figure should not be confused with the upama, a subspecies of rūpaka, delineated by Daņdin; in this latter case, the metaphorical identification is completed by a mention of the common property which justifies it.

GLOSSARY 171 upamārūpaka (II): (1) same as paramparita rūpaka. (2) V 4.3.31-32. (5) This is one of the two kinds of multiple figure (samsrsti) given by Vāmana (cf. utprekşâvayava).

upamāsamuccaya upamāsamuccaya, 'simile-conjunction': (1) same as samāna upamā. (2) R 4.32 (34).

upameyôpamā upameyôpamā, 'comparison of the compared': (1) same as anyonya upamā. (2) B 3.36 (37), V 4.3.15, U 5.14, M 136. (5) Another name for the same concept is ubhaya upama.

ubhayanyāsa ubhayanyäsa, 'introducing both': (1) a figure in which the statement of two general remarks suggests a parallel between them, which may in turn suggest a specific reference but in which there is no explicit comparison. (2) R 8.85 (86). (3) sakalajagatsādhāraņavibhavā bhuvi sādhavo'dhunā viralāh / santi kiyantas taravah susvādusugandhi- cāruphalāh (Rudrața: "Rare indeed are those genial souls whose dominion is spread throughout the world; how many trees are there sweet smelling and bearing tender fruit?"). (4) "When the lute is broken, / Sweet tones are remembered not; / When the lips have spoken, / Loved accents are soon forgot" (Shelley). (5) This figure is peculiar to Rudrata and seems to fill the classificatory gap occa- sioned by his definitions of arthântaranyāsa and drstânta: here we have adjunction of remarks general; in drstânta, of remarks specific; and in arthântaranyāsa, of a remark specific and its appropriate universal. The purpose of this figure is both illustrative and com- parative and may be seen as a continuation of prativastu (prativas- tûpamā) alamkāra and drstânta. Though the references seem to be general in both examples cited above, a particular (a beneficent king. a departed mistress) is obviously intended.

urjasvi, 'violent': (1) the expression of extraordinary self-assurance or ūrjasvi

arrogance. (2) B 3.7, D 2.294 (293), U 4.5. (3) apakartâham asmiti hrdi te mā sma bhūd bhayam / vimukheşu na me khadgah prahartum jātu vānchati (Dandin: "Let there be no fear in your heart from think- ing yourself an evil rogue; my sword never wishes to strike the backs

172 GLOSSARY

of those who flee from me!"). (4) "Nor Mike Fink along the Ohio and the Mississippi, half wild horse and half cock-eyed alligator, the rest of him snags and snapping turtle. 'I can out-run, out-jump, out-shoot, out-brag, out-drink, and out-fight, rough and tumble, no holds barred, any man on both sides of the river from Pittsburgh to New Orleans and back again to St. Louis. My trigger finger itches and I want to go redhot. War, famine and bloodshed puts flesh on my bones, and hardship's my daily bread'" (Carl Sandburg). (5) As the third in the trio preyas, rasavat, ūrjasvi, this figure may originally have meant "excess in the portrayal of a rasa", and this explanation is in fact adopted by Udbhata, though his example in no way differs from the one given. The other two writers seem to pair ūrjasvi with preyas (excess of animosity and excess of compliance). Mammața treats this trio, not under alamkāra, but in 'subordinated suggestion' (gunībhūtavyangyadhvani; 66ff.). He tries to reintegrate Ānandavardhana, who was not interested in figures except as they manifested a kind of imperfect dhvani, into the poetic tradition. Cf. rasavat and udātta.

ekâvali ekâvali, 'a single row': (1) a figure in which a series of statements is so arranged that a notion introduced as a qualification (direct object, etc.) in a preceding statement becomes the subject of the following qualification, and so on. (2) R 7.109 (110-11), M 198. (3) salilam vikāsikamalam kamalāni sugandhimadhusamrddhāni / madhu līnâliku- lâkulam alikulam api madhuraraņitam iha (Rudrața: "The stream is abloom with lotuses and the lotuses are replete with sweet-smelling nectar; the nectar is attracting bee swarms, and the bees are gently buzzing"). (4) "I come from the city of Boston, / The home of the bean and the cod, / Where the Cabots speak only to Lowells, / And the Lowells speak only to God" (Anon.). (5) Compare kāraņamālā, where a similar causal sequence is portrayed, and sāra, where a gradation of excellences constitutes the "necklace". Mālā ('garland') has of course been associated with many figures, notably upamā, as a series of (usually) concatenated comparisons. figure illustrates a rhetorical form only-that of superadded qualiti- The present

cation.

aucitya, 'appropriateness': (1) the appropriate correspondance of subject aucitya

and style, acting and mood; a blend of vigor and gentleness. (2) 173

AP 345.5. (5) This is one of the six śabdârthâlamkāra enumerated by the Agni Purāna in an unparalleled treatment (cf. abhivyakti, the sixth such figure). All six are elements of style rather than forms of speech and belong rather to that subject (guna). It is difficult to say precisely what is meant in the text, both due to the lack of examples and because the Agni Purāna also gives an entirely unique account of the gunas themselves. It seems safe to say that Dandin's and Vämana's list of ten has been differently sorted out, some now being called śabdârthâlamkāras, as kānti, some guņas, as śleșa. The catalytic factor may indeed be the dhvani theory (see abhivyakti), for the term aucitya is of extreme importance in that latter specula- tion (Dhvanyāloka, chap. 3); there seems to be no specific reference to such a concept among Dandin's ten gunas.

aupamya aupamya, 'comparative': (1) a generic term for those figures based ultimately on upamā ('simile') or describable in terms of the same structure (upameya, upamāna). (2) R 7.9, 8.1. (5) Rudrața divides arthâlamkāra into four subtypes: aupamya, vāstava (descriptive), atiśaya (hyperbolic), and śleșa (punning). In this, he improves upon Vāmana, who wanted all the figures involving meaning to be derived from upamā.

kānti, 'loveliness': (1) agreeable or pleasant utterance in appropriate kānti circumstances. (2) AP 345.4. (5) This is one of the six śabdârthâ- lamkāra of the Agni Purāna (see aucitya and abhivyakti). Kānti may be related to the guna "kānta" of Daņdin. See praśasti.

kāraņamālā, 'garland of causes': (1) a figure wherein an effect (a term so kāraņamālā introduced) is said to be the cause of a subsequent effect, and so on. (2) R 7.84 (85), M 186. (3) vinayena bhavati guņavān guņavati loko'nurajyate sakalah | abhigamyate'nuraktalı sasahāyo yujyate lakşmyā (Rudrața: "By just actions one attains virtue; the whole world delights in a virtuous man. When one is loved, he is never alone; a befriended man enjoys prosperity"). (4) "By the side of a murmuring stream an elderly gentleman sat. / On the top of his head was a wig, and a-top of his wig was his hat. / The wind it blew

174 GLOSSARY high and blew strong, as the elderly gentleman sat; / And bore from his head in a trice, and plunged in the river his hat. / The gentleman then took his cane which lay by his side as he sat; / And he dropped in the river his wig, in attempting to get out his hat. / His breast it grew cold with despair, and full in his eye madness sat; / So he flung in the river his cane to swim with his wig, and his hat. / Cool reflexion at last came across while this elderly gentleman sat; / So he thought he would follow the stream and look for his cane, wig, and hat. / His head being thicker than common, o'er-balanced the rest of his fat; / And in plumped this son of a woman to follow his wig, cane, and hat" (George Canning). (5) The figure is nothing but a string of causes enchainées.

kāvyadrșțânta kāvyadrsțânta, 'poetic example': (1) same as drsțânta. (2) U 6.8. (5) Udbhața has been reading Bhāmaha, where drstânta has its logical connotation only.

kāvyaliūga kāvyaliūga, 'poetic cause': (1) a figure in which a metaphorical relation of cause and effect is expressed conventionally either as intention or rationale. (2) M 174. (3) praņayisakhīsalīlaparihāsarasâdhigatair lalitaśirīşapuşpahananair api tāmyati yat / vapuși vadhāya tatra tava śastram upakşipatah patatu śirasy akāņde'yam adaņda ivâișa bhujah (Mammata; the subject phrase "may my arm fall" describes the con- ventional effect and riposte to the cause: the attack of the love god: "My body suffers from the blows of gay śirīşa flowers that you stole from the mocking games of her dearest friends! You have certainly shot these weapons in the hope of killing me. May my defenceless arm fall for once on your head!"). (4) "When he saw in their bright eyes the shadow of the registry office, he told them that the memory of his one great love would always prevent him from forming any permanent tie" (Somerset Maugham; the registry office is the cause of the convenient memory). (5) There is little ground for distinguishing this rather obscure figure from the ordinary hetu (q.v.). The main structural argument for the distinction is that the cause is here specified as poetic; for hetu, such a determination has always been implicit. Yet the figure has been rejected by several authors on the ground that it involves no element of vakrôkti. metaphorical utterance. Mammața's kāvyalinga may represent no

GLOSSARY more than an effort to take account of that hypothetical objection: 175 he does not himself define a figure hetu and indeed says, in discussing the figure kāraņamālā, "pūrvôktakāvyalingam eva hetuh" ('the figure hetu is indeed nothing but the previously mentioned figure kāvya- linga'). Mammata's three examples show quite forcefully that he intends the expression of the relation of cause and effect to be other than descriptive-definitely conventional and presumptive; the figure thus resembles leśa (I).

kāvyahetu Livyaheta, 'poetic cause': (1) same as smaraņa alamkāra. (2) U 6.7. (5) In the text, Udbhata also calls this kāvyalinga; Mammața uses the name kdvyalinga for another figure and calls this one smarana. Kāvyahetu is a jñāpaka hetu whose purpose is comparison. krama krama, 'series': (1) same as yathāsamkhya. (2) D 2.273, V 4.3.17.

ph 'stringing': (1) composition. (2) AP 342.31. (5) Gumphanā gomphanā is skill at managing the sequence of the narrative; it is paraphrased by the term racand. If the Sarasvatikanthabharana (2.118) can be allowed to have preserved the poetic tradition of the Agni Puruna, the term may mean only balanced composition and be similar to the fgure yathåsamkhya.

citra dm, 'glitter' (and duskara, 'difficult', krīda, 'play'): (1) names used variously by the different authors to cover four separate phenomena, but grouped together because of their basis in pure word play. (2) D 3.186, AP 343.22-31, R 5.1-33, M 4, 121. See also (5) and the termns there defined. (5) After the triumph of the rasa-dhvani theory, the term citrakavya comes to be used for the third and lowest kind of poctry, where mere verbal virtuosity precludes the expression of any (a) In the most obvious sense, -word play" refers to the comp/s- boa of various puzzkes and games. riddles and conundrums. ind the like None of the authors, except perhaps the Agru Purina. Bocs as far as calling this sort of thing poetry, but several treat of i becamne of its obvious function of entertaining the same aud:ence

GLOSSARY 176 for whom the poetry was destined and also because in vome manner it does demonstrate the virtuosity of the "poet" who wrote it. Dandin gives sixteen different types of conundrum in treating of prahelikā. The Agni Purana and Rudrata give approximately the vame liut of six games, but the former calls them citra (the eighth sahdálam- kāra of nine), and the latter krīdā, appending the topic to a discuwsion of citra (types [b] and [c] below), a śabdálamkāra. In both treatments, prahelikā is but one of six games given, the others being cyuta, gūdha (or gupta), praśna, and samasyā (variously subdivided to make six). A more important category of word play, however, is the various kinds of patterned verses, which probably have their origin in the figure yamaka ('cadence'). Three types can be distinguished, depend- ing upon the principle of repetition involved. (b) Included here are verses which, through a geometrical limitation of the sequence of their syllables, can be read in more than one way to give the same meaning. The most transparent example is the palindrome (pratilo- mânuloma), which specifies that the sequence of syllables be the same when read backwards. But the Indian authors have been ingenious in inventing other principles of total repetition, such as 'hop-scotch' (anulomaviloma), 'zig-zag' (gomūtrika), 'criss-cross' (muraja), 'double palindrome' (sarvatobhadra), `elephant-walk' (gajapada), and 'knight-at-chess' (turagapada). The geometric 'cart- track' (rathapada) is two palindromes separated by non-geometrical sequences and could be taken as an example of the next type. No specific name has been given to these geometrics. Dandin treats them as an extension of yamaka, along with (d), following in what he calls dușkara. The Agni Purāņa may intend this type by its cate- gory vikalpa duşkara, but some examples of it (sarvatobhadra) are included in bandha duşkara (c). Rudrața and Mammața treat both this and the next type (c) under the category citra, though Rudrata does not mix the examples indiscriminately, which perhaps implics that he was aware of a difference of principle. (c) When the principle of limitation is not applied to the entire sequence of syllables, but requires repetition only of certain strategi- cally placed syllables in terms of which the whole verse can be ar- ranged in imitation of natural objects, we have a type of word play most commonly known by the name bandha (Agni Purūna, Mam- mața); examples are: khadga ('sword'), cakra (wheel'), dhanu ('bow'), padma ('lotus"), musala ('pestle'), śara ('arrow'), śakti

GLOSSARY ('lance'), śūla ('spike'), and hala ('plough'). Rudrața apparently 177 considers pictorial verses citra par excellence, for they get first place in his account. He gives no specific name, however; type (b) above is also considered citrakāvya. Lastly (d), the principle of repetition may be located not in the verse at all, but in the individual syllable; that is, the place of the syllable is not specified, but rather its phonemic quality. Dandin and the Agni Purāna call this type niyama, and Dandin gives examples of verses composed of four or fewer vowels or consonants, including one tour de force the only consonant in which is the phoneme "n". The last three types depart in certain respects from the classic yamaka, which is a repetition of phonemically identical syllable sequences in specifically defined and symmetrically related parts of the verse. The first type (b) can be seen as a variation on the maha- yamaka (q.v.), inasmuch as the entire verse is somehow repeated; but the repetition is subjected to conditions which in turn limit the occurrence of syllables within the verse, and this is foreign to yamaka. The second type (c) involves a repetition only of specifically placed syllables and does not refer to symmetrically related parts of the verse. Type (d), of course, puts no restriction at all upon sequence. It may be seen from the preceding that not all authors agree either on the terminology or typology of citrakāvya. Daņdin, the earliest writer to deal with the subject, defines types (a), (b), and (d), calling them prahelikā, duşkara, and niyama. The Agni Purāņa seems to refer to all four under the names (a) citra, (b) vikalpa, (c) bandha, and (d) niyama, and groups the last three together as dușkara. Rudrața, in turn, considers (a) krīdā, (b) and (c) citra, but ignores (d). Lastly, Mammata, who uses only the term citra, treats indiscriminately of (b) and (c). Citra, as a category of poetry, is extremely important in the history of Indian poetic speculation. The growing contempt for the poet's virtuosity on the part of the critics probably reflects in part the increasing dependence of the poets on these devices, a fact that Sanskrit literary history has often remarked. From its origins per- haps ultimately in religious symbolism, citrakāvya has passed from one extreme (Dandin's view that it is a kind of recreation of the poet and his audience) to the other, becoming more and more a central issue opposing the poet to his audience (the critics). The development of the tantric religious systems may have accentuated this divorce,

GUAMARY 178 for the magical undertone, especially of the piional vene. a anulomaviloma, 'with the grain, ignoring the grain': (1) a type of wore apparent. play in which the syllables of a second half verse or iecond yerur repeat in a leapfrogging sequence those of the firu (half) vere (2) R 5.22-23. (3) samaranamahitopd yástanāmáripātā nahimatoyâpåstamānåritåpd ... (Rudrața: "Thoie who destroy de enemies of the Gods and cause to fall the enemies of those whote glory is decayed by having departed, the suffenngs of his caemies, being arrogance, were assuaged by applications of water borne from snow to a lover"). (4) The form is: A Ba Ca Da Ea Fa Ga Ha Ja A Ca Ba Da Fa Ea Ga Ja Ha. (5) Cf. pratilomånuloma; the tard viloma also means 'against the grain', but the connotation of the pre-verb vi- is simply 'irrespective of' rather than 'contrary lo' as conveyed by prati -. Loma means 'hair'; the idea conveyed is that of stroking hair in the way it naturally lies, or the inverse. ambuja, 'lotus': (1) same as padma. (2) AP 343.46. arthacitra, 'word play of sense': (1) another term for alankåra. (2) M 70 (5) Mammata is following Anandavardhana, who thinks that fgares of speech which do not evoke any rasa are mere word play, datin- guished from riddles and cadence (sabdocitra) only in that they involve the meanings of the words instead of their outward form (ntdn) They lack the poetic charm which alone derives from an approprisedy evoked rasa. Mammața mistakes this rhetorical remark for a chusi- fication of citra and reproduces it in that context. ardhabhrama, 'half-rotation': (1) a type of word play in which a vere. each of whose four pådas is written on a separate line, can be rend either in the normal way or as a helix, from outer verticals inwerds (2) D 3.80 (81), AP 343.39 (text reads ardhdbhyim in error), R 53 (18). (3): mã no bha va ta và ní kamt no da yã ya na mã nỉ nỉ bha yã da me yã mã mở t va ya me no ma yd na la. ("O Liebesgott, vor dem wir uns verneigen! Det Heer, die Grelterdt, Dapdin trägt den Sieg davon; wir mögen ein Vergthen begangen hahe cde nicht, so empfinden wir doch aus Furcht cine wermassiche Pai" (Böhtlingk). Or: "O God of Lovel your beloved, phalaas-Fke, is cartainly not for our misfortune! May we not be deomed secn!

GLOSSARY Our incompassible sufferings are from fear of you alone!") (4) The 179 form is: A B CDEF B 1 J K L G H M C J O P G O R D M K P S F T Q L E (5) Reading as a helix, that is, downwards on the first column, upwards on the eighth, then downwards on the second and upwards on the seventh, and so on, gives exactly the same sequence of syllables as reading from left to right in the normal way. Compare sarvato- bhadra, where the verse can be read backwards and forwards as well. kārakagūdha, 'concealment of the subject': (1) a grammatical riddle in which the subject at first reading is concealed by a false samdhi, but which another reading (dividing the words differently) reveals. (2) R 5.26 (30). (3) pibato vāri tavâsyām sariti śarāveņa pātitau kena (Rudrata; pātitau has no subject, but by reading śarāveņa 'with a dish' as śarāv eņa, 'arrows, O antelope', the subject is supplied; "While drinking water in this stream with a dish, by whom have been shot at you?"). (4) Read the example under kriyāgūdha on "Why did the raise her bill (razorbill) raise her bill?" (Oliver Onions; deformed for our purpose here). (5) Rudrata lists six games (krīda) of which this is one; cf. kriyāgūdha. kriyāgūdha, 'concealment of the verb': (1) a grammatical puzzle in which the verb at first reading is concealed by samdhi. (2) R 5.26 (30). (3) vāri śiśiram ramaņyo ratikhedād apurușasyêva (Rudrața; ramanyo, 'ladies' requires a verb; by reading apurușasyêva, 'as though of a non-man' as apur uşasy eva, 'drank only at dawn', it is supplied: "The ladies, exhausted from passion, the cool water as though of a non-man"). (4) "Why did the razorbill razorbill?" "So the sea- urchin could sea-urchin" (Oliver Onions). (5) See kārakagūdha. krīdā, 'play': (1) puzzles or conundrums. (2) R 5.24. (5) A cover term for six games: mātracyuta, binducyuta, prahelikā, kārakagūdha, kriyāgūdha, and praśnôttara. Rudrața distinguishes these six from citra, which are syllable arrangements and legitimate figures (cf. pratilomânuloma), by calling them "merely playful"; that is, they serve no function of embellishment and are not poetic. But Rudrața, following Dandin, treats of them presumably because they please and divert the same sophisticated audience for which the poetry was intended. The six games are wider in scope than our conun- drums, which term could usefully translate the third type, prahelikā;

GLOSSARY 180 types one and two depend on altering the written verse in a system- atic way, types four and five on false samdhi, and six is a question which answers itself. khadga, 'sword': (1) a verse or pair of verses whose syllables can be arranged, in terms of certain repetitions, in the visual form of a sword. (2) R 5.2 (6, 7), M 121. (3) mārâriśakrarāmébhamukhair āsāraramhasā | sārârabdhastavā nityam tadartiharaņakşamā / mātā natānām samghattah śriyām bādhitasambhramā / mānyâtha sīmā rāmāņām śam me diśyād umâdijā (Rudrața: "May Umā, the first born, show me favor-she who is praised most excellently and with the violence of a cloudburst by Siva, Indra, Rāma, and Gaņêśa, and who is able to remove their sufferings, mother of the devoted and collection of bounties, who cancels error, and who serves as the honorable model of women"). (4) The reader is referred to the appen- dix of Rudrata's text, where he will find the "picture" which this verse defines. (5) Such picture verses, generally referred to as bandha ('ligature', 'bond'), are first mentioned in the Agni Purāna and con- stitute probably the most original and indeed the most curious contribution of that text to poetics. The tantric symbolism of these verse arrangements is evident, but their actual function is not so. Some late writers, like Mägha, employ them, but it is clear that they have no other purpose than to demonstrate the virtuosity of the author. One chapter (the nineteenth) of the Śiśupālavadha contains a great number of these devices, and one may wonder if there is not some relation between the virtuosity so demonstrated and the concurrently rendered climax of the poem: the fight between Kṛşņa and Śiśupāla. gajapadapāțha, 'elephant-gait-version': (1) a verse which, when each of its four pādas is placed on a separate line, can be read either by vertical pairs of syllables from left to right, or in the normal way. This horizontal movement of two syllables at once is likened to the gait of the elephant, whose two legs are always in unison. (2) R 5.2 (16). (3): ye nã nã dhĩ nã và dhĩ rã nã dhĩ và rã dhī rã rã jan kim nā nā śam nā kam śam te nā śan kan te'šam te te jaḥ ("Those of your courtiers, O King, who preserve the force of Rudrața various epithets, are powerful, who eradicate obstacles and who do

GLOSSARY not preserve non-wisdom, why do they not notice your heavenly 181 grace, replete with many desires? [They do not observe that] your glory is not auspicious".) (4) The form is: A B B C B DC E B C D E C E E F (5) The same syllable sequence is obtained by reading the left vertical pair from top to bottom, then the second vertical pair, and so on. gupta, 'hidden': (1) probably the same as (kārakakriyā) gūdha. (2) AP 343.22 (26). (5) As always with the Agni Purāna, no examples are given, but the context suggests that Rudrata's two krīda are meant. gomūtrika, 'cow piss': (1) a verse the syllables of whose constituent pādas, when placed on separate lines, can be read either by zig-zag- ging from one line to the other, or in the regular way. (2) D 3.78-79, AP 343.36-38. (3): ma da no ma di rā kșī ņā ma pā ngā stro ja ye da yam ma de no ya di tat kşī na ma na ngā yāñ ja lim da de Daņdin ("Should the love-power of drunken-eyed women, armed with side- long glances, conquer me-if this my sin should finally be destroyed, I would give thanks to the Love God".) (4) The form is: ABCDEFGH A I C J EKGL (5) The same syllable sequence can be obtained by reading lower first vertical (A), upper second vertical (B), lower third vertical (C), and so on. This amounts to requiring that every other syllable be the same as the corresponding syllable in the next pāda. The Agni Purāna gives three other names for this type of limited verse: dhenu ('cow'), aśvapada 'horse track'), and jālabandha ('lattice'). The appropriateness of the more common name is evident. cakra, 'wheel': (1) a series of verses which can, in terms of certain signi- ficant repeated syllables, be arranged in the visual form of a wheel. (2) AP 343.47-54, R 5.2 (6-13). (3) See Rudrața's examples for khadga, musala, dhanu, śara, śūla, śakti, and hala, in that order. (5) The first half verse of each śloka begins with the same syllable (mä), and this constitutes the "hub". The first half verse itself is the spoke and the second half verse the part of the felloe to the right of the spoke, up to the next spoke. The syllable at the junction of the spoke and the felloe is thus part of three half verses and is

GLOSSARY 182 consequently the same in all three: last syllable of the preceding verse, last syllable of the first half verse, and first syllable of the second half verse following. In Rudrata's example, eight verses are given. constituting a wheel of eight spokes. cakrâbjaka, 'wheel-lotus': (1) presumably some combination of cakra and padma. (2) AP 343.55. (5) No examples and no similar type in the other texts. cyuta. 'fallen': (1) a group of conundrums which operate by dropping significant parts of the written Sanskrit sentence, such as vowel indicators, nasal vowel marks (anusvāra), final aspiration (visarga), and perhaps consonants ("r" in clusters). (2) AP 343.22, 28. (4) "There's a little old fellow and he has a little paint-pot, / And a paucity of brushes is something that he ain't got, / And when he sees a road sign, the road sign he betters, / And expresses of himself by eliminating letters. / Thus 'Through Road' / Becomes 'Rough Road' / And 'Curves Dangerous' / Is transformed to 'Curves Anger Us ... '" (Morris Bishop). (5) No Sanskrit examples are given, but in part cyuta is evidently the same as Rudrata's mātracyuta and binducyuta. The idea is that by dropping these discriminating ele- ments, another meaning is obtained. The possibility of this game, of course, depends on the fact that short "a" is inherent in all syllables and is "what is left" when superscripts are dropped. cyutadatta, 'dropped-added': (1) apparently a combination of cyuta and datta; perhaps certain discriminating elements are replaced by others. (2) AP 343.22, 30. (5) No example is given. Cf. cyuta, datta. turagapada, 'horse path': (1) a verse whose syllables, when arranged by pādas on separate lines, can be read either in the manner of the moves of a knight at chess or in the regular way. (2) R 5.2 (15). (3): se nā lī lī lĩ nã nã lĩ lĩ nã nã nã nã lĩ lĩ lĩ nā lī nā lī le nā lĩ nã lĩ lĩ lĩ nã nã nã nã lĩ Rudrața ("I praise the army whose leader is mighty in play, which is devoted -I, who am not acquainted with untruth, whose men are mounted in carts and keep together in various rows, who does not perpetrate meaningless deeds for his dependents, who has generals who assume the leadership of happy men, who has men of several sorts and no fools". (4) The form is:

GLOSSARY 5 183 2 6 8 4 10 3 11 7 (5) This amounts to the famous puzzle of moving the knight so 9

that it touches every square only once. I am indebted to V. Raghavan for suggesting the nature of the solution. It is beyond my powers to complete it. daņda, 'stick': (1) probably the same as khadga. (2) AP 343.37, 55. (5) No example. datta, 'given': (1) probably a group of conundrums which function by adding certain significant parts of the written Sanskrit sentence, as vowel indictors, nasal vowel marks (anusvāra), final aspiration (visarga), and perhaps consonants. (2) AP 343.22, 29. (4) "But the old fellow feels a slight dissatisfaction / With the uninspiring process of pure subtraction. / The evidence would indicate he's taken as his mission / The improvement of the road signs by the process of addition. / Thus 'Traffic Light Ahead' / Becomes 'Traffic Slight Ahead' / And 'Gas and Oil' / Is improved to 'Gasp and Boil ... '" (Morris Bishop). (5) No Sanskrit examples are given, but the context permits a comparison with cyuta and leads to the inference that datta is the reverse of this. Like cyuta, the games depend on the fact that the short vowel "a" is inherent in every syllable, and that graphically other vowels and vowel aspects are superscripts to that simple vowel. The second type, "consonantal" datta, presents more of a problem, but may refer to adding parts of ligatures, as "r" which is an obvious superscript. dușkara, 'difficult to accomplish': (1) a general name for various types of picture verses and geometric verses. (2) D 3.78, 83, AP 342.20, 343.32. (5) In general, dușkara is distinguished from puzzles and conundrums, which Dandin includes in prahelikā and the Agni Purāņa calls citra. Dușkara then refers to those extensions of yamaka in which the principle of repetition is not linear, or in which the limitation on occurrence applies only to certain letters (e.g., one vowel or consonant) or to certain places in the verse (picture verses). These two types are clearly delimited by Dandin (reference cited), who does not recognize any pictorial verses. (His gomūtrika may provide the key to the explanation of the origin of these latter: for Dandin, the zig-zag is clearly a geometrical verse with a graphic name; it may have encouraged others, more literal minded, to

GLOSSARY 184 explore further the field of graphic representation.) The Agni Purana apparently follows Dandin's usage of the term, but it makes an explicit distinction between three types of duskara, only the firgt and third of which have clear parallels in Dandin (verses employing only limited vowels or consonants and picture verses). The woond type (vikalpa) may refer to geometrical verses, for the palindrome is referred to, yet sarvatobhadra is included in the third type (pictare verses or bandha). dhanu, 'bow': (1) a verse whose syllables can, in terms of certain significant repetitions, be arranged in the visual form of a bow. (2) R 5.2 (9). (3) mām abhīdā śaraņyā mutsadaivârukpradā ca dhīh | dhīrā pavitrā samtrāsāt trāsīsthā mātar ārama (Rudrața: "O mother, save me from fear and lay off! you who dispense confidence, are a refuge. a giver of health always and uniquely enjoyable, wise, steadfast and holy"). (5) The first half verse is the "bow", the second, the"string". The syllables joining at each end are dhi and ma (second of the first half and last of the second; the Indian bow generally has an extension at the top above the junction with the string). For metrical reasons, Rudrața also refers to dhanu as bāņâsana (same meaning). niyama, 'limit': (1) a verse whose phonemic content is limited to certain vowels, consonants, or points of articulation. (2) D 3.83ff., AP 343.33-34. (3) śrīdīptī hrīkīrtī dhīnītī gīhprītī | edhete dve dve te ye neme deveśe [only two vowels]; nūnam nunnāni nānena nānanenânanāni nah / nānenā nanu nānūnenainenānānino ninīh [one consonant]; agā gām gāngakākākagāhakāghakākākahā / ahāhānga khagānkāgaka- gakhagakākaka [only gutterals]. All these examples are from Dandin: "They are not two overlords who prosper there, they are prosperity and beauty, shame and fame, wisdom and polity, celebrity and pleasure"; the last two examples are unclear). (5) One well-known example from literature of this virtuosity is the seventh chapter of The Ten Princes (also by Dandin) where no labials at all are employed. on the pretext that the narrator has wounded his lips in love-making. See (5) under duşkara, of which this is the first type. padma, 'lotus': (1) a verse whose syllables can, in terms of certain signi- ficant repetitions, be arranged in the visual form of a lotus. (2) R 5.21, M 121. (3) bhāsate pratibhāsāra rasâbhātâhatâvibha | bhāvitâtmā śubhā vāde devâbhā vata te sabhā (Mammața: "O essence of glory! your council is indeed brilliant, beautified by the eight moods, of unbeatable beauty, in which is revealed the ultimate soul of justice, clever in dispute, and similar to God"). (5) The syllable

GLOSSARY hha forms the center of the lotus. The two syllables following or 185 preceding bha constitute the "petals". The "center" recurs after every two "petals". One begins with bha, reads out along the first petal and back along the second to the center, then out along the third, back along the fourth, and so on, until one reaches the last petal, which should be the inverse of the first (here bhasate and te sabhā) and lead back to the center. In the present example, the petals at the end of each pāda and the beginning of the next are also inver- sions. pratilomânuloma, 'against the grain, with the grain': (1) a type of word play in which the syllables of a second half verse repeat in exact inverse order those of the first half verse. (2) AP 343.34-35, R 5.3 (17). (3) vedâpanne sa śakle racitanijarugucchedayatne'ramêre devâsakte'mudakso baladamanayadas todadurgâsavāse / sevāsargād udasto dayanamadalavakşodamukte savāde reme ratne'yadacche gurujanitacirakleśasanne'padāve (Rudrața: "A certain person, whose eyes know no pleasures, who gives directions for countermanding strength and who has ceased to desire serving others, delights in this virtuous man, accomplished in the Vedas, agreeable, who strikes down evil men and in whom is ingrained the struggle to eradicate his own suffering, who is devoted to the Gods and inherently capable of storming the bastions of sickness, who is free by having crushed the droplets of pride in giving, garrulous, spotless, accepting a fall from grace and devoted only to the trials born of attendance upon his master"). (4) The form is A Ba Ca Da / Da Ca Ba A. (5) Note that the pattern, as usual, is that of syllables (consonant plus vowel) rather than that of phonemes. See also anulomaviloma. This is not exactly a palindrome, since the meaning of the reverse reading is not the same; cf. "Madam, I'm Adam" and "Able was I ere I saw Elba" (James Joyce). praśna (praśnôttara), 'question' (or 'question-answer'): (1) a conundrum in which the same word answers several questions, but is used as a pun and is taken in a different sense for each question. (2) AP 343.22-24, R 5.26 (31-32). (3) udyan divasakaro'sau kim kurute kathaya me mrgāyâśu | kathayânindrāya tathā kim karavāņi kvaņita- kāmah | ahiņavakamaladalâruņiņa māņu phurattiņa keņa | jāņijjāi tarunīaņassa niddhā bhaņa ahareņa (Rudrața; the third question is of course in Prakrit. Aharena is the answer to all three: 'day' (ahar) and 'night' (ena), 'stupid' (ahare) 'make noise' ('na, samdhi for ana), 'by her lower lip' (adharena: "dh" becomes "h" in most Prakrits):

GLOSSARY 186 "What does the sun make when he rises? Tell me quickly, spy it out! Say, stupid, what should I do if I want to be loud? How do you tell when a girl is angry?"). (4) "What is that which will make you catch cold-cure the cold-and pay the doctor's bills?" (Robert Merry; answer: a draft). bandha, 'delimitation': (1) the generic term for those verses which can be arranged, in terms of certain significant repeated syllables, in the visual form of natural objects, as swords, wheels, axes, etc. (2) AP 343.33 (35-65), M 121. (3) (4) See śara, cakra, muraja, etc. (5) For both Rudrata and Mammata, pictorial verses represent citra par excellence; the older name of the Agni Purāna is retained only as a part of the name of each type, as khadgabandha. In the Agni Purāņa, moreover, bandha is one of three types of dușkara and is distinguished from citra, this last being a general name for conundrums, puzzles, and the like. Most probably this type of verse with obvious magical connotations grew out of the older geometrically arranged verses (palindromes, etc.), which are prom- inent in Dandin. The bridge may have been the geometric gomūtrika ('cow piss')-in Dandin simply a vivid name for a verse which can be read in a zig-zag fashion. The Agni Purāna significantly groups gomūtrika in bandha verses. The only instance I know in English of a verse that is what it means is: "Yet this I Prophesie; Thou shalt be seen, / (Tho' with some short Parenthesis between:) / High on the Throne of Wit" (John Dryden). binducyuta, 'dropping the anusvāra': (1) a type of word play in which one phrase, by dropping a nasal phoneme, is transformed into another phrase with another meaning. (2) R 5.25 (28). (3) kānto nayanânandī bālênduḥ khe na bhavati sadā (Rudrața: "The lovely young moon, delightful to see, is not always in the sky". By dropping a nasal, we get "bāle duhkhena" or "The lovely moon, delightful to see, young girl, is always accompanied by sadness"). (5) See cyuta. Rudrața gives examples for only two of the Agni Purāna's four types: this one and mātracyuta. mātracyuta, 'dropping the vowel sign': (1) a type of word play in which one phrase, by dropping the graphic syllabic modification indicating a vowel phoneme, is transformed into another phrase with another meaning. (2) R 5.25 (28). (3) niyatam agamyam adrśyam bhavati kile trasyato raņôpāntam (Rudrața: "In truth, for the fearful the environs of battle are unapproachable and their sight cannot be

GLOSSARY borne". By dropping the "i" in kila, we get kalatrasya toranôpantam or "For women, the environs of the city's gates are unapproachable 187 and their sight cannot be borne"). (5) Cf. binducyuta and cyuta. By dropping the vowel discriminator, a short "a" is obtained, since this vowel is considered inherent in the syllable sign itself. muraja. 'drum': (1) a verse whose syllables can be read either in a criss- cross form similar to the lacing of an Indian drum (mrdangam), or in the regular way. (2) AP 343.59, R 5.3 (19), M 121. (3): sa ra lā ba ha lā ra mba ta ra lā li ba lã ra và vā ra lā ba ha lā ma nda ka ra lā ba ha lā ma lã Rudrața, Mammața ("The autumn is full of the sounds of undulating armies of bees, thick and long; dense with geese; where kings are quick and āmala fruits are plentiful".) (5) The first and last lines can be read also on the four-syllable diagonals up and down, beginning and ending with the same syllables as the lines. The two internal lines must be taken in halves, but the same principle applies. musala, 'pestle': (1) a verse whose syllables can be arranged, in terms of certain significant repetitions, in the visual form of a pestle. (2) R 5.2 (8). (3) māyāvinam mahāhāvā rasāyātam lasadbhujā | jātalīlā- yathāsāravācam mahişam āvadhīh (Rudrața: "You, O mother, of great blandishments and gleaming arms, in whom joy is fulfilled, you have slain the buffalo demon, hiding in deceit, puffed up with pride, whose words did not correspond to the truth"). (4) See the appendix to Rudrata for the picture that this forms. (5) No mortar, oddly enough. rathapada, 'cart path': (1) a verse wherein the two even or the two odd pādas (but not both) are palindromes, thus producing the appearance of a cart track. (2) R 5.2 (14). (3): itîkșitā suraiś cakre yā yamāmamamāyayā mahişam pātu vo gaurī sāyatāsisitāyasā Rudrața ("Thus observed by the Gods, may Gauri, who without guile sent the buffalo demon to the nether world, Gauri, who has slain those demons who have destroyed property with their long arrows-may she protect you!") (4) The form of the second and fourth lines is:

GLOSSARY 188 ABCDDCBA and EFGHHGFE. (5) Two other padas are gaja and turaga. vikalpa, 'alternation': (1) a kind of duşkara. (2) AP 343.33, 34. (5) No example is given and there are no subdivisions; this may refer to geometrical verses-those which can be read by rearranging the syllables in a regular pattern (as palindrome). See duskara. The text reads "vikalpah prātilomyânulomyād evâbhidhīyate | prātilomyá- nulomyam ca śabdenârthena jāyate | anekadhâvrttavarņavinyāsaih śilpakalpanā" (343.33-34; 'Vikalpa gets its name from the backwards and forwards arrangement of syllables; the arrangement functions on both the levels of word and sense and is a technical construction of phoneme sequences repeated more than once'). vyährtârtha, 'whose meaning is spoken': (1) same as ekacchanna prahelikā. (2) R 5.25 (29). (5) See prahelikā. The name vyāhrtârtha probably intends that the paradox is made explicit in the riddle, and concerns the nature of the thing rather than words (see spastapracchannârtha). śakti, 'lance': (1) a verse whose syllables can be arranged, in terms of certain significant repetitions, in the visual form of a lance. (2) R 5.2 (12). (3) māhişākhye raņe'nyā nu sā nu nānêyam atra hi | himâtankād ivâmum ca kam kampinam upaplutam (Rudrața: "Was it she or another in the battle with the buffalo demon? Was it he, shuddering, who was consumed as though by the enemy of the snow?"). (4) See the appendix to Rudrata for the picture that this forms. śabdacitra, 'word play of words': (1) citra, properly speaking. (2) M 70. (5) See arthacitra. śara, 'arrow': (1) a verse whose syllables can be arranged, in terms of certain significant repetitions, in the visual form of an arrow. (2) R 5.2 (10). (3) mānanāparuşam lokadevīm sadrasa sannama / manasā sādaram gatvā sarvadā dāsyam añga tām (Rudrața: "O truly devoted. approach respectfully and honor the world Goddess with your mind and every effort from which anger has been driven out by service"). (4) See the appendix to Rudrata for the picture that this forms. (5) The first pada is the shaft, the second is the point, and the third and fourth are the feathers and bonds. śūla, 'spike': (1) a verse whose syllables can be arranged, in terms of certain significant repetitions, in the visual form of a spike. (2) R 5.2 (11). (3) mā mușo rājasa svâsūml lokakūțêśadevatām | tām śivāvāśitām siddhyâdhyāsitām hi stutām stuhi (Rudrața: "O violent one, do not delight in your own life: praise the Goddess-she who

GLOSSARY is diety to the world's high rulers, she who is obedient to Siva and 189 fulfilled with all success, she who is praised by all"). (4) See the appendix to Rudrata for the picture that this forms. (5) The first half śloka constitutes the shaft; six syllables suffice for the three prongs by reading forwards and backwards, each time adding the final syllable of the first half. samasyā, 'union': (1) a verbal game which consists in discovering the words of one verse which have been hidden systematically in a much larger verse. (2) AP 343.23, 31. (5) No examples are given, but the puzzle is well known, at least as far as the letters of a word are concerned, by the name "acrostic"; for example: "Dread monster, ruthless foe / Ever travelling to and fro / And causing tears of grief to flow, / The good, the loved, and those that be / Hale and strong, must yield to thee" (Robert Blackwell). sarvatobhadra, 'auspicious in all ways': (1) a verse, having the same number of lines as syllables, which can be read backwards and forwards both vertically and horizontally. (2) D 3.80 (82), AP 343.41, R 5.3 (20), M 121. (3) sāmāyāmāmāyā māsā mārānāyānārāmā / yānāvārārāvānāyā māyā rāmā mārāyāmā (Daņdin; this verse is also deemed to make sense when read backwards: by putting these two verses together on eight lines of one pāda each, the same double verse can be read backwards or forwards along the horizontals, or backwards or forwards along the verticals: "This lovely young lady, an extension of the fever of Love, a union of devices for inducing love, a snare made of the tinkling of her anklets, whose beauty is sorcery, is destined quickly to be the death of me, along with the moon"). (5) Cf. ardhabhrama, which is conceived as a half verse which can be so arranged. spastapracchannârtha, 'whose hidden meaning becomes clear': (1) same as ubhayacchanna prahelikā. (2) R 5.25 (29). (5) See prahelikā and vyāhrtârtha. hala, 'plough': (1) a verse whose syllables can be arranged, in terms of certain significant repetitions, in the visual form of a plough. (2) R 5.2 (13). (3) mātangânangavidhinâmunā pādam tam udyatam / tañgayitvā śirasy asya nipātyâhanti ramhasā(Rudrața: "Gaurī, lifting her foot and bringing it down on his head in the manner of a proud elephant, slew him violently"). (4) See the appendix to Rudrața for the picture that this forms.

190 chāyā chaya, 'shadow': (1) clever imitation of other styles or dialects. (- AP 342.21-25. (5) This is the first sabdâlamkara of the Agni Puräna four styles are enumerated : the vernacular (lokokti), scholarly jargos (chekôkti), children's speech (arbhakôkti), and the talk of drunk (mattôkti). chekânuprāsa chekânuprāsa, 'clever alliteration': (1) same as cheka, a type of anuprāsa. (2) U 1.3 (5). (5) Chekânuprāsa is considered by Udbhața to be an alamkāra separate from anuprāsa strictly speaking. jāti jāti, 'genus': (1) same as svabhāvôkti. (2) R 7.30 (31-33).

tattva tattva, 'reality': (1) an arthaślesa in which puns, in appearance descriptive adjectival qualifications of the subject, involve a second meaning as predicate nouns which metaphorize the subject. (2) R 10.20 (21). (3) nayane hi taralatāra sutanu kapolau ca candrakāntau te | adharo'pi padmarāgas tribhuvanaratnam tato vadanam (Rudrața; the parts of the girl's face are described by adjectives which are also names of jewels which compliment that particular feature: "For her eyes are glancing [two pearls of great price], her cheeks are beautiful as the moon [moonstones], her lower lip has the color of the lotus fis a ruby], and thus her face is the jewel of the three worlds"). (4) "'Before you had those timber toes, / Your love I did allow, / But then, you know, you stand upon / Another footing now!'" (Thomas Hood; spoken by a lady whose soldier-lover has returned from the wars minus both his legs: the pun on "footing" suggests that the lady would prefer a lover with a complete set of limbs. He "stands on another footing" in the direct sense because conditions have changed, but that "other footing" refers also to his wooden legs and heightens the force of the description). (5) The second meaning of this pun transforms a simple description into the figure rūpaka. The figure differs from avayava śleşa in that the pun is not merely on the qualification, but also on the status of the qualification as a gram- matical adjective. The implicit rūpaka of this type of śleșa, moreover, distinguishes it from avayava, where the pun suggests at most a flattering comparison.

GLOSSARY tadguņa 191

tadguna, 'having that thing's attribute': (1) a figure in which one thing is said to assume the quality or property of another thing, from which thereby it is made difficult to distinguish. (2) R 9.22-25. (3) navadhautadhavalavasanās candrikayā sāndrayā tirogamitāh | rama- nabhavanāny asankam sarpanty abhisārikāh sapadi (Rudrața: "Hidden by the thick moonlight, their garments white and newly washed, the women approach fearlessly the homes of their lovers"). (4) "Half a mile away across the young corn you saw her white sweater at the cliff's edge, and it seemed part of the whiteness of the screaming seabirds, of the whiteness of the awful glimpses of chalk where the turf suddenly ended in air, of the crawling whiteness of the waves far below" (Oliver Onions). (5) Rudrata considers two types: one in which the indistinguishables in fact share the property which makes them indistinguishable, and the other in which they do not but where the property of one imposes itself upon that of the other. This latter resembles pihita and seems to differ from it only in defini- tion: here the pre-eminent quality lends itself to the other subject; there, the pre-eminent quality hides the other. Tadguņa differs from bhrāntimat only as to the poetic end of the confusion. Here, presumably, the end is not to compare, but to portray the quality in a certain way. See note (5) to sambhāvyamā- nârtha atiśayôkti: such subtle distinctions are necessarily subjective and may not be entirely clear from the examples alone. See also atadguņa. tulyayogitā tulyayogitā, 'equal joining': (1) a figure in which several subjects sharing a property or mode of action, though in unequal degrees, are re- presented as equivalently endowed; the lesser subject is thus magni- fied. (2) B 3.26 (27), D 2.330 (331-32), V 4.3.26, U 5.7, M 158. (3) śeso himagiris tvam ca mahānto guravah sthirāh | yad alanghitama- ryādāś calantīm bibhrtha kșitim (Bhāmaha: "The primeval serpent, the Himalaya and you, O King, are weighty and firm; you three. surpassing all limitation, support the unstable world"). (4) " ... he had the harmlessness of the serpent and the wisdom of the dove" (Samuel Butler). (5) Tulyayogitā functions as a comparison, but the common property is predicated of the subject and object jointly. Dandin also gives a type of upamā based upon the conjunction of evident unequals. This figure can be subdivided as to its purpose

192 (ninda, stuti) and in relation to the grammar of its subject ( prastāva- aprastavabhaj. 'not relating to the subject of the utterance': (1) a type of bhaj) tulyayogita in which the concatenated terms are obliquely related to the intentional subject of the utterance. (2) U 5.7. (3) tvadañga- mārdavam drastuļı kasya citte na bhāsate | mālatīšašabhrllekhākadali- nām kațhoratā (Udbhata; the subject is the softness of Umā's body; the concatenated terms share the opposite property and stand as objects of comparison: "Who would not think the malati flower. the crescent moon, or the plantain tree harsh and rough when he had seen the softness of your body?"). (4) "The soundproof walls shut out all noise from the street, and, in the hushed atmosphere common to art galleries, cathedrals, and banks, Max's melodious drawl sounded less out of place ... " (Margery Allingham). (5) In aprastāvabhāj, the concatenated terms are the topical subjects of the utterance. The term appears in Mammata as aprākaraņika (with the same meaning). aprākaraņika, 'not relating to the subject': (1) same as aprastāvabhāj. (2) M 158C. nindā, 'blame: (1) a type of tulyayogitā which has blame or depreciation for its purpose. (2) D 2.330 (332). (3) sangatāni mrgâksīņām tadidvilasitāni ca | kşaņadvayam na tișthanti ghanârabdhāny api svayam (Dandin: "Neither affairs with young ladies nor the fleeting bolts of lightning last for more than a few moments-though begun profoundly [begun in the clouds]"). (4) "Death lays his icy hand on kings: / Sceptre and Crown / Must tumble down, / And in the dust be equal made / With the poor crooked scythe and spade" (James Shirley). (5) Cf. stuti tulyayogitā. prastāvabhāj, 'relating to the subject of the utterance': (1) a type of tulyayogitā in which the concatenated terms function as the intentional subject of the utterance. (2) U 5.7. (3) pāņdukşāmamı vadanam hrdayam sarasam tavâlasam ca vapuh / āvedayati [sic] nitāntaņı kşetriyarogam sakhi hrdantah (Mammața: "Your face, pale and wan, your soulful heart and indolent form betray an incurable disease in the soul"). (4) "The streets near the station were full of the smell of beer and coffee and decaying fruit and a shirt-sleeved populace moved through them with the intimate abandon of boarders going down the passage to the bathroom" (Edith Wharton). (5) Prastāvabhāj is to be distinguished from aprastāvabhāj. This term appears in Mammața as prākaraņika (with the same meaning).

GLOSSARY prākaraņika, 'relating to the subject of the utterance': (1) same as prastāvabhāj. (2) M 158C. 193 stuti, 'praise': (1) a type of tulyayogitā whose purpose is praise or apprecia- tion. (2) D 2.330 (331). (3) yamah kubero varuņaḥ sahasrâkșo bhavān api / bibhraty ananyavișayām lokapāla iti śrutim (Dandin: "The God of Death, the God of Wealth, the Lord Visnu, Indra himself, and you, O King, bear the unmistakable appellation "World- Protector"). (4) " ... inquisitive with the inquisitiveness of a despot and worried as with the responsibilities of a god" (G. K. Chesterton). (5) Cf. nindã.

dīpaka dipaka, 'enlightener': (1) a construction wherein several parallel phrases are each completed by a single (unrepeated) word or phrase; zeugma. (2) NŚ 16.40, 53-55, B 2.25-29, D 2.97-115, V 4.3.18-19, U 1.14, R 7.64-71, M 156-57. (3) sarāmsi hamsaih kusumaiś ca vrkşāh mattair dvirephaiś ca saroruhāņi / goșțhibhir udyānavanāni caîva tasminn aśūnyāni sadā kriyante (Bharata: "In this city, the ponds are filled with swans, the trees with flowers, the lotuses with drunken bees, and the pleasure groves with elite parties"). (4) "All things uncomely and broken, all things worn out and old, / The cry of a child by the roadway, the creak of a lumbering cart, / The heavy steps of the ploughman, splashing the wintry mould, / Are wronging your image that blossoms a rose in the deeps of my heart" (W. B. Yeats), (5) Zeugma is one of the universal devices of any stylized literary mode and is one of the four original figures of speech recognized by Bharata. The name dīpaka (from dip-, 'shine'), which has been translated as 'enlightener' or 'illuminator', is a more vivid formula- tion of the subjective effect of this alamkāra than is the static zeugma ('yoke'), which describes only the grammatical appearance or form of the figure. The word which ties together the various phrases by being at once a grammatical part of them all illuminates the entire phrase, or at least those parts of it which require that word in order to be understood. By multiplying the phrases dependent on the zeugma, the effect of illumination in increased. The effect may be compared to the Latin or German sentence in which the English speaker waits breathlessly for the verb. The figure zeugma has been classified in two ways by the Indian theorists: as to the place of the common word in the total construction, and by the grammatical role which the common word plays in the construction. The two

194 GLOSSARY modes are not exclusive, and several authors employ both. First of all, the shared word may occur at the beginning (di), in the middle (madhya), or at the end (anta) of the total construction. These words are usually interpreted to mean first, interior, or fourth quarters of the verse and not absolute first or last position in the verse (sce, madhya dipaka). It is interesting in this connection to compare the renaissance classification of zeugma into prozeugma, mesozeugma, and hypozeugma (Taylor, p. 132-33). All the writers except Bharata and Mammata reproduce this extrinsic triad, and it may represent the oldest view on the figure. Dandin, however, proposes a fourfold division by part of speech, in line with his several other uses of the same discrimination (cf. svabhāvôkti, vyatireka, viśeșôkti): the word shared by the several phrases may be either an adjective (guna), a verb (kriyā), a generic noun (jāti), or an proper name (dravya). Dandin does not abandon the threefold division, so his treatment may be said to produce a twelvefold zeugma. This classification by grammatical role is unknown only in the two earliest writers, Bharata and Bhämaha. Since the encyclopaedist Mammata uses it ex- clusively, we may presume that the older threefold division was no longer felt to be adequately diagnostic. However, the later writers simplify Dandin's four grammatical parts of speech into two-kriyā, 'verb' and kāraka or kartr, 'noun' (see kriyā dīpaka)-and Vāmana accepts only kriyā as legitimate. It should be remarked that the word "zeugma" is more commonly applied in English rhetoric to a defect of construction whereby a single word is related to two (or more) in such a way that the construction is not the same for the two (for example: "She came in a flood of tears and a bath chair", where "in" is used both modally and locatively). This is a special case of the more general figure. but in usage the two must not be confounded, as no defect is intended in our use of zeugma. Oddly enough, the Agni Purana, which represents an independent sāmkhya-oriented poetic tradition, ignores the figure dipaka. This is doubly curious, as that text in no way minimizes verbal figures-as opposed to figures founded on meaning (tropes)-and in view of the fact that dipaka is one of the four figures known to Bharata (with upamā, rūpaka, and yamaka). anta, 'final': (1) a type of zeugma in which the grammatical element shared by the several phrases occurs at the end of the entire construc- tion. (2) B 2.25 (29), D 2.102 (104-105), V 4.3.19, U 1.14, R 7.65

GLOSSARY (68. 71). (3) dūrād utkanthante dayitānām samnidhau tu lajjante | trasyanti vepamānāh šayane navapariņayā vadhvah (Rudrața: "They 195 long when their lovers are far, are shy when they are present, in bed are trembling and afraid-the newly-wed women"). (4) "The new moon behind her head, an old helmet upon it, a diadem of accidental dewdrops round her brow, would have been adjuncts sufficient to strike the note of Artemis, Athena, or Hera" (Thomas Hardy). (5) The examples also show kāraka dīpaka and kriyā. See ādi, madhya. adi, 'initial': (1) a type of zeugma in which the grammatical element shared by the several phrases occurs at the beginning of the entire construction. (2) B 2.25 (27), D 2.102 (98-101), V 4.3.19, U 1.14, R 7.65 (66, 69). (3) nidrâpaharati jāgaram upaśamayati madanadaha- nasamtāpam / janayati kāntāsamgamasukham ca ko'nyas tato bandhuh (Rudrata: "Sleep steals wakefulness away, calms the burn- ing passions of love, excites the pleasure of meeting the beloved; what else can be compared to it?"). (4) "Mrs. Pascoe stood at the gate looking after them; stood at the gate until the trap was round the corner; stood at the gate, looking now to the right, now to the left; then went back to her cottage" (Virginia Woolf). (5) Both examples also illustrate kāraka dīpaka (or dravya). ekârtha, 'integral': (1) a type of zeugma in which the tenor of the several conjoined phrases is complimentary. (2) D 2.112 (111). (3) haraty ābhogam āśānām grhņāti jyotişām gaņam | ādatte câdya me prāņān asau jaladharâvalī (Dandin: "The garland of rain clouds fills up the expanse of the sky, sequesters the flock of stars, and steals away my hopes"). (4) "A coyote sings more sweetly to me than any bird. He pushes the horizon back with his voice. He makes a gift of space. He says that something is still hidden. He reports escape. He acknowledges himself. He celebrates survival" (Jessamyn West). (5) Compare viruddhârtha dīpaka, where the tenor of the conjoined phrases is contradictory or contrary, and also ślesa dīpaka, where the phrases are neither contrary nor complimentary, but merely punned. The Sanskrit example also illustrates anta dīpaka and jāti dīpaka, the English, ādi and jāti. kāraka, 'nominal': (1) a type of zeugma in which the word common to the several phrases is a noun, usually the subject of the entire construction. (2) R 7.64 (69-71), M 156. (3) sramsayati gātram akhilam glapayati ceto nikāmam anurāgah | janam asulabham prati sakhe prāņān api mankșu mușņāti (Rudrața: "Passion exhausts the

GLOSSARY 196 body and makes the mind languid, but friend, when the loved one is unattainable, it quickly steals one's life away"). (4) "No profane hand shall dare, for mc, to curtail my Chaucer, to Bowdlerize my Shakespeare, or mutilate my Milton" (Anon., quoted in Burton Stevenson). (5) This item covers two of Dandin's four types: jati and dravya. The Sanskrit example also illustrates madhya dīpaka the English, di. Also called kartr ('agent'). kriyā, 'verbal': (1) a type of zeugma in which the word common to the several phrases is a verb. (2) D 2.97 (99, 104), V 4.3.18, U 1.14, R 7.64 (66-68), M 156. (3) kāntā dadāti madanam madanah samtāpam asamam anupaśamam | samtāpo maraņam aho tathâpi śaranam nrnām saîva (Rudrața: "The beloved woman imparts longing; longing, unquenchable and incomparable passion; passion, death. She is thus the only refuge of men"). (4) "Not poets alone, nor artists, nor that superior order of mind which arrogates to itself all refinement, feel this, but dogs and all men" (Theodore Dreiser). (5) Daņdin recognizes kriyā dīpaka as one of four types (see jāti, guņa, dravya), Vāmana as the only type, and the other authors, as one of two types (see kāraka). This classification by the gram- matical function of the common word (or phrase) is not known in the oldest texts (Bharata and Bhāmaha), but Dandin produces the standard fourfold division here as in other figures (svabhāvôkti, vyatireka, viśeşôkti). It becomes simplified later by the coalescing of jāti and dravya into kāraka and the dropping of guņa dīpaka. Vāmana's ignorance of other varieties is curious, but is probably due to his programmatic attempt to reduce all the figures to kinds of simile. A zeugma in which the shared word is a noun does not fit as well into the frame of comparison as does the verbal zeugma, since the figure then recounts only different aspects of one subject rather than the same aspect of two different subjects. The Sanskrit example also shows ādi dīpaka (the verb occurs not first, but in the first quarter stanza); the English shows madhya dīpaka. guņa, 'adjectival': (1) a type of zeugma in which the word common to the several phrases is an adjective of description. (2) D 2.97 (100). (3) śyāmalāḥ prāvrșeņyābhir dišo jīmūtapanktibhiḥ / bhuvaś ca suku- mārābhir navašādvalarājibhih (Dandin: "The sky is dark with great ranks of rain clouds, the earth with sweet shoots of new grass"). (4) "Beautiful lofty things: O'Leary's noble head" / My father upon the Abbey stage, before him a raging crowd / ... Standish O'Grady supporting himself between the tables / Speaking to a drunken

GLOSSARY audience high nonsensical words; / Augusta Gregory seated at her great ormolu table ... " (W. B. Ycats). (5) Both cxamples also 197 illustrate ādi dīpaka. Adjectival dīpaka is noted only by Daņdin and probably was not felt to be different from verbal (kriyā) dīpaka by later writers, inasmuch as the verb "to be" can be supplied in all such cases. jati, 'generic': (1) a type of zeugma in which the word common to the several phrases is a generic noun. (2) D 2.97 (98, 103, 105). (3) pavano dak şinah parņam jīrņam harati vīrudhām | sa evâvanatângīnām mānabhangāya kalpate (Dandin: "The southern wind wafts away the withered leaf of plants, calms the anger of modest women"). (4) "A book, like a person, has its fortunes with one; is lucky or unlucky in the precise moment of its falling in our way, and often by some happy accident counts with us for something more than its independent value" (Walter Pater). (5) Both examples also illustrate ādi dīpaka. Cf. dravya, where a proper (specific) noun supplies the unity of phrase. dravya, 'material': (1) a type of zeugma in which the word common to the several phrases is a proper name (or specific referent). (2) D 2.97 (101). (3) vişņunā vikramasthena dānavānām vibhūtayah / kvâpi nītāh kuto'py āsann ānītā daivatarddhayah (Daņdin: "Some- where the wealth and fortune of the Dānavas were taken by Vişnu victorious, from somewhere the success and prosperity of the Gods were brought"). (4) "Old Professor Huxtable, performing with the method of a clock his change of dress, let himself down into his chair; filled his pipe; chose his paper; crossed his feet; and extracated his glasses" (Virginia Woolf). (5) Both examples also illustrate ādi dīpaka. Cf. jāti, where a generic term occupies the common slot. madhya, 'mid': (1) a type of zeugma in which the grammatical element shared by the various phrases occurs in the middle of the entire construction. (2) B 2.25 (28), D 2.102 (103-104), V 4.3.19, U 1.14, R 7.65 (67, 70). (3) sramsayati gātram akhilam glapayati ceto nikāmam anurāgaḥ | janam asulabham prati sakhe prāņān api maīk su mușņāti (Rudrața: for the translation, see kāraka). (4) "Her presence brought memories of such things as Bourbon roses, rubies and tropical midnights; her moods recalled lotus-eaters and the march in 'Athalie', her motions, the ebb and flow of the sea, her voice, the viola" (Thomas Hardy). (5) The "middle" of the construc- tion is usually interpreted to mean the second or third quarter of a verse (cf. Rudrața's example for ādikriyādīpaka, 7.66). The English

GLOSSARY 198 example runs contrary to this convention in that the word "recalled" while standing third in its phrase, occurs in the first of the phrases united by the zeugma. The Sanskrit example also illustrates barakg dīpaka, the English, kriyā dīpaka. See ādi, anta. mala, 'garland': (1) a type of zeugma in which the related phraw :. only share a common word, but express a sequence of ideas in we a way that the following phrase recalls or refers to the former. (2) (2) D 2.108 (107), M 157. (3) śuklah śvetárcişo vrddhyai pakyah pañcaśarasya sah / sa ca rāgasya rāgo'pi yūnām rat yutsavaśri, ah (Dandin: "The waxing phase makes for the prosperity of the white- rayed moon, the moon, of the Love God, he, of passion, passion, for youths' happiness in festival of love"). (4) " ... while [the Wise Youth] remained tranquil on his solid unambitious ground, fitting his morality to the laws, his conscience to his morality, his comfort to his conscience" (George Meredith). (5) Both examples also show ādi dīpaka, in the sense noted under madhya (5). Various figures illustrate this enchainement of ideas: notably upamā, rūpaka, and nidarśanā. See also the figures kāraņamālā and ekâvalī, to which mālā dīpaka seems merely the added application of a zeugma. viruddhârtha, 'disparate': (1) a type of zeugma in which the tenor of the several conjoined phrases is contrary or contradictory. (2) D 2.110 (109). (3) avalepam anangasya vardhayanti balāhakāh karśayanti tu gharmasya mārutôddhūtaśīkarāh (Daņdin: "The rain clouds increase the arrogance of the Love God, but diminish the summer's heat-their showers blown about by the wind"). (4) "But in sooth Mr. Slope was pursuing Mrs. Bold in obedience to his better instincts, and the signora in obedience to his worser" (Anthony Trollope). (5) See ekârtha dipaka, in which the partial phrases are complimentary, that is, express compatible ideas. ślista, 'punned': (1) a type of zeugma in which the related phrases not only share a common word, but have appended puns expressing similarity. (2) D 2.114 (113). (3) hrdyagandhavahās tungās tamāla- śyāmalatvişaḥ | divi bhramanti jīmūtā bhuvi caite matangajāh (Dandin: "Carrying pleasant odors, lofty, hides as dark as the tamāla flower, the clouds roam in the sky, here on earth, elephants"). (4) "As lines so loves oblique may well / Themselves at every Angle greet: / But ours so truly Parallel, / Though infinite can never meet" (Andrew Marvell). (5) This should be compared with mālā dipaka, where the relation between the various phrases is based upon a real implication of meanings. The puns have nothing to do with the zeugma, but

GLOSSARY are simply additional modes of expressing relation (usually similarity) and can therefore reinforce the zeugma. The point may be that the 199 zeugma is entirely a formal device and may adjoin either legitimately consequential notions (mālā) or those whose similarity is entirely verbal (ślişta).

dușkara duşkara, 'difficult to accomplish': (1) a figure defined and treated under citra alamkāra.

drstânta drstânta, 'example': (1) the adjunction of a second situation which bears upon the same point as the first and where the purpose is entirely one of illustration. (2) U 6.8, R 8.94 (95-96), M 155. (3) kimcâtra bahunôktena vraja bhartāram āpnuhi / udanvantam anāsādya mahā- nadyah kim āsate (Udbhața: "What's the point of talking further? Go out and get yourself a husband! What fate will befall the great rivers if they do not fall into the ocean?"). (4) "Have you not in a Chimney seen / A sullen faggot, wet and green, / How coyly it receives the heat / And at both ends does fume and sweat? / So fares it with the harmless Maid / When first upon her Back she's laid; / But the kind experienced Dame / Cracks, and rejoices in the flame" (John Wilmot, Earl of Rochester). (5) The relationship is here between two specific situations, and the purpose of the rapprochement is clarification. That is why a rigorous parallelism of element and aspect is required. In prativastûpamā alamkāra, one common property is shared; the terms themselves need not be comparable. Note in Wilmot's example the duplication which borders on double-entendre: "sullen ... coyly ... heat ... fume", etc. Likewise, the intention of the speaker is not necessarily substantiation (where a doubt might arise), and in this drstânta differs from arthântaranyāsa, even though there is a general tendency to ccnsider the latter a relation between two remarks-a specific and its corresponding universal-which is more an explanation of the process of explanation than a condition of the relation between two terms which do clarify one another (cf. Daņdin's example for virodhavat arthântaranyāsa). Both are illustra- tions, but the word "illustration" is equivocal. All these figures (especially vāk yârtha upamā) differ from simile in that the comparative particle is lacking; but aside from this, all can be and usually are described in the same terms (subject, object, common property or

GLOSSARY 200 aspect). Rudrata carries this principle of explanation to the extreme of excluding all considerations of intention on the part of the speaker -a purely structural analysis-but there are so many figures to account for that this apparent simplification really serves only to confuse the many legitimate subtleties of analysis contained in Dandin, for example. The example may be thought of as introduced by the particle "so". avivaksita, 'not the intended subject': (1) a type of drstânta in which the subject term or phrase follows the object or example. (2) R 8.94 (96). (3) lokam lolitakisalayavişavanavāto'pi makşu mohayati tāpayatitarām tasyā hrdayam tvadgamanavārtâpi (Rudrața: "The forest wind, heavy with the fragrance of swaying buds, intoxicates everyone; the news of your coming quickens the pain in her heart"). (4) "Do you know these flashes of the absolute and unalterable rightness of a thing? One of them blinded Saul of Tarsus on the road to Damascus; something of the same kind blinded me for the whole iength of Lower Regent Street" (Oliver Onions). (5) Cf. vivakșita. vivakșita, 'intended subject': (1) a type of drstânta in which the term or situation intended as subject by the speaker precedes the term or situation intended as example or object. (2) R 8.94 (95). (3) tvayi drsta eva tasyā nirvāti mano manobhavajvalitam / āloke hi sitâmšor vikasati kumudam kumudvatyāḥ (Rudrața: "Her heart enflames with love as soon as he appears; at the sight of the moon, the lotus blooms"). (4) "But she was the bossiest human I'd ever seen, and my father indicated the same, but following his style in such matters. he couldn't say so direct but referred her case to a play by a man named Shakespeare, saying she needed taming" (Robert Lewis Taylor). (5) This figure is a good example of Rudrata's purely formal classification (see note on drstânta). vaidharmya, 'difference': (1) a type of drstânta in which the phrase serving as example (object) involves terms and aspect antithetical to the phrase standing as subject. (2) M 155. (3) tavâhave sāhasakarma- śarmanah karam krpāņântikam āninīşatah | bhațāh pareşām višarāru- tām agur dadhaty avāte sthiratām hi pāmsavah (Mammața: "You of the violent deed! at your challenge the enemy's soldiers-their sword-girt arms about to strike-perish utterly! So motes of dust are steady as long as the wind does not blow"). (4) "The righteous minds of innkeepers / Induce them now and then / To crack a bottle with a friend / Or treat unmoneyed men, / But who hath seen the Grocer / Treat housemaids to his teas / Or crack a bottle of fish-

GLOSSARY sauce / Or stand a man a cheese?" (G. K. Chesterton). (5) The 201 classification is refuted by Rudrata (8.86C), who would consider this a variation on samuccaya alamkāra. See also the remarks under drstânta vyatireka. sädharmya, 'similitude': (1) a type of drstânta in which the terms and aspect of subject and examples are in balanced concord. (2) M 155. (3) tvayi drsta eva tasyā nirvātt mano manobhavajvalitam | āloke hi himâmśor vikasati kusumam kumudvatyāh (Mammața; the example is the same as that offered by Rudrata for vivakșita drstânta). (4) "So shuts the marigold her leaves / At the departure of the sun; / So from the honeysuckle sheaves / The bee goes when the day is done; / So sits the turtle when she is but one / And so all woe, as I since she is gone" (William Browne). (5) This is, of course, drstânta itself, to be understood as a subtype only in the sense that Mammata recognizes an antithetical example, too (cf. vaidharmya). nidarśanā nidarśanā (I) (neuter in Daņdin and Vāmana), 'pointing to': (1) a figure in which a particular situation is translated into a general truth, and a moral is drawn which is based upon the mode of action and the ultimate tendency of that situation. (2) B 3.32 (33), D 2.348 (349-50), V 4.3.20, M 150. (3) ayam mandadyutir bhāsvān astam prati yiyāsati / udayah patanāyêti śrīmato bodhayan narān (Bhāmaha: "The dull red sun nears the western term, telling the wise that greatness is but the precedent of decline"). (4) "Poor HALL caught his death standing under a spout, / Expecting till midnight when NAN would come out, / But fatal his patience, as cruel the Dame, / And curst was the Weather that quench'd the man's flame. / 'Who e'er thou art, that read'st these moral lines, / Make love at home, and go to bed betimes'" (Matthew Prior). (5) This figure differs from arthântara- nyäsa in that the general truth is here expressed as the very meaning of the particular situation, not as another and more valid formula- tion of it. Some writers (Dandin, Vāmana) define the figure as a relation of two situations through a similar consequence: moral instruction in terms of the final cause. The second situation is seen not as extrinsic to the first, maintaining a relation of similitude to it, but it is, as it were, a universalization of the same idea. This view- point gives a rationale for the transition to the second type, described first by Mammata, which, in the modern writers (Candrâloka), supplants the other.

GLOSSARY 202 asat. 'unauspicious': (1) a type of nidarsana in which the instruction given is baleful. (2) D 2.348 (350). (3) atyuccapadádhyāsah patanā- yêty arthašālinam śamsat | āpāņdu patati patram taror idam bandha- nagrantheh (Vamana: "The withered leaf falls from its once firm stem on the trec, warning the attentive that the attainment of an exalted position is but a pretext for a greater fall"). (4) "When the leaves in autumn wither / With a tawny tanned face / Warped and wrinkled up together, / The year's late beauty to disgrace; / There thy life's glass may'st thou find thee: / Green now, grey now, gone anon, / Leaving, worldling, of thine own / Neither fruit nor leaf behind thee" (Joshua Sylvester). (5) Cf. sat. sat, 'auspicious': (1) a type of nidarsanā in which the instruction given is auspicious. (2) D 2.348 (349). (3) udayann eşa savitā padmeșv arpayati śriyam / vibhāvayitum rddhīnām phalam suhrdanugraham (Dandin: "The rising sun has wrought beauty in the lotus, conveying the results of success, pleasing to friends"). (4) "He thus became immensely Rich, / And built the Splendid Mansion which / Is called / 'The Cedars, / Muswell Hill'. / Where he resides in Affluence still / To show what Everybody might / Become by / SIMPLY DOING RIGHT" (Hilaire Belloc). (5) Cf. asat. nidarśanā (II): (1) negative illustration; a figure in which the illustrative example demonstrates in the negative the point originally made in the negative. (2) M 149. (3) kva sūryaprabhavo vamšah kva câlpavi- şayā matiḥ | titīrșur dustaram mohād udupenâsmi sāgaram (Kālidāsa, quoted by Mammata; the poet thus indicates his modesty before the task he has set himself: telling the true history of the Rāghava: "How can my feeble skill cope with a race born of the sun? I am about to cross the fearsome ocean in a rowboat"). (4) "I've heard it said my father was a bang-up poker player. But there wasn't much chance of his winning, not with those chaps. Mostly. they were serious, hard-working professional men-thieves, forgers, cutthroats. small-time river pirates and a backslid preacher or two-as interesting spoken a group as you would care to meet, but they could no more have gambled honest than they would have been comfortable in church" (Robert Lewis Taylor). (5) According to Mammața, this illustration through two negatives or opposites suggests a simile. Indeed, the positive understanding derived from the figure is a simile: "crossing the sea in a rowboat is like describing the Rāghava"; "those who gamble dishonestly resemble those who are uncomfort- able in church". Yet the point seems rather truistical and implicit

GLOSSARY in the notion of ilfustration anyway. The figure serves as the negative of both the figures nidarsana (I) and drstânta, depending on whether 203 illustration or instruction is intended. Cf. also vidaranã. mālā, `garland': (1) a type of nidarśanā in which several situations are expressed as equivalent to the situation standing as subject of the utterance. (2) M 149C. (3) dorbhyām titīrşati tarangavatībhujangam ādātum icchati kare hariņânkabimbam | merum lilanghayișati dhruvam eşa deva yas te guņān gaditum udyamam ādadhāti (Māgha, quoted by Mammata: "He who arrogates to himself the task of extolling your virtues, O King, is attempting to swim across the limitless ocean, is trying to grasp the moon in his hand or to climb the primeval mountain"). (4) "I'd love to waltz with you. I'd love to have my tonsils out, I'd love to be in a midnight fire at sea" (Dorothy Parker). (5) This figure is merely a repetition of nidarśanā II. parikara parikara, 'entourage': (1) a figure in which the adjectival qualifications or epithets of a thing are multiplied with a view to re-enforcing the distinctiveness of that thing. (2) R 7.72 (73-76), M 183. (3) upacita- pariņāmaramyam svādu sugandhi svayam kare patitam / phalam utsrjya tadānīm tāmyasi mugdhe mudhêdānīm (Rudrața; the excellence of the "fruit" abandoned is thus emphasized: "Abandoning this fruit, which has ripened in its own time and, sweet smelling, has fallen of itself into your hand, you now regret, O lovely-how uselessly!"). (4) "This wimpled, whining, purblind, wayward boy; / This senior-junior, giant-dwarf, Dan Cupid; / Regent of love-rimes, lord of folded arms, / Th'anointed sovereign of sighs and groans, / Liege of all loiterers and malcontents, / Dread prince of plackets, king of codpieces, / Sole imperator and great general / Of trotting paritors: -O my little heart !- / And I to be a corporal of his field, / And wear his colours like a tumbler's hoop!" (Shakespeare). (5) Rudrata gives four examples, each illustrating one of the four genera of meaning: jāti, kriyā, guņa, and dravya. The example cited is dravya; the English example, a potpourri. parivțtti parivrtti, 'exchange': (1) a figure expressing non-literally an exchange of ideas or things; "give and take". (2) B 3.40 (41), D 2.355 (356), V 4.3.16, U 5.16, R 7.77 (78), M 172. (3) śastraprahāram dadatā bhujena tava bhūbhujām | cirârjitam hrtam teşām yašah kumudapāņ-

GLOSSARY 204 duram (Dandin: "Striking blows with your sword, O King, your arm has seized the long possessed and lotus-pale glory of this earth princes"). (4) "'By whose direction found'st thou out this place" 'By love, that first did prompt me to enquire. He lent me counsel and I lent him eyes'" (Shakespeare). (5) The idea that one action entails both an acquisition and a loss is central to all the varieties of parivrtti mentioned by our authors. The differentiating criteria are extrinsic to the notion of exchange and concern only the thines exchanged; they may be similar or dissimilar (Vamana); if dissimilar better for worse or worse for better (Mammata, Udbhata); the ex- change may be auspicious or inauspicious (Udbhata), literal or figurative (Rudrata). Only Bhamaha, the earliest of the writers. diverges in any way from this sterile mechanism by asserting that the exchange must be seconded by drawing a moral (arthântaranyāsa, q.v.), as: "pradāya vittam arthibhyaḥ sa yaśodhanam ādita | satām višva- janīnānām idam askhalitam vratam" ('Giving wealth to all who ask, he obtains a treasure of glory: such is the inescapable duty of the just'). nyūna, 'deficient': (1) a type of parivrtti in which the item exchanged is inferior in quality or station to the item acquired. (2) U 5.16, M 172C. (3) netrôragavalabhrāmyan mandarâdriśiraścyutaih / ratnair āpūrya dugdhâbdhim yaḥ samādatta kaustubham (Udbhața; Vișnu is referred to: "He filled up the sea of milk with jewels fallen from the primeval mountain top where he wandered in the Snake King's caves; then he took from the sea a magic jewel of inestimable price"). (4) "Seeing my flesh must die so soon, / And want a head to dine next noon,- / Just at the stroke, when my veins start and spread, / Set on my soul an everlasting head" (Sir Walter Raleigh; in expecta- tion of his execution). viśișta, 'qualified': (1) a type of parivrtti in which the item exchanged is superior in quality or station to the item acquired. (2) U 5.16, M 172C. (3) latās tv adhvanyānām ahaha dršam ādāya sahasā dadaty ādhivyādhibhramiruditamohavyatikaram (Mammața; the wanderers are reminded of their distant loves: "The lovely creepers steal the sight of wanderers and give back pain, suffering, loss, sobbing, and confusion"). (4) "Even such is Time, that takes in trust / Our youth. our joys, all we have, / And pays us but with age and dust" (Sir Walter Raleigh). sama, 'equal': (1) a type of parivrtti in which the item exchanged and the item acquired are similar or of equal value. (2) U 5.16, V 4.3.16, M 172C. (3) (4) See parivrtti.

GLOSSARY parisamkhyā parisamkhya. 'dehmitation': (1 an alamkara identical with niyamavat 205 slea. (2) R 7.79, M 185. (5) The name most generally means 'enumeration'. but the logicians use the term in a special sense: exhaustive enumeration'; hence the application here, where a series of ideas are limited to special senses through an enumeration based upon mention of those limitations. paryāya paryāya (I). 'synonym': (1) same as aprastutapraśamsã II. (2) R 7.42 (43). (5) This term evidently suppletes the otherwise absent figure aprastutapraśamsā in Rudrata. The definition alone could be interpreted as the figure paryāyôkta, also absent in Rudrața. It is the inverse of bhāva II. paryāya (II): (1) same as viśeșa II. (2) R 7.44 (45-46), M 180-81. (5) A single thing is represented in several contexts, or several things in one context. The unique thing is generally a state of soul, such as happiness or valor, and the poetic force of the figure is presumably a function of a figurative denial of the law of the excluded middle. There is no obvious relation to the figure paryāyôkta or to the other variety of paryāya given by Rudrața. Mammața accepts both paryāya in sense two and paryāyôkta, establishing them as separate figures.

paryāyôkta paryāyôkta, 'periphrasis': (1) a figure in which a speaker conveys his intention without reference to the evident motive which prompts his utterance; oblique reference. (2) B 3.8-9, D 2.295-97, U 4.6, AP 345.18, M 175. (3) daśaty asau sahakārasya mañjarīm / tam aham vārayişyāmi yuvābhyām svairam āsyatām (Daņdin: "The cuckoo is eating the mango blooms; I'll go shoo him off. You two sit down here in peace"). (4) "'I should like that very much. Where shall we go?' I reflected a moment and answered, 'I hardly like to suggest a public bar, but I notice the Eagle is closc by, and though it is but a primitive place with a small bar and very hard chairs, it is a free house. The beer is said to be drinkable'. 'The only difliculty is that I am forbidden by my doctor to drink beer'. '1 understand that the Eagle frequently has whisky, rum, and gin, as well as beer'. 'I was hoping that we might find more congenial surroundings at your studio'. 'I'm afraid my studio is in process of cleaning. We should not be comfortable there. Why not come and try the lemonade

GLOSSARY 206 at the Eagle?'" (Joyce Cary; neither speaker has any money). (5) The figure might be called "pretext" (both examples illustrate this kind of oblique reference), but the term would be too limited for the Sanskrit cases, some of which would make good illustrations for Anandavardhana's dhvani. The specific alleging of a pretext is the figure leśa. Bhāmaha's example from the Ratnaharana shows a periphrasis without the alleging of a pretext: "grhesv adhvasu va nânnam bhuñjmahe yad adhītinah / na bhuñjate dvijās ... " ('We do not eat food, either at home or while travelling, which learned Brahmins have not tasted'). This is uttered by Krsna in order to avoid being given poisoned food. Similarly, the example from Dandin is spoken by a go-between who has arranged the rendez-vous and now takes her leave. All these cases share the common trait of making an oblique reference without that reference being based in any way upon simile or its elements. The figure samāsôkti also involves oblique reference, but to an object which is understood as the subject of a comparison and by means of a recitation of properties presum- ably common to both the implicit and explicit subject. Aprastuta- praśamsā reposes on other kinds of relation (such as cause-effect), but that relation is commonly evident in the examples. Paryāyôkta, however, depends upon nothing but context and connotation for its comprehension, and it need not serve even a descriptive (let alone comparative) end. It would seem that all cases of suggestion not otherwise classifiable should be referred to this head. Mammata gives paryāyôkta as well as paryāya II, establishing the two as differ- ent figures.

pihita, 'hidden': (1) a figure in which a quality or attribute is pictured as pihita

withstanding, and predominant over another quality which in the normal state of affairs would be the stronger. (2) R 9.50 (51). (3) priyatamaviyogajanitā krśatā katham iva tavêyam angeșu | lasadin- dukalākomalakāntikalāpeșu lakșyeta (Rudrața; the thinness of her aspect is obscured by her moon-like glamour: "The thinness which you suffer in separation from your lover is not easily perceived. for your limbs are effulgent with beauty gentle as the glimmer of the waning moon"). (4) "He that looks still on your eyes. / Though the winter have begun / To benumb our arteries, / Shall not want the summer sun" (William Browne). (5) Cf. tadguna. Unlike sambhāvyamānârtha, there is no transference of quality here, only

GLOSSARY cancellation as far as the onlooker is concerned. Pihita could be 207 mentioned as a special kind of ādhikya atiśayôkti: there the exaggera- tion of the force of the quality also minimizes another quality. It differs from atisayôkti in that the qualities in pihita are not contraries, but may be any two not normally concomitant. punaruktâbhāsa punaruktâbhäsa, 'appearance of redundancy': (1) a figure in which two homonyms are used in the same sentence in different senses. (2) U 1.3, M 122-24. (3) tanuvapur ajaghanyo'sau karikuñjararudhira kta- kharanakharalı / tejodhāma mahah prthumanasām indro harijişnuh (Mammata; tanu-vapur is not, as one might be led to believe, 'body-body'. but 'slight-bodied': "The lion, accustomed to conquer, slender bodied, first among beasts, the nails of his paws red with the blood of lordly elephants, a repository of splendor, majesty itself. the Indra of the ambitious"). (4) " ... for if I was a light of literature at all it was of the very lightest kind" (Samuel Butler). (5) Compare avrtti, where the repetition is not apparent, but real. For Mammata, this is the only alamkāra which involves both śabda and artha.

pūrva pürva (I), 'previous': (1) a figure in which the subject of comparison is said to precede in the order of creation or time the object to which it is compared. (2) R 8.97 (98). (3) kāle jaladakulâkuladaśadiśi pūrvam viyoginīvadanam | galadaviralasalilabharam paścād upajāyate gaganam (Rudrata; the face of the woman separated from her lover was created before the sky: "In the season when the ten directions of the sky are obscured by legions of clouds, the face of the separated wife appears first; only then does the sky release its load of incessant, flowing rain"). (4) "Be you not proud of that rich hair / Which wantons with the love-sick air; / Whereas that ruby which you wear, Sunk from the tip of your soft ear, / Will last to be a precious stone / When all your world of beauty's gone" (Robert Herrick; the beauty of the ruby will outlast the beauty of her hair). (5) There are a confusing number of figures and subtypes which involve in some way or other tampering with the normal time sequence. In the present case, since we are dealing with an implicit comparison, an assumption of the object's natural primordinateness is legitimate because, for the purposes of the comparison, it possesses reality in a

GLOSSARY

208 higher degree (that aspect of reality which is relevant-the common property). By inverting this assumption, we obtain an effect some- what like vatireka. Usually, however, the temporal inversion serves as a modality for exaggerating a quality of presumed effect (kārya- kāraņapaurvâparyaviparyaya atiśayokti), or it is simply one of the ways in which the causal process is distorted, for whatever effect (kāryânantaraja hetu and pūrva II). pūrva (II): (1) same as kāryânantaraja hetu (2) R 9.3 (4). (5) See pūrva I. Expressed as an inference, this figure is also noted by Rudrata under anumāna. prativastûpamā prativastûpamā, 'counterpart simile': (1) same as prativastu upamā. (2) V 4.3.2, U 1.22-23, M 154. (5) Prativastûpamā involves the repeti- tion of the common property; there need be no parallelism of terms within the two situations. At any rate, this figure differs from drstânta in that the emphasis is placed upon the repetition and not on the parallelism. Cf. also ubhayanyāsa. pratipa pratipa, 'against the grain': (1) a simile in which an affected pity or blame is directed at the subject of comparison in the presence of the object, or vice versa. (2) R 8.76 (77-78), M 201. (3) garvam asamvāhyam imam locanayugalena vahasi kim bhadre | santîdrśāni diśi diśi sarahsu nanu nīlanalināni (Rudrața, Mammața; interpreted differently, see below: "How can your eyes bear such insupportable arrogance? There are hundreds of blue lotuses in all the ponds"). (4) "Cupid and my Campaspe played / At cards for kisses-Cupid paid: / He stakes his quiver, bow and arrows, / His mother's doves, and team of sparrows; / Loses them too; then down he throws / The coral of his lip, the rose / Growing on's cheek (but none knows how); / With these, the crystal of his brow, / And then the dimple of his chin: / All these did my Campaspe win. / At last he set her both his eyes, She won, and Cupid blind did rise. / O Love! has she done this to thee? / What shall, alas, become of me?" (John Lyly: the speaker pities himself as he watches his rival, Cupid). (5) This figure seems to be a minor variation on vyatireka in which the outright distinction between the two terms compared is qualified or mediated by an expression of affected pity or blame. In the two examples given above, the lotuses and Cupid are depicted as exciting the envy of the

GLOSSARY subject with which they are implicitly compared. All of Mammata's 209 examples show an affectation of inferiority for or on the part of the object of comparison. Rudrata, however, though he gives the same example in one case. says that the inferiority attaches to the subject of comparison and thus amounts to an ironic affirmation of that sub- ject's comparability. He interprets the example already given as a taunt cast at the girl who is unable to bear the sight of the lotuses. Compare the following example: "The Model Boy of my timc-we never had but the one-was perfect: perfect in manners, perfect in dress. perfect in conduct, perfect in filial piety. perfect in exterior godliness: but at bottom he was a prig: and as for the contents of his skull. they could have changed place with the contents of a pie, and nobody would have been the worse off for it but the pie" (Mark Twain). pratyanika pratyanika. 'counter-attack': (1) a figure in which the object of comparison is represented as taking revenge for a set-back initially suffered at the hands of the subject. (2) R 8.92 (93), M 196. (3) yadi tava tayā jigişos tadvadanam ahāri kāntisarvasvam / mama tatra kim āpatitam tapasi sitâmšo yad evam mām (Rudrața: "The beauty of the cool- rayed moon. contesting my beloved's glance, has been claimed away; how does it happen then, O moon, that the sight of you consumes me?"). (4) "Arise, fair sun, and kill the envious moon, / Who is already sick and pale with grief / That thou her maid art far more fair than she" (Shakespeare). (5) This is a vyatireka with the tables turned. The term on whom the revenge is wreaked is ordinarily a neutral, but may also be the subject of comparison. The English example is perhaps wide of the mark in that the object of comparison itself is not only set back initially (by the girl to whom the remark is directed) but is also finally dishonored (by the sun); but the intent is the same as that of the Sanskrit: the speaker is disturbed by the suggestive beauty of the pale moon. Cf. pratīpa, where the object is determined as resentful, but where no element of riposte is present.

praśasti praśasti, 'flattery': (1) skill at employing words capable of melting the resistance of others. (2) AP 345.3-4. (5) Praśasti may be divided into premôkti ('speaking kindnesses') and stuti ('praise'). Kānti, the next śabdârthâlamkāra, is described as the use of such flattery in appropriate circumstances.

GLOSSARY

210 praśna praśna. `question': (1) a figure wherein a question or remark suggests ite answer or reply. (2) R 7.93 (95). (3) kim svargad adhikasukham lakşmīh / saurājyam adurbhikşam bandhusuhrtpanditaih samam satkāvyarasâmrtâsvādah (Rudrata; poetry is preferable to salvation: "Is the nectar-like taste of true poetry-a prosperous empire which knows no famine and the good fortune of relations, friends, and teachers alike-more agreeable than salvation?"). (4) "Was this the face that launch'd a thousand ships, / And burnt the topless towers of Ilium?" (Christopher Marlowe). (5) The rhetorical question. Rudra- ța calls this also uttara (q.v.), but since it is the reverse of the first type. we give it here. Praśna differs from parisamkhyā in that the answer alone is intended, not the nature of the concept. prahelikā prabelikā, 'riddle': (1) a puzzle, riddle, conundrum; a phrase, statement, or question constructed deliberately so that its meaning shall be misconstrued, but in some way intimating a solution to the difficulty thus created. (2) B 2.19, D 3.96-124, AP 343.22, 25-26, R 5.25 (29). (3) katham api na drśyate'sāv anvaksam harati vasanāni (Rudrața; answer: vayuh; "He is not at all visible though he seizes most obviously their clothing"; answer : the wind). (4) "What is that which is often brought to table, often cut, but never eaten?" (Robert Merry; answer: a pack of cards). (5) Prahelikā is first mentioned by Bhāmaha as an illegitimate extension of yamaka ('cadence"); serving no poetic purpose, this topic was apparently soon adopted into the growing rag bag of citrakāvya. Dandin, who gives the most extensive treatment (enumerating sixteen types), also avers that such puzzles serve only as recreation for scholars or critics and perhaps can be used to demonstrate the prowess of one writer over another-a sort of verbal jujitsu. The Agni Purāna and Rudrața both mention only two types of riddle: one in which the answer is already contained in the riddle phrase by a different reading of words (spastapracchan- nârtha), and the other, the most frequently met variety, in which an apparent paradox is proposed (vyāhrta). Both types can be found in Dandin, the first as ubhayacchanna, the second as ekacchanna. The example given above illustrates the second type. Since neither the names or the types found in Dandin are met with elsewhere, the sixteen varieties will be given here. ubhayacchanna, 'both concealed': (1) a conundrum in the form of a

GLOSSARY question in which the key to the solution is hidden, usually by a 211 grammatical device. (2) D 3.105 (122). (3) kena kah saha sambhūya sarvakāryeşu sannidhim | labdhvā bhojanakāle tu yadi drsto nirasyate (Dandin; the answer is "the hair" and is given by understanding kena as the instrumental not of the interrogative, but of the noun ka ('head"); it is considered improper to show one's hair during eating: "What with which coexisting in all their actions, and having come together during meal time, is, nevertheless, as soon as seen, put aside?"). (4) "What kind of cat is most valued in Sunday-school?" (Robert Merry; answer: a catechism). (5) The paradox here is entirely concealed because neither the answer nor a paradoxical reference which would suggest the nature of the solution is given explicitly in the statement; in fact, the first reading does not suggest that a paradox is even intended. In the examples, the key is given by a linguistic interpretation of one of the words of the statement (kena from ka; catechism as a kind of cat). Cf. vañcita and parușa. This and the following type (ekacchanna) represent a different kind of classification in that they necessarily involve the question form. Most Sanskrit riddles are not questions, unlike the English ones. The majority of the prahelika concern the mode of concealment (riddle-ness), whereas these two concern the mode of solution (given a question). ekacchanna, 'partly concealed': (1) a conundrum in the form of a question in which the key to the solution is partly given by a descriptive qualification. (2) D 3.104 (121). (3) na sprśaty āyudham jātu na strīņām stanamaņdalam / amanușyasya kasyâpi hasto'yam na kilâpha- lah (Dandin; intended is the gandharva-[amanușya] hasta, a name for the eranda tree: "He touches no weapon and not even the breast of woman; the hand of this unmasculine thing is nevertheless provided with fruits"). (4) "Who prolongs his work to as great a length as possible, and still completes it in time?" (Robert Merry; answer: the rope-maker). (5) The name "partly concealed" refers to the fact that a part of the truth must be given in order for the paradox to be understood. In the Sanskrit examples, the fruit (of the tree) is regarded as giving just enough clarity for the paradox to function. In the English, the word "length" fulfills the same function, for it candidly refers to the rope, too. This and the preceding type repre- sent a different level of classification and are opposed to one another in the sense that ubhayacchanna involves only an apparent paradox. Compare viśeșôkti alamkāra, virodha, and virodhâbhāsa.

namântarita, 'of another name': (1) a conundrum whose 212 effected by understanding that the statement in fact implies a name solution is and not the subject apparently intended. (2) D 3.102 (116). ādau rājêty adhīrâkși pārthivah ko'pi gīyate | sanâtanaś ca naîvâsou (3) rājā nâpi sanātanah (Daņdin; answer: the tree rājātana, i.e., rāiā accompanied by [sa] absence of [na] non- [a] posterity [tana]. This tree is neither a king nor eternal [sanātanah]: "There is a certain earth-lord, called a 'king' at first, who is eternal; but he is neither a king nor eternal"). (4) "What word is that to which if you add a syllable you will make it shorter?" (Robert Merry; answer: the word "short"). (5) In effect, the conundrum is here the obfuscation of a name. It differs from parusa prahelikā in that the statement of the conundrum conceals the name itself, which is then put in quotes, rather than simply concealing another possible interpretation (usually etymological) of the name. nibhrta, 'hidden': (1) a conundrum whose solution is hidden by a sequence of descriptive qualifications, each of which involves a parallelism of attribute between the hidden subject and the apparent meaning of the statement. (2) D 3.102 (117). (3) hrtadravyam naram tyaktvā dhanavantam vrajanti kāh / nānābhangisamākrsțalokā veśyā na durdharāhı (Daņdin; here the prostitutes are to be taken as "rivers", the penniless man as "mined mountains", the rich man as "the sea" (as the repository of jewels), the coquetry as "ripples", and so on. Instead of reading "what prostitute is not difficult to support?" the phrase now reads "what rivers are not to be bathed in [veśyā]?" The translation is: "Which prostitutes are not difficult to put up with, to whom men are attracted by various artifices and who abandon a man when they have taken his money in favor of one who has money?"). (4) "What is that which, supposing its greatest breadth to be four inches, length nine inches, and depth three inches, con- tains a solid foot?" (Robert Merry; answer: a shoe). (5) This conundrum differs from samānarūpa only grammatically: the latter should involve a similarity of form (nominal comparability), the former a similarity of qualification (adjectival comparability). parihārikā, 'avoiding': (1) a conundrum whose solution is effected by interpreting a series of words, in secondary senses, in such a way that each succeeding one refers to the preceding secondary sense. (2) D 3.104 (120). (3) vijitâtmabhavadveșigurupādahato janah ! himâpahâmitradharair vyāptam vyomâbhinandati (Daņdin: "vinā [inst.] garudena jita indras tasyâtmabhavah putrah arjunas tasya

GLOSSARY dveşī satruhı karnas tasya guruh pitā sūryas tasya pādaih kiranaih hatah santapitalı" and "himâpaho vahnih tasyâmitrāni jalāni teşām dharair 213 jaladharail meghaih vyāptam"; the example is written in what amounts to a code; translation would not clarify anything). (4) "When does a temperance lecturer say a grammar lesson?" (Robert Merry; answer: when he declines a drink). (5) "Drink" as a verb refers to "declines" in one sense, as a potion, in another. Strings such as the Sanskrit offers are unobtainable in English. This conun- drum is identical with a later dosa, neyârthatva. paruşa, 'hard': (1) a conundrum whose solution is given by applying an apparent or false etymology (which is not justified by any usage) to a word in the statement. (2) D 3.100 (113). (3) surāh surālaye svairam bhramanti daśanârcişā / majjanta iva mattās te saure sarasi samprati (Dandin; by taking surāḥ 'gods' as surā asti yeșām te surāh, the meaning 'drunkards' is obtained. This satisfies the paradox of the 'gods' being intoxicated: "The Gods wander freely in heaven showing their teeth, like drunken elephants in your celestial tank). (4) "What vessel is that which is always asking leave to move?" (Robert Merry; answer: can-I-ster). prakalpita, 'arranged': (1) a conundrum whose solution is given by under- standing the context in which the statement is made, and does not depend on any power of words to convey several meanings. (2) D 3.101 (115). (3) girā skhalantyā namreņa śirasā dīnayā drśā / tişthantam api sôtkampam vrddhe mām nânukampase (Daņdin; it would appear that this is addressed to an old lady, and therefore seems out of place; the solution is given by taking vrddhe as a name for Lakşmi: "You have no pity on me, aged lady, though I stand here agitated, with afflicted visage, bent head, and faltering voice"). (4) "'Now then: five crows were sitting in a tree .... A man came by with a gun. He shot one of them. How many were left in the tree?' 'That is no problem', said Sofie. 'There were four crows left, of course'. 'Yes, if they were such dumb crows as you .... Now, Hans, how many were left?' 'None', answered Hans, thoughtfully. 'Right- o !. There were none left .... But say, Hans, what do you think became of the others?' ' ... I suppose they flew away'" (O. E. Rol- vaag). (5) Vrddhā is not a pun, for Lakșmī bears that epithet in approximately the sense of 'the fully developed one'. The conundrum plays only on the legitimate connotations of the one word. Similarly, Rolvaag seems to be referring to a single crow, whereas in fact his question concerns the species crow and its social behavior. There

214 GLOSSARY is no pun in the striet sense (compare vancita prahelika); the puzzler simply phrases his statemient in such a way that the listener r. led to consider the wrong context for the statement. pramuita, 'distracted': (1) a conundrum in which the solution is hidden by a series of extremely difficult, archaic, or irregular words. (2) D 3.99 (111). (3) khâtayah kani kāle te sphātayah sphārhavalgavah candre sāk şād bhavant y atra vāyavo mama dhārinah (Dandin: "Young lady! [kani] your words [air-goers], full of sense, are sweet to my desire in this your lovely [moon-like] condition; my life [winds] is now secure"). (5) The best English equivalent for this puzzle is the rebus, a sequence of pictures and words whose puns, synonyms, and spatial arrangement are the key to comprehending the meaning. Compare this one from Robert Merry which involves only diagrams: "Mr. -wood being at the . of king of terrors, 10 mills for his quakers, and who, which and what. They odor for Dr. Juvenile Humanity, who to Dr. Hay preservers, and little devil behold scarlet his assistance; but B 4 he arrived, the not legally good changed color, taker and the was ct. for" (Mr. Dashwood, being at the point of death, sent for his friends and relatives. They sent for Dr. Childs, who inclosed a few lines to Dr. Barnes and imp-lo-red his assistance. But before he arrived, the invalid died, and the undertaker was sent for). Böhtlingk expresses indignation at this sort of thing, calling it too difficult for man or beast: "Ein solches Rätsel kann nur ein Inder lösen and wenn er uns die Lösung mittheilt, fragen wir, wie konnte Dandin ein Rätsel aufgeben, das ein normal gebildetes Gehirn nie und nimmer lösen wird?" Such scholarly impatience is too often met with-a regrettable ethno-centrism. Puzzles of this sort are common in all literate cultures, especially among the un- or under- educated; but when an Indian is obscure, the Western critic deems it his specific vice and never considers the purpose of the obscurity or its nature. Would Bohtlingk condemn all light verse? Comparative studies, such as the present one, may in some measure dispel this unwarranted prejudice which, in its way, is quite flattering to the Sanskrit language, for it would seem to imply that anything written in Sanskrit is ipso facto serious and great. vañcita, 'deceived': (1) a conundrum whose solution is obtained by taking one of the words in the statement of the conundrum in a secondary,

GLOSSARY but not lexically unjustifiable, sense. (2) D 3.98 (109). (3) kubjām 215 āsevamānasya yathā te vardhate ratih / naîvam nirvišato nārīr amarastrīvidambinīh (Daņdin; kubjā ('humpback') should be taken in the secondary sense of kanyakubja (the city), that is, Kanauj): "Your passion grows as you consort with the hunchback woman; you have no use for women who outdo the wives of the Gods"). (4) "Who was the first that bore arms?" (Robert Merry; answer: Adam; arrived at by taking "arms" in the sense of "upper extremi- ties"). (5) In these two examples, the obvious context suggests that kubja be taken in the sense of 'female hunchback', and that "arms" be taken as "firearms". The writer uses this evident primary sense to deceive us. vyutkrānta, 'out of order': (1) a conundrum whose solution is obtained by rearranging the words of the statement. (2) D 3.99 (110). (3) daņde cumbati padminyā hamsah karkaśakanțake | mukham valgura- vam kurvams tundenângāni ghattayan (Daņdin; the proper sense is given by reading "karkaśakante dande padminyā angāni ghattayan valguravam kurvan hamsas tundena (padminyāh) mukham cumbati": "Rubbing the members of the lotus on its stalk rough with spines, the swan kisses the face [of the lotus] with his beak, making a soft murmuring"). (4) "Was he short taken" (James Joyce; an example of the classical "anastrophe"). (5) This is a sort of anagram of words instead of letters. samkīrņa, 'mixed': (1) a riddle containing more than one of the riddles listed under prahelikā. (2) D 3.105 (123-24). (3) sahayā sagajā senā sabhațêyam na cej jitā | amātriko'yam mūdhaḥ syād akșarajñaś ca nah sutah (Daņdin; a mixture of nāmântarita (sahayā and sagajā taken as 'consisting of ha, ya, ga, and ja') and vañcita (senā in the unusual sense of sa-ina); but would this not more properly be inter- preted as pramușita? "If that army, with horse, elephant, and foot not be conquered, then our son, though he know the ineffable [akșara], is indeed unlettered and stupid"; or, "If the alphabet. with its 'h, y, g, j, i, n, bh, t' be not learnt, then our son. though he know letters, is indeed ."). (4) "In what ship, and in what capacity, do young ladies like to engage?" (Robert Merry: answer: in court ship, as marry-ners; the first illustrates nāmântarita, the second, parușa). samkhyāta, 'counted': (1) a conundrum whose solution is given by interpreting qualities in another way than that intended by the statement. (2) D 3.101 (114). (3) nāsikyamadhyā paritaś caturvarņa-

vibhūşitā asti kācit purī yasyām aştavarņáhvayā nrpāh (Dandin: "There is a certain city, adorned with the four castes, in the midst of the Nasikyas, whose kings are called 'eight fold'"; or: "There is a certain city with a nasal in the middle, surrounded by four phonemes, whose kings bear a title of eight phonemes"; the answer is Kānci, capital of the Pallavah). (4) "What has four eyes and can't see?" (Trad .; answer: Mississippi). samāgata, 'conjoined': (1) a conundrum whose solution is already contained in the statement of the conundrum and is revealed by a different reading of the constituent words (samdhi). (2) D 3.98 (108). (3) na mayā gorasâbhijñam cetaḥ kasmāt prakupyasi | asthānaruditair ebhir alam ālokitêkșane (Daņdin; the answer is obtained by reading na me āgo-rasâbhijñam, etc .: "I did not mark my mind with cow milk; why are you angry? Quit these inappropriate sobbings, my love of sidelong glances"; or: "My mind is not distinguished by a taste for sin ... "). (4) "What is that which is invisible, but never out of sight?" (Robert Merry; answer: by reading "in visible", the letters "j" or "s"). samānarūpa, 'of the same appearance': (1) a conundrum whose solution is given by interpreting several related words of the statement in appropriate secondary meanings. (2) D 3.100 (112). (3) atrôdyāne mayā drsțā vallarī pañcapallavā / pallave pallave tāmrā yasyām kusumamañjarī (Daņdin; the "creeper" is her arm, the "shoots" her fingers, and the "bud" her red fingernail: "I saw in the garden a creeper with five shoots, and on each shoot a scarlet flower bud"). (4) "Why is it profitable to keep fowl?" (Robert Merry; answer: for every grain they give a peck). (5) This type appears to be the same as vañcita, but with the added qualification that more than one word be taken in a secondary sense. The relation of the secondary senses then suggests the solution to the conundrum; as neither "grain" nor "peck" by itself is sufficient to enlighten the reader, their conjunc- tion alone suggests the secondary sense (relating to grain as a col- lective) in which they both are to be taken. samānaśabda, 'composed of the same words': (1) a conundrum whose solution is effected by substituting synonyms for identifiable mor- phemic elements of less than word length. (2) D 3.103 (118). (3) jitaprakrsțakeśâkhyo yas tavâbhūmisâhvayaḥ | sa mām adya prabhū- tôtkam karoti kalabhāșiņi (Dandin; the girl's lips are meant. Consider 'having the same name' (sâhvayah) as 'non-earth' (a-bhūmi). A synonym for bhūmi is dhara: substituting, we get a-dhara, which (not

GLOSSARY s0 morphemicized) is a common word for the lips (adhar-a is the 217

lower one): likewise, prakrstakeśa is to be taken as pra-vāla by the same process of substitution, and this is the name of a creeper: "surpassing the pravala creeper": "Something of yours, famed for having conquered excellent hair, and with the name of non-earth, now makes me awaken to desire"). (4) "Father plugs an abbreviation" (Robert Merry; this is so decoded as the name of a river: the "Pa- taps-co" in Baltimore). (5) Compare pramusita, where the etymologi- cal analyses are in theory legitimate (though our English example shows some of this type). sammūdha, 'confused': (1) a conundrum whose solution consists in prop- erly interpreting the intentional force of the statement, that is, in grasping fully its syntactical relationships. (2) D 3.103 (119). (3) śayanīye parāvrtya śayitau kāminau krudhā | tathaîva śayitau rāgāt svairam mukham acumbatām (Dandin; by taking tathaîva not as "in the same position" but as "turning over again", the statement is no longer paradoxical: "In the bed the two recumbent lovers turn over from anger; likewise recumbent they kiss each other's faces in passion"). (4) "How can five persons divide five eggs, so that each man shall receive one, and still one remain in the dish?" (Robert Merry; answer: one takes the dish with the egg). (5) This type is similar to prakalpita, but here the context in which the phrase is spoken does not provide the key; rather the reverse: the ability of the phrase to correctly represent the occurrence is at issue. One egg can remain in the dish when it is realized that this does not preclude the dish itself being appropriated. As in the Sanskrit example, the confusing word is "still", for the mind is, as it were, compelled to conceive of an entirely unchanged first condition: that the egg not only remains in the dish, but that the dish is still on the table. The mind is thus carried beyond the actual descriptive content of the phrase, where only one aspect of the original condition is unchanged, and falls into the syntactical blunder of overextending the prodosis in the apodosis. As in prakalpita, however, no power of words to convey two meanings is here employed.

preyas preyas (I), 'more agreeable': (1) the expression of affection in an extra- ordinary way. (2) B 3.5, D 2.275 (276). (3) adya yā mama govinda jātā tvayi grhâgate | kālenaîşā bhavet prītis tavaîvâgamanāt punah (Bhãmaha, Dandin; spoken by Vidura: "The joy which I feel at

your coming today, Govinda, will return only when you come again!"). (4) "Dear Mrs. A., / Hooray, hooray, / At last you are deflowered. / On this as every other day / I love you. - Noel Coward" (quoted by Wolcott Gibbs: telegram sent to Gertrude Lawrence on the occasion of her marriage). (5) See rasavat, ūrjasvi. preyas (I1): (1) that quality of a work of art by which descriptive situations clicit and sustain in every way appropriately the basic mood (rasa) of the work. (2) U 4.2. (3) iyam ca sutavāllabhyān nirvišeșā sprhāvatī/ ullāpayitum arabdhā krtvêmam kroda ātmanab (Udbhaļa; Pārvati is fondling a young fawn: "Paying no heed to the difference and made loving through tenderness for all offspring, she began to hum and took it to her breast"). (4) "There will be a rusty gun on the wall, sweetheart, / The rille grooves curling with flakes of rust. / A spider will make a silver string next in the darkest, warmest corner of it. / The trigger and the range-finder, they too will be rusty. / And no hands will polish the gun, and it will hang on the wall. / Fore- finger and thumbs will point absently and casually toward it. / It will be spoken among half-forgotten, wished-to-be-forgotten things. / They will tell the spider: Go on, you're doing good work" (Carl Sandburg: the title of this poem is "A. E. F." and its mood is śānta). (5) Udbhata incorporates several elements of the rasa theory into the traditional corpus of alamkāra. His pretext is the figure rasavat, which from the time of Bhamaha could be determined in any passage in which a rasa was evident or pre-eminent. The two figures closely allied to rasavat, namely ūrjasvi and preyas, which originally meant only 'arrogance' and 'compliance' and so contrasted with rasavat (as reposing upon the ego and not upon bhāvas common to all and especially the audience), are reinterpreted as special cases of rasavat: excessive demonstration of any rasa, and the present adapta- tion of situation and mood. We need not go into the precise and sometimes subtle analysis of situation (vibhāva, anubhāva, etc.); suffice it to say that the present figure can be explained (and is by later writers) as a composition whose situational elements are consist- ent and evocative of the proper final mood.

bhava bhāva (I), 'emotion': (1) a figure wherein the visible effect of an emotional state, together with its apparently unrelated cause, suggests the nature of that emotional state, which, in turn, explains the relevance of the cause. (2) R 7.38 (39). (3) grāmataruņam taruņyā navavañjulama-

GLOSSARY njarīsanāthakaram / paśyantyā bhavati muhur nitarām malinā mukhacchaya (Rudrata; the girl's troubled look [visible effect], 219 with its cause [the flower in the youth's hand], suggest her emotion [despair], since the cause is seen to be a secret signal that the rendez- vous must be postponed: "The girl's face is troubled as she looks upon the village youth with a new garland of hibiscus in his hand"). (4) " ... we had a small game, / And Ah Sin took a hand: / It was Euchre. The same / He did not understand; / But he smiled as he sat by the table, / With a smile that was child-like and bland. / ... / But the hands that were played / By that heathen Chinee, / And the points that he made, / Were quite frightful to see, - / Till at last he put down a right bower, / Which the same Nye had dealt unto me. / Then I looked up at Nye, / And he gazed upon me; / And he rose with a sign, / And said, 'Can this be? / We are ruined by Chinese cheap labour', / And he went for that heathen Chinee" (Bret Harte: Nye's sigh [visible effect], with its cause [right bower in wrong hand] suggest his emotion [thirst for revenge], since the cause is seen to be an indication of dishonest play). (5) This figure is testimony to the efforts of the stricter poeticians to include suggestion [dhvani] within the realm of figure as traditionally conceived. Bhāva is more limited in scope than rasavat alamkāra, aiming only at suggest- ing a specific, temporally limited emotion, rather than a mood [rasa] which would be a general characteristic of the work itself. It is a kind of periphrasis of the soul; but since we know from the Dhvanyāloka that the soul can only be obliquely referred to, bhāva is hard to differentiate from the notion of suggestion itself. The figure differs from aprastutapraśamsā both in that the intimated subject is there capable of representation, and in that the means of suggestion are similitudes of subject and explicit object. See paryāyôkta. bhāva (II): (1) a figure in which a literal truth is expressed for the purpose of conveying a hidden intention. (2) R 7.40 (41). (3) ekākinī yad abalā taruņī tathâham asmin grhe grhapatiś ca gato videśam / kim yācase tad iha vāsam iyam varākī śvaśrūr mamândhabadhirā nanu mūdha pāntha (Rudrața; the girl is inviting the wayfarer to bed: "I am alone and weak and innocent, and my husband has left this house for a far country; why do you ask refuge here? Don't you realize that my mother-in-law is about, deaf and blind? Stupid traveller!"). (4) "'Good night. If I get to talkin' and tossin', or what not, you'll understand you're to-' 'Yes, I'll wake you'. 'No

220 don't yu', for God's sakc" 'Not? 'Don't yu' touch me' Whurl I do?' 'Roll away quick to your side it don't last but a toute Oh, just don't let your arm or your laig touch me if I gu to jumpm' around. I'm dreamin' of Indians when I do that. And if saytheng touches me then, I'm liable to grab my knife nght in my derp- (Owen Wister: the Virginian wishes to avoid sharing s bed with the drummer). (5) Cf. parydyôkta, where the hidden intention a con- veyed by a remark with no apparent relevance. This hgure, bke vyajókti, conveys a type of irony, but does not involve spenking the opposite of one's real intent.

bhávika bhävika, 'expressive': (1) the coherence of the entire work in a clear and realistic unity. (2) B 3.53-54, D 2.364-66, U 6.6. (3) (4) No exampkes are offered by Bhamaha or Dandin. (5) This curious alamtdre, defined as a quality [guna] of the entire work, perhaps represents the extreme interpretation of the extent of the figure of speech. Here is clearer than usual the etymological sense of alamkdra: a 'making adequate' of the work of art. Bhavika is perceived in such aspects as the relevance of the various parts of the story to one another, the clarification of difficult contexts by an emphasis on a chain of events the suitability of the story to be represented in the form chosca. clarity of language, and so on. Dandin cxplains that this figure is a function of the poet's intention or desire (abhiprâya) and can be seca as a competent rendering of that unity in the work. For a speculative treatment of the figure, see the Introduction, pp. SOff.

Merintimat Mristimat, 'confused': (1) a figure in which one thing, usually the obint of comparison, is mistaken for another, usually the subject of compe- rison. (2) R 8.87 (88). M 200. (3) pâlayati tvayt vanm nhemt radhūmamālinių kakubhah | pašyanti důyante ghenaswdsettan hamsåb (Rudrața: "While you. O King, protect the carth, the moue- tain peaks are blackened by the smoke of so many secridces thet the geese, looking on, mistake them for the rain clouds and are snd") (4) " ... her eyes in heaven / Would through the airy region stvem so bright / That birds would sing and think it were eot nt" (Shakespeare). (5) No element of doubt is edmitted by thoee defnins this term; confusion is complete to the point of misenke in this the figure differs from samdaye and from ripate. As in praitee and

GLOSSARY pratyanika. Rudrata decrees that the two things confounded are 221 subject and object of an implied simile; Mammata is more imprecise. Of course, any terms capable of being so confused must be similar, so the question is really one of whether the aspect of similitude is the predominant feature or not-the motive, as it were. As Rudrata's example shows, this need not be the case, for there the alleged con- fusion of sacrificial smoke and clouds is subordinate to the flattery of the king, which is the obvious motive.

mata mata, 'deemed': (1) a figure in which one thing, conventionally well known and functioning as the subject of comparison, is taken to be another thing (the object) by a specially qualified observer. (2) R 8.69 (70-71). (3) madirāmadabharapāțalam alikulanīlâlakâlidha- mmillam | taruņīmukham iti yad idam kathayati lokah samasto'yam / manye'ham indur eşah sphuțam udaye'runarucih sthitaih paścāt | udayagirau chadmaparair niśātamobhir grhīta iva (Rudrața: "What the common herd deem to be a maiden's face, flushed with the first ecstasies of drink and framed in hair whose braids are dark as a cloud of bees, I know to be the ruddy-glowing risen moon, attacked by the black vapors of night hidden like secrets behind the eastern mountain"). (4) "The fairness of that lady that I see / Yond in the garden roamen to and fro, / Is cause of all my crying and my woe. / I noot wher she be woman or goddess; / But Venus is it, soothly, as I guess" (Chaucer). (5) Like the figures pratīpa, pratyanīka, and bhräntimat, this figure may be thought of as an implicit simile. on the principle that what is confusable is comparable. But the intention of the speaker may or may not be to compare: in the second example, the point of the rapprochement lies in the striking antithesis rather than in the similitude of the two things. As a good example of how those writers go astray who classify only by formal criteria. the Sanskrit example, which is a simile, reads also 'what looks like a girl's face to the unwashed is the moon to me'.

milita mīlita, 'fused': (1) a figure wherein two qualities or aspects of the same thing are said to be indistinguishable, one imposing itself on the other. (2) R 7.106 (107-108), M 197. (3) madirāmadabharapātalaka- polatalalocaneșu vadaneșu / kopo manasvinīnām na lakşyate kāmibhiḥ prabhavan (Rudrata: "The anger caused by teasing lovers cannot be

222 GLOSSARY seen on the faces of the spirited women whose cheeks arc flushed and whose eyes are bloodshot from devotion to spiritous liquor"). (4) "The hippopotamus's day / Is passed in sleep; at night he hunts; / God works in a mysterious way-/ The Church can sleep and feed at once" (T. S. Eliot). (5) If this figure is to be distinguished from ordinary atiśayôkti, the difference is most probably that the confusion of the two states through a property is to be understood literally and not as exaggeration. If so, the figure would be an asserted samsaya: it is not that the redness of her cheeks due to wine imposes itself figuratively on the redness of passion; one really can't tell from the redness alone what causes it. Mīlita differs from pihita in that the states confused are similar, not disparate.

mudrā mudrā, 'seal': (1) the ability of the poet to express his intentions clearly. (2) AP 342.26. (5) This is the second śabdâlamkāra of the Agni Purāņa and can also be called śayyā ('bed'). Presumably, what is meant is the old idea of sahitya as the unity of word and meaning, the expression of just the right idea in just the right way. In the Sarasvatīkanțhâbharana (2.125), the term is described as the able rendering of the context of an idea.

yathāsaņkhya yathāsamkhya, 'each to each': (1) a figure consisting of ordered sequences of terms, such as nouns and adjectives or subjects and objects of comparison, so arranged that item one of the first sequence matches item one of the second, item two of the first matches item two of the second, and so on. (2) B 2.89 (90), D 2.273 (274), V 4.3.17, U 3.2, AP 346.21, R 7.34 (36-37), M 164. (3) dhruvam te coritā tanvi smitêkșaņamukhadyutiḥ / snātum ambhaḥ pravișțāyāh kumu- dôtpalapankajaih (Daņdin: "As you entered the water to bathe, you certainly stole the beauty of your smile, eyes, and face from the red lotus, the blue lotus, and the white lotus"). (4) "The Piazza, with its three great attractions-the Palazzo Pubblico, the Collegiate Church, and the Caffe Garibaldi: the intellect, the soul, and the body-had never looked more charming" (E. M. Forster). (5) Dandin gives as alternate names for this figure krama and samkhyāna; Vāmana uses only krama. In the Agni Purāņa, yathāsamkhya is considered a guna rather than an alamkāra!

GLOSSARY yamaka, 'doubled' or 'restraint': (1) a figure in which a part of a verse, yamaka 223 specified either as to length or position or both, is repeated within the confines of the same verse, usually in such a way that the meaning of the two readings is different; word play (one of the meanings usually given to paronomasia). (2) NŚ 16.59-86, B 2.9-20 (definition, 17), D 3.1-72, V 4.1.1-7, AP 343.12-17, R 3.1-59, M 117-18. (3) na te dhīr dhīra bhogeșu ramaņīyesu samgatā | munīn api haranty ete ramanī yeșu samgatā (Bhamaha; second repeats fourth pāda: "O steadfast, your consciousness is not acquainted with delightful pleasures; they seduce even the sages with whom a beautiful woman is involved"). (4) From an anonymous poem, entitled "London, sad London", written during the early part of the Civil War: What wants thee, that thou art in this sad taking? A King. What made him first remove hence his residing? Siding. Did any here deny him satisfaction? Faction. Tell me whereon this strength of Faction lies. On lies. What did'st thou do when the King left Parliament? Lament. What terms would'st give to gain his company? Any. But how would'st serve him, with thy best endeavour? Ever. (5) Yamaka, one of the four original alamkāras (with upamā, rūpaka, and dīpaka), resembles a kind of paronomasia which goes by that somewhat inexact name "word play" in English. But because of its highly developed and formal character in Sanskrit, I have chosen a more appropriate rendering: "cadence". The etymology of the word can be traced either directly to the root yam- ('restrain') or to the derived form yama- (which can mean 'twin') from the same root. The application of either is obvious. Yamaka is a pun spelled out. Except for the earliest writer, Bharata, it is specifically stated that the meanings of the repeated portions must be different, and this involves either splitting the words differently (different morpheme boundaries) or taking each word as a double-entendre. Either way, a kind of pun is recalled (see ślesa): in the former, bhinnapada, in the

224 GLOSSARY latter, abhinnapada (different or identical words). The term "parono- masia" can be used loosely either of word play or of punning (double- entendre): when the two meanings of an identical sequence of words are obtained simultaneously, we have ślesa or pun; when thcy are obtained sequentially, we have yamaka or 'cadence'. In English poetry, the figure yamaka is generally restricted to light verse and doggerel: "But from her grave in Mary-bone / They've come and boned your Mary" (Thomas Hood; referring to body- snatchers; a parivrtti yamaka). Occasionally, it serves a satirical purpose, as in Joyce Cary's " ... watercolour ... slaughtercolour ... mortarcolour ... scortacolour ... tortacolour ... thoughtacolour ... ", a samastapāda yamaka based on the movie industry's overuse of the suffix "-color" ("technicolor, vistacolor", etc.). Nowadays, in America at any rate, yamakas are found most frequently in advertis- ing jargon, where it is apparently believed that they awaken curiosity: "For news you can depend on, depend on the Chicago Daily News". This is also a parivrtti yamaka. The figure yamaka is closely related to, and probably the progenitor of, several other figures, notably lāțânuprāsa and the various kinds of citra. The former involves repetition of contiguous words (like Bharata's yamakas), and the latter depend on principles of repetition other than the obvious linear one (such as zig-zag, palindrome, hop-scotch, etc.). The figure has been minutely subdivided, especially by Bharata, Dandin, and Rudrața, but all the distinctions relate only to the scope and place of the repetition in the verse (first pāda, beginning of first pāda, first half of first pāda, etc.). For this reason, I have not attempted to give an English example for each variety. English verses employ yamakas in no such consistent fashion (all would be classified as samuccaya yamakas by the precious), and most of the English yamakas are not found in verse, anyhow. All the infinite varieties reduce to the same uniform notion: repetition of word spans with different meanings. Udbhata alone of the early writers does not mention yamaka, though he devotes much thought to an elaborate classification of lāțânuprāsa. The other writers differ largely as to the degree to which the analysis is carried. Bhāmaha proposes only five types, Vāmana a half-dozen, and Mammata, though admitting the infinity of possible types, illustrates only five. Bharata, however, describes ten varieties, which is rather surprising considering that he finds only four types of upamā. Daņdin's elaborate classification differs from

GLOSSARY Rudrata's only in detail and in the fact that he proposes little special 225 terminology to cover the multitude of types. Rudrata gives a technical name for each variety, only a very few of which seem to have been traditional (see samudga yamaka). Both writers distinguish yamakas whose scope is the entire pāda (quarter verse) from those whose scope is only a fraction of a pāda (samastapādaja, pādaîkadeśaja). In the former category are included half śloka and whole śloka yamakas (samudgaka, mahāyamaka). (See also mukha, samdamśa, āvṛti, garbha, samdastaka, puccha, pankti, parivrtti, and yugmaka yamakas). The latter variety is, of course, infinite, and both writers classify as to where and how much of the pāda is involved (ādimad- hyânta yamakas, etc.) and also as to how many pādas exhibit yama- kas, and how many times the yamaka itself is repeated. Many of these varieties have no name. Rudrata, the supreme technician, divides "partial" yamaka into those whose scope is the half pāda and those whose scope is a quarter or a third of a pāda (see antâdika, vaktra, ādimadhya, etc.). All writers profess to be incapable of dealing with yamaka in its entirety. After the triumph of the dhvani theory, yamaka comes to be con- sidered the type par excellence of citrakāvya, the lowest of the three varieties of poetry, which embodies nothing of poetic value and displays mere verbal virtuosity. (See citra.) Yamaka and citra are, however, distinguished from one another very sharply by the tradition. Cf. the consistent yamakas of Raghuvamśa 9th canto. akşara, 'syllable': (1) a type of yamaka in which the repeated elements are syllables which occur either once, at the beginning of each pāda, or in such a way that the entire pāda or verse is composed of doubles. (2) V 4.1.2. (3) nānākāreņa kāntābhrūr ārādhitamanobhuvā / viviktena vilāsena tatakșa hrdayam nrņām (Vāmana: "The eyebrows of a lovely girl strike into the heart of a man with varied coquetry, each of their shapes delightful to the Love God"). vividhadhavavanā nãgagargharghanānāvivitatagagnānāmamajjajjanānā (Vāmana; un- clear in meaning). (5) Most yamakas are repetitions defined in pādas or parts of pādas, rather than in single syllables, but compare pādâdi yamaka, where only the location of the repetition is specified. anta, 'end': (1) see pādânta. antâdika, 'final-initial': (1) a type of yamaka in which the repeated elements are the last half of the first pāda and the first half of the second. (2) R 3.23 (24). (3) nārīņām alasam nâbhi lasannābhi kadambakam / paramâstram anagasya kasya no ramayen manah (Rudrața: "Whose

226 GLOSSARY heart would not be delighted by the army of young women, langorous, trepid, with gay navels-indeed the great weapon of the Love God?"). (4) The form is: xA Ax xx xx. (5) Cf. vyasta, samasta. ardhaparivrtti. 'alternation by halves': (1) a type of yamaka consisting of a samasta plus an ādyanta yamaka. (2) R 3.34 (35). (3) sasāra sūkam darpeņa kamdarpeņa sasārasā / śarannavānā bibhrāņā nâvi- bhrāņā saram navā (Rudrața: "The autumn now appears, replete with ducks and the proud God of Love, her new cart bearing grass to the accompaniment of the chirpings of birds"). (4) The form is: AB BA CD DC. (5) The samasta yamaka accounts for the internal repetition (BB, DD), the ādyanta for the enveloping repetition (A .. A C .. C). avyapeta, 'uninterrupted': (1) a type of yamaka in which the repeated sequences are contiguous. (2) D 3.1 (4-18), AP 343.12. (3) ramaņī ramaņīyā me pāțalâpāțalâmśukā | vāruņîvâruņībhūtasaurabhā saura- bhâspadam (Dandin: "My beloved is lovely as the western sky illuminated by the setting sun, her gown flecked with pink and sweet smelling"). (4) The form, for this example only, is: AAx BBx CCx DDx. (5) Cf. vyapeta yamaka. ādi, 'beginning': (1) see pādâdi. ādimadhya, 'initial-mid': (1) a type of yamaka in which the repeated elements are the first and second third of each pāda. (2) R 3.52 (53). (3) sa raņe saraņena nrpo balitāvalitârijanaḥ | padam āpa damāt svamater ucitam rucitam ca nijam (Rudrața: "The King, whose enemies were enveloped by his strength, attained that station in battle by means of his vehicle and through his self-restraint which was appropriate in his own view and pleasant"). (4) The form is: AAx BBx CCx DDx. (5) Compare ādyanta, type (b), and madhyânta. Only Rudrața divides the pāda into thirds for purposes of defining the scope of yamaka. adyanta, 'initial-final': (1) a type of yamaka in which the repeated elements are: (a) the first half of the preceding pāda and the last half of the following pāda, (b) the first and last third of each pāda, (c) the first and last quarter of each pāda, or (d) the first and last half of each pāda. (2) R 3.32, 3.50 (54), 3.44 (46), M 118 (365). (3) dīnā dūnavişādînā šarâpāditabhīśarā | senā tena parāse nā rane pumjīvitêrane (Rudrata; type (c) is illustrated: "Oh man, the army, afflicted, whose leader was desperate and in whom the wound of fear was produced by showers of arrows, was defeated by him in a battle which excited the lives of men"). (4) The form is: AxxA

GLOSSARY BzzB CxxC DxxD. (5) Type (a) is not illustrated; it is the inverse of 227

antâdika. amredita, 'reiterated': (1) a type of yamaka in which the word at the end of each pāda is repeated. (2) NŚ 16.79 (80), AP 343.16. (3) vijrmbhi- tam nihśvasitam muhur muhuh katham vidheyasmaranam pade pade / yathā ca te dhyānam idam punah punar dhruvam gatâte rajanī vinā vinā (Bharata: "Again and again yawning and sighing; why is your remembrance fixed on these several situations? Just as your medita- tion returns again and again, so indeed the night passes without her, without her"). (5) This yamaka of Bharata is peculiar in referring to the repetition of words; in fact, the later definitions of yamaka specify that the two readings obtained by repetition must have a different word base, and that the scope of the repetition is usually the pāda or a fractional part thereof (that is, must be entirely independent of the semantic content of the utterance). Bharata is less strict on this point; āmredita yamaka resembles lāțânuprāsa, except that the repetition of words is confined to the end of the pāda. āvalī, 'necklace': (1) a type of yamaka showing varied types of repetition within the confines of single pādas. (2) B 2.9 (14). (3) sitâsitâkşīm supayodharâdharām susammadām vyaktamadām lalāmadām | ghanā- ghanā nīlaghanā ghanâlakām priyām imām utsukayanti yanti ca (Bhämaha: "The great dark heaps of clouds come to and cause to pine my love, thick locked, with eyes both light and dark, her body with such lovely breasts and lips, by whom passion is manifested, giver of ornament"). (4) The form, for this example only, is: AAxxBB xCxCxC DDxDDx xxxxEE. (5) This type of random repetition shows the close relation between alliteration (anuprāsa) and cadence (yamaka). Later writers tend to extend the scope of the repetition to larger parts of the verse and thus lessen the possibility of such free play. āvrti, 'covering': (1) a type of yamaka in which the repeated elements are the first and fourth pādas. (2) R 3.3 (6), M 118C. (3) mudâratāți samarâjirâjitah pravrddhatejāḥ prathamo dhanuşmatām ! bhavān bibhartîha nagaś ca medinīm udāratādīsamarājirajitah (Rudrața: text and commentator agree in taking "beautiful ... " as a nominative masculine, agreeing with "mountain", though the sense would seem better served by interpreting it as an accusative feminine, with "earth". No metrical change would be occasioned. "With pleasure crushing your enemies, unconquered on the field of battle, of mature glory, first among archers, you, O King, support here the world as

228 GLOSSARY does the Himalaya, beautiful with even rows of lofty tādi trees"). (4) The form is: A x x A. (5) Compare mukha, garbha, etc. ekântarapāda. pādas with one (pāda) intervening': (1) a type of yamaka in which the repeated elements are corresponding parts of non- adjacent pādas. (2) V 4.1.2. (3) udvejayati bhūtāni yasya rājñah kušūsanam | simhâsanaviyuktasya tasya kşipram kušâsanam (Vāmana: "The evil government of a king who afflicts mankind will quickly be transformed into residence among the kuśa grass for him, when his lion-throne is lost"). (4) The form is: xx xA xx xA. (5) The scope of the repetition is not specified; the example shows half pādas. Naturally, the repetition could occur at the beginning or in the middle of non-adjacent pādas as well. kāncī, 'Conjeeveram': (1) a type of yamaka in which the repeated elements are located severally at the beginning and end of each pāda, or, in the manner of a combination of madhya yamaka and ādyanta yamaka, are the first and last quarters and second and third quarters of each pāda. (2) NŚ 16.66 (67), AP 343.15, R 3.44 (47). (3) yāmam yāmam candravatīnām dravatīnām / vyaktâvyaktā sārajanīnām rajanīnām (Bharata: "The going, going of the golden rivers; appear and dis- appear the nights, like butter"). yā mānītânītâyāmā lokâdhîrā dhīrâ- lokā | sênâsannâsannā senā sâram hatvâha tvā sāram (Rudrața: "Your army, supported by haughty women, greatly expanded, expelling the sufferings of mankind, terrible to see, is ready and near. Having felled the enemy, your army thus reports to you"). (4) The forms are: AAxBB CCxDD, etc. and ABBA CDDC EFFE GHHG. (5) See madhya and ādyanta yamaka. The name kāñcī probably refers to the city, but may be intended only in the sense of 'girdle'. garbha, 'womb': (1) a type of yamaka in which the repeated elements are the second and third pādas of the verse. (2) R 3.7 (8). (3) yo rājyam āsādya bhavaty acintah samud ratârambharatah sadaiva / samudra- tāram bharata sa daivapramāņam ārabhya payasy udtâse (Rudrața: "The king who, having attained empire, becomes careless, indulgent. and spends his time in making love is, as it were, attempting to cross the ocean by swimming and, by the will of fate, finds himself abandoned in the water"). (4) The form is: x A A x. (5) Cf. āvțti yamaka. cakraka, 'circular': (1) a type of yamaka in which the last half of each pāda is repeated by the first half of the following påda, and in which the first half of the first pāda is repeated by the last half of the fourth

GLOSSARY pāda. (2) R 3.30 (31). (3) sabhājanam samānīya sa mānī yah sphuțann 229 api | sphutam na pihitam cakre hitam cakre sabhājanam (Rudrața: "He is haughty who, though brilliant, gathering his courtiers, does good in the kingdom, not obviously, but covertly, evidencing pleas- ure"). (4) The form is: AB BC CD DA. (5) In effect, this is a com- bination of all the kinds of antâdika yamaka listed in R 3.23-30. Cf. vamśa, samasta, vyasta, madhya. cakravāla, 'circle': (1) a type of yamaka in which the last part of each pāda is the same as the first part of the following pāda. (2) NŚ 16.72 (74), AP 343.16 (alternate reading). (3) śailās [talais; alternate reading], tathā śatrubhir āhatāhatā | hatāś ca bhūyas tv anupumkhagaih khagaih | khagaiś ca sarvair yudhi samcitāś citāś | citâdhirūdhā nihatās talais talaih (Bharata: "Struck again and again by the enemy arrows, true to their shafts, collected and heaped up by birds who grasp at the death of the stricken"). (5) The example shows a triple yamaka in each pāda, but the definition does not make this obligatory. If the reading talaih is accepted, then this type is the same as the preceding except that the scope of the repetition is not specified (the interior of each pāda is not involved). caturvyavasita, 'determined by fours': (1) a type of yamaka in which all four pādas are the same. (2) NŚ 16.81 (82), AP 343.17. (3) vāraņānām ayam eva kālo vāraņânāmayam eva kālaḥ / vāraņānām ayam eva kālo vā raņānām ayam eva kālaḥ (Bharata: "This is the season of vāraņa flowers, the time when elephants have no diseases, when death stalks his enemies and suggestion is of battles"). (5) Note that minor variations in samdhi do not vitiate the identity of the yamaka. Each pāda, though phonemically identical, has a different morphemic analysis; the first and third, which appear the same morphemically, are assigned to different homonyms: vārana (a 'flower' and 'enemy'). This is the same as pankti yamaka. dușkara, 'difficult': (1) a yamaka involving greater limitation on occurrence and scope than is usual. (2) D 3.38ff. (3) (4) See, for example, mahāyamaka. See also under citra. pañkti, 'series': (1) same as caturvyavasita. (2) R 3.10 (12). parivrtti, 'exchange': (1) a type of yamaka in which the fourth pāda repeats the first and the third repeats the second; this is, in effect, a combination of āvrti and garbha yamakas. (2) R 3.13 (14). (3) mudā ratâsau ramaņī yatā yam smarasy ado'lam kurutena vodhā / smarasyado'lamkurute'navodhām udāratâsau ramaņiyatāyām (Rudra- ta: "The girl whom you have married is devoted to you from pleasure

GLOSSARY 230 fand not from hope of gain]. You think on this too much; it is appropriate in matters of the heart: excess of love always ornaments a proud woman"). (4) The form is: A B B A. (5) Cf. yugmaka pādamadhya. 'middle of the pāda': (1) a type of yamaka in which a vamaka. repetition occurs in the middle of one, or cach päda, or in which one or both of the middle (second and third) pādas contains the repetition. (2) B 2.9 (12), D 3.1-2 (5-6), V 4.1.2, AP 343.15. (3) mcghanādena hamsānām madano madanodinā | nunnamānam manah strīņām saha ratyā vigāhate (Daņdin; here the yamaka occurs in the second pāda: "Love, together with his wife, Passion, enters the anger-freed minds of women by means of the thunder, which also drives out the pride of the swans"). (5) The scope of the repetition is not prescribed: compare madhya yamaka of Rudrața. Bharata recognizes pādâdi and pādânta yamakas, but not pādamadhya. The equivocation in the meaning of "middle" is shared by the two other terms. Rudrata attempts to specify the scope of all repetitions, but the earlier writers are sometimes satisfied by indicating merely the place of the repetition -- here the middle of the pāda, or the middle pāda. pādayor ādimadhyântayamakāni, 'yamakas involving two pādas': (1) same as pādâdi, pādamadhya, and pādânta yamakas, except that the scope of the repetition is two adjacent pādas, rather than one pāda alone. (2) V 4.1.2. (3) bhramara drumapuşpāņi bhrama ratyai piban madhu| kā kundakusume prītiḥ kākum dattvā virauși kim (Vāmana: "Bee, go roam from flower to tree-flower for pleasure drinking honey; what joy is there in the jasmine bloom? Why do you make such a noise, murmuring?"). (4) The form, for this example only, is: Axxx Axxx Bxxx Bxxx. (5) For Vāmana, a pādâdi yamaka has only the form AAxx BBxx CCxx DDxx; hence this variation in which the word pāda appears in the dual. pādasamudgaka, 'pāda-casket': (1) a type of yamaka where the repeated elements are the first and second half of the same pāda. (2) R 3.36 (37-39). (3) rasāsāra rasâsāra vidā raņavidāraņa | bhavatârambhava- târam mahīyatam ahīyata (Rudrața: "O essence of the earth! O shower of moods! O exterminator in battle! By you, wise and enter- prising, the enemy, sent to earth, has been slain"). (4) The form is: AA BB CC DD. (5) Rudrața gives two other examples, in which not all pādas show repetition (AA xx BB xx; xx AA xx BB). Compare samudga yamaka.

GLOSSARY

pādâdi (1), 'beginning of the pāda': (1) a type of yamaka in which each 231

pāda begins with the same word. (2) NS 16.77 (78). (3) vişnuh srjati bhūtāni vişnuh samharate prajāh | vişnuh prasūte trailokyam vişnur lokâdhidaivatam (Bharata: "Vișņu creates all things; Vișņu destroys men; Visnu evolves the three worlds; Visnu is the all high God!"). (5) Like āmredita yamaka, this early type violates the rule requiring a different morphemiccontent for each repetition. Compare lāțânuprāsa. pādâdi (II): (1) a type of yamaka in which a repetition occurs at the beginning of one, or of each pāda, or in which the first pāda contains the repetition. (2) B 2.9-10 (11), D 3.1-2 (4), V 4.1.2, AP 343.15. (3) mânena mānena sakhi praņayo bhūt priye jane | khaņditā kanțham āślişya tam eva kuru satrapam (Dandin: "Friend, don't show affection for your lover by such anger! Though betrayed, you must embrace him and make him ashamed"). (5) See pādamadhya. Vāmana's example shows each pāda beginning with a different yamaka (note the difference from pädâdi I, where the same word, not repeated, begins each pāda). The scope of the repetition is not specified. pādânta (I), 'end of the pāda': (1) a type of yamaka in which each pāda ends with the same word. (2) NŚ 16.63 (64). (3) dinakşayāt samhrta- raśmimandalam divîva lagnam tapanīyamaņdalam / vibhāti tāmram divi sūryamaņdalam yathā taruņyāh stanabhāramandalam (Bharata: "The burnished sun seems fixed in the western heaven at the close of day, its orb of rays restrained. The copper-colored sun disk shines in the sky like a young girl's heavy breast"). (5) See pādâdi I. pādânta (II): (1) a type of yamaka in which a repetition occurs at the end of one, or of each päda, or in which the first pāda contains the repeti- tion. (2) B 2.9-10 (12), D 3.1-2 (7), V 4.1.2, AP 343.15. (3) aranyam kaiścid ākrāntam kaiścit sadma divâukasām / padātirathanāgâ- śvarahitair ahitais tava (Dandin: "Some of your enemies, shorn of soldiers, chariots, elephant, and horse, have retired to the forest, some to the seats of the Gods"). (5) See pādâdi II. pādâbhyāsa, 'repetition of the pāda': (1) same as samastapāda yamaka. (2) B 2.9 (13), D 3.53 (57-66). (5) Bhāmaha gives an example illustrating Rudrața's samdaștaka yamaka (second and fourth pādas identical). Dandin illustrates pāda repetition in all possible loci, including triple repetitions. pādâikadeśaja, 'relating to parts (only) of pādas': (1) a generic term for all those kinds of repetition whose scope is less than pāda length. (2) R 3.2 (20-55). (5) The opposed term is samastapādaja ('referring to the entire păda'). This division classifies all yamakas in the most

GLOSSARY 232 general way, and each alternative is itself minutely subdivided. Partial pāda yamakas are either half pāda, one-third pāda, or one- fourth pāda in length. For examples, see antâdika, ādimadhya, vaktra yamakas. puccha, 'tail': (1) a type of yamaka where the repeated elements are the third and fourth pādas. (2) R 3.10 (11). (3) uttungamātangakulâkule yo vyajesta śatrūn samare sadâiva / sa sāram ānīya mahâri cakram sasāra mānī yamahâricakram (Rudrața: "The insolent King, slayer of Death, approached the circle of the enemy carrying his best wide- spoked discus; he triumphed in the battle over his enemies confused with herds of lofty elephants"). (4) The form is: x x A A. (5) For Dandin, this is an unnamed subtype of samudga yamaka (D 3.63). Cf. mukha yamaka. pratiloma. 'against the grain': (1) same as pratilomânuloma citra; palin- drome. (2) D 3.73 (74-77). (5) This is the last yamaka treated by Daņdin before going on to duskara (gomūtrika, etc.). On the differ- ence between citra and yamaka see citra. prahelikā, 'conundrum': (1) a bad yamaka. (2) B 2.19. (5) Bhāmaha probably is referring to repetitive verses in which the principle of repetition is not linear, as gomūtrika, muraja, turagapada, and the like. These are not 'conundrums' properly speaking, and the use of "prahelikā" is misleading. For a discussion of the difference between these extensions of yamaka, see citra, prahelikā. bhañga, 'analysis': (1) juncture. (2) V 4.1.3 (4-6). (5) A category in the analysis of yamaka; according to Vāmana, the yamaka is successful if the repeated elements of the verse show different word junctures for each reading. He gives three illustrations: (a) śrnkhalā, 'chain'; where the juncture occurs between different (successive) syllables for each reading, that is, not between the same syllable for both readings. For example: kali-kāma-dhug and kalikā-madhu. (b) parivartaka, 'exchange'; where the juncture varies only by one consonant, attach- ing that consonant by one reading to the preceding word and, by the other, to the following. For example: -dhug-arhitam and -madhu- garhitam. (c) cūrņa, 'mixture'; where the juncture by one reading breaks up a consonant cluster which is, by the other reading, within a word. For example: -unmukta-śukti-mīnām- and -unmukta-śuk- timīnām -. Yamakas which do not involve these charming junctures are, in fact, dull and uninspired in that both readings repose upon the same words, distinguishable only as puns. Nevertheless, certain puns also show junctures of this sort.

GLOSSARY madhya. 'middle': (1) a type of yamaka in which the repeated elements 233 are cither the last and first halves of the second and third pādas, or the sccond and third quarters of each pāda. (2) R 3.27 (28), 3.44 (45). (3) samastabhuvanavyāpiyaśasas tarasêhate | rasêha te priyam kartum prānair api mahīpate (Rudrața: "The earth desires quickly to do you favor, even with its life, O great King! Your glory penetrates the whole world"). (4) The form is: xx xA Ax xx. (5) The other madhya, whose scope is only the quarter pāda instead of the half pada, would have the form xAAx xBBx xCCx xDDx. The literal appropriateness of the name is revealed by a cursory examina- tion of the schemata. Compare pādamadhya yamaka, where the scope is not so exactly determined. madhyânta, 'mid-final': (1) a type of yamaka in which the repeated elements are the second and last thirds of each pāda. (2) R 3.52 (55). (3) asatām ahito yudhi sāratayā ratayā / sa tayôruruce ruruce paramêbha- vate bhavate (Rudrata: "O widely renowned, this person, in battle the nemesis of evil men, is pleased with you, possessor of fine ele- phants, because of your pithiness, and out of pleasure"). (4) The form is: xAA xBB xCC xDD. (5) The example has no yamaka in the first pāda. Compare ādimadhya and ādyanta yamakas. mahāyamaka (I), 'great yamaka': (1) a type of yamaka consisting of four identical pādas, each of which is itself a yamaka of half-pāda scope. (2) D 3.70 (71). (3) samānayāsa mānayā samānayâsamânayā | samānayā samānayā samāna yā samānayā (Dandin: "O unequalled! O uniformly active! With this universal arrogance, join me! O haughty! Arrogance which is beautiful and decisive!"). (4) The form is: AA AA AA AA. mahāyamaka (II): (1) same as ślokâbhyāsa yamaka. (2) R 3.16 (18-19). mālā (I), 'garland': (1) a type of yamaka in which each half pāda is itself a complete yamaka; it consists, that is, of repeated quarter pādas. (2) R 3.40 (43). (3) bhītâbhītā sannâsannā sênā senâgatyâgatyā / dhīrâdhīrâha tvā hatvā samtrāsam trayasvâyasvā (Rudrața: "O steadfast! O expeller of misery! The opposing army with its leader, drawn up in front and fearful, miserable and near, having advanced because it had no other choice, thus informs you: 'Having reduced us, please save us from fear!' Your coming is their only remaining hope"). (4) The form is: AABB CCDD EEFF GGHH. (5) In effect, this is a combination of vaktra and śikhā yamakas. mālā (II): (1) a type of yamaka in which one consonant is repeated, and the syllabic vowel is varied. (2) NŚ 16.83 (84). (3) hali balī hali

234 GLOSSARY

mālī khelī mālī salī jalī | khalo balo balo mālo musalī tv abhirakșatu (Bharata: the translation is approximate; "May Baladeva protect you: with a plow, strong, from Hala, garlanded, shaking, having a different garland, a sal flower, and water; cruel, strong, Bala, a forest"). mukha, 'face': (1) a type of yamaka in which the repeated elements are the first and second pādas. (2) R 3.3 (4). (3) cakram dahatâram cakranda hatâram | khadgena tavâjau rājann arinārī (Rudrața: "O King, the wives of your enemies, struck by your sword, quickly scourging the mass of your enemies in battle, moan"). (4) The form is: A A x x. (5) Compare puccha, samdamśa, āvrti, garbha, and samdastaka yamakas for other types of whole pāda repetitions. yugmaka, 'pair': (1) a type of yamaka in which the second pāda repeats the first, and the fourth the third. (2) R 3.13 (15), M 118 (361). (3) vinâyam eno nayatâ sukhâdinā vinā yamenônayatā sukhâdinā / mahājano'dīyata mānasād aram mahâjanodī yatamānasādaram (Rud- rața: "This noble person, suffering no blemish, alone, yet destroying the arrogance of his enemies, extirpating those who disrupt festivals, is quickly carried away by Death, defying the doctors, who has approached, hungry for life, the great reaper, consumer of happiness"). (4) The form is: A A B B. (5) In effect, this is a combina- tion of mukha and puccha yamakas. Compare parivrtti, which is a combination of garbha and āvrti (A B B A). vamśa, 'flute': (1) a type of yamaka in which the first half of each succeed- ing pāda repeats the second half of the preceding pāda. (2) R 3.27 (29). (3) grīşmeņa mahimânīto himānītoyaśobhitah / yaśo'bhitaḥ parvatasya parva tasya hi tan mahat (Rudrața: "The grandeur of the mountain, beautified by the melting snow, has been wrought by the summer heat, and its glory is all about, for its festival is of great extent"). (4) The form is: xA AB BC Cx. (5) This is a combination of madhya and samasta yamakas. A cakraka yamaka completes the circle by having the last half of the fourth pāda repeat the first half pāda. vaktra, 'mouth', 'origin': (1) a type of yamaka in which the first half pāda of each pāda is a yamaka consisting of repeated quarter pādas. (2) R 3.40 (41). (3) ghanāghanâbhinīlānām āsthām āsthāya śāśvatīm / calâcalâpi kamale līnâlīnām ihâvalī (Rudrața: "Moving and motion- less on the lotus adheres a wreath of bees, dark as the rain cloud, perpetually in place"). (4) The form is: AAxx BBxx CCxx DDxx. (5) Cf. śikhā and mālā yamakas.

GLOSSARY vikranta. 'overstepped": (1) a type of yamaka in which the second and 235 fourth padas are the same. (2) NS 16.70 (71), AP 343.16. (3) sa pürvam vārano bhūtva dvišrnga iva parvatah | abhavad dantavaikalyād risrnga iva parvatah (Bharata: "That elephant was formerly like a two-peaked mountain, but now, for the loss of his tusks, is like a peakless mountain"). (4) The form is: x A x A. (5) Bharata, as usual. allows entire words to be repeated. Compare samdastaka yamaka. vyapeta. 'interrupted': (1) a type of yamaka in which the repeated se- quences are not contiguous. (2) D 3.1, 19 (20-32), AP 343.12. (3) karo titāmro rāmāņām tantrītādanavibhramam / karoti sêrșye [or sêrşyamı (D. T. Tatacharya)] kānte ca śravaņôtpalatādanam (Daņdin: "The red-lac'd hands of the beloved coquettishly strum the lute; and when her lover makes her angry, they disarrange the flowers at her ear"). (4) The form, for this example only, is: Axx xxx AxX xxx. (5) Dandin gives thirteen examples without introducing special terminology. They illustrate various possible loci of the interrupted sequences and could be classified in the same way as uninterrupted vamakas (see pādâdi, pādamadhya, pādânta). vyapetâvyapeta, 'interrupted-uninterrupted': (1) a type of yamaka in which a contiguous repeated sequence is itself repeated in a non- contiguous place. (2) D 3.33 (34-36), AP 343.12. (3) kâlam kālam anālakșyatāratārakam īkșitum / tāratâramyarasitam kālam kālama- hāghanam (Dandin: "What woman is able to look at the great black rain clouds in their season, hiding the myriads of stars and, death- like and fearsome, resounding in the great distances"). (5) Dandin gives two other examples illustrating different loci for this non- contiguous repetition of contiguous repetitions. See vyapeta yamaka. vyasta, 'separated': (1) a type of yamaka in which the repeated elements are the second half of the third pada and the first half of the fourth pāda. (2) R 3.23 (25). (3) paśyanti pathikāh kōmaśikhidhūmaśikhām iva | imām padmâlayâlīnām layâlīnām mahâvalīm (Rudrața: "The passersby observe a great column of bees, thickly entwined, above the lotus pond, like a tongue of smoke from the fire of Love"). (4) The form is: xx xx xA Ax. (5) The name probably signifies only that this is half of a samasta (complete) yamaka. The same pattern in the first two pādas is called antâdika. śikhā, 'tuft': (1) a type of yamaka in which the last half of each pāda is a yamaka consisting of repeated quarter pādas. (2) R 3.40 (42). (3) yāsām citte māno'māno nārīr bhūyo'ram tā rantā / sārapremā sannc mnnā jayetaîvânantā nantā (Rudrața: "May the virtuous man,

236 GLOSSARY the able lover. obedient and always agreeable, be quickly victorious over women. who are so numerous and so near and in whose minds is immeasurable pride"). (4) The form is: xxAA xxBB xxCC xxDD. (5) See vaktra and mālā yamakas. ślokabhyasa. 'repetition of the śloka': (1) a ramaka in which the repeated clement is the entire verse. (2) D 3.67 (68-69). (3) vināyakena bhavatā vrttopacitabūhuna svamitrôddhārinâbhītā prthvîyam atulâśritā | vināyakena bhavatā vrttôpacitabāhunā / svamitrôddhâriņā bhītā prthvī vamatulaśrita (Dandin: "This incomparable earth, confident, is devoted to you who remove obstacles. whose powerful arms are round. and who upholds his friends. This earth, like death, has taken refuge and is afraid of your present enemy who is leaderless, whose arms are engaged in laying waste, and who seized both property and friends"). (5) This is the same as mahāyamaka II. samdamsa. 'tongs : (1) a type of yamaka in which the repeated elements are the first and third pādas. (2) R 3.3 (5). (3) sannārībharanômāyam ārādhya vidhušekharam | sannârîbharaņo'māyas tatas tvam prthivīm jaya (Rudrata: "Having pleased Siva, moon on his brow, who resorts to Umã. the support of chaste women, you will certainly conquer the carth. for the elephants of your enemies have been lost, and you are free from disease"). (4) The form is: A x A x. (5) See samda- staka. samdașta (I). 'pinched': (1) a type of yamaka in which the word at the beginning of each pāda is repeated. (2) NŚ 16.75 (76), AP 343.16 (alternate reading). (3) paśya paśya raņam asya me guņān yena yena vaśagām karoti mām / yena yena hi mamâiti darśanam tena tena vaśagām karoti mām (Bharata; should undoubtedly be emended to ramanasya me: "Look, look at the virtues of my lover, by which he makes me do his will! For by them I have my view of him, and by them he makes me do his will!"). (5) This is the opposite of āmredita yamaka and is one of Bharata's peculiar yamakas which makes reference to units of meaning, rather than to sequences of phonemes of necessarily different meaning. samdașța (II): (1) a type of yamaka in which the first part of the succeed- ing păda repeats the last part of the preceding pāda for the three pāda junctures within the verse. (2) D 3.51 (52). (3) upôdharāgâpy abalā madena sā | madenasā manyurasena yojitā | na yojitâtmānam anangatāpitām / gatâpi tāpāya mamâsa nêyate (Dandin: "Though she had been made amorous with drink, the young woman would not join her love-sick self to me, as she was inspired by the very

GLOSSARY soul of fury because of my unfaithfulness: her leaving did not 237 make me suffer too much"). (4) The form. for this example only, is: xxA AxB BxC Cxx. (5) The scope of the repetition is unspecificd; otherwise, this vamaka resembles vamśa yamaka. samdaştaka. 'pincers': (1) same as vikrānta yamaka. (2) R 3.7 (9). (5) The earlier version of Bharata allows a certain amount of word repeti- tion: Rudrata, of course, proscribes it. Mammata offers an example (118, 364) which the commentator calls a samdastaka yamaka, but it shows repetition only of the last half of the second and fourth pādas (xx xA xx xA). The figure is mentioned in B 2.10, but is not illustrated. samasta. 'entire': (1) a yamaka in which the first half of the second and fourth padas repeat the last half of the first and third, respectively. (2) R 3.23 (26). (3) pusyan vilāsamı nārīņām sannârīņām kulak șayam / ā kalpam vasudhāsāra sudhâsāra jagaj jaya (Rudrața: "O essence of the earth! O shower of nectar! May you rule the earth till the end of the age, encouraging the coquetry of women and promoting the familial ruin of your crushed enemies!"). (4) The form is: xA Ax xB Bx. (5) In effect, this is a combination of antâdika and vyasta yamakas. samastapāda (I), 'all pādas': (1) same as pādânta I. (2) B 2.9 (15). samastapāda (II): (1) a type of yamaka in which the same repetition occurs at the end of each pāda. (2) V 4.1.2. (3) natônnatabhrūgati- baddhalāsyām vilokya tanvīm šaśipeśalâsyām / manah kim uttāmyasi cañcalâsyām kṛtī smarâjñā yadi pușkalā syām (Vāmana; first two pādas in third person, third in second, and fourth in first: "Looking at this slender girl whose face is gentle as the moon, where the dance is formed by the rising and the falling of her brow, why do you lose heart? The fickle authority of Love should be recognized in her. if I am any judge"). (4) The form, for this example only. is: xxxA xxxA xxxA xxxA. (5) This differs from samastapāda 1 only in not referring to words. It differs from pādânta (of Vāmana) in that there a different yamaka occurs at the end of each pāda (not the same word in all pādas). samastapādaja, 'referring to whole pādas': (1) a generic category for those yamakas whose scope is not less than an entire pāda. (2) R 3.2 (4-19). (5) Samastapādaja is opposed to pādâik adešaja yamaka; for illustrations, see mukha, pankti, samudgaka. mahāyamaka, etc. It is the same as pādâbhyāsa yamaka. samuccaya, 'collection': (1) a yamaka which is composed of several

238 GLOSSARY

randomly related vamakas. (2) M 118. (5) No example is given. samudga, 'round box or casket': (1) a type of yamaka in which the entire half śloka is repeated. (2) NŚ 16.68 (69), B 2.10, D 3.53 (54-56), AP 343.16 (the text reads samsarga), R 3.16 (17). (3) nanāma loko vidam ānavena mahī na câritramud āradhīram / na nâmalo'kovidamā- navênam ahīnacāritram udāradhīram (Rudrața: "The people, festive and not without spotlessness, honor with praises the wise man who does not delight at the salvation of his enemies but afflicts their consciences, and whose master is the arrogance of the stupid, whose behavior is integral, who is noble and steadfast"). (4) The form is: A/A. (5) Samudga shows to best advantage the requirement of all yamakas that the two readings involve different meanings and different word separations. In repeating the entire half śloka, this type demonstrates its close affinity with ślesa, where it would be expected that the two meanings emerge from a single reading. Samudga(ka) is the only type of yamaka (aside from the threefold distinction into ādi, madhya, and anta) whose name is consistently and universally used. Compare samdasta, also found in many texts, but varying widely in acceptation.

yāvadarthatā yāvadarthatā, 'correspondence to the needful': (1) absence of either superfluity or inadequacy in either word or content. (2) AP 345.6. (5) This is the fifth sabdârthâlamkāra of the Agni Purāna. See aucit ya and abhivyakti. In principle, there appears to be little differ- ence between this term and the śabdâlamkāra "mudrā"; however, there may be intended here simply a reference to prolixity and its contrary defect, terseness, and not to the adequacy of the word to the sense, as such.

yukti yukti, 'junction': (1) a figurative usage. (2) AP 342.29-30. (5) This is the fourth sabdârthâlamkāra; it is defined literally as the joining of a word and a meaning which appear to be unrelated one to the other ("ayuktayor iva mitho vācyavācakayor dvayoh | yojanāyai kalpamānā yuktir uktā manīşibhih"). The examples given in the Sarasvatīkanthâ- bharana (2.98 ff.) suggest that this term may signify, like Dandin's guņa "samādhi", the use of a word in a secondary sense-the function called in mīmāmsā "rūdhi” (cf. Mammața, chap. 2, kā. 18). Like Kuntaka's vakrôkti, the scope of this figuration is sixfold: relating

GLOSSARY to words, meanings, sentences, intentions, contexts, and entire 239 compositions.

rasavat rasavat, 'expressing a mood': (1) a figure in which is clearly expressed a mood or rasa-usually śrngāra, the amorous. (2) B 3.6, D 2.275 (280-92), U 4.3-4, M 66 (123C.). (3) mrtêti pretya samgantum yayā me maraņam matam | sâisâvantī mayā labdhā katham atrâiva janmani (Dandin; śrngāra rasa: "She whom I thought had gone beyond to death, who made me want to follow her, did it happen that I won her to me in this present life, my Vāsavadattā?"). (4) "Everywhere the vastness and terror of the immense night which is roused and stirred for a brief while by the day, but which returns, and will remain at last eternal, holding everything in its silence and its living gloom" (D. H. Lawrence; perhaps śānta rasa). (5) This figure be- comes crucial for the tenants of the dhvani theory, who want to es- tablish the autonomous expression of rasa vis-à-vis the figures of speech. Should there be a figure which itself is the expression of a rasa, the contrary would a fortiori be proven, and the rasa would be subordinated to the general notion of the figure. The outcome of the argument allows rasavat as the general term for those figures which contain a touch of rasa, but where the rasa is not the major end of the poet employing that figure. Rasa as the proper end is pure dhvani and not related to any figure (Dhvanyāloka 2.5).

rūpaka rūpaka, 'having the form of': (1) metaphorical identification. A figure in which the subject of comparison is identified with its object by a specific process of grammatical subordination. (2) NŚ 16.56-58, B 2.21-24, D 2.66-96, V 4.3.6, U 1.11-13, AP 344.22-23 (the definitions of both Bhāmaha and Dandin are repeated), R 8.38-56, M 139-45. (3) tadidvalayakakşyāņām balākāmālabhāriņām / payomucām dhvanir dhīro dunoti mama tām priyām (Bhāmaha; the lightning is charac- terized as a bracelet, the cranes as a garland: "The roaring of the great clouds, wearing a bracelet of lightning and a garland of cranes, frightens my love"). (4) "The flute of morning stilled in noon- / noon the implacable bassoon- / now twilight seeks the thrill of moon, / washed with a wild and thin / despair of violin" (e. e. cummings). (5) Rūpaka, one of the four original alamkāras, is considered by all writers to be a development of upamā and is, in

principle, a simile or comparison. While in upamā the two terms (subject and object of comparison, or upameya and upamāna) are literally compared, usually through the use of a comparative particle such as "like" or "as". in rūpaka that comparison is expressed as an identification of the two. No longer is her face "like" the moon, it is the moon. We use the word identification in the root sense: the "making identical" of what is different. By virtue of this identi- fication, the behavior and properties of the moon (or object) are ipso facto transferred to the face (or subject), and propositions which are literally true of the moon are now understood as applying, in the poet's intention, to the face exclusively (the moon is brilliant; her face-moon is brilliant). Of course, this transfer of property constitutes the aim of the poet who uses rūpaka, as well as allied figures, such as utpreksā; but the manner and the technique of the transfer (metaphor) differ for each of the figures. This formal differ- ence constitutes the distinctive, or essential feature of each figure and naturally preoccupies the writers on the subject. By "identification" is meant the specific characterization of one thing (the subject of comparison) as another (the object). Both terms must be mentioned specifically (thus distinguishing rūpaka from samāsôkti), and the relation between them must be immediate and substantial rather than through an aspect or a property, as in utprekșā or upamā. The proper grammatical form through which this substantial identification is expressed is ideally the compound noun, with the object of comparison following (as, face-moon; see samastarūpaka). Since the object alone is grammatically free, the syntax of the sentence relates to it exclusively; from this total subordination-morphological as well as syntactic-of the subject (face-), the suggestion of identification or fusion derives. It is not the compounding of the terms which defines the rūpaka, but rather specifically the subordination of the subject to the morphologically free object. This is shown most directly by the interpretation of inverse compounds (where the object is grammatically subordinated to the subject, as moon-face) as upamās (similes) showing ellipsis of the common property (tertium). (Cf. R 8.21ff. For exceptions. see samasta rūpaka.) Two other varieties of grammatical subordina- tion must be allowed as rūpakas, since they accomplish the same expressive end : simple predication and the use of a subjective genetive. Both Dandin and Rudrata define an uncompounded (asamastarū- paka) metaphor where the object is simply predicated of the subject

GLOSSARY (her face is the moon). The identification is so evident here as not 241 to require explanation; the difference between upamā and rūpaka is also best perceived in this case. The grammatical subordination is that of the imposition of the object on the subject. A subjective genitive is often employed in English to express an identification; although our language permits compounding, the use of this device is restricted, compared to Sanskrit at any rate, and prepositional phrases having the object of comparison in the syntactically free position are generally used instead (the face of time, rather than the time-face). The irreversibility of this construction is even clearer than in the case of direct compounding: careful consideration may be required to distinguish the upamā "lotus-eye" from the rūpaka "eye-lotus", but the rūpaka "the lotus of her eye" simply cannot be turned into a simile "the eye of her lotus". For one reason, the genitive is no longer subjective. A comparative idea must somehow be introduced into the genitive (as, "her eye of lotus hue") where the notion of the common property is made explicit ("of lotus hue" is "lotus-hued"). Since rūpaka is an identification of two things, the classifications of this figure have generally followed the line of specifying the degree of identification or the kind of (non-) parallelism involved in the fusion. The notion of "degree" implies and is always discussed in terms of a standard metaphorical type, usually referred to as samasta-[vastu]visaya or complete metaphor, wherein not only the two terms themselves (the major terms) are identified, but several aspects or parts of each major term are likewise identified with parallel aspects of the other major term, resulting in a total image. A metaphor in which only some of the subordinate aspects of the two major terms are identified-in which, in other words, some of the aspects associated with the subject of comparison are not meta- phorically identified-is considered a partial metaphor (ekadešavi- varti). Dandin, in turn, enumerates several different types of partialness as the non-metaphorized term or terms is or are the subject of comparison itself (major term, avayavi), the subordinate aspects of the subject term (avayava), only some of the subordinate aspects (vişama), or, finally, the major and some of the subordinate aspects (ekânga). The other terms in all of these types, which are understood from a comparison with the complete metaphor already described, are, in fact, metaphorical identifications. The elements of the image are implicit. Dandin also considers two types of avayava rūpaka,

24 GLOSSARY for the subsidiary aspect of the subject of comparison may be identified with aspects of the object which subtend a very different relation to cach other than do the aspects of the subject (ayukta rüpaka): for example, in terms of the standard total image of the girl's face (subject) and the lotus pond (object), the smile may be identified with the moonlight and the eyes with the "lotuses". Now, if the lotuses are that type which bloom only during the day, they will be related to the moonlight in a quite different way than the eyes are to the smile of the girl. This approaches what we mean by mixed metaphor, but more often the latter is a kind of utprekşā. Rudrața, admitting the same complete metaphor in his descriptive model, classifies it in a somewhat different way. Instead of allowing certain aspects of the total image to be treated (almost at random) in terms of the subject only and outside the metaphor, Rudrata distinguishes only the two cases where all or some of the subsidiary aspects are drawn out in metaphor (sâvayava), or none at all are (niravayava). The latter case is not the same as Dandin's avayavi, since Dandin requires that the total image be present at least implicitly in the descriptive qualifications of the subject. Rudrata intends that only the major term be mentioned. The sâvayava category is then divided into three types, not according to which subordinate aspects of the subject are or are not identified, but as to the nature of the aspect vis-à-vis the aspect with which it is identified. The aspects of the subject and the object may be inherent qualities of the subject and object respectively (sahaja), or they may be accidental (āhārya), or those of one may be inherent and those of the other accidental (ubhaya). The classification recalls the ayukta rūpaka of Daņdin in that it introduces a philosophical discrimination, but here the criterion of consistency is not primary. Rudrața goes on to consider several types of niravayava rūpaka, that is, metaphorical identifications not involving subsidiary meta- phors. Of course, the simplest type is metaphor itself: one subject and one object so identified (śuddha). But there are certain cases of metaphor involving more than one identification which are not to be taken as expressing a total image. The several metaphors are not comprehended in a relation of subordination, as were those analyzed in terms of a major term and its aspects. Three such "compound" but not "complex" metaphors are recognized by Rudrata: the same subject of comparison may be identified with a number of different objects of comparison, giving a 'garland' of metaphors (mālā);

GLOSSARY the object of comparison of the first metaphor may become the subject 243 of comparison of the following, and so on, giving a chain of meta- phors (rasana); lastly, the object of comparison of one metaphor may itself imply a second, completely independent, metaphorical identification which, as it were, grows out of the first (paramparita, or "continuous" metaphor). Mammata considers this last a separate type, not a subtype of niravayava rūpaka. Otherwise, his classifica- tion is a simplified version of Rudrata's. Dandin, as is his wont, completes the discussion of rūpaka by considering the implications of other figures for this figure. Among the rūpakas so qualified are viruddha (the figure virodha), hetu, ślișta (the figure śleşa), upamā, vyatireka, āksepa, and rūpaka itself. This last is a three-member metaphor differing from the types of complex and compound metaphors in that the three terms are identified with each other as such and imply no relation of subordination or qualification. Further categories proposed by Dandin and illustrat- ing various principles are (sa)viśesana rūpaka, where the metaphors are grammatically adjectival, that is neither compounded (samasta) nor predicated (asamasta); samādhāna rūpaka, where an inconsistency in the metaphor is alleged and explained; and the curious tattvâpa- hnava or "denial of identity", which at first glance appears to be the very inverse of rūpaka or "identification". See that term. The other writers propose classifications which adumbrate those more fully outlined by Dandin or Rudrata. The definitions of rūpaka are remarkably uniform, yet some authors (Daņdin, Udbhața, Rudrața) emphasize the negative aspect of identification and thus relate the figure explicitly to upamā, while others (Bhāmaha, Vāmana) consider the identification in a positive way. The rūpaka is, for the former group, a simile with the difference between the two things suppressed (tirobhūtabheda) and, for the latter group, is simply an identification (tattva) of the two things. ayukta, 'unrelated': (1) a type of complex rūpaka in which the objects of the subsidiary metaphors are not mutually related in terms of some well-known cliché or image which in fact suggests a contrary: mixed metaphor. (2) D 2.78. (3) idam ārdrasmitajyotsnam snigdhane- trôpalam mukham (Dandin; "moonlight" and "day lotus" are not usually connected in popular usage: "This face is radiant with its moonlight of tender smilcs and the soft day lotus of its eye"). (4) "'De sun is made of mud from de bottom of de river; / De moon is made o' fox-fire, as you might disciver'" (Owen Wister; fire and

GLOSSARY 244 mud). (5) Cf. yukta rūpaka and ekanga rupaka. See also samkīrno where the "mixed metaphor" is more generally delined and is not limited to an opposition of ideas. avayava (I). 'part: (1) a complex rupaka in which the subsidiary aspects of the identified subject and object are themselves explicitly identified and become members of the total metaphor, but in which the principal metaphor is only implicitly expressed, the subject alonc being mentioned. (2) D 2.72 (71). (3) akasmād eva te candi sphuritā. dharapallavam / mukhamı muktāruco dhatte gharmambhahkanaman- jarīh (Dandin; her lip is metaphorically a blossom, her beads of sweat are blossoms, but her face is just her face: "Suddenly your face. O cruel one! the blossom of its lower lip bursting forth, gives to the garland of beads of sweat the appearance of pearls"). (4) "In striving to avoid that terrible Charybdis of a Slope she was in great danger of falling into an unseen Scylla on the other hand, that Scylla being Bertie Stanhope" (Anthony Trollope; Mrs. Bold is not here meta- phorically identified with Ulysses). (5) Cf. avayavi. avayava (II): (1) a complex metaphor showing identification of subsidiary aspects. (2) R 8.41-42 (43-45). (3) (4) See the terms mentioned under (5). (5) The figure is subdivided into three types, depending upon how necessarily related to the principal metaphor are the subordinate metaphors: sahaja (the sub-aspects of both are inherent qualities), āhārya (they are accidental qualities), and ubhaya (the sub-aspects of one are inherent, of the other, accidental). The text reads sâvayava. avayavi, whole': (1) a complex rūpaka in which certain subsidiary aspects of the principal subject are mentioned descriptively, while it alone is described metaphorically. (2) D 2.74 (73). (3) valgitabhru gala- dgharmajalam ālokitêkșaņam / vivrņoti madâvasthām idam vadana- pankajam (Daņdin; the face is a lotus; but the brows, sweat, and eyes are but themselves: "The lotus of her face betrays intoxication with its fluttering brows, dripping beads of sweat, and inviting glances"). (4) "But two men in an aeroplane are twins in a womb. The very pulse of one must be the pulse of both, their senses, glances, thoughts, such a unison of co-operation as the former world never saw (Oliver Onions). (5) Cf. avayava, the inverse case. aślișta, 'not punned': (1) a rūpaka in which the descriptive qualifications apply to one or the other of the metaphorically identified terms (usually the object) and not to both, in the manner of a pun. (2) M 145. (3) niravadhi ca nirāśrayam ca yasya sthitam aniva titakautu-

GLOSSARY kaprapañcam | prathama iha bhavān sa kūrmamūrtir jayati caturdaśa- 245 lokavallikandalı (Mammaļa; here the qualifications apply primarily to Visnu, not to the "fourtcen-world-lotus-root": "May the Lord Vişnu in the form of a tortoise be victorious, infinite and independent, source of inexhaustible amazement, the root of the vine of the four- teen worlds!"). (4) "Give me ... / My scrip of Joy, immortal diet; / My bottle of Salvation; / My gown of Glory, hope's true gage; ! and thus I'll take my pilgrimage" (Sir Walter Raleigh; "diet" applies to the scrip, not the joy, and "gage" to glory, not the gown). (5) Cf. ślista. This subtype does not concern the metaphor properly speaking, but only the descriptive qualifications appended thereto. This aside, aślista is indistinguishable from the genus metaphor itself. asamasta, 'uncompounded': (1) a rūpaka in which the subject and object of identification are expressed as grammatically independent words; identification by predication. (2) D 2.68 (67). (3) angulyah pallavāny āsan kusumāni nakhatvișah / bāhū late vasantaśrīs tvam nah pratya- kşacāriņī(Dandin:"Your fingers are new sprouts, the beams from their nails are flowers, your arms are vines; indeed, you are the beauty of spring in visible form"). (4) "And she balanced in the delight of her thought, / A wren, happy, tail into the wind" (Theodore Roethke). (5) The object is predicated of the subject (or parenthetically predicated of it) instead of standing as final member in a compound word. Cf. samasta and the discussion thereunder. This case is distinguishable from upamā (simile) only in the absence of the comparative particle (iva, 'like'), but later writers have also distin- guished similes without the comparative particle (cf. dyotakalupta upamā), as "dawn-eyed". But the two cases, even in their similarity. are necessarily distinct, for the rūpaka is by nature uncompounded (reposing upon a predication), while the upama is always compounded (with the object preceding). The case "gauze of evening" (Virginia Woolf; cf. samasta) is equivocal, but because it does not involve a predication explicitly, I have considered it under samasta rūpaka. Asamasta is the same as Rudrața's "first" rūpaka (R 8.38. 39). asamāsa, 'not a compound': (1) same as asamasta. (2) R 8.38 (39). äkșepa, 'challenge': (1) a rūpaka in which the adequacy of the metaphor is challenged by pointing out in the subject a property which the object does not in fact possess. (2) D 2.91. (3) mukhacandrasya candratvam ittham anyôpatāpinah / na te sundari samvādi (Daņdin: "The qualities of the moon, O Lovely, do not entirely correspond with those of your face-moon which makes others suffer"). (4)

246 GLOSSARY "Re-enter LION and MOONSHINE LYSANDER: This lion i, a very fox for his valour. THESEUS: True, and a goose for hi discretion. DEMETRIUS: Not so, my lord; for his valour cannot carry his discretion; and the fox carries the goose. THESEUS: His discretion, I am sure, cannot carry his valour; for the goose carries not the fox. It is well: leave it to his discretion, and let us listen to the moon" (Shakespeare). ābārya. 'adventitious': (1) a complex rūpaka in which the subsidiary metaphors repose upon accidental properties of the subject and object of the principal metaphor. (2) R 8.42 (44). (3) vikasitatārākumude gaganasarasy amalacandrikāsalile / vilasati śaśikalahamsah prāvr- dvipadapagame sadyah (Rudrata; the moon and stars are only oc- casional aspects of the sky; the lotus and the swan are occasional aspects of the lake: "At the close of the oppressive rainy season, the great swan of the moon diverts himself in the lake of the sky, of blooming lotus-stars, of ripples of clear moonbeams"). (4) "Up in the heavenly saloon / Sheriff sun and rustler moon / Gamble, stuck in the sheriff's mouth / The fag end of an afternoon" (James Michie; sun, moon, and afternoon are occasional properties of the heavens; sheriff, rustler, and cigarette are similarly occasional properties of a saloon). (5) Āhārya is a subtype of sâvayava rūpaka contrasted with sahaja and ubhaya, q.v. The basis of this classification seems to be the logical notion of inherence (samavāya). If a term (concept, thing, property, or any mixture thereof) is related to another so that the destruction of the second entails the destruction of the first, the relation of the first to the second is said to be inherent. This does not imply that the relation itself is eternal or cannot be subject to modification. Blue is said to be an inherent property of the pot, not because the pot is necessarily blue, but because the destruction of the pot cannot be conceived without destroying its color. (Note that the reverse is not the case, by which we mean that the pot is not inherently related to blue; the relation of inherence is not neces- sarily reciprocal.) Similarly, the relation of a lotus to its parts (cf. the example under sahaja) is inherent, while the relation of the sky to its members (sun, moon, and stars), though eternal, is not inherent, since their absence is perfectly compatible with the continued exist- ence of the sky (at night, at new moon, and during the day; or when the sky is overcast). Of course, scientific astronomy would take a different view, but poets are allowed a certain license. spami, 'simile': (1) a type of rūpaka in which the property which under-

GLOSSARY 247 lies the metaphorical identification of the subject and object is made explicit for each of the terms. (2) D 2.88 (89). (3) ayam ālohitacchāyo madena mukhacandramāh | sannaddhódayarāgasya candrasya prati- garjati(Dandin: "The moon of her face, slightly flushed from drink, dis- putes with the moon, ruddy above the eastern hills"). (4) "She was a limpet, with the sensitive side of her stuck to a rock, for ever dead to the rush of fresh and beautiful things" (Virginia Woolf). (5) Cf. vyatireka rüpaka. The difference between simile and metaphor is here exemplified. Both repose upon a similitude (shared property or aspect), but metaphor adds a second dimension by proposing an identification of the two similar things. Simile is thus in principle a realistic figure, while metaphor is necessarily figurative. But as the various subclassifications of simile show, the similitude may be exaggerated beyond any probability (cf. catu, tattvâkhyāna); it is, however, that element of exaggeration which defines the figure as simile: in metaphor, no exaggeration is possible because the two things have become one. As far as the common property is concerned, there is no distinction between the figures. That is why, throughout this work, figures whose specificity relates to the common property may be exemplified by metaphors, even where the original calls for a simile, and vice versa. This was generally recognized by the Indian authors themselves, who usually classify simile in reference to the elements of similitude (among which is the common property), while metaphor is classified almost exclusively in reference to the com- plexity of the identification involved. ubhaya, 'both': (1) a complex rūpaka in which the subsidiary metaphors repose upon the inherent properties of one, and the accidental properties of the other of the two terms in the principal metaphor. (2) R 8.42 (45). (3) alikulakuntalabhārāḥ sarasijavadanāś ca cakra- vākakucāḥ | rājanti hamsavasanāḥ samprati vāņīvilāsinyah (Rudrața; the bees, lotuses, etc. are accidental concomitants of the tank; the tresses, faces, etc. are inherent properties of the girls: "The forest maidens-the river Väni-are lovely with their lotus-faces and chign- ons of bees, their breasts of nightingales and clothes of swans"). (4) "A steamer, probably bound for Cardiff, now crosses the horizon, while near at hand one bell of a foxglove swings to and fro with a bumble-bee for a clapper" (Virginia Woolf; the bumblebee is ac- cidentally related to the foxglove; the clapper is inherent in the bell). (5) This figure is a type of sâvayava rūpaka, contrasted with sahaja and āhārya. For the meaning of "inherence", see āhārya (5).

GLOSSARY ekadeśavivarti, ekadeśi, 'partial': (1) a complex rupaka, some of the aspects 248 of whose subject, subordinate or principal, are not metaphorically identified with the corresponding aspect of the object, but are de- scribed literally. (2) B 2.22 (24), U 1.12, R 8.41 (56), M 141. (3) yasya ranântahpure kare kurvato maņdalâgralatām | rasasammukhy api sahasā parāūmukhī bhavati ripusenā (Mammața; the battle- harem and sword-vine are metaphors, but neither the king nor his enemy are specified as lover or rival: "When the King puts his hands to the vine of his sword in the harem of battle, his enemy's army, though at first filled with passion, is quickly chased away"). (4) "America, for many of us, used to mean a very large apron, covered with a pattern of lozenges, edged by a frill, and chastely suspended by a boundary tape round the ample waist of Canada" (E. M. Fors- ter). (5) All the authors emphasize that the mere incompleteness of the total metaphor is not the only consideration: those aspects not specified must in fact be articulating parts of the metaphor and be readily inferrible as such. "Mixed" metaphor is definitely not de rigeur. Ekadeśavivarti is the opposite of samastavastuvişaya; Bhāmaha and Udbhata know only this single distinction for all of rūpaka, which then amounts to "complete" and "partial". In the more subtle classifications of Dandin and Rudrata, the term is the equivalent of several others or has several subclassifications. Cf. avayava, avayavi, ekâñga. inga, 'one member': (1) a simple (but potentially complex) rūpaka in which one aspect is treated metaphorically, the whole and the other aspect are treated descriptively. (2) D 2.76 (75). (3) madapāțala- gandena raktanetrôtpalena te / mukhena mugdhaḥ so'py eșa jano rāgamayah krtah (Daņdin; the "eye-lotuses" are an aspect of the face, but the face itself and its other aspects (flushed cheeks) are not metaphorically identified with anything else, though the whole is probably meant to imply a lotus pond: "Anyone bewildered by your face, its cheeks pink with drink and its eye-lotuses tender with affection, is turned into a passion"). (4) "It was a night so beautiful that your soul seemed hardly able to bear the prison of the body. You felt that it was ready to be wafted away on the immaterial air. and death bore all the aspect of a beloved friend" (Somerset Maugham). (5) We say potentially complex because of the possibility that several subsidiary aspects be metaphorically identified in this way. According to Dandin the

GLOSSARY tryanga, etc. (two-membered, three-membered). As the figure stands, 249 however, only one metaphor is explicit and therefore must be simple (śuddha); we may sec this type as complex by definition (as Dandin certainly docs): the fourth possibility in the quartet samastavastu (principal and subordinate), avayavi (principal), avayava (subordin- ate). ekânga (part of the subordinate). If so, it differs from avayava in that there all the subordinate aspects described must be made subjects of the metaphor. Here some must be left literal, as, in Maugham's example, the soul or the night. tattvâpahnava, 'denial of the real': (1) a type of rūpaka in which the existence of the real subject is denied, and the object is affirmed in place of it. (2) D 2.95 (94). (3) nâitan mukham idam padmam na netre bhramarāv imau / etāni kesarāņy eva nâitā dantârcișas tava (Dandin: "This is no face, but a lotus; these are not eyes, but bees; these are flower filaments and not the gleams of her teeth"). (4) "What looks like a stone wall to a layman is a triumphal arch to a corporation lawyer" (Finley P. Dunne). (5) Formally, this figure is just the inverse of tattvâkhyäna, where the subject is affirmed in place of the object, but affirmation is classified as simile, denial as metaphor. This difference testifies again to that more basic distinc- tion between simile and metaphor: the former is in principle realistic, the latter is in principle figurative (cf. upamā rūpaka). Denying the existence of the object of comparison does no more than reaffirm the "pre-similar" nature of things, whereas denying the existence of the subject strikes at the very heart of all actuality (which, for the purposes of poetics, we may take as a complex of subjects, attributes, and actions). niravayava, 'without members': (1) a non-complex metaphor; that is, one showing no metaphorical identification of subsidiary aspects or elements. (2) R 8.41,46. (3) (4) See (5). (5) Cf. sâvayava. This figure is subdivided into four types: śuddha (one metaphor). mālā (several independent metaphors), raśanā (several interdependent metaphors), and paramparita (a single metaphor with a multiple predicate or object). See these terms for examples. paramparita, 'continuous': (1) a type of non-complex rūpaka in which the main metaphor is continued by a second which is grammatically subordinate to the object of the first and which gives a rationale for its identification. (2) R 8.46-47 (51), M 145. (3) smaraśabaracd- payaştir jayati janânandajaladhišašilekhā | lāvanyasalilasindhuh sakalakalākamalasarasîyam (Rudrața; the girl is metaphorically

250 identified with a bow; this, in itself, is unclear until the bow is siid to belong to the love-hunter: she is the love-hunter's bow: "This lovely maiden conquers all-she is a lotus pond of all the arts. a river of beauty, a crescent moon on the lake of carthly joys, the bow of the love-hunter"). (4) "Our caresses, our tender words, our still rapture under the influence of autumn sunsets, or pillared vistas, or calm majestic statues, or Beethoven symphonies, all bring with them the consciousness that they are mere waves and ripples in an unfathom- able ocean of love and beauty" (George Eliot). (5) The term param- parita is defined with differing emphases by the two authors who use it: Rudrata draws attention to the formal peculiarity of the com- pound word which expresses the metaphor; namely, that it contains one subject but two objects of comparison (love, subject: hunter. object; bow, object), this second object referring to a second subject which is outside the compound: the girl. This verbalistic account seems to avoid the main point, which is the relation of inclusion or extension obtaining between the object of the principal metaphor and the second metaphor (bow? which bow? love-hunter's bow). Mam- mața, with uncharacteristic insight, fixes upon this conceptual rela- tionship and expresses it as that of condition and conditioned (only by knowing that the bow is that of the love-hunter does it make sense to equate it with the girl.) Paramparita is classified as a type of niravayava rūpaka, a non- complex metaphor without subordinate metaphorical identifications: that is, the relation of subordination discussed above does not involve the relation of a whole to its parts, which is what is intended by the term "complex". Cf. sâvayava, niravayava. Analytically, paramparita is the inverse of avayava, for the metaphor that in avayava would be the whole (love-hunter) is here the subordinate (at least grammatic- ally), and the part (girl-bow) is the principal. But this is not the point at issue, for the form of paramparita is not merely the inverse of the form of sâvayava (where the principal, neatly spelled out, is accom- panied by metaphorically identified aspects, neatly spelled out and inserted at appropriate places in the larger idea: an architectonic of metaphor); rather, in paramparita the subordinate metaphor (which is the principal of sâvayava in meaning) is both a grammatical and conceptual element of an aspect (the object) of the principal meta- phor. Instead of being founded upon a part, it is a part: it is inte- grated in what is analytically its own consequence. Paramparita is apparently identical with the upamārūpaka alam-

GLOSSARY kära of Vamana, which is defined as a rūpaka (metaphor) founded 251 upon an upama (simile). The examples, however, are the same: Vämana probably takes "rajanipuramdhrirodhratilakah šaśī "('the moon is the beauty mark of the night-woman') to mean: 'of the night, which resembles a woman'. It is upon this idea of similitude (though expressed as a metaphor) that the identification of moon and beauty mark is predicated. An alternative explanation would be: "the night with the moon is like a woman with her beauty mark", though one fails to see in this case just how this figure differs from simple metaphor (which is also founded upon a similitude). Vāmana is eager to reduce all figures to a basic simile, so perhaps his contor- tions have no other rationale. This integration of metaphor into metaphor should be distin- guished from rūpaka rūpaka, which is only a triple metaphor (identification of three terms). bhedabhāj, 'distinct': (1) another name for aślista rūpaka. mālā, 'garland': (1) a type of non-complex metaphor in which the same subject is successively identified with a number of objects, each suggesting a different property or aspect. (2) U 1.13, R 8.46 (49), M 144. (3) kusumâyudhaparamâstram lāvanyamahôdadhir guņani- dhānam / ānandamandiram aho hrdi dayitā skhalati me śalyam (Rudrața: "Alas, that lady has shot an arrow into my heart! She is the ultimate weapon of the God of Love, an ocean of beauty, a treasure-trove of qualities, a palace of pleasure!"). (4) "This royal throne of kings, this scepter'd isle, / This earth of majesty, this seat of Mars, / This other Eden, demi-paradise, / This fortress built by Nature for herself / Against infection and the hand of war, / This happy breed of men, this little world, / This precious stone set in the silver sea, / ... This blessed plot, this earth, this realm, this England, / This nurse, this teeming womb of royal kings ... " (Shakespeare). (5) Cf. niravayava; most of the common figures (upamã, vyatireka) are capable of this extension, since they are all analyzable into the same elements: thing compared, object of comparison, common property, etc. Mammața also gives an example for a garland of paramparita metaphors. No new principle is involved. yukta, 'related': (1) a type of complex rūpaka in which the objects of the subsidiary metaphors go together, that is, are mutually related in terms of some well-known cliché or image. (2) D 2.77. (3) smitapuşpôjjvalam lolanetrabhrīgam idam mukham (Daņdin; bees

252 GLOSSARY and flowers, with which the cyes and the smile are identified, are thus mutually related: "This face is radiant with its smile-flower and the bees which are her roaming cyes"). (4) "Saratoga perhaps deserves our greater homage, as being characteristically democratic and American .... Let us, then, make Saratoga the heaven of our aspiration, but let us yet a while content ourselves with Newport as the lowly earth of our residence" (Henry James). (5) Cf. ayukta rūpaka. In Dandin's classification, this term is to be taken as a subtype of dvyanga (or tryanga) rūpaka, which is itself an extension of ekânga, q.v. That is to say, when more than one sub-aspect of an image is made the subject of a metaphorical identification, these identifications may be classified as to their mutual compatibility. We may not apply this classification to those metaphors in which the principal identification is explicit, since presumably in that case all the subsidiary objects will be aspects of the principal, and thus a fortiori go together. raśanā, 'rope': (1) a series of rūpaka in which the object of identification of the preceding metaphor becomes the subject of the following, and so on. (2) R 8.46 (50), M 145C. (3) kisalayakarair latānām karakamalaih kāminām jagaj jayati | nalinīnām kamalamukhair mukhêndubhir yoşitām madanah (Rudrața; the bud-hands of the creepers, then the hand-lotuses of the girls: "The Love God conquers the entire world with weapons which are the bud-hands of creepers, with the hand-lotuses of lovers, with the lotus-faces of the blue lotus plant, with the face-moons of maidens"). (4) "Nor doth this wood lack worlds of company, / For you, in my respect, are all the world" (Shakespeare). (5) This figure is the inverse of raśanā upamā. rūpaka, 'metaphor': (1) a rūpaka in which the object of a simple meta- phorical identification is itself taken as the subject of a further meta- phor; a triple metaphor. (2) D 2.93. (3) mukhapankajarange smin bhrūlatānartakī tava | līlānrtyam karoti (Dandin; on the stage which is a lotus which is her face: on the face-lotus-stage: "On the stage of your face-lotus, the player of your brow-creeper acts out a divertissement"). (4) "In those mirrors, the minds of men, in those pools of uneasy water, in which clouds forever turn and shadows form" (Virginia Woolf; in those mind-mirror-pools). (5) If the two identifications are independently taken, a mālārūpaka is formed. I think, however, that in Virginia Woolf's example, the "pools" should be taken with the "mirrors" rather than with "minds", since it is their reflecting surface which is to be emphasized by the

metaphor. This figure is a rasana rūpaka in which the term taken as 253 both subject and object is expressed only once. It differs from paramparita in that only a term is metaphorically subordinated to the principal metaphor. not another metaphor. viruddha. 'obstructed': (1) a rūpaka in which the metaphorical identifica- tion takes place in the presence of and despite characteristics sufficient to distinguish the two terms. (2) D 2.83. (3) na mīlayati padmāni na nabho py avagāhate | tvanmukhêndur mamâsūnām haraņāyaîva kalpate (Dandin: "The moon of your face does not cause the lotuses to close and does not swim amongst the clouds, but it does seem uniquely able to enslave me!"). (4) "You are a tulip seen to-day, / But, dearest. of so short a stay / That where you grew scarce men can say" (Robert Herrick). (5) Viruddha differs from vyatireka rūpaka in that the identification overrides the distinction. The "obstruction", specifically, is the non-performance by the subject of the metaphor of an act which is characteristic of the object: the moon causes lotuses to close; a tulip grows in a garden. viśeșana. 'qualification': (1) a rūpaka which descriptively qualifies another word. (2) D 2.82 (81). (3) haripādah śirolagnajahnukanyājalâmśukah i jayaty asuranihsankasurânandôtsavadhvajah (Daņdin; Vișņu's foot is said to be wreathed with "Ganges water-gauze"; note that the Ganges is said to have sprung from a footprint of Visnu: "May the foot of Visnu be victorious: the banner of the festival of joy of the Gods who were made fearless [by that foot] of the hosts of demons, [the banner] being the water-gauze of the Ganges attached to the mast [of Vișņu's foot]"). (4) "Books were on his shelves by Wells and Shaw; on the table serious sixpenny weeklies written by pale men in muddy boots-the weekly creak and screech of brains rinsed in cold water and wrung dry-melancholy papers" (Virginia Woolf). (5) These metaphors differ from samastavastuvişaya rüpaka in that a relation other than that of whole to part underlies the subordina- tion of one to the other component metaphor. In the examples, it is that of substantive and qualification. Of course, this formal distinc- tion does not affect the meaning or subject matter of the metaphor. and it should be possible to express the matter as a samastavastuvisaya by stating the terms in that relation. vişama, 'uneven': (1) a complex rūpaka in which the principal identifica- tion is made explicit together with only some of the subordinate identifications; the remaining aspects are treated descriptively and unmetaphorically. (2) D 2.79 (80). (3) madaraktakapolena man-

GLOSSARY 254 mathas tvanmukhênduna | nartitabhrūlatenalam marditum bhuvana- trayam (Dandin: "face-moon" is explicit. along with "brow-creeper" but "cheeks flushed with drink" is descriptive and does not continue the metaphor: "The God of Love has punished the world enough with the moon of your face, its cheeks flushed with drink. its brow- creeper dancing"). (4) "Let us by reverend degrees draw near, / I feel the Goddess here. / Lo I. dread Sack. an humble priest of thine / First kiss this cup thy shrine. / That with more hallowed lips and inlarg'd soul / I may receive the whole" (Anon., "In Praise of Sack": "Sack Goddess" should be taken as the principal; the drinker and his cup are identified with the priest and the shrine, but neither the "priest's" soul nor his lips is metaphorically identified). (5) This figure amounts to a combination of avayavi and ekânga rūpakas. It is a samastavastuvişaya rüpaka with some of the subordinate aspects not metaphorically identified. vyatireka. 'distinction': (1) a rūpaka in which a distinction is drawn between the metaphorically identified object and its real counterpart; a literal re-evaluation of the metaphor itself. (2) D 2.88 (90). (3) candramāļ pīyate devair mayā tvanmukhacandramāh | asamagro'py asau saśvad ayam āpūrnamandalah (Dandin: "The moon[beams] are drunk by the Gods, [the beams from] your face-moon I drink; the former is often less than full, the latter ever and always perfect"). (4) "I will be Paris, and for love of thee, / Instead of Troy, shall Wertenberg be sacked" (Christopher Marlowe; Faustus is speaking). (5) Vyatireka differs from viruddha rūpaka in that the distinction is specified for both the subject and object, giving the subject a positive content (Wertenberg), whereas, in viruddha, the inadequacy alone of the subject is shown, hence the name "obstructed". Cf. upamā rūpaka. This subfigure is vyatireka alamkāra expressed in the form of a metaphor, just as upamā rūpaka was a simile ex- pressed as a metaphor. In the more formal classifications which follow Dandin, such combinations would be relegated to the etcetera category samkīrņa, q.v. vyasta, 'separate': (1) same as asamasta. the term samastavyasta. (2) D 2.68. (5) Used only in śuddha, 'simple': (1) a non-complex rūpaka without subsidiary metaphors of any kind. (2) R 8.46 (48), M 143. (3) kah pūrayed aśeșān kāmān upaśamitasakalasamtāpaḥ | akhilârthinām yadi tvam na syāh kalpadru- mo rājan (Rudrața: "Who would fulfill the numberless wishes of your suitors, O King, were you not the veritable tree of desire through

GLOSSARY which all suffering is assuaged"). (4) "Christ, whose Glory fills 255 the Skies. / Christ, the true, the only Light" (Charles Wesley). (5) The point is not that the metaphor is uncompounded (cf. asamasta), but that the image coined by the metaphor is not extended by other metaphors or metaphorical usages dependent in some way upon that first metaphor. This is the type par excellence of niravayava metaphor; according 1o Rudrața, mālā. rašanā, and paramparita are variously defined combinations of suddha rūpakas. All are called niravayava ('lacking members') for this reason. Sce these terms. ślita. 'punned': (1) a rūpaka to which certain descriptive qualifications are appended which are to be taken differently for each of the two terms of the metaphor. (2) D 2.87, M 145. (3) rājahamsôpabhogâ- rham bhramaraprārthyasaurabham / sakhi vaktrâmbujam idam tava (Dandin; for the lotus, "swan-enjoyed" and "bee-beloved"; for the face. "king-enjoyed" and "suitor-beloved": "Friend, the lotus of your face is worthy of being enjoyed by the Great King [the swan] and so sweet as to attract a lover [bee]"). (4) "A Woman is a book, and often found / To prove far better in the Sheets than bound: / No marvel then why men take such delight / Above all things to study in the night" (Anon.). (5) Mammata takes ślista to be a category in the determination of paramparita rūpaka, along with mālā. There is no clear reason why puns may not serve in other metaphors as well. sakala, 'whole': (1) same as samastavastuvişaya. (2) D 2.70 (69). samkirņa, 'mixed': (1) a complex rūpaka in which a principal subject and several of its aspects are metaphorically identified with objects having no mutual relation. (2) R 8.52-55. (3) lakşmīs tvam mukham indur nayane nīlôtpale karau kamale | keśāh kekikalāpo daśanā api kundakalikās te (Rudrața: "You are Lakșmī; your face is the moon, your eyes, blue lotuses, your hands, white lotuses, your tresses, a peacock's tail, and your teeth, jasmine buds"). (4) "One man is a living soul, but two men are an india rubber milking machine for a beer engine, and three men are noses off and four men are an asylum for cretins and five men are a committee and twenty-five are a meeting and after that you get to the mummy house at the British Museum, and the Sovereign People and Common Humanity and the Average and the Public and the Majority and the Life Force and Statistics and the Economic Man brainless, eyeless, wicked spawn of the universal toad sitting in the black bloody ditch of eternal night and

GLOSSARY 256 croaking for its mate which is the specter of Hell" (oys far. the principal, mankind, has its parts, one man, two men, threr men, etc., severally identified with uninterrelated objccts (', In ; :. a: usage, a "mixed metaphor" has a much wider appucato Fe majority of current examples would probably fall ifito o"f s. gories than rupaka; for example: "The crowned heads , ; E .: :e were quaking in their boots", where "heads" i an czanp.c f vakrôkti II (lakşaņā) and "quaking ... " an example of utpre, ;. But the essential point-a lack of parallelism in multiple figuret ..: predication-is well illustrated by the present case. The eleares: instance is rupaka, for all the terms-subject, object, and aspects-are there necessarily explicit; in utprekşa and vakrokti II ulak;anh the subject is implied by a sort of shorthand. Some mixed metaphers may not be vicious, as the examples show. Cary's mixture serves a specific poetic purpose in that it increases the emphasis of the se- quence itself. In ayukta rūpaka, a more explicit opposition in the objects of identification is required. making the metaphor mere mixed than this. samasta, 'compounded': (1) a rūpaka in which the subject and object of identification are compounded into a single grammatical word, the subject preceding. (2) D 2.68 (66). (3) bāhulatā pūņipadmam caranapallavah (Dandin; three separate examples: "Arm-vines, hand- lotuses, foot-buds"). (4) "The weeds wear moon mist mourning veils" (Carl Sandburg). (5) English is sufficiently like Sanskrit to permit a valid illustration of this grammatical point. A rūpaka is the identification of a subject (literal) with an object (figurative). This identification may be accomplished by the simple assertion of an identity |(see asamasta), but it may also repose upon an implia- tion grounded upon a grammatical feature of the language. namchy. that the last member of a compound alone has a direet relation to the rest of the sentence (it carries the case termination, plural termination, etc.). The first member has syntactical reahty only through the second and thercfore is subordinated to and is taken when possible as an equivalent of the sccond. Through grammatral identification, concrete identification is implied. Now, into this syntactic framework the poet inserts words frecly and particularty seeks those expressions whose intuitive or logical structure differs from that, implied and ready made, of the grammar. Such is "moon mist mourning veils", where the literal and primary fact, "mist", is grammatically subordinated, in the manner described above, to

GLOSSARY that idea which is only a figurative and intuitive qualification of 257 "mourning veil". Grammatically, the primary fact is now "veil", and through this confusion of primacy and subordination on differ- ent levels, the identification of the two terms is accomplished, In the phrase "dear tiger-lily, fanged and striped" (Conrad Aiken), the literal fact and the grammatically defined primary term coincide in "lily". Here is no instance of identification, because the consist- ency of the various levels reasserts a fortiori the literally acceptable subordination of the first term: "lily like a tiger". For this reason, the Indian aestheticians consider such compounds upamā (simile), not metaphor. The degree of subordination is limited by the in- tuitive possibilities of the terms involved. The compound "milk- machine" is neither a rūpaka nor an upamā, for the subordination is merely teleological and can have nothing to do with the representa- tion of the terms themselves. Which is to say, in Pāninean phrase- ology, that tatpurusa compounds can be figurative only when they are karmadharaya, where both members have the same case relation- ship (samānâdhikaranatva). We have, however, one skew case in English, as in Virginia Woolf's phrase "gauze of evening": there is not much question that this is a metaphor in the Indian sense, as it identifies the evening (subject) with gauze (object). It does not mean "the gauze belonging to evening"; it is the equivalent of "evening- gauze", but since English does not offer the same facility for com- pounding as Sanskrit, poets employ this "of" of identification (see Twain's example under samastavastuviaya). A case relationship implying subordination is used to indicate equivalence. We may ordain for English a karmadhārayagarbhatatpurușasamāsa. Such an "of" of identification must be carefully distinguished from the other "ofs": a sine qua non is that the grammatically independent term (gauze) is the representation of the term thereto subordinated (eve- ning), as "moon mist mourning veils". In the example: "And there the lion's ruddy eyes / Shall flow with tears of gold" (William Blake), "tears of gold" does not satisfy this criterion; rather, "gold" is a representation of "tears", and we have upamā, not rūpaka. samastavastuvișaya, 'referring to the entire thing': (1) a rūpaka in which the subject of identification and its several parts or aspects are identified in rigorously parallel fashion with the object and its several parts; complex metaphor. (2) B 2.22 (23), U 1.12, R 8.41- 55, M 140. (3) jyotsnābhasmacchuraņadhavalā bibhratī tārakâsthīny antardhānavyasanarasikā rātrikāpālikîyam | dvipād dvīpam bhramati

GLOSSARY 258 dadhatī candramudrākapāle nyastam siddháñjanaparimalam lāňcha- nasya cchalena (Mammata; the principal subject "night" is identified with the principal object "beggar woman"; their respective parts are the subjects of various subordinate identifications: "moonlight- ash smeared body", "stars-beads", "moon-bowl", etc .: "The beggar woman of the night, white with the smeared ash of moonbeams. who wears a necklace of stars and shows great interest in hidden things, wanders from island to island gathering in the begging bowl of the moon fragrances and magic ointments, dark appearing"). (4) "We see bursting upon us the domes and steeples of Saint Paul, giant young chief of the North ... carving his beneficent way with the tomahawk of commercial enterprise, sounding the war whoop of Christian Culture, tearing off the reeking scalp of sloth and supersti- tion" (Mark Twain). (5) In distinguishing this metaphor from eka- desavivarti, the important thing is that none of the aspects of the subject be descriptively treated (not identified metaphorically). Samastavastuvișaya is the type of complex rūpaka whose many variations herein described are simply the non-identification of one or another characteristic part: avayava (the principal subject- object), avayavi (the subordinates), ekânga (the principal and some subordinates), and vișama (some subordinates). All these terms are types of ekadeśavivarti, but only Dandin goes beyond the basic distinction to elaborate types of partiality. The relation of the sub- metaphors to the main metaphor is that of a part or aspect to the whole. For a discussion of other kinds of subordination, see paramparita and viśeşana rūpaka. samastavyasta, 'compounded and separate': (1) two rūpakas in the same expression, one of which is in the form of a compound word, the other not. (2) D 2.68. (3) smitam mukhêndor jyotsnā (Daņdin: "The smile is the beam of her face-moon"). (4) "Walled in by towering stone-/ Peaked margin of antiquity's delay" (Allen Tate). (5) This is paramparita rūpaka defined in a purely formal fashion, without reference to the kind of relation existing between two metaphors. samādhāna, 'justification': (1) a rūpaka in which a justification is offered for a previously expressed inconsistency in the metaphor. (2) D 2.92. (3) mukhêndur api te candi mām nirdahati nirdayam / bhāgyadoșān mamaîva (Dandin: "Even the moon of your face, O cruel girl, consumes me fierily! Such is the defect of my fate!"). (4) "LION: ... For all the rest, / Let Lion, Moonshine, Wall, and lovers twain / At large discourse while here they do remain. THESEUS: I wonder

GLOSSARY 259 if the lion be to speak. DEMETRIUS: No wonder, my lord. One lion may, when many asses do" (Shakespeare; the player Snug has introduced himself as the Lion: a theatrical metaphor). (5) In ākșepa rūpaka, a property sufficient to distinguish subject from object is pointed out, and the metaphor itself is cast into doubt; here an attempt is made to override that suspicion and thereby to save the original metaphor. In viruddha, the same difference is expressed negatively (non-performance by the subject of an action proper to the object), in vyatireka, positively (the proper action of the subject is also given), but in both, the mode is the indicative; in samā- dhāna and ākepa, the adequacy of the metaphor has yet to be decided. samāsa, 'compound': (1) same as samasta. (2) R 8.40. (5) All types of rūpaka given by Rudrata, except asamāsa, are examples of samāsa. saviśeșaņa, 'with a qualification': (1) another term for višeaņa rūpaka. sâvayava, 'having parts': (1) another name for avayava II hetu, 'cause': (1) a rūpaka in which a cause is given for the identification of the subject and object; or, in which the common property under- lying the identification is expressed in the form of a cause. (2) D 2.86. (3) gāmbhīryeņa samudro'si gauraveņâsi parvatah / kāmadatvāc ca lokānām asi tvam kalpapādapah (Dandin: "You are the ocean for depth, O King, and the mountain for weightiness; having granted boons to all men, you are the tree of desire!"). (4) "LYSANDER: This lion is a very fox for his valour. THESEUS: True, and a goose for his discretion" (Shakespeare). (5) Cf. hetu upamā.

lāțânuprāsa lāțânuprāsa, 'Gujarati alliteration': (1) same as lāțīyā, a type of anuprāsa. (2) U 1.8-10. (5) Udbhața considers it a separate alamkāra.

leśa leśa (I), 'trace': (1) a figure in which a pretext is alleged to cover an embarrassing or otherwise unpleasant situation. (2) D 2.265 (266- 67). (3) ānandâśru pravrttam me katham drstvaîva kanyakām | akși me pușparajasā vātôddhūtena kampitam (Daņdin; the lover is ashamed to admit his tears of joy: "Why should I cry for joy at the sight of that maiden? My eyes are only bothered by some wind-blown flower pollen"). (4) "'I was a-stannin' heah, an' de dog was a- stannin' heah; de dog he went for de shell, gwine to pick a fuss wid it; but I didn't; I says, "Jes' make youseff at home heah; lay still whah you is, or bust up de place, jes' as you's a mind to, but I's got

260 GLOSSARY business out in de woods, I has!"'" (Mark Twain; the darky's "business" is a pretext for his hasty withdrawal; the scene is the siege of Vicksburg). (5) The figure is rejected by Bhāmaha (2.86) as not involving vakrôkti. leśa (II): (1) a figure in which a quality is portrayed as involving a defect, or vice versa. (2) R 7.100 (101-102). (3) hırdayam sadaîva yeşām anabhijňam guņaviyogadulıkhasya | dhanyās te guņahīnā vidagdha- goșthīrasâpetāh (Rudrata: "They whose simple minds are unaware of the pains of ignorance are indeed blessed, for they feel no compul- sion to attend the assemblies of the wise"). (4) " ... bankruptcy itself can be a pleasure. I'd been bankrupt four times. There's nothing like a good smash for getting rid of small worries, the things that don't matter, but peck a man to death" (Joyce Cary). (5) The Sanskrit example shows a defect implying a virtue, the other a virtue implying a defect: "Every cloud has a silver lining". leśa (III): (1) same as vyājastuti. (2) D 2.268-72.

vakra vakra, 'crooked': (1) an artha ślesa in which underlying the double- entendre is a further equivocation of mood (rasa). (2) R 10.9. (3) ākramya madhyadeśam vidadhat samvāhanam tathângānām / patati karah kāñcyām api tava nirjitakāmarūpasya (Rudrața; taking the nouns as names of countries gives one rasa (vīra), as parts of the female anatomy, another (śrngāra); "Overcoming Madhyadeśa [grasping her waist], sending your army against [caressing] the Añgas [her limbs], your hand has fallen even on Kāñci [on her girdle], O you who has conquered Kāmarūpa [bested the Love God]"). (4) "License my roving hands, and let them go / Before, between, behind, above, below. / Oh, my America, my Newfoundland, / My kingdom's safest when with one man manned" (John Donne). (5) Here the equivocation is thought of on two levels: literal (the word itself having two meanings) and contextual (the two literal meanings implying contrary or contradictory emotions). As such, there are four meanings here being punned, and the form presents a nice problem to the dhvani theorist: is the ślesa here subordinate to the expression of the rasa, or vice versa?

vakrôkti (I), 'evasive speech': (1) a figure in which a rejoinder is appended vakrôkti

to a certain remark, either by the speaker or another, in such a way

GLOSSARY that a second meaning is rcad into that remark. (2) R 2.14-17, 261 M 103. (3) aho kenêdršī buddhir dāruņā tava nirmitā / triguņā śrūyate buddhir na tu dārumayī kvacit (Mammața; also a pun on the philosophical sense of "buddhi": in the threefold manifestations of sattva, tamas, and rajas; "Hey, who made your will so pitiless [dāruņā]? Will is said to be threefold, but never, ever made of wood [dāru]"). (4) "'I mean: what do I tell Jane? She's down at the apart- ment by now. She called me. She was worrying that bad about you'. 'What's she want?' 'It has all different names', Captain Dyer said. 'They use like a bed for it. You ought to know, sonny boy'" (James Gould Cozzens). (5) This is one of the rare figures which are essentially dialectical, involving the development of an idea from thesis to antithesis. We are here very close to our own idea of irony if we take this to mean reference to something through its opposite, or, at any rate, its other. It differs from irony in requiring that the thesis, that is, the conventional formulation which is to be referred to through its opposite, be explicit. Although vakrôkti can be ex- pressed through a pun (see ślesa vakrôkti), it differs from pun in requiring that the pun be a reinterpretation of something previously mentioned, not simply a self-contained play on words. kāku, 'intonation': (1) a type of vakrôkti in which the rejoinder is not stated, but is conveyed through an ironic inflection of the voice. (2) R 2.16 (17), M 103. (3) gurujanaparatantratayā dūrataram deśam udyato gantum / alikulakokilalalite naîşyati sakhi surabhisamaye'sau (Mammața; meaning: "he will surely go": "He intends to go to a far land out of obedience to his teacher; surely he will not leave in the sweet-smelling season gay with the sounds of bee and cuckoo!"). (4) "'I don't want to interfere', she said, using the tone and the phrase to mean its exact opposite" (Margery Allingham). (5) This is vakrôkti in the sense that the conversation proceeds on the basis of the under- standing, not the remark. There is no reply because it is not neces- sary. The phrase is its own rejoinder and conveys its reinterpretation through an inflection rather than mere verbal stuff. Stated linguis- tically, the two senses of kāku vakrôkti are carried one by the seg- mental morphemes, the other by the suprasegmental; in śleşa vakrôk- ti, they are carried by two different sequences. That irony is con- sidered a type of vakrôkti in this sense, shows that the Indian writers were aware of it as a function of the continuum of speech, not, as might appear on the face of it, as a static relation of two super- imposed meanings, like ślesa. This is a profound view of irony and

262 GLOSSARY shows how the idea of a figure can be treated quite independently of the expressive modes which it may entail. A conversation is here reproduced. not in its accustomed time continuum, but quite simultaneously. indifferent to time. ślesa 'double-entendre': (1) a type of vakrôkti in which the second sense is conveyed through a rejoinder based on a pun or a play on words. (2) R 2.14 (15). M 103. (3) kim gauri mām prati rușā | nanu gaur aham kupyāmi kām prati | mayity anumānato ham jānāmy atas | tvam anumānata eva satyam (Rudrata: "SIVA: Why are you angry with me. Gauri? SHE: Why am I deemed a cow [gaur] by you? With whom should I be angry? HE: I know by inference [anumānato] that you are angry with me. SHE: indeed, you are anumānato [anumā- nata: not devoted to Umā, another name for Gauri]"). (4) "POLO- NIUS: My honourable lord, I will most humbly take my leave of you. HAMLET: You cannot sir take from me anything that I will more willingly part withal: except my life, except my life, except my life" (Shakespeare). (5) As in śleșa arthântaranyāsa, a second topic is intimated through a pun, but here that duality is comprehended in the form of repartee-a conversation (or monologue) on two planes. vakrôkti (II): (1) a figure wherein a word is used in a figurative sense based on similitude; metonymy of resemblance. (2) V 4.3.8. (3) samsthānena sphuratu subhagaḥ svarcișā cumbatu dyām (Vāmana; "kiss" is used in a figurative sense: "May the blessed one be brilliant with excellence; may he kiss the day with his darting rays"). (4) "She stood breast-high among corn, / Clasp'd by the golden light of morn" (Thomas Hood). (5) The Indian writers distinguish several types of figurative usage (lakșanā), depending on the kind of relation- ship existing between the literal and secondary senses. For example, the relation of proximity underlies "the grandstand roared with approval"; that of whole-part in "Washington was scandalized by it". Vāmana specifies that the relation be understood in this figure as being restricted to similitude only: though light cannot "clasp" anything, the meaning is obtained here through the striking simil- arity of that act with the literal effective manifestation of the light. The figure utprekșā, as described by Daņdin, appears identical with this one, but involves a totally different orientation. This one figure, sauf erreur, marks the only effort of an Indian ālamkārika to define a poetic figure in the referential usage of single words (trope, metonymy) which plays such a large role in our own rhetorics. See the Introduction.

GLOSSARY 263 vākovākya vākovākya, 'dialogue': (1) a figure containing a remark and a reply. (2) AP 342.32-33. (5) Vākovākya may be divided into rjúkti and vakrôkti. The latter is the well-known alamkāra of Rudrața and Mammata consisting of a reply which consciously mistakes the con- tent of the question; irony. Rjûkti is defined as 'inherent' (sahaja) speech, perhaps conversation. Again the scope of the figures is demonstrated (cf. bhāvika). Vākovākya constitutes the sixth śabdá- lamkāra. Other figures involving conversation are praśna and uttara.

vāstava vāstava, 'descriptive': (1) a generic term for those figures which are neither comparative nor hyperbolic. (2) R 7.9-10. (5) Rudrata intends those figures which are more rhetorical than poetic, involving arrangements and sequences of terms, or descriptive intimations. The list is given in 7.11-12. See aupamya, atiśaya, and ślea.

vidarśanā vidarśanā, 'making apparent': (1) a figure in which a similitude is sug- gested by attributing to one subject a property which is characterized as really belonging to another. (2) U 5.10. (3) vinôcitena patyā ca rūpavaty api kāminī / vidhuvandhyavibhāvaryāh prabibharti viśobha- tām (Udbhața: "A girl without a husband, though she be beautiful, offers a sight rivalling in ugliness the moonless night"). (4) "In phrases full of the audible equivalents of Capital Letters, he now went on to assure Mr. Stoyte ... " (Aldous Huxley). (5) Just as one woman cannot carry the ugliness of something else, so capital letters cannot be an attribute of the spoken word; nevertheless, the adjunc- tion suggests comparability-of the woman and the night, of audible and visible sententiousness. But for the example which Udbhata offers, this figure would be indistinguishable from nidarśanā II of Mammata, for the definitions are almost equivalent. In Mammata, the rapprochement is via a similar or common result, and two distinct verbs underlie the contrast: here there is but one verb, which does not apply literally to the sense expressed, and the rapprochement is simply with the object of that verb, as taken figuratively. This figure is perhaps a "portmanteau" of nidarśanā. A second type is mentioned by Udbhata in the definition, wherein the two situations are in fact related, but no example is offered.

GLOSSARY 264 He may be referring to the more common nidarsana I. The examples also suggest that vidarśanā is similar to utpreksā, but the figurativeness of the ascription is literally pointed out. vinôkti vinôkti. 'speech including the word "without"': (1) a figure in which two things are presented as excluding one another. (2) M 171. (3) arucir niśayā vinā śašī śaśinā sâpi vinā mahattamah / ubhayena vinā manobhavasphuritam naîva cakāsti kāminoh (Mammata: "The moon without the night has no brilliance; the moonless night is fearful and dark: without both, no thought of love comes to the minds of lovers"). (4) "Through want of [passion] she had sung without being merry, possessed without enjoying, outshone without triumphing" (Thomas Hardy). (5) Vinôkti is an obvious inversion of sahôkti ("speech including the word 'with'"). vibhāvanā vibhāvanā, 'manifesting': (1) a figure in which an effect is realized in the absence of its normal or conventional cause, thus implying another, unusual cause. (2) B 2.77 (78), D 2.199-204, V 4.3.13, U 2.9, AP 344.27, R 9.16 (21), M 162. (3) apītamattāh śikhino diśo'nutkanțhi- tâkulāḥ / nīpo'viliptasurabhir abhrastakalusam jalam (Bhāmaha; a description of the rainy season: "The peacocks are drunk without having imbibed, the skies are confused without being in love, the nīpa tree is sweet smelling without being anointed, the water is corrupt without having sinned"). (4) "O world invisible, we view thee, / O world intangible, we touch thee, / O world unknowable, we know thee, / Inapprehensible, we clutch thee!" (Francis Thompson; presumably a description of religious man). (5) Outwardly, this figure resembles a simple virodha (contradiction); its peculiarity reposes in the fact that the two terms contradicted are cause and effect, not any two coexistent properties. According to Dandin and the Agni Purāna, vibhāvanā is of two sorts, depending on whether another (less common) cause is implied (kāranântaram), or whether the nature of the thing itself provides the explanation (svābhāvika). Rudrața, finally, divides the former into vibhāvanā of the subject (abhidheya) and vibhāvanā of an aspect of the subject (vikūra). kāraņântaram, 'another condition': (1) a type of vibhāvana in which a less usual cause is to be inferred as explanation of the seeming contradic- tion. (2) D 2.199 (200), AP 344.28, R 9.16 (17). (3) apītakşībakā-

te sry s cear phcet having beec cieaned the wiier are -Wr ze was beard and pet i: mas noc toid: My Srait is fTa. cic Cudeck Ticaborne: wrimen jast berore tis execation at the gaiata. sarraness': (l) a type of riontiond in which the aure of the thing so described is implied as the explanation of the seering contradiction. (2) D 2199 (OD) AP 344.28. R 9.30 (IL sararas tara sundari (Dandin; her anger is the "canse" of ber parad- oxical appearance; she doesn't love him: "Your glance is dark without eyeshade, your brow is arched with no one bending it, your lip is full colored and not a touch of lipstick). (4) -Lo. lo! how brave she decks her bounteous bower, With silken curtains and gold coverlets, Therein to shroud her sumptuous belamour, ! Yet neither spins nor cards, nor cares nor frets, But to her mother Nature all her care she lets" (Edmund Spenser; the lilies bloom "naturally"). (5) See kāranántara.

virodba virodha, 'contradiction': (1) a figure in which contradictory properties are expressed of the same subject; the affirmation of the excluded middle. (2) B 3.24 (25), D 2.333-39, V 4.3.12, U 5.6, AP 344.28, R 9.30-44, M 166-67. (3) upāntarūdhôpavanacchāyāšītâpi dhūr asau / vidūradešān api vah samtāpayati vidvişah (Bhāmaha; the king's "cool" sovereignty inflames his enemies: "The royal sovereignty, cool in the shade of the city's environing groves, consumes his enemies, though they be in a distant land"). (4) "I find no peace, and all my war is done; / I fear and hope, I burn, and freeze like ice; / I fly aloft, yet can I not arise; / And nought I have, and all the world I seize on" (Sir Thomas Wyatt; "Description of the con- trarious Passions in a Lover"). (5) This figure is elaborately catalo- gued by Rudrata, probably following Dandin's sketchy and un- named sixfold division, into fourteen types based on the common- place distinction jāti, guņa, kriyā, dravya ('genus', 'attribute', 'act', 'substance'). (Rudrata asserts, however, that one of the fourteen

266 GLOSSARY types -- jati dravya-is theoretical only.) The contradiction may exemplify any of these categories, or may involve terms from any two of them. (For an example of multiple or complex virodha, see guņakriyā virodha.) Mammata copies this scheme. Virodha differs from pihita and like figures in emphasis only. Where the intention is to show two incompatibles as equally valid, and where no implicit decision is made, we have virodha; if the intention is to exaggerate one of the incompatibles at the expense of the other, we have hyper- bole, which uses contradiction as a means. Like all the subtle distinctions of alamkāraśāstra, the properly understood intention of the poet is the ultimate clue. For this reason, it might be added, the body of poetic distinctions can be a powerful adjunct in coming to grips with the poet's often murky thoughts. kriyā, 'verb': (1) a type of virodha where the incompatibility is that of two actions of the same subject. (2) D 2.334, R 9.36, M 167. (3) bālamrgalocanāyāś caritam idam citram atra yad asau mām | jadayati samtāpayati ca dūre hrdaye ca me vasati (Rudrața; "soothes" and "inflames": "How can this doe-eyed girl perform such wonders? She both soothes and inflames me from a distance, yet she resides in my heart!"). (4) "It's bad when they think well of you, Cokey, because you get to think too much of yourself. And it's bad when they think badly of you because you may get to think badly of them. Take your mind off your work" (Joyce Cary; "think well" and "think badly"). guņa, 'attribute': (1) a type of virodha where the incompatibility is that of two (adjectivally stated) attributes of the subject. (2) R 9.35, M 167. (3) satatam musalâsaktā bahutaragrhakarmaghațanayā nrpate | dvijapatnīnām kațhināh sati bhavati karāh sarojasukumārāh (Mam- mata; "harsh" and "soft": "The hands of the Brahmin ladies, busily attached to the duties of the home, forever making foods and sauces, have become rough; yet in your presence, King, those hands are lotus soft!"). (4) "Behold a critic, pitched like the castrati. Imperious youngling, though approaching forty" (Theodore Roeth- ke). (5) Dandin gives an (unclassified) example which, though based on guna, does not show virodha in any accepted sense: "tanumadhyam prthuśroni raktâustham asitêkşaņam / natanabhi vapuh strīņām kanı na hanty unnatastanam" ('Thin waisted and ample thighed, red lipped and black eyed, with depressed navel and raised breasts-who would not be striken by the body of a woman?'); compare: "This will celebrate the occasion-a curious sense rising in her, at once freakish

GLOSSARY 267 and tender, of celebrating a festival, as if two emotions were called up in her, one profound for what could be more serious than the love of man for woman, what more commanding, more impressrve, bearing in its bosom the seeds of death; at the same time these lovers, these people entering into illusion glittering eyed, must be danced around with mockery, decorated with garlands" (Virginia Woolf). Both examples show only a situation of contrasts, not contradiction (D 2.336). guņakriyā, 'attribute-verb': (1) a type of complex virodha where incom- patibility is shown between actions and qualities or traits of the same subject. (2) R 9.39, M 167. (3) sā komalâpi dalayati mama hrdayam (Rudrața: "O how that tender one causes my heart to break!"). (4) "Of course I don't deny that if I were thoroughly unscrupulous I could make a sensation. It would be rather amusing to show the man with his passion for beauty and his careless treatment of his obligations, his fine style and his personal hatred for soap and water, his idealism and his tippling in disreputable pubs; but honestly, would it pay?" (Somerset Maugham). jāti, 'genus': (1) a type of virodha where generic incompatibility is shown in the same subject. (2) D 2.335, R 9.37, M 167. (3) ekasyām eva tanau bibharti yugapan naratvasimhatve | manujatvavarāhatve tathâiva yo vibhur asau jayati (Rudrața; "manliness" and "lionhood" in speaking of a form of Visnu: "May he be victorious, the Lord who unites in one body humanity and lionhood, manliness and porcinity"). (4) "Now I am well aware that pettiness and grandeur, malice and charity, hatred and love, can find place side by side in the same human heart" (Somerset Maugham; "pettiness and grandeur"). (5) The logical category "genus" is never given abstract substantiation, but is always considered to be a peculiar kind of attribute. In this sense. the important distinction is between genus and guna, not between genus and species (dravya). This turns out to be largely a matter of morphology: if a pot is red, then redness is a generic attribute of its color. dravya, 'particular thing': (1) a type of virodha where the incompatibility is that of particular individual things or ideas. (2) D 2.337, R 9.34. M 167. (3) atrêndranīlabhittişu guhāsu śaile sadā suvelâkhye i anyonyânabhibhūte tejastamasī pravartete (Rudrața; tejas ['energy'] and tamas ['indolence'] paired with "glimmering" and "shadow": "The glimmering and the shadow, ever contending, persist on the sapphire-set walls of the caves of Mount Suvela"). (4) "[Theodore

268 GLOSSARY Roosevelt] is an interesting combination of St. Vitus and St. Paul" (J. Morely, quoted by Frederick Lewis Allen). (5) These categories are ill translated by our own: witness the example "energy" considered as a category of substance. virodha (II): (1) same as asamgati. (2) V 4.3.12. (5) The subject of virodha has been fully developed in relation to the causal process. Various contradictions or distortions of the normal relation of cause and effect are exemplified by the figures anyonya, pūrva, vyāghāta, asamgati, samādhi, and ahetu. An ordinary virodha concerns aspects of one subject and does not extend necessarily to relations between many subjects, nor is it based on any notion of the propriety of such relationship, as implied, for example, by the notion of cause and effect. virodha (III): (1) an artha ślesa in which a positive term or thought is directly contradicted by its negative, and resolution of the contradic- tion is effected by understanding one (usually the negative) as a pun. (2) R 10.5. (3) samvardhitavividhâdhikakamalo'py avadalitanālikaḥ so'bhūt | sakalâridārarasiko'py anabhimataparânganāsangah (Rudra- ta; the first reading suggests that the king is both an enjoyer of the wives of his enemies and uninterested in the wives of others, but dāra (first "wife") also means "sword": "Although he has nurtured various excellent lotuses, he has destroyed lotuses [fools]; although he is an amateur of all his enemies' wives, he disdains the embraces of others' women [swords]"). (4) "In those old days, the Nymph called Etiquette / (Appalling thought to dwell on) was not born. / They had their May, but no Mayfair as yet, / No fashions varying as the hues of morn" (C. S. Calverly; "May" without "Mayfair" appears to be a contradiction, until Mayfair is understood as a [fashionable] quarter of London). (5) This figure is virodha alamkāra expressed through ślesa. See also virodhâbhāsa. This could better be called paradox.

virodhâbhäsa, 'appearance of paradox': (1) an artha ślesa in which two virodhâbhāsa

terms appear to contradict each other, and in which the contradiction is resolved by understanding one or the other as a pun. (2) R 10.22. (3) tava dakşiņo'pi vāmo balabhadro'pi pralamba eșa bhujah : duryod- hano'pi rājan yudhisthiro'stîty aho citram (Rudrața; the king's arm is both "right" and "left", but "left" also means "unpropitious" (for his enemies); Pralamba and Balabhadra, Duryodhana and Yudhi-

GLOSSARY 269 șțhira are enemies in the Rāmâyana and the Mahābhārata: "Your arm is left, although it is right [clever], Pralamba [pendant] although Balabhadra [strong], although Duryodhana [difficult to fight], it is Yudhişthira [firm in battle]"). (4) "Until yesterday I had no idea that there were any families or persons whose origin was a Terminus" (Oscar Wilde; Lady Bracknall refers to Jack's having been abandoned, as an infant, in the cloak-room of Victoria Station). (5) This figure differs from virodha III only in that the apparent paradox is borne by the meanings of the words, rather than by a formal (positive- negative) contrast. Both are appearances of paradox, and both are resolved in the same way.

viśeșa viśeșa (I), 'difference': (1) a figure in which a quality or thing is represented without, or in separation from its natural substratum or basis. (2) R 9.5 (6), M 203. (3) divam apy upayātānām ākalpam analpagu- ņagaņā yeşām / ramayanti jaganti girah katham iha kavayo na te vandyāh (Rudrata; the poet's voice is represented here in separation from his body: "Why are the poets now gone beyond to heaven not honored here-poets whose voices will delight the worlds and mani- fest numberless qualities till the end of time?"). (4) " ... when the door opened, and a somewhat forbidding countenance peeped into the room. The eyes in the forbidding countenance looked very earnestly at Mr. Pickwick, for several seconds, and were to all appearance satisfied with their investigation; for the body to which the forbidding countenance belonged, slowly brought itself into the apartment, and presented the form of an elderly individual in top- boots-not to keep the reader any longer in suspense, in short, the eyes were the wandering eyes of Mr. Grummer, and the body was the body of the same gentleman" (Charles Dickens). (5) Cf. adhika II, where the substratum is exceeded or surpassed by the condition which it limits. In the following example from Shelley, the sub- stratum is actually left behind: "My heart, for my weak feet were weary soon, / Did companion thee". viśeșa (II): (1) a figure in which a single quality or thing is represented as inhering in, or being in conjunction with, several natural substrata or bases. (2) R 9.7 (8), M 203. (3) sā vasai tujjha hiae sā cciacchīsu sāa vaaņesu / ahmārisāņa sundara oāso kattha pāvāņam (Mammața; spoken by a jealous wife about her rival: "She resides in your heart, in your eyes and voice; where is there place for sinners like us?").

270 GLOSSARY (4) "Mine eye and heart are at a mortal war, / How to divide the conquest of thy sight; / ... My heart doth plead that thou in him dost lie, / A closet never pierced with crystal eyes, / But the defendant doth that plea deny, / And says in him thy fair appearance he;" (Shakespeare; the same idea as in the Prakrit example with the added fillip that the various "residences" of "her" are considered contestants; cf. the figure pratyanīka). viśeşa (III): (1) a figure in which a single cause is represented as producing, in addition to its usual effect, another, literally impossible effect. (2) R 9.9 (10), M 203. (3) likhitam bālamrgâksyā mama manasi tayā śarīram ātmīyam / sphuțam ātmano likhantyā tilakam vimale kapolatale (Rudrața; applying cosmetics produces not only the tilakam but also a mark on the young man's soul: "As she paints the bright tilaka mark on her spotless cheek, her whole form is graven in my mind"). (4) "Why I tie about thy wrist, / Julia, this silken twist; / For what other reason is't / But to show thee how, in part / Thou my pretty captive art? / But thy bond-slave is my heart ... " (Robert Herrick). (5) Compare asamgati, where the only effect produced is literally impossible. The present case seems to be the combination of an ordinary hetu with asamgati.

viśeșôkti viśeşôkti (I), 'giving a difference': (1) a figure in which a deficiency (a negative attribute), either natural or occasional, is pointed out in such a way as to magnify or emphasize the capability of its subject. (2) B 3.22, D 2.323. (3) ekas trīņi jayati jaganti kusumâyudhah / haratâpi tanum yasya śambhunā na hrtam balam (Bhāmaha; though incorporeal, Love conquers all with his flowered arrows: "One Love God has conquered the three worlds, and though Siva stole his body away, his force remains"). (4) "The eyes in front of their [women's] face are not used for seeing with, but for improving the appearance. Hang the thickest veil in front of them and a girl of seventeen will still see the other woman, through two doors and a brick wall, with the various organs of perception existing in her skin. which changes color, her breasts which tingle, and her brain which performs evolutions of incalculable direction and speed" (Joyce Cary; though they cannot see, yet they are aware). (5) The present case differs from vibhāvanā in showing only a deficiency in the ability of the subject as a means to exaggerating the power of a given cause which realizes its effect in a normal way. But the emphasis

GLOSSARY in vibhavana rests entirely upon a negated cause as the cause of just 271 that "impossible" effect. In other words, we wonder, in vibhāvanā, that the effect should come at all, but here we wonder at how it has come about, or, to put it another way, the proximate cause is denied in vibhāvanā, but in viśeșôkti, only a condition is denied. Dandin classifies visesôkti into five sorts, using the common four- some-jāti, guņa, kriyā, and dravya (cf. virodha)-and adding a fifth, called hetu, which is "qualified deficiency", much as viśitá- dvaita is "qualified non-dualism". Compare viśeșôkti II, the exact reverse of this term. kriyā, 'verb': (1) a type of viśesôkti in which the subject's deficiency is that of an action usually contributing to a given effect. (2) D 2.323 (326). (3) na baddhā bhrukuțir nâpi sphurito daśanacchadah | na ca raktâbhavad drstir jitam ca dvisatām kulam (Daņdin; the king invests his enemies without the slightest sign of anger: "His brows unlower- ing, his lips not athrob, his glance uninflamed-yet he conquers the hosts of his enemies!"). (4) "So forth those joyous Birds did pass along / Adown the Lee, that to them murmur'd low, / As he would speak, but that he lack'd a tongue, / Yet did by signs his glad affec- tion show, / Making his stream run slow" (Edmund Spenser; the stream is unable to speak, yet he shows his affection). guņa, 'attribute': (1) a type of viśeșôkti in which the subject's deficiency is that of a specific (adjectival) attribute usually associated with that effect. (2) D 2.323 (324). (3) na kathoram na vā tīkşnam āyudham puspadhanvanaḥ | tathâpi jitam evâsīd amunā bhuvanatrayam (Daņdin: "The Love God's weapon-his bow of flowers-is neither hard nor sharp, yet with its aid he has subdued the three worlds!"). (4) "The turtle lives 'twixt plated decks / Which practically conceal its sex. / I think it clever of the turtle / In such a fix to be so fertile" (Ogden Nash). jāti, 'genus': (1) a type of viśeșôkti in which the subject's deficiency is that of a generic characteristic usually associated with the given capacity or effect. (2) D 2.323 (325). (3) na devakanyakā nâpi gandharvakulasambhavā | tathâpy esā tapobhangam vidhātum vedhaso py alam (Dandin: "Though not one of the heavenly chorus, not even a daughter of the Gods, she could still disrupt the penances of Brahma!"). (4) "'I know I'm not a great novelist', he will tell you. 'When I compare myself with the giants I simply don't exist .... All I want people to say is that I do my best .... And after all, the proof of the pudding is in the eating: The Eye of the Needle sold

272 GLOSSARY thirty-five thousand in England and cighty thousand in America, and for the scrial rights of my next book I've got the bipgest terms I've ever had yet'" (Somerset Maugham; though not a preat novelist, his books sell). (5) Sec the remark on jati virodha; jati here means 'class'. dravya, 'substance': (1) a type of visesokti in which the subjeet's deficieney is that of particular things which usually accompany its capacity or effect. (2) D 2.323 (327). (3) na rathā na ca mātañgā na hayā na ca pattayah | strīņām apângadrstyâiva jīyate jagatām trayam (Daņdin; though unaccompanied by the accoutrements of war, Love conquers all: "No chariots! No elephants or horses! No infantry! With sidelong glances alone do the ladies conquer the three worlds!"). (4) "From thence to Heaven's brideless hall, / Where no corrupted voices brawl; / No conscience molten into gold; / Nor forged accusers bought and sold; / No cause deferred; nor vainspent jour- ney; / For there Christ is the King's Attorney, / Who pleads for all without degrees, / And he hath angels, but no fees" (Sir Walter Raleigh; though it lack the accoutrements of a courtroom, yet justice is done). hetu, 'cause': (1) a type of viśesôkti in which the subject's deficiency is not total, but qualified; the deficient attribute is present but in an unusual or improbable form. (2) D 2.329 (328). (3) ekacakro ratho yantā vikalo vișamā hayāḥ | ākrāmaty eva tejasvī tathâpy arko nabhastalam (Dandin; references are to the disc of the sun, the seven days of the week, and the dawn, said to be anūru, 'lacking thighs': "The burning sun still courses through the heavens, though his chariot has but one wheel, his steeds are uneven, and his driver maimed"). (4) "As I walked along the winding road ... I mused upon what I should say. Do they not tell us that style is the art of omission? If that is so I should certainly write a very pretty piece" (Somerset Maugham; the author's style is present, but its form is a bit unusual). viśeșôkti (II): (1) a figure wherein no effect obtains despite the presence of an effective cause. (2) U 5.4, M 163. (3) nidrānivrttār udite dyura- tne sakhījane dvārapadam parāpte | ślathīkrtâśleșarase bhujamge cacāla nâlinganato'ngana (Mammata; the causes for awakening are present, but the girl, exhausted by lovemaking, remains asleep: "The sun has risen to put an end to sleep, the troop of friends awaits outside the door, yet the lady departs not from a loose embrace in the arms of a lover who has tamed her passion"). (4) "But Adam could not sit

GLOSSARY 273 down to his reading again, and he sauntered along by the brook and stood leaning against the stiles, with eager, intense eyes, which looked as if they saw something very vividly; but it was not the brook, or the willows, or the fields or the sky" (George Eliot). (5) This figure is just the inverse of viśesôkti I, where the effect is present, the con- tributing cause absent. But, as Mammata shows, a clever scholiast can turn black into white; he cites the example, originally from Bhämaha, which we have given under viśeșôkti I and interprets it: "Deprivation of body is usually followed by deprivation of power; this does not happen in the case of Love, whose power is unimpaired." By taking as the effective cause the very deficiency which other writers allege, the same example can be made to prove contraries. The influence of Sanskrit logic, where double negatives are the preferred modes of exposition (e.g., concomitance is the non-existence of mutual absence), upon Mammata is evident, to the detriment of his poetics. Udbhața and Mammata subdivide into several classes: the ex- planation for the non-operation is given (nimittadrsti) or left to inference (nimittâdrsti). Mammata adds in logical fashion a third category-where the explanation is quite beyond us (acintyanimitta) -but he gives the same example as for nimittadrsti! acintyanimitta, 'inconceivable cause': (1) a type of viśesôkti II whose distinctiveness is questionable. (2) M 163. (5) See viśeşôkti II. The example offered is identical with that of anuktanimitta viśesôkti (see nimittâdrsți). anuktanimitta, 'whose cause is not expressed': (1) same as nimittâdrsti viśeşôkti. (2) M 163. uktanimitta, 'whose cause is expressed': (1) same as nimittadrsti viśeşôkti. (2) M 163. nimittadrsți, 'evidence of cause': (1) a type of viśesôkti II in which an explanation is given for the unexpected non-operation of the cause. (2) U 5.5, M 163. (3) ittham visamsthulam drstvā tāvakīnam vicesțitam/ nôdeti kimapi prastum satvarasyâpi me vacah (Udbhața; the reason why he didn't speak is given; it was Parvati's stumbling gait: "Seeing your hesitating gait, no words arose to put a question, though I was eager"). (4) "Miss Thorne declared that she was delightedito have Mrs. Bold and Dr. Stanhope still with her; and Mr. Thorne would have said the same, had he not been checked by a yawn, which he could not suppress" (Anthony Trollope). (5) The figure is called uktanimitta in Mammața.

274 GLOSSARY nimittâdrsti, 'non-evidence of cause': (1) a type of viśesôkti II in which no explanation is given for the unexpected interruption of the cause. (2) U 5.5, M 163. (3) (3) maharddhini grhe janma rūpam smarasuhr- dvayah / tathâpi na sukhaprāptih kasya citrīyate na dhīh (Udbhața; fate is alleged to be the unspoken explanation for the girl's un- happiness: "Though born in a family of great wealth, beautiful and young and befriended by Love, she was not happy. Who does not marvel at it?"). (4) "'A fine painter, Mr. Esdaile, it's a pleasure to work for him', the old man ran on; and I did not reply that in my experience few pleasures in the world lasted quite so long. I was thinking of other things, the nature of which you may guess at" (Oliver Onions). (5) This and the previous type illustrate a topic which preoccupies certain logicians, since it is relevant to the con- sideration of what constitutes an inferrible proposition (thesis). A cause, or condition is not sufficient in itself to account for an effect: the absence of counteracting causes or conditions is just as necessary; thus the absence of certainty is also an element in the inferribleness of propositions. Certainty causes the operation of the syllogism to be fruitless. The figure is called anuktanimitta by Mammata. viśeşôkti (III): (1) a striking identification of non-similar things, obtained by expressing one of the terms on the distinctive level of reality appropriate to the other. (2) V 4.3.23. (3) dyūtam hi nāma purușa- syâsimhâsanam rājyam Mrcchakațika, quoted by Vāmana); vyasanam hi nāma sôcchvāsam maraņam (Vāmana: "Gambling for some men is a kingdom without a throne!" "Sin is breathing death!"). (4) "Fame is a food that dead men eat" (Henry Austin Dobson; food and fame are ontologically different entities: food is expressed on the ideal plane appropriate to fame). (5) This figure is just the inverse of vyatireka, where two otherwise similar things are said to be subject to a point of difference which is generally figurative or at least irrelevant. Here, two otherwise different things are said to be subject to a point of similarity, which is just as figurative. Despite his examples, Vāmana's definition is misleading, for his terms resemble those used by others for vyatireka ("ekaguņahānikalpanāyām šeşa- sāmyadārdhyam"), but his definition of vyatireka is correspondingly skew (in which the pre-eminence of the subject is expressed). The issue is thus neatly avoided, since the pre-eminence is expressed precisely by pointing out a defect in the other term.

GLOSSARY 275 vişama vişama (I), 'dissimilar': (1) a figure wherein the relationship between two things, which may exist either in fact or in the mind of the speaker, is considered incongruous. (2) R 7.47-50, M 194. (3) rūpam kva madhuram etat kva cêdam asyāh sudāruņam vyasanam / iti cintayanti pathikās tava vairivadhūm vane drstvā (Rudrața; flattery of the victorious warrior: "'How can such beauty exist in such sinful and harsh condition?' Thus do the travellers wonder who see in the forests the wives of your vanquished enemies"). (4) "Did He smile His work to see? / Did He who made the lamb make thee?" (William Blake; a tiger is referred to). (5) The relation under consideration may not in fact exist, as: "He had the passion of Romeo in the body of Sir Toby Belch" (Somerset Maugham); the incongruity is given figurative force. vişama (II): (1) a figure in which the causal process is described as functioning in an incongruous way: either an outclassed cause accomplishes its effect, or a potent cause somehow fails to accomplish its effect. (2) R 7.51 (52-53). (3) tvadbhrtyâvayavān api sodhum samare kşamā na te kşudrāḥ / asidhārāpathapatitam tvam tu nihanyā mahêndram api (Rudrața: "Your vile enemies can't withstand even the little fingers of your dependents in battle; and you, O King, would slay even Indra should he fall beneath your sword!"). (4) "Ah, how can those fair Eyes endure / To give the wounds they will not cure!" (John Dryden). (5) This figure resembles the two types of viśesôkti, but, unlike them, does not repose upon the idea of deficiency. In viśeșôkti, a cause must be specifically unable to produce that effect; here no judgment is implied beyond the mere improbability of the event. vișama (III): (1) a figure in which a cause is described as producing the contrary effect of the one desired. (2) R 7.54 (55), M 194. (3) utkanthā paritāpo raņaraņakam jāgaras tanos tanutā / phalam idam aho mayâptam sukhāya mrgalocanām drstvā (Rudrața: "Longing, sorrow, and regret, wakefulness and loss of weight-this is the result I have gained when I sought her out for pleasure"). (4) "But thence I learn, and find the lesson true, / Drugs poison him that so fell sick of you" (Shakespeare; drugs only aggravate the sickness of love). (5) Cf. viśeşa III and abhāva hetu. vişama (IV): (1) a figure in which a cause and its effect are described as possessing incompatible qualities. (2) M 194. (3) sadyah karasparśam avāpya citram raņe raņe yasya krpāņalekhā | tamālanīlā saradindu-

276 GLOSSARY pāņdu yaśas trilokyâbharaņam prasūte (Mammata; the sword is said to be dark blue and its effect, glory, to be the color of the autumn moon: "Always in his hand, in battle after battle, that thin sword, dark as a blue lotus, engenders in him glory worthy of the three worlds and pale as the autumn moon"). (4) "My mother bore me in the southern wild, / And I am black, but O, my soul is white!" (William Blake). (5) An example of cause and effect differing as to mode of action is: "The walls were of old plaster ... marked with the blood of mosquitoes and bed-bugs slain in portentous battles long ago by theologians now gone forth to bestow their thus uplifted visions upon a materialistic world" (Sinclair Lewis). We are to believe that the fight with bedbugs has prepared the theologian for his later calling.

vyatireka vyatireka (I), 'distinction': (1) a figure wherein two notoriously similar things are said to be subject to a point of difference; usually the subject of comparison is stated to excel the object, surpassing the norm of its own comparability; hence, an inverted simile. (2) B 2.75 (76), D 2.180-98, V 4.3.22, U 2.6, R 7.86-89, M 159. (3) kuvala- yavanam pratyākhyātam navam madhu ninditam hasitam amrtam bhagnam svādoḥ padam rasasampadaḥ / viam upahitam cintāvyājān manasy api kāminām caturamadhurair līlātantrais tavârdhavilokitaiḥ (Vämana: "The lotus forest repudiated! The new springtime has been put to shame! Honey has become a mockery, the state of sweet satisfaction is ended, and poison, in the guise of longing, has been put into the minds of lovers by your playful, passionate, coquettish sidelong glance"). (4) "Eyes, that displace / The neighbour diamond, and outface / That sunshine by their own sweet grace" (Richard Crashaw). (5) The most extensive anatomies of this figure are given by Dandin and Mammata and rest upon the same sorts of criteria. Dandin divides first, into those dissimilitudes whose common property is expressed (śabdôpādāna), and second, into those where it is implicit (pratīyamāna). Mammața follows suit, but subdivides the former category as to whether a verbal or nominal similitude is expressed (śabda, artha). Both authors admit formal criteria depending upon the number, character, and scope of the differential qualifications which express the dissimilitude (eka, ubhaya, ādhik ya, hetu). In the Sanskrit texts considered, there is but one example offered of a vyatireka whose function is to extol the object at the

GLOSSARY expense of the subject of comparison. Cf. vyatireka II. 277 artha, 'implicit': (1) a type of vyatireka in which the comparative particle (iva. 'like') is present, expressing a similitude of action or behavior. (2) M 160. (3) asimātrasahāyo'pi prabhūtâriparābhave / nivânya- tucchajanavat sagarvo'yam mahādhrtih (Mammața: "Armed only with a sword, in the midst of his enemies' rout, King Mahādhrti is not swelled up with pride like the rest of the vain world"). (4) " ... and art, the most nervous of the three, musn't be pushed aside like a butterfly" (E. M. Forster). (5) In śabda vyatireka, the comparative particle expresses a nominal similitude, that is, a similitude of two things through a common property. This same distinction was made for simile (upama). It pervades the thinking of the Indian aesthe- ticians: cf. vākyârthôpama. Here the form of vyatireka is overlaid, but the distinction (artha-sabda) concerns only the terms in respect of their comparability. In Forster's example, an injunction forbids us to push aside art like we push aside a butterfly (note the repetition of the verbal idea when the figure is spelled out; cf. vākyârtha), but the difference in our attitude is meaningful only insofar as the terms are comparable. In ākșipta vyatireka, the comparative particle is not expressed. The form is artha in the commentary ahetu, 'without the cause': (1) same as nimittâdrsti vyatireka. (2) M 160. (5) See sahetu vyatireka. ākșipta, 'implied': (1) a type of vyatireka in which the force of the distinc- tion is carried by circumlocutions which avoid the comparative particle entirely. (2) M 160. (3) iyam sunayanā dāsīkṛtatāmarasaśriyā / ānanenâkalankena jayatîndum kalankinam (Mammața: "This girl of lovely eyes and beauty which enslaves the day lotus-she conquers the spotted moon with her unblemished face"). (4) "Bacchus must now his power resign-/ I am the only God of wine!" (Henry Carey; instead of: Bacchus is unlike me, for I am a drinker). (5) In pratīya- mãna, it is the similitude which is understood, not the comparative particle. Cf. artha, śabda-the other members of Mammața's triad. ākșepa, 'refusal': (1) a type of vyatireka in which the underlying similitude is admitted only tentatively, as a prelude to showing the difference (refuting the similitude). (2) D 2.186 (187). (3) sthitimān api dhīro'pi ratnānām ākaro'pi san | tava kakșām na yāty eva malino makarâlayah (Dandin: "Though it is steadfast and a mine of jewels, the ocean cannot be compared to you, O King, for it is dark colored"). (4) "Her sacred beauty hath enchanted heaven; / And, had she liv'd before the siege of Troy, / Helen, whose beauty summon'd Greece

GLOSSARY 278 to arms, / And drew and thousand ships to Tenedos, / Had not been nam'd in Homer's lliads" (Christopher Marlowe; Tamburlaine praises his dead Zenocrate). (5) Here the interest resides in the form the differentiation takes: we have only concessive similitude, so that the distinction has the aspect of a refutation of that similitude. Usu- ally, the similitude is unflinchingly admitted. Marlowe, however allows Helen her virtues only so long as the accidents of time are respected: the similitude of Zenocrate and Helen is in fact a fiction whose only purpose is to lead the unsuspecting reader to a more forceful perception of the pre-eminence of the former. In the Sanskrit example, the similitude is introduced by the word api, the discrimina- tion by eva. This has the advantage of making the concession explicit. ādhikya. 'superabundance': (1) a type of vyatireka in which both the pre-eminence of the subject of comparison and the defectiveness of the object in respect of the criterion of differentiation are stated. (2) D 2.192 (191). (3) abhrūvilāsam asprstamadarāgam mrgêkșanam | idam tu nayanadvandvam tava tadguņabhūșitam (Dandin: "The doe's glance is unrelieved by coquettish brows, untouched by love's heady passion : your two eyes redeem this lack"). (4) "Mr. Irwine was like a good meal o' victual, you were the better for him without thinking on it, and Mr. Ryde was like a dose o' physic, he gripped you and worreted you, and after all he left you much the same" (George Eliot; two preachers are being distinguished; note the subordinate comparisons). (5) In bhedamātra, the bare distinction is made between the two things, and there is no expression of pre-eminence since the distinction is entirely circumstantial. Note the subordina- tion of the similes to the synchisis: a common "samsrsti" in English. ubhaya, 'both': (1) a type of vyatireka in which the point of difference is made explicit for both things being distinguished. (2) D 2.184 (183). (3) abhinnavelau gambhīrāv amburāśir bhavān api / asāv añjanasañ- kāśas tvam tu cāmīkarad yutih (Daņdin: "You, O King, and the ocean both are deep and limitless, but it has the appearance of collyrium and you that of gold!"). (4) "Old black rooks flapping along the sky and old black taxicabs flapping down the street" (Joyce Cary). (5) Cf. eka vyatireka, where the difference of one term only is given. eka, 'one': (1) a type of vyatireka in which the point of difference is made explicit for only one of the things being distinguished. (2) D 2.182 (181). (3) dhairyalāvaņyagāmbhīryapramukhais tvam udanvatah | guņais tulyo'si bhedas tu vapușâivêdrśena te (Dandin: "You are indeed similar to the ocean in steadfastness, kindness [saltiness] and depth;

GLOSSARY however. you differ as to your aspect"). (4) "'I like the front 279

window'. I said. 'for the view. As good as a Rolls. Better; higher and not the same responsibility not to run over the poor'" (Joyce Cary; in a London bus). (5) There is, of course, no mystery about how the distinction applies to the other term (it being only the con- verse); the point is that for that term, the appropriate qualification is only suggested (the ocean has another form; the driver of the Rolls has a responsibility). Cf. ubhaya. gamyamāna, 'being understood': (1) same as pratīyamāna vyatireka. jati, 'genus': (1) a type of vyatireka in which an instance, usually figurative, (2) V 4.3.22C.

is distinguished from its own genus by pointing out the sense in which it is figurative. (2) D 2.198 (197). (3) aratnâlokasamhāryam ahāryam sūryaraśmibhih | drstirodhakaram yūnām yauvanaprabhavam tamah (Dandin: "The passionate darkness born of youth which obstructs the vision of the young is impenetrable even to the clear brilliance of jewels and cannot be dispelled by the rays of the sun"). (4) "He had merely meant to express his feeling that the streams which ran through their veins were not yet purified by time to that perfec- tion, had not yet become as genuine an ichor, as to be worthy of being called blood in a genealogical sense" (Anthony Trollope; here common blood is distinguished from noble blood, but the principle is the same). (5) This amounts to specifying the two senses of a legitimate double-entendre, insofar as one meaning is literal and the other figurative. Cf. ślesa, where the duplicity of meaning is founded upon a pun, not a double-entendre. drstânta, 'example': (1) a type of vyatireka in which the distinction is cast in the form of proposition and example, and where the example involves the contrary property or aspect. (2) U 2.7. (3) śīrņapa- mņâmbuvātâśakașțe' pi tapasi sthitam | samudvahantīm [Umām] nâpū- rvam garvam anyatapasvivat (Udbhața: "Umā was not overborne by unprecedented pride like other ascetics, whose privations were made difficult by a diet of withered leaves, water, and air"). (4) "I may be nothing but an old failure, having muffed just about everything l ever put my hand to; I seem to have the Midas touch in reverse" (Saul Bellow). (5) Udbhata says that this figure is the same as vaidharmya drstânta. This is a curious point, because Udbhața is the first known writer to mention drstânta as a figure, and among the later writers, only Mammata recognizes a vaidharmya subtype. Of course, all the figures ultimately go back to sources unknown

280 GLOSSARY to us (in the sense, at least, that the writers never tell us when they are innovating), but here is an cxample of a notion refuted before it is stated. The issue is not major, for it reduces to a decision as to which of the two figures (drstânta, vyatireka) here com- bined is primary, and that seems to depend on the intention of the speaker. In the example from Saul Bellow, the distinction seems more important than the illustration (he is not looking around for un mot juste at any rate). nimittadrsti. 'whose cause is evident': (1) a type of vyatireka in which a cause explains the pre-eminence of the subject. (2) U 2.6. (3) padmaņ ca niśi nihśrīkam divā candram ca nisprabham / sphuracchā- yena satatam mukhenâdhah prakurvatīm [Umām] (Udbhața; Umā's face surpasses both the lotus and the moon because it is beautiful both by day and by night: "Uma, conquering with her ever- radiant face the lotus, at night unlovely, and the moon, dull by day"). (4) "Return sweet Evening ... / Not sumptuously adorned, nor needing aid, / Like homely featured Night, of clustering gems; / A star or two, just twinkling on thy brow, / Suffices thee" (William Cowper). nimittâdrsti, 'whose cause is not evident': (1) a type of vyatireka in which the cause of the subject's pre-eminence is only hinted at. (2) U 2.6. (3) sā gaurī śikharam gatvā dadarśômām tapaḥkrśām / rāhupītapra- bhasyêndor jayantīm dūratas tanum (Udbhața; the explanation-the eclipse-is not given: Rähu is unable to obscure the moon for long, whereas Uma's pale and intense tapas is constant; "Gauri, gone to the mountain peak, saw Uma, frail through penance but conquering from afar the beauty of the pale, eclipsing moon"). (4) "I think that I shall never see / A billboard lovely as a tree. / Perhaps unless the billboards fall, / I'll never see a tree at all" (Ogden Nash; why trees are preferable to billboards is left to our imagination). (5) Udbhața subdivides vyatireka only in this way (cf. nimittadrsti); now, any distinctive qualification can be taken as a cause of dis- similitude or pre-eminence, but it seems that we should take Udbhata more literally than that, and see his cause to be an effective and not merely formal cause. Eka and ubhaya vyatireka concern distinctive qualifications only as forms. In hetu vyatireka, the discrimination is itself given the form of a cause; in the present case, the cause is taken to mean the explanation of that discrimination. pratiyamāna, 'being understood, implicit': (1) a type of vyatireka in

GLOSSAR Y which the similitude which underlies the discrimination is not 281

mentioned explicitly, but where the distinetion is spelled out. (2) D 2.189 (190-98). (3) tvanmukham kamalam cêti dvayor apy anayor bhidā / kamalanı jalasamırohi tvanmukham tvadupāšrayam (Daņdin; their loveliness is not mentioned: "Your face and the lotus differ in this: the lotus grows in the water, and your face appears only on you"). (4) "A man's profession is not like his wife which he must take once and for all for better or for worse without proof before- hand" (Samucl Butler). (5) This figure is the same as gamyamāna of Vamana. Dandin distinguishes vyatireka into two broad classes- this and śabdôpādāna (explicit) and then subdivides cach type further. As examples, of ādhikya, bhedamātra, jāti. Cf. atišayôpamā, which differs from this type only in asserting that the difference given is unique. bhedamätra, 'the difference only': (1) a type of vyatireka in which the bare and circumstantial difference of the two comparable things is stated. (2) D 2.192 (190). (3) tvanmukham kamalam cêti dvayor apy anayor bhidā | kamalam jalasamrohi tvanmukham tvadupāśrayam (Daņdin; see the translation under pratīyamāna). (4) "When a man bleeds inwardly, it is a dangerous thing for himself; but when he laughs inwardly, it bodes no good to other people" (Charles Dickens). (5) This figure is to be distinguished from ādhikya, in which the differentiation of the two terms involves the assignment of pre- eminent and defective status, respectively. Here the differentiation is entirely circumstantial, eschewing such judgements. Cf. catu upamā, which simile consists in overlooking the difference. mālā, 'garland': (1) a type of vyatireka in which the subject of comparison is distinguished from several objects in terms of several properties; a series of vyatirekas having the same subject. (2) M 160C. (3) haravan na vişamadrsțir harivan na vibho vidhūtavitatavrşah | ravivan na câtiduhsahakaratāpitabhūh kadācid asi (Mammața: "Unlike Śiva, you have but two eyes, O King. Unlike Krsna, you have not overthrown the great bull demon (viz. vice personified as a bull). Unlike the sun, you do not beat down upon the earth with rays [hands] of great violence"). (4) "-But your voice, -never the rush- ing / Of a river underground, / Not the rising of the wind / In the trees before the rain, / Not the feet of children pushing / Yellow leaves along the gutters / In the blue and bitter fall, / Shall content my musing mind / For the beauty of that sound" (Edna St. Vincent Millay). (5) This category is introduced by Mammata only to show

GLOSSARY 282 the futility of such classifications; cf. artha, sahda, aksipta, hetu, and elements of the structure of vyatireka. vyasta, 'separate': (1) same as eka vyatireka. (2) R 7.86 (87). (5) Cf. śabda, 'literal': (1) a type of vyatireka in which the comparative particle samasta. is present expressing a similitude of two nominal ideas through a common property. (2) M 160. (3) jitêndriyatayā samyagvidyāvr- ddhanişeviņah | atigādhaguņasyâsya nâbjavad bhangurā guņāh (Mam- mața: "Your virtues [garlands] are not fragile like those of the lotus, for your qualities are deep rooted and you honor the aged sages by your great self-control"). (4) "Men's muscles move better when their souls are making merry music, though their merriment is of a poor blundering sort, not at all like the merriment of birds" (George Eliot). (5) Cf. artha; the distinction is the same as that applied to simile (cf. ārtha, sābda) and concerns only the interpreta- tion of the comparative particle. For a full discussion, see vākyâ- rthavrtti. In the example offered under artha vyatireka, "like" ex- presses a verbal similitude ("pushed aside"), here a nominal similitude ("poor blundering sort"). The distinction is linguistic and not absolute: the former could be expressed nominally ("as light as"), the second verbally ("blunders"). The form śābda is found in the commentary. śabdôpādāna, 'literal': (1) a type of vyatireka in which the similitude which underlies the differentiation of the comparable things is spelled out. (2) D 2.180-88. (3) tvanmukham pundarīkam ca phulle surabhigandhinī | bhramadbhramaram ambhojam lolanetram mukham tu te (Dandin: "Your face and the lotus are full blown and sweet smelling, but the flower is surrounded by wandering bees and the face has a coquettish glance"). (4) "All human things are subject to decay, / And, when Fate summons, Monarchs must obey: / This Flecknoe found, who, like Augustus, young / Was called to Empire and had govern'd long: / In Prose and Verse was own'd without dispute / Through all the realms of Non-sense, absolute (John Dryden). (5) Cf. pratīyamāna, where the similitude is only implied. See kalpitôpamā. ślesa, 'double-entendre': (1) a type of vyatireka in which the similitude on which the distinction is founded is punned upon. (2) D 2.186 (185), M 160. (3) akhandamaņdalah śrīmān paśyâișa prthivīpatih / na niśākaravaj jātu kalāvaikalyam āgatah (Mammața: "Regard that great Lord of prosperous court [whose disc is full] and who embodies

GLOSSARY good fortune [beauty]; unlike the full moon, he need never suffer 283

decrepitude [enter the waning phase]"). (4) "a politician is an arse upon / which everyone has sat except a man" (c. e. cummings). (5) In Dandin's example, the qualifications which serve to discriminate the two things are also punned: "tvam samudras ca durvārau mahā- sattvau satejasau | ayam tu yuvayor bhedah sa jadâtmā pațur bhavān" ("You and the ocean, O King, are indomitable [uncrossable], of great character [containing many substances], violent [stormy]; this, however, is the difference between you: the ocean is cold [stupid] souled; you, however, are acrid [keen witted]"). Compare: "When I am dead, I hope it may be said: / 'His sins were scarlet, but his books were read'" (Hilaire Belloc). sadṛśa, 'similar': (1) a type of vyatireka in which the distinction is itself cast in terms which suggest comparable aspects of the two things. (2) D 2.192 (193-95). (3) candro'yam ambarôttamso hamso'yam toyabhūşanam / nabho nakşatramālîdam utphullakumudam payah (Dandin; the four terms are distinguished, but all are expressed as ornaments of the distinguishing feature: "The moon is an ornament of the sky, the swan adorns the lake; the sky has a necklace of stars, the lake is abloom with lotuses"). (4) "That punctual servant of all work, the sun, had just risen, and begun to strike a light on the morning of the thirteenth of May, one thousand eight hundred and twenty-seven, when Mr. Samuel Pickwick burst like another sun from his slumbers, threw open his chamber window, and looked out upon the world beneath" (Charles Dickens). (5) This is instead of stating the distinction as an adjunction to the description of the similarity of the two things (cf. śabdôpādāna). samasta, 'conjoined': (1) same as ubhaya or ādhikya vyatireka. (2) R 7.86 (88). (5) Cf. vyasta. Rudrața's classification stops at this distinction. His intention is probably closer to the ādhikya of Daņdin, since there it is a question explicitly of pre-eminence and inferiority, not just of distinctive qualifications applied to both terms. sahetu, 'including the cause': (1) same as nimittadrsti vyatireka. (2) M 160. (5) Mammata distinguishes four types, according to whether the cause for the discrimination is given for both, for one or the other, or for neither of the two compared things. He offers examples only for the first category. hetu, 'cause': (1) a type of vyatireka in which the distinctive qualification is cast in the form of a cause (of that difference). (2) D 2.186 (188). (3) vahann api mahīm krtsnām sašailadvīpasūgarām | bhartrbhāvād

GLOSSARY 284 bhujangānam seșas tvatto nikrsyate (Dandin: "Although you both support the whole world with its islands and mountains and seas. you surpass the great Serpent who is only the first among snakes"). (4) "In Germany the Reformation was due to the passionate convic- tion of Luther. In England it was due to a palace intrigue" (E. M. Forster). (5) Cf. nimittâdrsti. vyatireka (II): (1) a vyatireka in which the expression of difference focuses upon the pre-eminence of the object of comparison. (2) R 7.89 (90). (3) kşīņah kşīņo'pi śašī bhūyo bhūyo vivardhate satyam / virama prasīda sundari yauvanam anivarti yātam (Rudrața: "The waning moon grows less and less, but will become again complete. Be kind, O lovely, and leave off anger; youth once gone does not return"). (4) "Still do the stars impart their light / To those that travel in the night; Still time runs on, nor doth the hand / Or shadow on the dial stand; The streams still glide and constant are: / Only my mind / Untrue I find Which carelessly / Neglects to be / Like stream or shadow, hand or star" (William Cartwright). (5) Though this could easily be included in the first variety of vyatireka, it is worthy of note that of the thirty or forty examples offered in the different texts, only this one of Rudrata fails to exhibit the usual exaggeration and consequent flattery of the subject of comparison (in the present case, the mind). Of course, in English poetry, such instances are if anything more frequent than the flattery, and several have been included as examples of subtypes of vyatireka I where the point at issue is formal and does not contest with the intention of the speaker (cf. hetu, śabda). Rudrața's late arrival does, however, pose one interesting problem of interpretation. If the intention of the speaker is flattery, the vyatireka amounts to an elliptical simile, for the differ- ential qualifications all operate to the disadvantage of the object or the advantage of the subject, and thus reinforce their basic similitude by cancelling the natural relation of the subject (which shares the property of comparison to a lesser degree) and the object (to a greater). However, once the discrimination of the two terms is divested of this function, the character of the figure assumes only secondary importance, for it then acquiesces in the natural mode of expression of its own terms and becomes either a matter of fact distinction or an anti-simile, as above. It cannot be said that the intention of the poet is to compare mind and stream, etc., in the same sense as that in which moon and face are habitually compared; his intention is rather to express the peculiarity of the subject, not

GLOSSARY ironically (cf. asadharana upama), but factually. In this effort, the 285 underlying comparability of the distinguished terms is only a con- dition or a means to the end of that distinction, and far from assum- ing primary importance (through irony), it is effective only as it remains just below the level of interest, as the basses in an orchestra. vyāghāta väghata. 'obstacle': (1) a figure in which a cause, for some reason obstructed in its operation, is carried through to its effect in another way by another cause which is a modified version of the obstruction. (2) M 206. (3) drśā dagdhamı manasijam jīvayanti drsâiva yāh | virūpâkşasya jayinīs tāļı stuve vāmalocanāh (Mammața; Love, consumed by the fire of Siva's glance, is reanimated by the glances of the young ladies: "I praise their lovely, conquering eyes in whose glances the God of Love again lives, whose body was consumed by another's glance"). (4) "Catherine rushed to congratulate him, but knew not what to say, and her eloquence was only in her eyes. From them, however, the eight parts of speech shone out most expressively, and James could combine them with ease" (Jane Austen). (5) The cause, as it were, is shown getting the best of its hindrance by taking a form whereby that hindrance can be turned to its own ends. This figure is neither viśesôkti, because the first cause, though obstructed, does contribute to the final effect, nor is it asamgati, because a mediating cause is introduced to explain the effect. vyāghāta (II): (1) same as viśeșôkti II. (2) R 9.52 (53). (5) There may be a slight difference in that the absence of obstruction is specifically alleged.

vyāja vyāja, 'pretence': (1) an artha ślesa in which praise is understood from apparent blame, or vice versa. (2) R 10.11. (3) tvayā madarthe samupetya dattam idam yathā bhogavate śarīram | tathâsya te dūti kṛtasya śakyā pratikriyânena na janmanā me (Rudrața: spoken by a girl to her messenger after learning that her lover found the messenger a bit too pliable and not uniquely concerned with arranging the rendez-vous: as apparent praise, bhogavate is taken as 'lover' and pratikriyā as 'recompense'; but as blame, as 'snake' and 'vengeance'. The snake image is further suggested by the messenger's returning with [the lover's] teeth wounds on her lips; "You have gone for my sake and given your body to my lover [serpent], O procuress; I will

GLOSSARY 286 never be able in my life to repay you [get even with youl"). (4) "When other fair ones to the shades go down / Still Chloe, Flavia, Delia, stay in town: / Those ghosts of beauty wandering here reside, / And haunt the places where their honour died" (Alexander Pope; the pun is on two kinds of death). (5) Both examples show apparent praise which is blame. This ślesa differs from vyājastuti alamkāra only in that a pun is at the bottom of the irony. It is perfectly possible, of course, to express such pretence of praise without actually employing puns. vyājastuti vyājastuti, 'deceptive eulogy': (1) a figure in which apparent blame con- ceals real praise or appreciation. (2) B 3.30 (31), D 2.343-47, V 4.3.24, U 5.9, R 10.11 (12-13), M 169. (3) pumsah purāņād ācchidya śrīs tvayā paribhujyate / rājann ikșvākuvamśyasya kim idam tava yujyate (Dandin; Śri is the wife of Vişņu: "O King, you have stolen Śri [prosperity] from her former spouse and enjoyed her! Can this deed be condoned in the scion of the Iksvāku clan?"). (4) "The poor man's sins are glaring; / In the face of ghostly warning / He is caught in the fact / Of an overt act-/ Buying greens on Sunday morning" (T. L. Peacock; really a defense of the lower classes). (5) The earlier writers consider only the case of blame concealing praise (although Dandin in his unique interpretation of the figure aprastutapraśamsā, considers it to be blame concealed as praise of something irrelevant). But Rudrata and Mammata extend the figure and take account of the other possibility-that of praise concealing blame: "tvayā madarthe samupetya dattam idam yathā bhogavate śarīram | tathâsya te dūti kṛtasya śakyā pratikriyânena na janmanā me (Rudrața; for translation, see vyāja); or: "Lament him, Mauchline husbands a', / He often did assist ye; / For had ye staid whole weeks awa', / Your wives they ne'er had missed ye" (Robert Burns; "praise" of the gay deceased). Rudrata groups this figure with those based upon a pun. See vyāja.

vyājôkti, 'pretexť': (1) same as leśa I. (2) V 4.3.25, M 184. vyājôkti

śabda, 'word': (1) a generic term for those figures whose poetic effect śabda

GLOSSARY 287 depends upon some aspect of their grammatical, phonemic, or metrical form, and not upon the idea they convey. (2) B 1.15-16, D 3.186, V 1.1.1, 4.1.1 (vrtti), U 5.12, AP 342.18-19, R 2.13, M 103C. (5) See artha. The common examples of śabdâlamkāra are anuprāsa (alliteration), yamaka (word play, cadence), certain kinds of punning (śleşa), prehelikā (conundrum), and the citra, or verses arranged in imitation of visual forms. Only the Agni Purāna deviates significantly from this canonical list, by adding six figures which have no parallel except in much later works (Sarasvatīkanthâbharana). They are (a) chāyā, imitation of specific verbal styles, (b) mudrā, the poet's ability to translate his intentions into plausible situations. (c) ukti, use of injunctions, (d) yukti, figurative usage, (e) gumphanā, ability to construct the narrative, and (f) vākovākya, conversation. These notions, which have to do with the technical skills of the poet, are matched by certain stylistic figures which the Agni Purāna treats under śabdârthâlamkāra and which relate to the congruence and appropriateness of the different aspects of the poem to one another (see praśasti, kānti, aucitya, samkșepa, yāvadarthatā, and abhivyakti). These irregular notions of poetic style have been incorporated into the traditional treatment of the figures. Anuprāsa and the others are treated as the seventh, eighth, and ninth sabdâlamkāra. The five figures ākșepa, aprastutastotra, samāsôkti, apahnuti, and paryāyôkta are considered subtypes of abhivyakti. The arthâlamkāra are the usual ones, mainly based on simile. A fusion of several different systems is indicated also by the fact that the Agni Purāna, despite this incorporation of style into the figures, treats at length (though again with innovations) the traditional topic of style (kāvyaguņa). One figure (yathāsamkhya) finds its way into this category. This early attempt at broadening the notion of verbal figure was not, however, viewed with favor by later writers. With the triumph of the dhvani school, an eclipse of the śabdâlamkāra is evident, at least among the poeticians. Mammata goes so far as to view even yamaka as a citra. The previously elaborate classifications of yamaka and anuprāsa are reduced or are even ignored entirely (Rasagañgâdhara). This is doubly curious, since the poetry written during this period (eleventh-sixteenth centuries) is, for the most part, alleged to depend upon such verbal devices.

sabdârtha, 'word-sense': (1) a generic term used to indicate those in- śabdârtha

288 GLOSSARY determinate figures whose poetic effect cannot be assigned exelusively to either their formal character or to the meaning they convey, but rather involves both. (2) AP 345.1, M 124. (5) Sec sabda, artha. This category in the Agni Purāna includes six figures: prasasti, kānti, aucitya, samkșepa, yāvadarthatā, and abhivyakti. All are aspects of literary style concerned with the integration of the various elements of a poem, and the most interesting is the last: abhivyakti. Here are treated the various kinds of meaning-direct, connotational, and secondary-which a word or phrase may convey, and five modes of relating the suggested meaning to the major sense of the poem are outlined. All five happen to be traditional alamkāras: ākșepa, aprastutastotra, samāsôkti, apahnuti, paryāyôkta, and in this sense they are taken as examples of dhvani. The category dhvani is not otherwise noted, and we may here be in the presence of one of the forerunner theories involving this term. Out of this may have grown the dhvani which denied all association with figures of speech as expressively necessary. It is easy to see how this notion may have arisen out of an examination of those figures of speech which do convey a second meaning, not literally expressed, as abbreviated metaphors (samāsôkti). It is, however, just as possible in theory to take the Agni Purāna as a conservative text attempting to re- integrate an already proposed dhvani (the kārikās of the Dhvanyāloka may have been approximately contemporary) into a traditional structuring of the figures.

śayyā śayyā, 'bed': (1) same as mudrā. (2) AP 342.26.

ślișța ślișta, 'punned, conjoined': (1) a variant form of śleșa.

śleșa śleșa, 'adhesion' or 'conjoined': (1) paronomasia; pun: double-entendre: the simultaneous expression of two (or more) meanings. (2) B 3.14. D 2.310, V 4.3.7, U 4.9, R 4.1, 10.1, M 119, 147. (3) akrstâmala- mandalâgrarucayah samnaddhavak şahsthalāh soșmāņo vraņino vipa- kşahrdayaprônmāthinah karkašāh | udvrttā guravaš ca yasya vašinah śyāmāyamānânanā yodhā māravadhūstanās ca na dadhuh ksobham sa vo'vyāj jinah (Vamana; all the descriptive adjectives have one sense when taken with the noun "warriors", another when taken with

GLOSSARY "women's breasts": "May he alone conquer you who is unaffected 289

by soldiers and the breasts of lovely women, brilliant with their spotless drawn scimitars [the nipples of whose spotless circles gleam when revealed], whose chests are girded [whose chests are full blown], lively [heaving], scarred [wounded], striking fear into the hearts of their enemies [producing agitation in the hearts of their rivals], rough [firm], arrogant [swollen], stolid [heavy], and whose faces are dark with rage [whose tips are dark]"). (4) "Beneath in the Dust, the mouldy old Crust / Of Moll Batchelor lately was shoven, / Who was skill'd in the Arts of Pyes, Custards and Tarts, / And every Device of the Oven. / When she'd liv'd long enough, she made her last Puff, / A Puff by her Husband much prais'd; / And here she doth lie, and makes a Dirt Pye, / In hopes that her Crust may be rais'd" (Anon .; the puns here relate the kitchen and the tomb). (5) The confusion of terms is here significant and oddly appropriate to the concept involved; the first two writers to define pun, Bhāmaha and Dandin, prefer the term ślista, perhaps to distinguish double- entendre from the well-known guna (stylistic quality) "śleșa" (usually defined as the appropriateness of word and sense) which dates at least from the Nātyaśāstra. The two terms thus contrasted are only grammatically different, the former being a past participle, 'con- joined', and the latter a derived noun, 'conjunction'. The first writer to confuse the issue was Vāmana, who uses "śleșa" only. Udbhata follows Bhāmaha's lead, but does not treat the gunas in the part of his work we have. The Agni Purāna describes the guņa "śleşa", but ignores the alamkāra. Rudrața does not consider the guņas, but calls the alamkāra "śleșa". Mammața, to complete the confusion, refutes the guna "śleșa" (he accepts only three guņas), but follows Rudrata in describing śleșa alamkāra. On the level of the idea, we fare little better. It is obvious that all the writers are talking about the same thing, but their definitions, though having some elements in common, vary widely. Ślesa is subdivided in extremely different ways. Its relation to the other figures is only partially described. Śleşa is the associating figure par excellence: it is found in conjunc- tion with a long list of other figures, notably upamā, rūpaka, arthân- taranyāsa, vyatireka, vakrôkti, ākșepa, dīpaka, sahôkti, hetu, etc. With these figures, the ślesa occupies the subordinate place and serves as the means whereby the idea of the figure is expressed. For ex- ample, in upamāśleșa the pun replaces the common property as that

290 GLOSSARY element in terms of which the subject and object are compared ("Why is a woman like a hinge? Because she's a thing to adore"; M. E. W. Sherwood). It would appear that ślesa is at least compatible with every figure except svabhāvôkti and embodies more fully than any other figure the universality of the poetic idea. It comes closer than any other figure to the essence of poetry, variously expressed as "vicchitti", "vaicitrya", "vakrôkti", etc. ('strikingness', 'charming- ness', 'circumlocution'), inasmuch as no expression containing a śleșa can ever be regarded as a mere statement, an apodiction. Theoretically, too, the idea of a pun, in which two different meanings inhere in the same phonemic span, is an employment of language which transcends the character of language, at once demonstrating the limitations of non-poetic speech and using those limitations to another purpose. Punning in Sanskrit is not the weak-kneed and self-conscious crudity that it is in most Western languages. A pun involving only one word or a single idea is an object of indifference to the Sanskrit writer; it becomes interesting only when whole ideas and complicated concepts are double-entendus. A person who knows only English or Greek simply cannot appreciate the elegance and refinement of punned verse as written in India, the reading of which resembles nothing so much as a continuous re-awakening of one's own apprehensive powers. Puns are very sharply distinguished from the pointless word play which deforms so much light verse in English and which, in Sanskrit, is relegated to the category yamaka, or cadence. Ogden Nash, for example, abounds in usages which depend upon the associative power of partial or incomplete repetition ("In spite of her sniffle / Isabel's chiffle"), or Thomas Hood's: "But from her grave in Mary- bone / They've come and boned your Mary". These are not puns, because the expression of the ideas is not simultaneous. All that happens is that two ideas totally unrelated are here juxtaposed through the extraneous (and even forced) similarity of the words which carry them. They are nascent puns, if you like, or inchoate puns which the author has left unclothed with their bare unrelatedness showing. Simultaneity of expression is at once the formal criterion and the condition of delectation of double-entendre. It is the one feature which all writers are able to agree upon. Formally, then, ślesa is most characteristically differentiated from yamaka; but in terms of its content (the two ideas conjoined), the figure which is most often taken as crucial is rūpaka (metaphor).

GLOSSARY 291 In fact, both Bhāmaha and Vāmana define śleșa in relation to rūpaka, for has this latter not been already defined as the imposition of an object onto a subject-that is, the identification as though there were coalescence in reality of a subject of description with a similar object, as eye-lotus or face-moon? The answer is again obvious and is the same answer: ślesa may be taken as a rūpaka with simultaneous expression of subject and object. The treatment of ślesa in terms of rūpaka is, however, far too narrow to do justice to the concept of double-entendre, for since rūpaka itself is an identification only of similars, this view seems to imply that ślesa, too, is at bottom a comparison (upamā) or can, at any rate, be described in the same context (cf. Bhāmaha's definition "upamānena yat tattvam upameyasya sādhyate ... "). But it is simply not true that every ślesa rests on an implicit comparison, for there are puns, in the extreme case, which involve words of different form classes or even do not involve the same words (as: "The moon- shine's bright on my Old Kentucky Home", or: "Focus, where the Sons raise Meat" as the name of a ranch). It is, of course, true that many double-entendres (especially the best) do repose upon such an implicit comparison; for example: "But then, you know, you stand upon / Another footing now" (see asambhava ślesa); but the concept is not wide enough. Hence later writers follow Dandin in not committing themselves on the subject: a pun is simply 'two senses expressed in the same words' ("anekârtham ekarūpânvitam vacaḥ"). The dhvani writers are understandably preoccupied with śleâ- lamkāra; as defined, it borders precariously upon the domain of suggestion. In fact, one of the most important functions of śleșa is to suggest (vakra, etc.). We cannot enter here into the niceties of dhvani theory, but in general it can be remarked that the suggested sense of that system is not, strictly speaking, expressed (in words) at all; not only is the most characteristic dhvani [rasadhvani] "in- expressible", but even those evocations of discrete ideas and things which are considered subordinate are not denotatively comprehended in the utterance as the two meanings of a pun always are. The distinction here rests upon a theory of signification which all ālamkārikas are far from sharing and which is not entirely germane to the subject of poetic figures. Hence we leave the matter at that, remarking only that here again ślesa occupies a crucial position in the history of aesthetic speculation. Śleșa cannot ultimately be defined in relation to the content of

292 GLOSSARY the figure (that is, the idea it cxprevses), since, as has been noted. ślesa can be associated with almost any other higure not merely in the sense of adjunction of two figures, but as an cssenhal elament in the expression of that other figure's idea. That is why mow later writers resort to what amounts to a formal, or grammatica, Jan: fication of subtypes. The earliest writer, Bhamaha, who treats slista as a variety of metaphor (rupaka), gives two independent classifications; the ienta; of the subject and the object can be expressed in any of three aspeets: adjectival qualification (guna), mode of activity (kriya), or essence (nāma). (These correspond to the grammatical triad adjective, verb, and noun, but Bhämaha's intention is not to equate them with the formal categories, at least in the sense that he treats primarily the aspects of the thing thereby expressed.) Secondly, Bhāmaha notes that ślista can be associated with three other alamkūras: sahôkti, hetu, and upamā. Oddly, he gives only three examples for these six types, each exemplifying one term of each triad (although the triads have no relation to one another). Dandin also gives two classifications, and one of them recognizes the formal aspect of ślesa, for the phoneme span which expresses the double-entendre may consist of the same words for both senses (abhinnapada), or of different words (bhinnapada). (Compare "another footing now", using the same words, with "moonshine's bright", using different words.) This distinction is essentially the same as the more popular later version artha-sabda, but the elabora- tion which this has received has somewhat clouded the issue (for example, see varnaśleșa, where it is not at all clear that vibhau would be taken as an example of bhinnapada śleşa by Daņdin, despite its being considered sabdaśleşa by Rudrața). We might say. in modern parlance, that the "morpho-phonemics" of the two spans are either the same or different. As an aside, it should be noted that most writers (beginning with Udbhata) specifically exempt intena- tion as a consideration in determining this difference. It is mevitabk that many compounds will be differently accented as they are taken in different senses, but unless the analysis also reveals a morpheme difference in the line (doşa-kara, vs. dosakara), these compounds are not considered different. Dandin's other classification is also formal, but is cxtnnse to the double-entendre strictly speaking: it assumes a standard form for the pun, which is on qualitications of two (given) nouns. These

GLOSSARY two (hence unpunned) nouns may construct with one verb [abhin- 293 nakriya], with two verbs which are different in sensc [aviruddhakriyā], or with two verbs which are contrary in sense [viruddhakarman]. Neither the nouns or the verbs are punned; the puns reside only in the adjectives pertaining to the nouns. Having thus introduced the verb into the system, Dandin proceeds to state the possible relations between the two senses of the pun: they may be entirely different, but consistent (avirodhin), contradictory (virodhin), they may be two aspects of the same concept (niyamâkseparūpa), or, finally, an aspect and the concept itself (niyamavat). Dandin is the only writer who shows metaphysical ability in the organization of his classifications; this one is extremely interesting in laying bare the generic kinds of oppositions which any two meanings may have: they will either be indifferent or generically related; if indifferent, then either compatible or contradictory, if generic, then either as species or as genus and species. Needless to say, this categorization was not followed by later writers, whose interests become more and more formal and verbal. But it does have an interesting parallel in the ten types of artha ślesa which Rudrata describes. These, however, relate to the kinds of situations in which ślesa can be used and suggest a context similar to that of the dhvani theory, rather than involving the struc- ture of meaning itself. Vāmana does not subdivide śleșa. Udbhața is the first to use the terms śabda and artha ślesa, probably in the sense described above, but since no explanation is contained in the text, we have only the opinions of commentators to go on, all of whom were posterior to the writers (notably Rudrata) who developed śabda and artha ślesa in great detail. There is always a tendency in such cases to read back into the text the later opinions. With Rudrata, the idea of double-entendre reaches its apogee. He recognizes the distinction of ślesa into śabda and artha, but raises each to the status of an independent subject. Chapter Four of his work is devoted to the former and Chapter Ten to the latter. More- over, artha ślesa represents one of the four large categories into which artha alamkāras are divided, the others being 'descriptive' (vāstava). 'comparative' (aupamya), and 'hyperbolic' (atiśaya). The division implies that artha ślesa is both comparative and hyperbolic (since "descriptive" is neither comparative nor hyperbolic) and as such represents for Rudrata the most poetic as well as the most interesting case. Such an inference is not entirely without plausibility, inasmuch

GLOSSARY 294 as an artha śleșa does bring together two ideas in such a way that one heightens the effect of the other (hyperbole). This point is made clearer in the discussions which accompany the individual figure, It should be emphasized that mere punning, that is, punning which is not based on the meanings involved and which docs not aim at an end to which those meanings are relevant, is definitely unacceptable here. In this sense, too, the concept of artha (ślesa) has acquired a positive content; in Dandin, it was a residual category (a-bhinna- pada): that pun which cannot be explained by grammatical exegesis. The ways in which the two meanings can complement one another are ten: (a) they can be qualified in the same way (aviśesa), (b) or not (virodha); (c) the second may constitute flattery of the first (adhika); (d) the pun may suggest a further contrast of mood (vakra); (e) one may be flattery, the other reproof-ironic praise (vyāja); (f) the second meaning may be risqué (ukti); (g) the two meanings may be similar but take contradictory qualifications (asambhava); (h) the second meaning may augment the force of an adjectival description (avayava), or (i) the force of the names themselves (tattva); (j) and, lastly, if they cannot be qualified in the same way (case b), the contradiction in qualification may be a function of the meanings of the terms, rather than a question of mere negation (virodhâbhāsa). Rudrața is not less inventive with regard to śabda śleșa. Given that the two meanings must relate to a different morphemic analysis of the common span, that difference can be specified in terms of the kinds of morphemes (form classes) which are thus confused. Rudrața gives eight examples, starting with the syllable (varna; this alone is sub-morphemic), stem (pada), gender-indicating suffix (linga), verb root (prakrti), nominal affix (pratyaya; excluding inflec- tions), inflection (vibhakti; both nominal and verbal), and number- indicating suffix (vacana). It is clear that the subdivisions of artha are not incompatible with those of śabda, but it would indeed be rare for an example to be found showing both grammatical and contextual precocity to such a degree. Nothing within the realm of human experience is beyond the power of the Sanskrit language, but we feel that here, at least, we are straining at the limits. Mammața repeats Rudrața's classification of sabda slesa, but he no- tices only one "śleşa" in his chapter on arthâlamıkāra. Many of the oth- ers, however, have been treated as separate figures (e.g., virodhâbhāsa).

GLOSSARY 295 sbhinnakriya, 'identical verb': (1) a type of paronomasia in which the verb does not participate in cither sense of the double-entendre, but satisfies both indiscriminately. (2) D 2.314 (316). (3) vakrāh svabhāvamadhurāņ šamsantyo rāgam ulvaņam | drśo dūtyaś ca karșanti kāntābhih preşitāh priyān (Dandin: "Sidelong [devious], by nature affectionate and making declaration of heated passion, both glances and go-betweens are attractive when sent by the beloved to the lover"). (4) "Two dozen cows, knee-deep in grass, / I saw, and twenty-seven goats, / And heard a hundred sparrows pour / Upon a bank ten thousand notes. / And, though I've seen the golden notes / That rich men pour in city banks, / And know the sparrow's note is 'cheep', / I lifted up my heart in thanks" (Samuel Hoffenstein). (5) Here the verbs karsanti and "pour" go with either of the subjects whose qualifications are punned upon: in the English, the "bankers" or the "sparrows" pour "notes" into or upon "banks". Cf. viru- ddhakriyā and aviruddhakriyā, where each of the senses of the pun demands a different verb. Note that though identical, the verb itself is not here punned upon; compare Harry Graham's reproachful "Baby roused its father's ire / By a cold and formal lisp. / So he placed it on the fire / And reduced it to a crisp. / Mother said, 'Oh, stop a bit! / This is overdoing it!'" where the pun is the verb. abhinnapada, 'same words': (1) a type of paronomasia in which the morph- emic analysis implicit in both meanings is the same. (2) D 2.310 (311). (3) asāv udayam ārūdhah kāntimān raktamandalaḥ | rājā harati lokasya hrdayam mrdubhih karaih (Daņdin: "This moon [King], mounted on the eastern mountain [having attained prosperity], lovely, ruddy orbed [whose courtiers are devoted], delights the hearts of men with his soft rays [low taxes]"). (4) "Did you hear about the moth who chewed a hole in the rug because he wanted to see the floor show? And then he covered it over because he didn't want to see the whole show?" (Lennart Rydfors). (5) Dandin does not men- tion the distinction artha| śabda ślesa, but the present two types could easily be argued into it. On the other hand, the differentiae are stated in such a way as to make possible a division such as we have discussed under artha śleșa, where two different "etyma", though the "same" word, would not exemplify artha ślesa in the strict sense, being homonyms. artha, 'meaning': (1) a type of ślesa in which the double-entendre reposes upon a legitimate duplicity of meanings within those meanings properly ascribed to the word and not upon some accidental gram-

296 GLOSSARY matical or formal feature. (2) U 4.9, 10, R 10.1, M 147. (3) svayam ca pallavâtāmrabhāsvatkaravirājinī | prabhātasamdhyêva [bhagavatī] Udbhata: the word kara is taken in two legitimate senses: 'hand' of Parvati and 'ray' of the dawn: "She is brilliant like the dawn onrushing. her hands [rays] red gleaming like new buds"). (4) "There was a young lady from Wantage / Of whom the town clerk took advantage. / Said the borough surveyor: / 'Indeed you must pay 'er. / You've totally altered her frontage'" (Anon .; "frontage" in two senses). (5) We may distinguish three levels of double-entendre or paronomasia: (a) in which two occurrences of the same etymon differ as to context. as here: (b) in which two etyma have the same phonemic shape: pun. properly speaking; and (c) in which two words. differing as to phonemic shape, are used in such a way as to suggest cancellation of that difference; plays on words, such as Ogden Nash has popularized. The same distinction may be seen in Sanskrit most clearly where type (a) is artha ślesa. It would seem that Sanskrit. so rich in natural puns, does not feel the need to distort words to obtain unnatural ones (type c); nevertheless, the same kind of discrimination can be seen first, in those puns which require a different accentuation, as muktâśrīh (on the first syllable a bahuvrīhi meaning 'by whom ugliness was dispelled'; on the last, a tatpurusa meaning 'beauty of pearls'), and secondly, in those playful puns where the word divisions of one sense are not the same as those of the other, as in the pun of Kālidāsa 'u mêti mātrā tapaso nișiddhā paścād Umâkhyām sumukhī jagāma" ("from her mother's warning her "don't, don't (do tapas)", she came to be called Umā'). These types are generally grouped indiscriminately under śabda śleșa. aviruddhakriyā, 'unopposed verbs': (1) a type of paronomasia in which separate verbs accompany each of the senses of the double-entendre. (2) D 2.314 (317). (3) madhurā rāgavardhinyaḥ komalāḥ kokilāgirah / ākarņyante madakalāḥ ślişyante câsitêkșaņāh (Daņdin: "Lovely, inspiring passion, soft and low from sipping drink, voices of night- ingales are heard and dark-eyed girls are embraced"). (5) The classi- cal form of the pun in Sanskrit is a series of adjectives or qualifica- tions which apply equally to one or another of two given subjects. Given this form, the present distinction should be understood as an attempt to include the verb within the frame of reference of double- entendre, since the two nouns may or may not be accompanied by different verbs and, if they are, the verbs may or may not be opposed in meaning. It should be noted that in these types, the verb is not

GLOSSARY 297 itself punned and hence, as far as the pun is concerned, it provides only an extrinsic, though relevant, classification. avirodhin. 'non-contradicting : (1) a type of paronomasia in which the double-entendre is suscitated by the application of a single qualifica- tion to two subjects in such a way that the qualification is understood differently for each, but without implying a contradiction. (2) D 2.315 (321). (3) mahībhrd bhūrikațakas tejasvī niyatôdayah | dakşah prajāpatiś cāsīt svāmī saktidharas ca sah (Dandin: "The King [mountain], possessing many camps [valleys], glorious, of consistent prosperity [having regular heights], clever (Daksa by name], protector of men [the father of men], Lord, and bearing a spear [powerful]"). (4) "Them young gells are like th' unripe grain; they'll make a good meal by-and-by. but they're squashy as yet" (George Eliot). (5) See virodhin. ākşepa. 'objection': (1) see under ślesa ākșepa. upamā. simile : (1) see under śleșa upamā. kriyā. 'verb': (1) a type of paronomasia in which the double-entendre resides in a verb; a pun on verbs. (2) B 3.14 (19). (3) unnatā lokada- vitā mahāntah prājyavarșiņah / śamayanti kșites tāpam surājāno ghanā iva (Bhamaha: "Good kings, like rain clouds, calm the suffer- ing of the earth -- lofty, beloved of men, great, and giving copiously of their rain"). (4) "How beastly the bourgeois is / especially the male of the species-/ Nicely groomed, like a mushroom / Standing there so sleek and erect and eyeable- / and like a fungus, living on the remains of bygone life" (D. H. Lawrence). (5) Bhāmaha offers three examples for six types enumerated; this one serves also for upamā śleșa. As far as the verb is concerned, we must interpret loosely: the verbal idea here includes the object of the verb as well. These examples should be compared with those of guna ślesa ('adjective'). Bhāmaha, the first writer to discuss śleșa as a figure, does not seem too ready to force these equivocations into sharp categories of mode and means: the simple adjectives in our examples above are not puns, because they can apply with equal justice to both subjects in the same sense, though they do not mean exactly the same thing as qualifications of the two subjects-that difference is to be sought in the kinds of things the subjects are rather than in a consideration of the kinds of usage and context the adjectives themselves are capable of sustaining. Bhāmaha seems more interested in defining this difference than in specifying the source and scope of the equivocation itself.

298 GLOSSARY guņa, 'property, adjective': (1) a pun on adjectives. (2) B 3.14 (18). (3) chāyāvanto gatavyālāļı svārohāņ phaladāyinah | mārgadrumā mahāntaś ca pareșām eva bhūtaye (Bhamaha: "Shady [possessing lustre], free of snakes [suspicious characters], firm rooted [independ- ent]. and producing fruits [benefits], both the roadside trees and great men live for the benefit of others"). (4) "As lines so loves oblique may well / Themselves at every Angle greet: / But ours so truly Parallel, / Though infinite can never meet" (Andrew Marvell). (5) Bhāmaha's example is also used for upamā śleșa. See the discus- sion under kriyā ślesa. The examples show how subtle can be the difference between pun and metaphorical usage. "Chāyā" has two senses here which are merely the two ends of a continuum: "shade, complexion, aspect, lustre"; similarly, "oblique" with the added overtones of "contrast". As puns, these meanings would have to be considered different. nāman, 'name, noun': (1) a pun in which a noun figures as the locus of the double-entendre. (2) B 3.14 (20). (3) ratnavattvād agādhatvāt svamaryādâvilanghanāt | bahusattvâśrayatvāc ca sadrśas tvam udanvatā (Bhāmaha; the terms "having jewels", etc. are expressed as abstract nouns in -tva: "You resemble the ocean for possessing jewels, for depth, for keeping within your own limits, for containing many substances [much strength of character]"). (4) "You tell 'em, whalebone, you've been around the ladies" (S. J. Perelman). (5) Cf. kriyā and guņa śleșa. This example serves also for hetu śleșa. Bhāmaha, in specifying the grammatical classification of the pun, lays the groundwork for the elaborate stratifications of Rudrata and Mammata where every conceivable form class (root, suffix. gender, number, etc.) is shown to be capable of expressing a double- entendre. See śleşa. niyamavat, 'restricted': (1) a type of paronomasia in which a double- entendre is explicitly limited to its further or irregular sense. (2) D 2.314 (319). (3) nistrimśatvam asāv eva dhanuşy evâsya vakratā śareşv eva narendrasya mārgaņatvam ca vartate (Dandin: "Sharpness is but in his sword, deviousness [being curved] but in his bow, supplic- ation [seeking a target] only in his arrows"). (4) "Who could continue to exist where there are no cows but the cows on the chim- ney-pots; nothing redolent of Pan but pan-tiles: no crop but a stone crop?" (Charles Dickens). (5) The figure is the same as parisamk hyd alamkāra. This is a sort of pun a fortiori, in which the distinction implicit in the pun is made explicit and the pun is thereby literally

GLOSSARY 299 annulled. Since. however, of the two senses, it is the less obvious, obseure. or figurative which is aflirmed, it is readily seen that the denial of the double-entendre is a mere formal device whereby the more obvious, clear. and primary sense is ironically intended. This type is contrasted with niyamâkseparūpa ślesa, wherein the primary meaning is broadened to include the secondary. niyamâkseparupa, 'having the form of an objection to limitation': (1) a type of paronomasia in which the double-entendre is explicitly extended to another, more current, sense. (2) D 2.315 (320). (3) padmānām eva dandeșu kanțakas tvayi rakșati | athavā drśyate rāgimithunâlinganesv api (Dandin: "As long as you arc King, thorniness/rivalry/horripilation exists only in the stalks of lotuses; but then. it could also be seen in the embraces of love-making couples"). (4) "'As you are doubtless aware', began Marcel, drawing on his praline, 'Harper's Bazaar not only prognosticates the mode but frequently publishes news of consuming interest-you will pardon the play on words-to gourmets'" (S. J. Perelman). (5) The name of this figure implies that the pun is an objection to the scope of meaning of a word, the extended meaning thereby offered being the rest of the double-entendre. Now, the example from S. J. Perel- man does not formulate an objection as such, but since the purpose of the parenthesis is to mark off clearly the second meaning of "con- suming" and to focus the attention of the reader on that extension, I believe it legitimately mentioned here. The parenthesis can also be taken as an ironic objection. pada, 'word': (1) a type of verbal pun (śabda ślesa) in which the constituent words of a phrase are differently separated to give different senses to the phrase. (2) R 4.5, M 119C. (3) prthu-kārttasvara-pātram (prthuka-ārtta-svara-pātram) bhūşita-nihśeşaparijanam (bhū-ușita) deva / vilasatkareņu-gahanam (bila-satka-reņu-gahanam) samprati samam āvayoḥ sadanam (Mammața: "Our residences are quite similar, O King: furnished with extensive gold [furnished with the pained howling of children], where the entire court is adorned with jewels [where all of the residents sleep on the bare ground]. where young elephants bathe and play [where the mice emerge from their holes and play in the dust]"). (4) "Only Seven: A pastoral Story after Wordsworth: I marvell'd why a simple child, / That lightly draws its breath, / Should utter groans so very wild / And look as pale as Death. / Adopting a parental tone, / I ask'd her why she cried; / The damsel answered with a groan, / 'I've got a pain

300 GLOSSARY inside!' / ... / Postscript: / To borrow Wordsworth's name wa wrong, / Or slightly misapplied; / And so I'd better call my sobg, / 'Lines after Ache-Inside'" (H. S. Leigh). (5) Rudrata's example not only shows compound words which can be analyzed differently, but also sequences of grammatically independent words showing different junctures (śarīram adaḥ; śarīra-madaḥ). This type of pun, in Rudrata and Mammata, assumes the word as one of the possible loci of śabda ślesa and contrasts with other topics such as letter (varņa), root (prakrti), gender (liñga), etc. Hence the word pada is to be understood in its strictly grammatical sense: that whose final is an inflection; that is, since pada is here distinguished from inflection (vibhakti), it should be taken as the grammatical añga (Pāņ. 1.4.13, 14). prakrti, 'root element': (1) a type of paronomasia in which the locus of the double-entendre is the primary and uninflected root. (2) R 4.28, M 119C. (3) ayam sarvāņi śāstrāņi hrdi jñeșu ca vaksyati | sāmart- hyakd amitrāņām mitrāņām ca nrpâtmajah (Mammața: "He holds in his heart and [speaks] to the wise all the śāstras; he makes [destroys] the ability of both his friends and his enemies"). (4) "'So they ought to be your best friends', I said. 'With all that money. You stick to them, old man. Clasp them to your soul with bands of steel-say steal ... '" (Joyce Cary). (5) The double-entendre here consists in a doubt as to which of two roots a given inflected form should be assigned: in the Sanskrit, vaksyati can be taken as the third singular future active of both vah 'to carry' and vac 'to speak'. It should be noted for purposes of comparison with other types of ślesa (such as linga, vacana, etc.) that no equivocation attaches to any of the inflectional elements of the form (future, singular, etc.). As to the English example, though it is not as convincing, we may reason as follows: "steel" and "steal" are nominal forms of two roots "to steel" and "to steal" which are not identical in all forms (as. "steeled" and "stolen"). The two roots are, however, accidentally identical in the present case (cf. Skt. krt, nominal of both kr. 'to do' and krr 'to cut'), as well as in their infinitives. We would be hard pressed to find in English, with its analytical structure, so revealing an example as vakşyati. pratyaya, 'affix': (1) a type of paronomasia in which the double-entendre depends upon the phonemic identity of two aflixes. (2) R 4.26, M 119C. (3) pramathanivahamadhye jatu cit tvatprasādād aham ucitarucih syăn nanditā sā tathā me (Mammața; may be read either

"aham ucitarucih syām nanditā; sa tathā me" or "aham ucitaruciķ; 301 srat nandita sa tatha me". in which case nandita is to be taken as an -r stem: "Sometimes in the midst of your attendant hosts, by your grace. O Siva. I am an enjoyer who has become appropriately delighted [likewise, I, as it were, occupy the state of your mount Nandi]"). (4) "A dripping pauper crawls along the way / The only real willing out-of-doorer / And says, or seems to say / 'Well, I am poor enough-but here's a pourer!'" (Thomas Hood). (5) As in the other types of sabda ślesa. the point is here that the locus of the pun is restricted to the formal element in question: -t or -m alone being equivocal and not the stem syā- 'am/is'. Likewise, the element -tã is both the nominative singular feminine of the past participle and the nominative singular masculine of an agent noun, but the root is the same: nand- 'to enjoy'. The English example, as usual, is not as finely adjusted to the point at issue. for the root "pour" is not the same as the root "poor"; but the example does involve as well an equivocation based on the affix, namely the "-er" of the comparative degree and the "-er" of the agent noun. So we offer it here and hope that any more crucial example will be brought to our attention by the kind reader. bhāșā, 'language': (1) a kind of paronomasia in which the double- entendre depends upon the apparent or verbal (I hesitate to say "phonemic") identity between two languages. (2) R 4.10, 16, M 119C. (3) akalañkakula kalālaya bahulīlālola vimalabāhubala | khalamaulikīla komala mangalakamalālalāma lala (Rudrața; it is asserted that this can be read in six languages: samskrta, prākrta. māgadhi, paiśācī, sūrasenī, and apabhramśa; in Sanskrit, it is but a string of vocatives: "O thou, be delighted! of spotless family, re- pository of the arts, stumbling from too much gay sport. strong in your spotless arms, a thorn in the heads of the corrupt, loveable. bearing the beauty mark of the lotus of good fortune!"). (4) "1 don't know what I'm doing mucking about with a lot of French authors at this hour, anyway. First thing you know I'll be reciting Fleurs du Mal to myself ... and I'll say off Verlaine too; he was always chasing Rimbauds" (Dorothy Parker). (5) In this category, the meanings need not be different; and this shows to what extent slesa was a question of technical manipulation of forms rather than (always) directed at some intelligent purpose. Rudrața gives examples for two-, three-, four-, five-, and six-language puns. Curiously, bhāsā ślesa is included in the eight kinds of sabda

302 GLOSSARY ślesa. all the other of which localize the double-entendre in some formal element of the same language. Completeness demands that this possibility be accounted for: the language is itself a form. This would then represent the other pole to artha ślesa, where no formal element can be ascribed as the locus of the double-entendre: language is grasped as meaning alone. bhinnapada. 'different words : (1) same as pada śleşa. (2) D 2.310 (312). rūpaka. 'metaphor : (1) for rūpaka sleșa. see under rūpaka. (2) D 2.313. linga. 'gender': (1) a type of paronomasia in which the double-entendre depends upon the phonemic identity of gender-indicating suffixes. (2) R 4.8. M 119C. (3) devi mahī kumārī padmānām bhāvanī rasāhārī | sukhani rāja tiro hitam ahimānam tasya saddhārī (Rudrața; as mas- culines, the first line must be read as nominative-n stems, as feminines, as stems in -i: other types of punning are here evidenced, mostly pada śleşa: for the king. rāja ('be brilliant"), for the girl, rājati ('illuminates'). Note also that, for the king, tasya is also a vocative from the obscure root tas- ('slay'): "Be brilliant! Slay the enemy, proud as a snake. O playful. festal, destroyer of evil, engenderer of prosperity. depository of passion, bringer of pleasure, upholder of the just!" or: "The earth, the queen, young, happy, source of lotuses and of spices, illuminates him who has assumed pride and upholds him!"). (4) " ... I'm a pris'ner, gen'l'm'n. Con-fined, as the lady said" (Charles Dickens). (5) Of course, our English example stretches a point, since the grammatical category in question does not exist. Contextually. however, the sense of the illustration is clear, since the word "confined" in feminine syntax means quite a different thing. vacana, 'number': (1) a type of paronomasia in which the double-entendre depends upon the phonemic identity of numerical suffixes. (2) R 4.28(30), M 119C. (3) āryo'si taromālyah satyo'natakukşayah stavāvācyah / sannābhayo yuvatayah sanmukhyah sunayanā vandyah (Rudrața; for singular, āryo'si, etc .; for plural, āryo (from pl. of āri) 'sita-romālyah (from pl. of romāli), etc. Note that the example also involves liūga śleșa and pada śleșa; I think it is impossible to find vacana śleșa "pure", as it were: "You are a noble [the wives of your enemies], the ornament of strength [the hair on whose navels is black], truthful [faithful], destroyer of the lands of the unbent [with round bellies], who cannot be praised in words [silent before compliments], fearless of those he has destroyed [with beautiful navels], who consorts with women [being young women], first among

GLOSSARY the respectable [whose faces are beautiful], lcader of good govern- 303 ment [of beautiful eyes], praiseworthy [are to be seized]"). (4) "Ben Battle was a soldier bold. / And used to war's alarms; / But a cannon- ball took off his legs, / So he laid down his arms" (Thomas Hood; we might claim here an equivocation between dual and plural. At least it is a numerical idea which is punned upon). (5) Strictly speak- ing, no illustration is possible from English since any pun involving plural-singular contrast (e.g., "rose", "rows") would also involve a different etymon. Here the point is that the same word lends itself to two numerically different references. Perhaps a pun could be found involving the word "sheep" (sing.) and "sheep" (pl.). varņa, 'syllable, letter': (1) a type of paronomasia in which the locus of the double-entendre is restricted to a single phoneme. (2) R 4.3, M 119C. (3) alankāraḥ śankākaranarakapālam parijano višīrņângo bhrīgī vasu ca vrsa eko bahuvayāḥ / avasthêyam sthāņor api bhavati sarvâmara- guror vidhau vakre mūrdhni sthitavati vayam ke punar amī (Mammața; the pun is restricted to the syllable -au of vidhau, which happens to be the locative singular of both vidhi ('fate') and vidhu ('moon'): "His only ornament is a human skull which inspires fright; his attendant is Bhrngi, the one who has consumed his own limbs; his wealth is one old bull. Such is the condition of Siva, Lord of all the Gods! Since inscrutable fate [the curved moon] sits on his brow, what indeed do we mere mortals amount to?"). (4) "Oh, the moon shines bright on my old Kentucky home" (Anon .; or: moonshine's). (5) This category shows a delicacy of interpretation which is truly astounding: at first glance, this seems to be an instance of pratyaya ślesa, where the affix is quite accidentally a single letter. But this analysis does not stand examination, because in the case of vidhau. the pratyaya is not punned at all, since it is locative singular in both cases (we say that the locative singular of both -i and -u stems is -au). Neither can the pun be attributed to the roots, for they are not the same (one is -i stem, the other -u stem). It might appear then that we have a case of vibhakti ślesa, where two inflections have the same form. But this interpretation, too, is invalid, for it cannot be said that the locative singular of two form classes is two inflections. We are left then with the recondite category pun on nominal form class, which in this unique instance involves only one syllable: hence the classification. Our English example is of course fictitious, since it is, strictly speaking, a case of vibhakti ślesa (the "-s" being both present singular of verbs and contraction of the nominal verb),

304 GLOSSARY but since it involves but a single letter, I have included it herc. No pun of this finesse can be conceived in English, since the only basis for a distinction of form classes in English nouns is the formation of the plural, and we could not find two words in different form classes which had the same plural. vibhakti, 'inflection': (1) a type of paronomasia in which the double- entendre depends upon the phonemic identity of two inflections. (2) R 4.28, M 119C. (3) sarvasvam hara sarvasya tvam bhavacche- datatparah | nayôpakārasāmmukhyam āyāsi tanuvartanam (Mam- mata; spoken by a thief to his son, hara and naya are imperatives; spoken by a devotee of Siva, hara is vocative and naya- a nominal stem: "O Siva, thou art the whole of everything [steal everything], you whose intention is to destroy the bonds of being [you must intend to break through the walls to everyone's treasure]. You have attained that state of soul which consists of an ensemble of help and counsel [bring into play every device and countermeasure and you will exhaust the efforts of others"]. (4) "Will there never come a season / Which shall rid us from the curse / Of a prose that knows no reason / And an unmelodious verse: / ... / When there stands a muzzled stripling, / Mute, beside a muzzled bore: / When the Rudyards cease from Kipling / And the Haggards Ride no more?" (J. K. Stephen). (5) This type may be understood in two ways: either as defined, or as a pun on an inflection common to verbs and nouns. In the first case, which is probably that of Rudrata, vibhakti is to be understood as a specific type of pratyaya (affix); in the second (Mammața), pratyaya ślesa will be understood as limited to the general affixes of the nominal form class, whether inflections or not, while vibhakti will involve the contrast between inflections (neces- sarily) of verb and noun. The names thus do not exactly specify the scope of the pun intended. Another favorite example stems from the identity of the genitive plural present participle and the third singular middle imperative -tām, as, jīvatām ('may he live' and 'of those living'). viruddhakarman, 'opposed verbs': (1) a type of paronomasia in which a separate verb accompanics each of the senses of the double-en- tendre, and which verbs are moreover opposite in meaning. (2) D 2.314 (318). (3) rāgam ādaršayann eşa vāruņīyogavardhitam ! tirobhavati gharmâmśur angajas tu vijrmbhate (Dandin: "Having revealed a passion [redness] born of association with intoxicating liquors [the western sky], the God of Love shows his full power;

GLOSSARY the sun, however, sinks from sight"). (4) "At Thieves I bark'd, at 305

lovers wagg'd my tail, / And thus I pleas'd both Lord and Lady Frail" (John Wilkes; "pleas'd" is the double-entendre; "bark'd" and "wagg'd", which are opposite in sense, accompany each meaning of "pleas'd"). (5) This figure contrasts with aviruddhakriya, which has two verbs which are not opposite in sense, and abhinnakriyā. which has one verb only. virodhin, 'contradicting': (1) a type of paronomasia in which a double- entendre is suscitated through the resolution of an apparent con- tradiction between a noun and a qualification appended on it. (2) D 2.315 (322). (3) acyuto'py avrşacchedī rājpy aviditaksayah devo'py avibudho jajñe śankaro'py abhujangavān (Daņdin: "Although he is Visnu [not deviating from the right way], he has not slain the demon Vrsa; although the moon [King], he does not know diminu- ation [never suffers decline]; although a God [King], he is not a God [not surrounded by wise men]; although Siva [appeasing], he is not possessed of snakes [questionable friends]"). (4) "Now as they bore him off the field, / Said he, 'Let others shoot, / For here I leave my second leg, / And the Forty-second Foot!'" (Thomas Hood). (5) Nobody has forty-two feet on his leg. The point here is that no pun would be understood were it not for the juxtaposition of incompatibles which serves then as the efficient cause of the double- entendre. Usually, in Sanskrit punning, the duplicity of meaning is suscitated by the adjunction of two different things which are said to have the same qualifications (see avirodhin). There is no con- tradiction between the thing and the qualification taken singly. See aviruddhakriyā for notes on form. vyatireka, 'distinction': (1) see śleşa vyatireka. (2) D 2.313. śabda, 'word': (1) a type of paronomasia in which some point of gram- matical interpretation is involved in distinguishing the meanings of the double-entendre. (2) U 4.10, M 119. (3) prabhātasamdhyèva ... muktāśrīs tvam eva pārvatī (Udbhața; applying to Pārvatī. muktâśrīh, a bahuvrīhi with udātta accent on the first member: apply- ing to the dawn, an appositive tatpurusa, "beauty of pearl", with ac- cent on the stem of the final member: "Parvati, by whom ugliness was abandoned [beautiful as a pearl], is like the onrushing dawn"). (4) "This is how it came to pass that their children were white and puny; they were suffering from home-sickness" (Samuel Butler; here is a pun on the compound form "sea-sickness": ordinarily the compound form "home-sick" does not express a cause, but the ab-

306 GLOSSARY sence of the antidote). (5) If the pun depends on no determinable formal equivocation, but only on a legitimate duplicity of meaning assignable to the word (lexical item, etymon) itself, we have artha śleșa (q.v.). For a full discussion of this difference, see artha ślesa. Mammața takes the distinction śabda|artha ślesa from Udbhața and then finds that Rudrata's eight grammatical categories of ślesa can be taken as subdivisions of śabda ślesa. Thus do systems grow. He treats śabda śleşa in ullāsa 9 (kārikā 119) and artha śleșa in 10 (kārikā 147). sahôkti, 'expression of concomitance': (1) sahôkti alamkāra accompanied by śleșa. (2) B 3.17 (18). (3) (4) (5) See guņa śleșa. hetu, 'cause': (1) a type of paronomasia in which the double-entendre accompanies hetu alamkāra. (2) B 3.17 (20). (3) (4) (5) See nāma śleșa.

saņśaya samśaya (I) 'doubt': (1) same as samdeha. (2) R 8.59, 61. (5) See samdeha. saņśaya (II): (1) a figure in which two similar but discriminable things are said to be subject to a doubt concerning their respective nature or mode of action. (2) R 8.65 (66). (3) gamanam adhītam hamsais tvattah subhage tvayā nu hamsebhyah / kim śaśinah pratibimbam vadanam te kim mukhasya śaśī (Rudrata; in the first half śloka, the mode of cause and effect is suspended: who taught whom? In the last half, the question is: the beauty of which term is the model for that of the other? "Did the swans teach you how to walk, O lovely, or did you teach the swans? Is your face modeled on the moon, or the moon on your face?"). (4) "Such was the mutual love and reciprocal respect that this worthy man had to his flock, and his flock to him, that it was hard to judge whether he delighted more in having such a people, or they in having such a pastor" (Carl Sandburg). (5) This figure differs from the ordinary samdeha only in that it is not qua comparability that the two similar things are doubted. Here they are confused through some other aspect of their relation (which is, of course, many sided), as, for example, cause-effect (which is cause and which effect?), if they happen to be related as to cause and effect (either in fact or in the poet's imagina- tion). For other exaggerations of nature and mode of action, cf. adhika and vişama. The present type differs from them inasmuch as it is expressed by means of a doubt, not declaratively.

GLOSSARY 307 samsrsți samsrsți, 'combination': (1) a multiple or compound figure. (2) B 3.48- 51, D 2.363 (samkīrņa, 2.359), V 4.3.30-33, U 6.5, M 207. (3) (4) See following discussion for examples. (5) The three terms samsrsti, samkirna, and samkara, used in gencral to refer to the combination or compounding of other figures of specch, have such a variety of particular acceptations in the various authors that it is impossible to separate them clearly. In general, the phenomenon of the multiple figure was considered, from Dandin onwards, in two aspects: the two or more constituent figures could be related somehow, or they could be purely extrinsic to one another. Only Bhāmaha and Vāmana do not make this distinction in some form; their term samsrsti is therefore to be taken as referring to the genus of multiple alamkāras. That they do use the term samsrsti and not one of the other terms, probably testifies to the originality of this term, the others being used ad hoc by later writers. Dandin, however, in making the distinction between related and independent figures in conjunction, employs the term samkīrna in the generic sense of Bhamaha's samsrsti. The word samsrsti does appear in Dandin's definition of samkīrņa ("nānâlamkārasamsrşțih samkīrņam") and suggests that the use of samkīrna is idiosyncratic and well within the normal variation of otherwise univocal technical terminology (Dandin, for example, refers to svabhāvôkti as svabhāvâkhyāna (2.4) for metrical reasons). Within samkīrņa, then, Dandin enumerates two types: mutually related (angângi), and independent (samakakșatā), depending on whether or not one figure implies the other in the sense that the other cannot be grasped as a figure without the first. Udbhata recognizes this distinction, but applies to it the two terms which up to now appear to be but mere stylistic variations of one another: samsrsti (for unrelated figures) and samkara (from the same root -kf as samkīrņa; for related figures). Within the latter term, Udbhata finds four subtypes: Dandin's familiar angângi (renamed anugrāhyângu- grāhaka) as well as an ekaśabdâbhidhāna (where the two figures overlap in the total expression), samdeha (where they coalesce entire- ly), and a type in which both śabda and artha alamkāras (śabdârtha- vivarti) are mixed. Mammata repeats the entire discussion of Ud- bhata except for his omission of the subtype ekaśabdabhidhāna. Rudrața, however, introduces the ultimate element of confusion by using the term samkara in the generic sense, as did Dandin, and not mentioning samsrsti at all. His two types, vyaktâmśa and avya-

308 GLOSSARY ktâmsa, seem to refer to the same distinction as Dandin intended by samakakșatā and angângi, but his definitions of them are analogical and do not permit precise rendering. The former is said to resemble a mixture of rice and sesame, where the components remain discrete, the latter to resemble a mixture of milk and curds, where the com- ponents fuse in an indissoluble agglutinate. Whatever use is made of this terminology, all the writers agree on regarding ślesa or pun as the compoundable figure par excellence. Dandin's remark to this effect (2.363) is probably to be taken in context with his previous discussion of samkirna. Most of the ex- amples offered by the various writers involve ślesa as one of the figures. angâūgi (-tva; -bhāvâvasthāna), 'related as whole and part': (1) a type of complex alamkāra (samkara) in which the several constituent figures are mutually interrelated in the sense that the figurativeness of some cannot be understood apart from the others. (2) D 2.360 (361), U 5.13, M 208. (3) ākşipanty aravindāni mugdhe tava mukhaśriyam / kośadaņdasamagrāņām kim eşām asti dușkaram (Daņdin; the śleșa of the second half verse depends on the upama of the first: if the lotuses were not "disputing" the beauty of her face, then their "buds" and "stalks" would not be taken as weapons: "treasury" and "army": "The lotuses are attacking the beauty of your face! Is this difficult for them, whose means are buds [treasury] and stalks [army]?"). (4)"He[ the biographer] will row out over that great ocean of material, and lower down into it, here and there, a little bucket, which will bring up to the light of day some characteristic specimen, from those far depths, to be examined with a careful curiosity" (Lytton Strachey; the samāsôkti of "bucket", "specimen", and "depths" cannot be understood apart from the rūpaka "ocean of material"). (5) A subordination of implication only is to be understood by this term: the two figures concerned are formally distinguishable (cf. samdeha) and occupy different places in the total phrase (cf. avyaktâmsa. ekaśabdâbhidhāna). They are not, however, entirely independent -- a mere concomitance of figures-as in samakakșatā: should one of the figures (the angin) be deleted, the other (the anga) no longer makes any sense as a figure. The terms samkara and samsrsti are used by Udbhata and Mammata in the sense of this distinction, but Dandin, who does not use the term samkara, views both as types of samsrsti. For the use of the terminology, see samsrsti. The figure is called anugrāhyânugrāhaka by Udbhaļa.

GLOSSARY gnugrāhyânugrāhaka, 'related as facilitated and facilitating : (1) same as 309 añgâñgi samsrsti. (2) 1J 5.13. avyaktâmśa, 'whose parts are not manifest': (1) a type of complex alamkara (samkara) in which the component figures are so expressed as to make impossible a decision as to the location of each in the phrase. (2) R 10.25 (28-29). (3) ālokanam bhavatyā jananayaná- nandanêndukarajālam (Rudrața; in this case, the commentator asserts that the compound can be taken in one way as an upamā, in another as a rūpaka: "Your glance is a net of moonbcams, delight- ing the eyes of men"). (4) "Even when the President, the Rev. Dr. Willoughby Quarles ... than whom no man had written more about the necessity of baptism by immersion, in fact in every way a thoroughly than-whom figure ... " (Sinclair Lewis; the last phrase can be taken as an arthântaranyāsa or a rūpaka). (5) As a karma- dhāraya ("jananayanânandana evêndukarajālam"), the figure in the Sanskrit example is a rūpaka; as a locative tatpurușa ("jananayanā- nandana indukarajālam"), the attributive relation to the subject āloka- nam can, it is said, be interpreted as a simile lacking the comparative particle (iva). It may be said that this latter interpretation seems to be at variance with Rudrata's own treatment of both simile and metaphor. "Indukarajālam" by itself would not be an upamā, since the upamāna follows the upameya (R 8.21); neither would the predi- cative assertion "ālokanam ... indukarajālam", since this is the clearest type of rūpaka (metaphor) definable in Rudrata's system (8.38-39). This difficulty does, however, parallel Bhāmaha's appar- ently inconsistent definition of the figure upamārūpaka, where the example of the upama is just such a rūpaka. It may be that the formal discriminations were not rigorously applied to those cases where. despite the predicative identification of upameya and upamāna. the idea of comparison was judged to be uppermost in the mind of the poet. If this irregularity is allowed, the example then indeed illustrates the definition of avyaktâmsa, for both interpretations are valid (neither can be refuted), and both apply to exactly the same word sequencc. In the English example, if the phrase "than-whom figure" is taken as a predicate to Dr. Quarles, we have a rūpoka; however, the phrase can also be taken as a general summary of the preceding descriptive passage (arthântaranyāsa). Compare ekasa- bdâbhidhăna, where the two figures only overlap: here they coalesce entirely. This type is the same as the sansdeha of Udbhata and the aniścaya of Mammața. Rudrața gives another example which the

310 GLOSSARY commentator interprets as an angângi samsrsti; that is, a simple relation of implication exists between the two figures, not physical coalescence. Rudrata's definitions of this dichotomy (vyaktám- śâvyaktâmśa) are only analogical (see samsrsti), and much leeway must be allowed in their interpretation. ekaśabdâbhidhāna, 'naming by the same words': (1) a type of complex alamkāra (samkara) where the two constituent figures overlap as to the words which express them. (2) U 5.12. (3) mâivam evastha sacchāyavarņikācārukarņikā / ambhojinîva citrasthā drsțimātrasu- khapradā (Udbhata; Gauri is here compared to a lotus as delighting the eyes only (upamā); another common property is offered which is, in fact, a pun referring to the golden earrings of Gauri and to the seed pods within the lotus (ślesa); the two figures share the word ambho- jinī ('lotus'): "Don't remain so, dazzling and giving pleasure only to the eyes, like a lotus, earrings [seed pods] of beautifully shaded gold"). (4) "But the man who fell in love with Rozzie was the poor little Peter Pan who wanted to creep back into his mother's womb and be safe and warm and comfortable for the rest of his life. Isola- tionist, with navel defence" (Joyce Cary; same analysis except that the figures are rūpaka and śleșa). (5) This type is recognized as such only by Udbhata, who distinguishes it from samdeha samkara: here the two figures are partially different and wholly identifiable; in samdeha, they are expressed in wholly the same words and are not identifiable. To the naive observer, this type would appear to be more akin to samsrsti than to samkara, but the fact that there is partial coalescence is sufficient to make the two figures constitutive and "interdependent". vyaktâmśa, 'whose parts are evident': (1) a type of complex alamkāra (samkara) in which the component figures are separable as to the words which express them. (2) R 10.25. (3) abhiyujya lolanayanā sādhvasajanitôruvepathusvedā | abalêva vairisenā nrpa janye bhajyate bhavatā (Rudrața; an upamā in the last half, a pun in the first: "When attached, its leaders unsteady [eyes rolling], sweat and great trembling [trembling at the thighs] produced suddenly, the enemy army, like a young girl, O King, is enjoyed by you"). (4) "As lightning, or a taper's light, / Thine eyes, and not thy noise, waked me" (John Donne; upamă in the first line, vyatireka in the second). (5) The example here is identical to that offered by Udbhata to illustrate the exact inverse (see ekaśabdâbhidhāna). But Rudrața takes the śleșa as confined to the word which is actually punned, while

GLOSSARY 211 Udbhata apparently considers that the pun extend to the aor! which are said to be alike by virtue of the pun (here ' abaleva vaieena and "thine eyes"). If Rudrata is correct, then wyah tum.a tha; V: equated with samakaksata; it should, however, be noted that the distinctions implied by the two terms are not the same: iyahtam.d may mean 'whose parts are not in doubt' (see avyaktamia) wherca, samakaksatā means 'whose parts are not subordinate' (see anganzr) See samsrsti for a discussion of these classifications. sabdârthavarti, 'involving both (figures of) word and sense': (1) a type of multiple alamkāra (samisrsți) in which are mixed both sabda and artha alamıkāras. (2) U 5.12, M 210. (3) spastollasatkiraņakesara- süryabimbavistīrņakarņikam atho divasâravindam ślistāstadigdala- kalāpamuk hâvatārabaddhândhakāramadhupāvali samcukoca (Mam- mața; contains both rūpakas (an arthâlamkāra) and anuprūsa (a śabdâlamkāra): "The lotus of the day closes its vast pericarp. the sun's orb and its filaments the clearly dancing rays, with a swarm of bees-the great darkness-clinging to the face of the eight direc- tions"). (4) "Glory be to God for dappled things- / For skies of couple-colour as a brinded cow; / For rose-moles all in stipple upon trout that swim; / Fresh-firecoal chestnut-falls: finches' wings: Landscape plotted and pieced-fold, fallow, and plough; / And all trades, their gear and tackle and trim. / All things counter. original. spare and strange; / Whatever is fickle, freckled (who knows how?) With swift, slow; sweet, sour; a dazzle, dim; / He fathers-forth whose beauty is past change: / Praise him" (Gerard Manly Hopkins: anuprāsa and upamā). (5) This type would exemplify samkara or complex alamkāra, for the same word span may share two or more figures. It differs from avyaktâmsa in that the basis of interpretation is not such that one figure, once defined, necessarily excludes the other: here both are conjointly possible and discrete because the figures do not refer to the same definand (one refers to phonemic patterns, the other to morphemic contrasts and usages). samkara, 'intermixture': (I) a multiple alamkāra. (2) U 5.11. 13. R 10.24-29, M 208-210. (5) See sanisrsi. The term is often used to signify complex alamkāra as opposed to compound alamkāra. samkīrņa, 'commingled': (1) a multiple alamkāra. (2) D 2.359-303 (5) See samsrsti. samdeha, 'doubt': (1) same as avyaktâmša samkara. (2) U 5.11. (5) Also called aniścaya in Mammața. samakakşatā, 'equality': (1) same as vyaktâmsa samkara or samsrsți.

312 GLOSSARY (2) D 2.360 (362). (5) In Dandin, the term is opposed to angáñgi, in the sense that no relation of implication exists between the two figures; in Rudrata, however, it is opposed to avyaktâmśa, in that the parts of the phrase expressing the several figures are physically separable. The two ideas are therefore of differing extent, but in intent they are the same. saņkara samkara, 'intermixture': (1) another word for samsrsti. saņkīrņa samkirņa, 'commingled': (1) another word for samsrsti.

saņkșepa samksepa, 'abbreviation': (1) terseness, condensation of meaning into few words. (2) AP 345.6. (5) This is one of the śabdârthâlamkāra of the Agni Purāņa.

saņdeha samdeha, 'doubt': (1) a figure in which the speaker hesitates to identify which of two similar things is which; the expression of a similitude through the affectation of an inability to decide the relative identity of two things-the subject and object of comparison. (2) B 3.42 (43), V 4.3.11, U 6.2-3, R 8.59-64, M 138. (3) kim idam līnâlikulam kamalam kim vā mukham sunīlakacam / iti samśete lokas tvayi sutanu sarovatīrņāyām (Rudrața; cf. samšaya: "As you descend to the lake, O Lovely, onlookers are bound to wonder if that is a face set in hair of darkest hue or a bee-girt lotus"). (4) "Mr. Smith, suddenly meeting the lovely young thing, may not be sure whether his feet are treading a polished studio floor or whether they have little Mercury wings on them that waft him through the empyrean ." (Oliver Onions). (5) Rudrața names this figure samśaya. departing from tradition slightly (the word also means 'doubt'). perhaps because he enlarges the scope of the figure beyond simile (see samšaya II). Daņdin (2.358) mentions the figure only to assert its identity with his samsaya upamā. The figure has two forms in the early writers, judging by the examples offered; this difference is recognized by Rudrata and Mammața, who subdivide into those doubts about which no attempt is made to decide (aniścaya). following Dandin and Vämana, and those in which the doubt is

GLOSSARY either resolved or in which a discrimination is offered which would 313

permit its resolution (niścayânta, niścayagarbha), following Bhāmaha Udbhata, though following Bhāmaha, distinguishes those doubts which are feigned and those which are real, perhaps anticipating bhrāntimat alamkāra; however, a "feigned" doubt for him involves the suggestion of another figure, usually upamā (simile), and in this respect he returns to the opinion of Dandin. I append here Udbhata's example for "real" doubt, since none of the other writers seems wil- ling to limit the figure in this way: "haste kim asya nihśeșadai- tyahrddalanôdbhavah / yaśahsamcaya eşa syāt piņdībhāvo'sya kim kṛtah / nābhipadmasprhâyātah kim hamso nâișa cañcalaḥ / iti yasyâ- bhitah śankham aśankistârjavo janah" ('Can this be in [Śiva's] hand the accumulation of glory born of his destroying all the demons' loves? Why has it become spherical? Can it be a swan filled with desire for his lotus-navel? But it is motionless! So does the stupid fellow wonder about the conch shell'). Or: "He thought he saw an Elephant, / That practised on a fife: / He looked again, and found it was / A letter from his wife. / 'At length I realize', he said, / 'The bitterness of Life!'" (Lewis Carroll). aniścaya, 'undecided': (1) a type of samdeha in which the resolution of the doubt is neither intimated nor given. (2) R 8.59 (60). (3) asyāḥ sargavidhau prajāpatir abhūc candro nu kāntipradah śrīgārâi- karasaḥ svayam nu madano māso nu puspâkarah / vedâbhyāsajadah nu vişayavyāvrttakautūhalo nirmātum prabhaven manoharam idam rūpam purāņo munih (Kālidāsa, quoted by Rudrața: "The God Prajāpati presided at her birth! Or was it the moon, giver of beauty? Or was it Love, made of pure affection? Or the springtime, the mine of flowers? How could the ancient sage-his mind dulled by recitation of the Veda, his curiosity turned away from outward things -conspire to create this excellently lovely form?"). (4) "God bless the King, I mean the faith's defender; / God bless-no harm in blessing-the Pretender:/ But who Pretender is, or who is King, / God bless us all-that's quite another thing" (John Byrom). (5) In niścayagarbha, the doubt is implicitly resolved by proposing a reason why the two things cannot be the same. In niścayânta, this is put positively and one learns what, in fact, they are. Note that neither example can be taken as a simile: in the first case, the inten- tion is not to compare Prajäpati with the moon, etc., but, quite the contrary, to say that only the moon, etc. are qualified to serve as tutelary deities at Urvaśi's birth (this might be called a samdeha

314 GLOSSARY vyatireka). Byrom likewise expresses only the insecurity of man's state and is unsure which "tutelary deity" guides him. niścayagarbha, 'pregnant with decision': (1) a type of samdeha in which the doubt is implicitly resolved by advancing a discriminating qualification which makes it impossible that the two things in question can be the same. (2) R 8.61 (62-63). (3) etat kim śaśibimbam na tad asti katham kalankam anke'sya | kim vā vadanam idam tat katham iyam iyatī prabhâsya syāt (Rudrata: "Can that be the moon's orb? Impossible; there are no spots upon it! Can it then be a face? If so, how could it be so brilliant?"). (4) "Women ben full of Ragerie, / Yet swinken not sans secresie / Thilke Moral shall ye understand, / From Schoole-boy's Tale of fayre Irelond: / Which to the Fennes hath him betake, / To filch the gray Ducke fro the Lake. / Right then, there passen by the Way / His Aunt, and eke her Daughters tway. / Ducke in his Trowses hath he hent, / Not to be spied of Ladies gent. / 'But ho! our Nephew', (crieth one) / 'Ho', quoth another, 'Cozen John'; / And stoppen, and laugh, and callen out,- / This sely Clerk full low doth lout: / They asken that, and talken this, / 'Lo here is Coz, and here is Miss'. / But, as he glozeth with Speeches soote, / The Ducke sore tickleth his Erse-root: / Force-piece and buttons all-to-brest, / Forth thrust a white neck, and red crest. / 'Te-he', cry'd Ladies; Clerke nought spake: / Miss star'd; and gray Ducke crieth Quake. / 'O Moder, Moder' (quoth the daughter) / 'Be thilke same thing Maids longen a'ter? / 'Bette is to pyne on coals and chalke, / Then trust on Mon, whose yerde can talke'" (Alexander Pope; "Imitation of Chaucer"). (5) The discrimination is given, but the proper identification is not made: cf. niścayânta. niścayânta, 'conclusive': (1) a type of samdeha in which the doubt is resolved by the proper identification of the two similar things. (2) R 8.61 (64). (3) kim ayam harih katham tad gaurah kim vā harah kva so'sya vrsah / iti samšayya bhavantam nāmnā niścinvate lokāh (Rudrata: "Is he Krșņa? Why is he not dark? Is he Śiva? Where is his bull? In this way onlookers at first doubting now at last conclude who you are, O King"). (4) "Surely she lean'd o'er me-her hair / Fell about my face .... / Nothing: the autumn-fall of leaves" (Dante Gabriel Rossetti). (5) Cf. aniścaya and niścayagarbha samdeha. bhedânukti, 'non-expression of the difference': (1) same as aniścaya. (2) M 138. bhedôkti, 'expression of the difference': (1) same as niścayagarbha. (2) M 138.

GLOSSARY 315 sama sama. 'together': (1) a figure in which is depicted an appropriate conjunc- tion of events, persons, or qualities. (2) M 193. (3) dhātuh šilpátiša- vanikaşasthānam eşā mrgâkșī | rūpe devo'py ayam anupamo datta- patrah smarasya | jātam daivāt sadrsam anayoh samgatam yat tad etat | śrn̄gārasyôpanatam adhunā rājyam ekâtapatram (Mammața; a marriage is described: "This doe-eyed maiden is the veritable touchstone of the creator's skill; in beauty, the King is incomparable and the given vessel of Love itself! That their union should come about through fate means that the kingdom of Love has been brought under one umbrella!"). (4) "Happy, happy, happy Pair! / None but the Brave, / None but the Brave, / None but the Brave deserves the Fair" (John Dryden). (5) In origin, this figure may simply be an adverb mistaken for a name in an enumerative verse. The Agni Purāņa affirms that the arthâlamkāra (q.v.) are eight in number (344.2-3); the chapter is devoted to definitions of them, and from that text, we learn that the eight are svarūpa (3), sādrśya (5), utprekșā (25), atiśaya (26), viśeșôkti (27), vibhāvanā (28), virodha (29), and hetu (30). The introductory verse enumerating the eight gives, how- ever, only seven, ignoring viśeşôkti. After hetu, the words "samam astadhā" ('together, eightfold') close the half verse, presumably to fill out the metre. Later commentators and writers, apparently ignoring the body of the text, have raised this superfluous adverb (samam) into the eighth arthâlamkāra, and it is included in all later anthologies, starting with Mammata. The Sarasvatīkanthâbharana, which follows the Agni Purāna closely, does not mention samâlam- kāra. There is a sāmya, but it refers to intimated similes. If this interpretation is correct, we have here an excellent instance of the eclectic resilience of the later encyclopaedists. None of the printed texts of the Agni Purāna, not even the so-called critical edition of S. M. Bhattacharya, seem to have noticed this anacoluthon. samādhi samādhi, 'conjunction': (1) a figure in which a desired effect is accomplished by the coincidental intervention of another and quite irrelevant cause. (2) M 192. (3) mānam asyā nirākartum pādayor me patişyatah / upakārāya diştyêdam udīrņam ghanagarjitam (Mammața: "As I fell at her feet to beg respite from her wrath, to my aid came a great exploding thunderclap"). (4) "I asked professors who teach the meaning of life to tell me what is happiness. / And I went to

316 GLOSSARY famous executives who boss the work of thousands of men. / They all shook their heads and gave me a smile as though I was trying to fool with them. / And then one Sunday afternoon I wandered out along the Desplaines river / And I saw a crowd of Hungarians under the trees with their women and children and a keg of beer and an accordion" (Carl Sandburg). (5) This figure differs from vyāghāta in that the intervening cause is unrelated to the original (obstructed) motive. Samadhi is the same as samāhita of Bhāmaha and Daņdin. samāsôkti samāsôkti, 'concise speech': (1) a figure in which the descriptive qualifica- tions of an explicit subject suggest an implicitly comparable object to which they likewise apply. (2) B 2.79 (80), D 2.205-213, V 4.3.3, U 2.10, AP 345.17, R 8.67 (68), M 148. (3) skandhavān rjur avyālaḥ sthiro'nekamahāphalaḥ / jātas tarur ayam côccaih pātitaś ca nabha- svatā (Bhāmaha; the description of the tree suggests the picture of a noble man fallen on hard times: "A great tree has grown here, with many limbs, erect and without flaws, firm and bestowing many great fruits; now it has fallen from its high place, uprooted by the wind"). (4) "The young man ... reached down the boughs where the scarlet beady cherries hung thick underneath, and tore off handfull after handfull of the sleek cool-fleshed fruit. Cherries touched his ears and his neck as he stretched forward, their chill finger-tips sending a flash down his blood. All shades of red, from a golden vermillion to a rich crimson glowed and met his eyes under a darkness of leaves" (D. H. Lawrence; the description of the tree suggests the young man's incipient passion). (5) This figure is easily confused with aprastutapraśamsā, and any attempt to distin- guish sharply the two figures is rendered fruitless by an historical examination of the relation between them. Though each writer distinguishes them in his way, none follows exactly his predecessor, and the same concept is likely to end up on both sides of the defini- tion at one time or other. There are two criteria involved in the distinction, from which only Dandin deviates significantly. The first is that samāsôkti tends to repose on the identity of descriptive qualifications of two terms: one explicit, the upamāna; one implicit, the upameya. There is no emphasis placed on the implicit term, thus making it into the explanation of the occasion of the remark itself. Aprastutapraśamsā, on the other hand, generally imposes upon the terms such an emphasis, and it can function through relations other

GLOSSARY 317 than identity of qualification. such as similitude or causality. We might say that in aprastutaprasamisa the remark is always motivated, while in samasôkti the suggestion of the other situation is subordin- ated to the requirements of clever characterization. The case is perhaps clearest in Bhämaha, the first writer to define both terms: "yatrôkte gamyate'nyo'rthas" ('where another meaning is understood in a given expression'; samāsôkti) and "adhikārād apetasya vastuno'- nyasya ya stutih" ('mention made of a subject irrelevant to the topic'; aprastutapraśamsā). Evidently exegesis is still required. Udbhața and Mammata conform in general to this pattern, but the latter complicates the discussion by defining a type of aprastutapraśamsā based on samāsôkti. This is perfectly reasonable and justified by the distinction given above: where the identity of descriptive qualifi- cations serves (in lieu of other possible relations) to suggest the real subject of the motivated remark, we have conjunction. Rudrața's definition is straightforward, but since he does not accept the figure aprastutapraśamsā, his samāsôkti can be made to cover both cases: the object of comparison serves to suggest the subject of comparison. Vāmana, in his attempt to reduce all figures to a basic simile, is forced to modify the present distinction inasmuch as it is indifferent to the nature of comparison (cf. Rudrata's definition). For him, the subject of comparison is unmentioned in samāsôkti, but is partially mentioned (that is, through its attributes only) in aprastuta- praśamsā. Thus samāsôkti covers the two terms as used by the other writers, and aprastutapraśamsā is unparalleled (cf. aprastutapraśamsā II), but resembles the more usual samāsôkti. It is Daņdin's treatment, however, which diverges farthest from the norm. He defines samā- sôkti in the way Mammata defines tulya aprastutapraśamsā, and his aprastutapraśamsā is etymologically interpreted as a kind of vyāja- stuti (concealed praise) which could be called "concealed blame" (cf. aprastutapraśamsā III). Only Daņdin subdivides samāsôkti, so the peculiar character of his general definition must be born in mind when examining these subtypes. The qualifications will either apply equally to both subjects (tulyâkāraviśeșaņa), or some to both and some to the implicit subject only (bhinnâbhinnaviseșaņa). Thirdly, the qualifications may cease to apply to either subject in time (apūrva). The encyclopaedist Ruyyaka has stated the opposition between aprastutapraśamsā and samāsôkti most rigorously. Accepting the usual version of aprastutapraśamsā, he understands the other figure

318 GLOSSARY as its exact inverse: "prastutād aprastutapratītau samāsôktir uktā |; adhunā tadvaiparītyenâprastutāt prastutapratītāv aprastutaprasamsô- cyate" ('samāsôkti is defined as the understanding of what was not intended, on the basis of what was; now in opposition to that, aprastutapraśamsā is defined as the understanding of what was intended on the basis of what was not'; p. 132). This recapitulates rather forcefully and with perhaps too much antithesis what was stated earlier: in aprastutapraśamsā the purpose of the utterance (its prastuta) is precisely what is not said; in samāsôkti, for the most part, that implicit reference should not be understood as the main topic itself. The implicit aprastuta serves only as a flattering back- ground; for example, when a king is described in terms also appro- priate to the Diety: "he is of clever mien [four faced]". In these terms, then, samāsôkti appears as a simile in which the upamāna is implicit. apūrva, 'unprecedented': (1) a type of samāsôkti in which the qualifica- tions will cease to apply to either subject at some future time. (2) D 2.213 (212). (3) nivrttavyālasamsargo nisargamadhurâśayah / ayam ambhonidhih kastam kālena pariușyati (Daņdin; the reference is to the eventual demise of a liberal benefactor: "This great ocean where serpents [evil men] perish and whose fund of fresh water [affection] is elemental will in time, alas, dry up!"). (4) "There was a lean and haggard woman, too-a prisoner's wife-who was watering, with great solicitude, the wretched stump of a dried-up, withered plant, which, it was plain to see, could never send forth a green leaf again-too true an emblem, perhaps, of the office she had come there to discharge" (Charles Dickens; in this example, the qualifica- tion [sending forth a green leaf] has already ceased to apply, but the function of the periphrasis remains the same-to indicate termination through parallelism of change). tulyâkāraviśeșaņa, 'whose qualifications apply in the same respect': (1) a type of samāsôkti in which the qualifications apply in the same respect to both the expressed and understood subjects. (2) D 2.208 (209). (3) rūdhamūlah phalabharaiḥ pușņann anišam arthinah | sāndracchāyo mahāvrkşah so'yam āsādito mayā (Daņdin: "I have seated myself under this great tree, deep rooted and thick shading, which ever provides for the needy with its harvests of fruit"). (4) "And then he dwelt for a while on the wife of a man called Socrates, who he didn't bother to place, though I judged he had something to do with the mule train, as therc was a fair number of foreigners amongst them" (Robert Lewis Taylor).

GLOSSARY hhinnabhinnaviseşana, 'whose qualifications apply both differently and 319

in the same respect': (1) a type of samasôkti in which some of the qualifications apply only to the understood subject. (2) I 2.208 (210). (3) analpavițapâbhogah phalapușpasamrddhimān | sôcchrāyah sthairyavān daivād eşa labdho mayā drumah (Daņdin; a bencfactor is thus referred to: "My fortune is that I have found this tree of expansive foliage, rich with fruit and flower, lofty and steadfast"). (4) "'Is the gray mare made over to anybody?' inquired Mr. Weller anxiously. George nodded in the affirmative. 'Vell, that's all right', said Mr. Weller. 'Coach taken care of, also?' 'Con-signed in a safe quarter', replied George, wringing the heads off half a dozen shrimps, and swallowing them without any more ado. 'Wery good, wery good', said Mr. Weller. 'Alvays see to the drag ven you go downhill. Is the vaybill all clear and straight for'erd?' 'The schedule, sir', said Pell, guessing at Mr. Weller's meaning, 'the schedule is as plain and satisfactory as pen and ink can make it'" (Charles Dickens; the references are to the legal liquidation of a bankrupt coachman's property. The first two remarks apply only to the lawsuit; the last two ("downhill", "vay-bill") apply to the legal "journey" the coach- man is about to make).

samāhita samāhita (I), 'composed': (1) a figure wherein a similitude serves as the basis for the total identification of two similar things in the mind of a particular person; a consciously formulated metaphor. (2) V 4.3.29. (3) tanvī meghajalârdrapallavatayā dhautâdharêvâśrubhiḥ śūnyêvâ- bharanaiḥ svakālavirahād viśrāntapuspôdgamā | cintāmaunam ivâsthi- tā madhulihām śabdair vinā laksyate caņdī mām avadhūya pādapati- tam jātânutāpêva sā (Vāmana; Purūravas addresses the creeper as though it were Urvasi: "The thin creeper, its leaves wet with the rain, is like her lip wet with tears; like her dress without jewels, the creeper sleeps through its flowerless season without the sound of bees like my love lost in thought; the cruel thing spurns me fallen at its feet like my love whom I hurt"). (4) "Sleepily she cuddled up in the fold of his left arm, her cheek against his heart, though a hard hand which seemed to be pounding against a wall was trying to wake her up again; she would just let it go on pounding all it pleased. She had to sleep some more! ... But now mother was here. Hurriedly she was transferred into her mother's arms and squeezed almost to a pan- cake. She had to gasp for breath; nevertheless she snuggled into

320 GLOSSARY her arms as closely as she could, for she felt, oh, so sleepy! ... But no peace here, either! Here, too, a hand pounded against a wall. Were they tearing down the house?" (O. E. Rolvaag; the parents' chest; are represented as a wall). (5) Though the element of confusion r present objectively, the figure differs from bhrantimat in that the confusion is subjectively willed, or is at least not the concomitant of error. The identification of the two things here differs from rūpaka, since the mode of the identification is volitional, not conven- tional; a stratum of explicit consciousness is overlaid on the identifi- cation. samāhita (II): (1) same as samādhi. (2) B 3.10, D 2.298 (299).

samuccaya samuccaya, 'accumulation': (1) a figure consisting of the multiplication of descriptive adjuncts to a thing or mood. (2) R 7.19 (20-26), 7.27 (28-29), 8.103 (104), M 178-79. (3) sukham idam etāvad iha sphārasphuradindumandalā rajanī / saudhatalam kāvyakathā suhrdah snigdhā vidagdhāś ca (Rudrața: "It is so pleasant here! The night is brilliant with a glimmering moon; here on the palace roof poetry is read; friends are kind and clever"). (4) "As [the Public] sat, listening to his speeches, in which considerations of stolid plainness succeeded one another with complete flatness, they felt, involved and supported by the colossal tedium, that their confidence was finally assured" (Lytton Strachey; Lord Hartington is quite a bore). (5) Rudrata gives many examples showing not different kinds of accumulation. but differences in the kinds of things accumulated. In general. his distinctions follow the canonical fourfold pattern of jāti, kriyā. guņa, and dravya. The accumulation of states, events, or adjuncts is. of course, designed to re-enforce the tone of a description, as in parikara alamkāra, where the qualifications or epithets alone of a thing are multiplied. Formally, samuccaya is the same as dipaka alamkāra, but in fact the motives underlying their use are diametrically opposed: here the emphasis is on the quantity of adjuncts; in dipaka. it is on the single word which bears the syntactical force of the whole sentence.

sahôkti sahôkti 'speech containing the word "with"': (1) a figure in which two separate things or ideas are represented as conjoined or occurring at once. (2) B 3.38 (39), 3.17, D 2.351 (352-56), V 4.3.28,U 5.15,

GLOSSARY AP 344.23, R 7.13-18, 8.99 (102), M 170. (3) bhavadaparādhaih 321

sārdham samtāpo vardhatetarām tasyāh / kşayam eti sā varāki snehena samam tvadiyena (Rudrata: "Together with your unfaith- fulness, her suffering grows and grows; the unfortunate girl is ruined along with your love for her"). (4) "Hark, hark! the lark at heaven's gate sings, / And Phoebus gins arise, / His steeds to water at those springs / On chaliced flowers that lies; / And winking Mary-buds begin / To ope their golden eyes: / With everything that pretty is, / My lady sweet, arise; / Arise, arise!" (Shakespeare). (5) Accom- paniment is, of course, one type of association, and this latter notion can always suggest a comparison (as in the second example). Some writers, notably Vāmana and the Agni Purāna, have emphasized the comparative aspect of sahôkti; others have preferred to leave its definition purely formal (as in the first example). Rudrata allows both. The many definitions and examples which Rudrata gives are not distinguished by any special terminology, and in fact the discriminations are quite extrinsic and depend on the actual relation- ship of the things said to be accompanying one another. One example refers to two things which, in fact, do physically accompany one another (breasts and attractiveness), another to two things which are always complimentary (man's unfaithfulness and woman's sorrow), the third to two things which are merely contemporary, never con- tigent. Dandin's three examples show contemporaneous association, either through qualities (as: "saha dīrghā mama śvāsair imāh sam- prati rātrayah", 'the nights are long, and my sighs'), or through actions (as: "yānti sārdham janânandair vrddhim surabhivāsarāh", 'the spring days are fulfilled, along with the peoples' joys'). Mam- mața has taken the obvious step of inventing a figure vinôkti ('exclu- sion').

sādṛśya sādrśya, 'similitude': (1) a generic term for simile (upamā) and those figures based upon it. (2) AP 344.5 (6-24). (5) See aupamya, a term used by Rudrata with the same intent. The term here refers to upamā, rūpaka, sahôkti, and arthântaranyāsa.

sāmānya sāmānya, 'identity': (1) same as ananyatva atiśayôkti. (2) M 202.

322 GLOSSARY

sāmya sāmya (I), 'likeness': (1) a figure in which the subject of comparison is represented as fulfilling the function of the object of comparison. (2) R 8.105 (106). (3) abhisara ramaņam kim imām diśam aindrīm ākulam vilokayasi / śaśinah karoti kāryam sakalam mukham eva te mughde (Rudrata: "Go to your lover! Why are you gazing distract- edly at the eastern sky? O foolish girl, your face is accomplishing the work of the moon!"). (4) "O saw ye not fair Ines? / She's gone into the West, / To dazzle when the sun is down, / And rob the world of rest" (Thomas Hood). (5) Cf. pihita alamkāra, where one quality superimposes itself upon another. Here the two subjects are not confused, though their functions make them interchangeable. sāmya (II): (1) a figure in which the subject and object of comparison differ only modally. (2) R 8.107 (108). (3) mrgam mrgânkah sahajam kalañkam bibharti tasyās tu mukham kadācit / āhāryam evam mrganābhipattram iyān aśeşeņa tayor viseșah (Rudrața: "The rabbit- marked moon is inherently spotted; her face, however, is only oc- casionally marked with lines of musk. This is the only difference between them"). (4) "Far more welcome than the spring; / He that parteth from you never / Shall enjoy a spring forever" (William Browne). (5) In these examples, one of the terms compared is represented as existing continuously, while the other exists only temporarily. The figure is thus distinguished from catu upamā in that the difference there is simply overlooked, and from atiśaya upamā in that the difference is there reduced to the bare fact of separate existence. All three figures agree in ignoring the concrete difference between the things compared, that is, their sharing the common property in different degrees.

sāra sāra, 'pith': (1) a figure wherein is expressed a concatenated series such that each succeeding term expresses a characteristic improvement in relation to the preceding. (2) R 7.96 (97), M 190. (3) rājye sāram vasudhā vasumdharāyām puram pure saudham / saudhe talpam talpe varânganânangasarvasvam (Rudrața: "In the kingdom, the earth is best, on the earth, the capital, in the capital, your palace, in the palace, a bed, in the bed, the entire wealth of the Love God-that beautiful woman"). (4) "The mouth seemed formed less to speak then to quiver, less to quiver than to kiss, (some might have added) less to kiss than to curl" (Thomas Hardy). (5) This figure amounts to a

GLOSSARY 323

sequence of related comparatives; here the comparative degree of the adjective connects the series of terms as, in ekávali, they were connected through the notion of qualification and, in kāraņamālā, through the relation of cause and effect. See also the various similes and metaphors extended into sequences (mālā). sūkșma sükșma (I), 'subtle': (1) a figure in which an intention or idea is said to be conveyed through a gesture, glance, or means other than language. (2) B 2.86, D 2.235, 260 (261-64), M 189. (3) padmasammīlanād atra sūcito niśi samgamah / āśvāsayitum icchantyā priyam angaja- pīditam (Dandin: "Putting a lotus on, she indicates a rendez-vous that night, desiring to comfort her love-lorn favorite"). (4) "The voice was Southern and gentle and drawling; and a second voice came in immediate answer, cracked and querulous :- 'It ain't again. Who says it's again? Who told you, anyway?' And the first voice responded caressingly :- 'Why, your Sunday clothes told me, Uncle Hughey. They are speakin' mighty loud o' nuptials'" (Owen Wister). (5) The Sanskrit example is literal, the English attributes an intention to an inanimate object. This figure is one of several dealing with communication (see paryāya, vakrôkti, leśa). It may seem curious that a figure of speech could be defined in an area which is specifically non-verbal or extra-verbal. In the first place, the figure is a representation of such an occurrence in words and may be com- patible with figurative usage. Secondly, taken in the context of the other figures indicating an intention of the speaker, the present figure has a certain place in the system and may not be ignored. sūkșma (II): (1) same as utprekșā of Daņdin. (2) R 7.98 (99).

smaraņa smaraņa, 'recollection': (1) a figure conveying a recollection based on similitude. (2) M 199. (3) nimnanābhikuharesu yad ambhah plāvitam caladrśām laharībhiḥ / tadbhavaih kuharutaih suranāryah smāritāh suratakaņharutānām (Mammața: "The water of the waves laps in the deep navels of the swift-glancing nymphs, reminding one of goddesses making love and the throaty murmurings of their pas- sion"). (4) "Her presence brought memories of such things as Bour- bon roses, rubies and tropical midnights; her moods recalled lotus- eaters and the march in 'Athalie'; her motions, the ebb and flow of the sea; her voice, the viola" (Thomas Hardy). (5) The point of

GLOSSARY 324 this figure is that one of the relations upon which recollection can be based is similitude (the others are contiguity, contemporaneity, etc.), and such recollection constitutes an implicit simile. Another name for this figure is kāvyahetu. svabhāvôkti svabhāvôkti, 'telling the nature (of a thing)': (1) a figure in which a natural or typical individual is characterized. (2) B 1.30, 2.93, D 2.8 (9-13), U 3.5, R 7.30 (31-33: jāti), M 168. (3) dhūlīdhūsaratanavo rājyasthi- tiracanakalpitâikanrpāḥ / krtamukhavādyavikārāḥ krīdanti suni- rbharam dimbhah (Rudrata: "The children are intent at their play, bodies gray with dust, one among them chosen to occupy the station of a king, assuming grave miens and voices"). (4) "On the day when I first learned of my father's journey, I had come back with two companions from a satisfactory afternoon in the weeds near Kay's Bell Foundry, shooting a slingshot at the new bells, which were lying out in the yard and strung up on rafters. Struck with rocks, they made a beautiful sound, although it seemed to upset Mr. William Kay, the proprietor. His sign, "Maker of Church, Steamboat, Tavern and Other Bells", hung over the doorway of his barnlike shop and had a row of little brass bells swinging beneath, squat and burnished, but these were hard to hit, and if you missed them, you were apt to hit one of the men working inside, and this was what seemed to upset Mr. William Kay most of all. So toward the end of the afternoon he pranced out with a double barreled shotgun loaded with pepper and blistered Herbert Swann's seat as he zigzagged to safety through the high grass" (Robert Lewis Taylor). (5) Both examples describe children at play. The poetic nature of this figure constitutes one of the longest standing disputes of the alamkāraśāstra. In fact, the oldest writer, Bhāmaha, specifically objects to this figure on the ground that it does not involve vakrôkti or the figurative turn of phrase essential in any poetry. This point of view is taken up in more detail much later by Kuntaka (Vakrô- ktijīvita 1.11). Dandin, however, followed by most of the other alamkāra writers, has included svabhāvôkti, although in a much discussed verse (2.363), he seems to distinguish that figure from the rest of the alamkāras. The poetic basis of the figure is probably to be sought in the genre called jāti: short verses, extremely condensed yet full of minute detail, each one attempting to seize the instantane- ous totality of a certain event, or an individual as wholly characteristic

GLOSSARY of a genus. such as a hawk diving on a scrap of food or a village 325 woman drawing water from the well (Subhūşitaratnakoşa, verscs 1148-92). Here the charm lies precisely in the completeness of the description within the limits imposed by the verse, and not on any figurative usage. Yet that charm is undeniable. Dandin includes in his definition of the figure not merely that it is descriptive, but that that description reflects the 'different aspects' (nanâvastha) of the thing. Rudrata. likewise. is far from thinking svabhavokti mere description: his criteria specify that the rendering be appropriate to the time and characteristic of the thing (avasthôcita). Some con- texts are simply more interesting than others; his description of the young woman being made love to. though clinically exact, illustrates this appropriateness of subject, detail, and occasion perfectly. Dandin gives four examples of svabhāvôkti, as the description emphasizes one of the four metaphysical categories jāti ('type'), kriyā ('act"), guņa ('attribute'), or dravya, ('individual'). The distinction is arbitrary in the sense that any good example of svabhavokti will express that perfect blending of type, action, attribute, and individual which we tend to call "characteristic unity". But the emphasis may vary from one case to another. A speculative discussion of the rationale for including this figure within figurative usage will be found in the Introduction, pp. 42ff. kriyā, 'verb': (1) a type of svabhāvôkti in which the behavior of the subject is emphasized. (2) D 2.13 (10). (3) kalakvaņitagarbheņa kaņthenâghūrņitêkșaņaiḥ | pārāvatah paribhramya riramsuś cumbati priyām (Daņdin: "A pigeon wanders about with eyes rolling and soft cooings coming from its throat and, desiring its mate, kisses her"). (4) "Save yow, I herde nevere man so synge / As dide youre fader in the morwenynge. / Certes, it was of herte, al that he song. / And for to make his voys the moore strong, / He wolde so peyne hym that with bothe his yen / He moste wynke, so loude he wolde cryen, / And stonden on his tiptoon therwithal, / And strecche forth his nekke long and smal" (Geoffrey Chaucer). (5) See jāti, guņa. dravya. guņa, 'attribute, adjective': (1) a type of svabhāvôkti in which the contex- tual or descriptive attributes of the subject are emphasized. (2) D 2.13 (11). (3) badhnann angeşu romāñcam kurvan manasi nirvrtim | netre câmīlayann eșa priyāsparśah pravartate (Dandin: "The touch of the beloved provokes a tingling in the limbs, happy release in the mind, a closing of the eyes"). (4) "The open hills were airy and

326 GLOSSARY clear, and the remote atmosphere appeared, as it often appears on a fine winter day, in distinct planes of illumination independently toned, the rays which lit the nearer tracts of landscape streaming visibly across those further off; a stratum of ensaffroned light was imposed on a stratum of deep blue, and behind these lay still remoter scenes wrapped in frigid grey" (Thomas Hardy). (5) Guna svabhāvókti is to be distinguished from jāti svabhāvôkti, where the constitutive or generic aspects of the subject are emphasized. jāti, 'genus': (1) a type of svabhāvôkti in which the generic aspects of the subject are emphasized. (2) D 2.13 (9). (3) tundair ātāmrakuțilaih pakşair haritakomalaiḥ | trivarņarājibhiḥ kanțhair ete mañjugiraḥ śukāh (Dandin: "With beaks bent and reddish, with wings yellow and soft and throats bearing three-colored stripes, the parrots call softly"). (4) "She was an egregious fowl. She was huge and gaunt, with great yellow beak, and she stood straight and alert in the manner of responsible people. There was something wrong with the tail. It slanted far to one side, one feather in it twice as long as the rest. Feathers on her breast there were none. These had been worn entirely off by her habit of sitting upon potatoes and other rough objects .... Her legs were blue, long, and remarkably stout" (Owen Wister). (5) See kriyā, guņa, dravya. dravya, 'material': (1) a type of svabhāvôkti in which a typical individual is characterized. (2) D 2.13 (12). (3) kanthekālah karasthena kapālenêndušekharaḥ | jațābhiḥ snigdhatāmrābhir āvir āsīd vrşadhva- jah (Dandin: "Śiva comes into view atop his bull, with dark blue neck, a skull in his hand, the moon on his brow, and matted locks, reddish and greasy"). (4) "And on her legs she painted buskins wore, / Basted with bands of gold on every side, / And mails between, and laced close afore: / Upon her thigh her scimitar was tied, / With an embroidered belt of mickle pride; / And on her shoulder hung her shield, bedecked / Upon the boss with stones, that shined wide. As the fair moon in her most full aspect, / That to the moon it mote be like in each respect" (Edmund Spenser; of Radigund). (5) This type of svabhāvôkti is doubtless designed to cover that large category of exceptional individuals whose attributes in Indian story have become standardized. The stotra literature abounds in such descrip- tive verses of gods and godlings. Presumably the description of ordinary individuals would fall into the category jāti svabhāvôkti.

GLOSSARY svarūpa 327

svarūpa. 'nature : (I) probably the same as svabhāvôkti. (2) AP 344.3-4. (5) The figure is divided into sāmsiddhika (nija) and naimittika (agantuka). It is the first arthâlamkāra of the Agni Purāna. hetu betu. 'cause': (1) a figure in which an effect is described along with its cause. (2) D 2.235-59, AP 344.29, R 7.82 (83); refuted by B 2.86, M 186. (3) aviralakamalavikāsah sakalâlimadaś ca kokilânandah / ramyo'yam eti samprati lokôtkanthākarahı kālaḥ (Rudrața; descrip- tion of the springtime and its several effects: "The delightful season progresses; men fall in love, nightingales rejoice, drunken bees hover about the unbroken spread of lotus blooms"). (4) "Beneath this slab / John Brown is stowed. / He watched the ads / And not the road" (Ogden Nash). (5) This is the most controversial alamkāra. It would seem to be nothing but literal description, like the figure svabhāvôkti, and it has been rejected by Bhāmaha and Mammata for that reason, for they feel that an alamkāra must repose upon some figurative usage (Mammata does in effect resuscitate the figure hetu as kāvyalinga, q.v.). But, as usual, such objections miss the point: those authors who accept hetu are far from thinking it mere literalism, judging by the examples which they give. All involve some striking, though not necessarily deformed or unnatural (cf. vyāghāta, asamgati, etc.) instance of the cause-effect relation. Though the cause of John Brown's death is given literally, it touches upon other issues which strike a responsive chord in the reader's mind, and he is pleased. It would be said by Anandavardhana that in this instance, the figure hetu was nothing but a means to the expression of a dhvani (sugges- tion) regarding the ubiquity of billboards, etc. In such considerations may be said to reside the alamkāratā of the figure hetu. Daņdin gives an extensive inventory of examples based, oddly enough, on the philosophical analysis of cause: it is a cause either of knowledge (jñāpaka) or of work (kāraka); it may increase (upabrhaņa) or diminish (nivrtti); the cause may be non-existent (abhāva), simulated (prāpya), immediate (tatsahaja), or mediate (dūrakārya). Within the figure, the cause may be appropriate (yukta) or inappropriate (ayukta) to its effect. This last amounts to a mixed metaphor. abhava, 'non-existence': (1) a type of hetu in which a certain cause is absent, and the effect is the appropriate contrary. (2) D 2.246 (247-52). (3)gatah kāmakathônmādo galito yauvanajvaraļı / kșato mohaś

328 GLOSSARY cyutā trsnā krtam punyasrame manab (Dandin: "Gone is the fascina- tion of the tales of love, slipped away the fever of youth; error is ended, desire lost; my mind is fixed on the holy retreat"). (4) "'And yet you incessantly stand on your hcad- / Do you think, at your age, it is right?' / 'In my youth', Father William replied to his son, / 'I feared it might injure the brain; / But, now that I'm perfectly sure I have none, / Why, I do it again and again'" (Lewis Carroll). (5) Dandin, with careful scholarship, illustrates the four kinds of non-existence which are recognized by the logicians: prāgabhāva (non-existence preceding becoming), pradhvamsâbhāva (non-existence following cessation), anyonyâbhāva (difference), and atyantâbhāva (unqualified non-existence). Both the examples given are of pradham- sâbhāva. ayukta, 'inappropriate': (1) a type of hetu wherein the cause is expressed figuratively and in such a way that, beyond the figure, the cause is improperly related to the effect. (2) D 2.253 (258). (3) rājñām hastâravindāni kudmalīkurute kutaḥ / deva tvaccaraņadvandvarāga- bālâtapah sprśan (Daņdin; lotuses do not normally close during the day, but the hand-lotuses of the court do, in the presence of the Sun-King; "King, why do the hand-lotuses of your court close when they have been touched by the red suns of your lac-anointed feet?"). (4) "To Daisies, not to shut so soon: / ... / Stay but till my Julia close / Her life-begetting eye, / And let the whole world then dispose / Itself to live or die" (Robert Herrick; the "inappropriateness" is only in the poet's wish that the "sun" of Julia's eye should cause the daisies to bloom into the night). (5) See yukta. In these two cases, the figurative expression, in the sense of a non-natural representation, does not attach to the cause itself, but only to the terms which serve as cause and effect. In this, they are different from such figures as vyāghāta, asamgati, etc., where the nature of the relation of cause to effect is expressed differently. The "inappropriateness" of ayukta consists in the wrong effect being said of that cause, and it is excused by both cause and effect serving as objects of comparison to other terms, onto which the relation of cause-effect is thereby transferred. These terms, as subjects of comparison, can support the non- literalness or inappropriateness inherent in the basic expression. upabrhaņa, 'augmentation': (1) a type of hetu wherein the modality of the cause is increase or augmentation. (2) D 2.237 (236). (3) ayam āndolitapraudhacandanadrumapallavaḥ / utpādayati sarvasya prītim malayamārutah (Daņdin; "produces" pleasure: "The southern wind

GLOSSARY 329 brings pleasant change to all, swaying the buds of the bursting sandal trees"). (4) "Here lies the body of Mary Ann Lowder, / She burst while drinking a seidlitz powder" (Anon.). (5) Cf. nivrtti (diminution); these are the two types of nirvrtya 'transformation'), which is, in turn, a sub-variety of kāraka hetu (see hetu). käraka, 'productive': (1) a type of hetu wherein the effect described has the form of a work or thing. (2) D 2.235. (3) See under upabrhana, nivrtti, vikārya, or prāpya. (4) "There was a young belle of old Natchez / Whose garments were always in patchez. / When comment arose / On the state of her clothes, / She drawled, 'When Ah itchez, Ah scratchez'" (Ogden Nash). (5) The two principal subspecies of hetu are this and jñāpaka, wherein the effect is knowledge; for example: "The Grizzley Bear is huge and wild / He has devoured the infant child. / The infant child is not aware / He has been eaten by the bear" (A. E. Housman). Aristotle's two orders are meant. kāryânantaraja, '(cause) produced after the effect': (1) a type of hetu in which the order of cause and effect is reversed. (2) D 2.253 (257). (3) paścāt paryasya kiraņān udīrņam candramaņdalam / prāg eva hariņâksīņām udīrņo rāgasāgarah (Daņdin; first came woman's passion, then the moon: "Last arose the moon's full orb, scattering its rays, but first of all, the passion-ocean of young women, doe eyed"). (4) "For the leaf came / Alone and shining in the empty room; / After a while the twig shot downward from it; / And from the twig a bough; and then the trunk, / Massive and coarse; and last the one black root. / The black root cracked the walls ... " (Conrad Aiken). (5) This is the third term in the trilogy dūrakārya (effect produced long after cause), tatsahaja (produced simultane- ously), kāryânantaraja (effect produced before the cause). Compare pūrva, which is the same as kāryânantaraja except that the end of the inversion need not be exaggeration of a quality. In Aiken's example, the exaggeration is used simply to assert an introspective mood. jñāpaka, 'causing to know': (1) a type of hetu in which the effect has the form of an idea or of information learned. (2) D 2.235 (244-45). (3) gato'stam arko bhātîndur yānti vāsāya pakşiņah / itîdam api sādhv eva kālâvasthānivedane (Daņdin: "The sun has set, the moon smiles down, the birds have gone to rest; this is enough to tell the time of day"). (4) "Roy's club was sedate. In the ante-chamber were only an ancient porter and a page; and I had a sudden and melancholy feeling that the members were all attending the funeral

330 GLOSSARY of the head waiter" (Somerset Maugham). (5) Scc kāraka hetu. dūrakārya, 'whose effect is at a distance': (1) a type of hetu in which the cause (or condition) is materially separated from its effect; whercin the two terms imply different levels of reality or differ as to sub- stratum. (2) D 2.253 (255). (3) tvadapângâhvayam jaitram anangá- stram yad angane / muktam tad anyatas tena so'py aham manasi kşatah (Dandin; the lady shoots her glances elsewhere, and my heart is wounded: "O woman, the well-aimed arrow of Lovc-that goes by the name of your sidelong glance-was shot elsewhere, but I, here, am struck in the heart!"). (4) "I sent thee late a rosy wreath, / Not so much honouring thee / As giving it a hope that there / It could not wither'd be; / But thou thereon didst only breathe, / And sent'st it back to me; / Since when it grows, and smells, I swear, / Not of itself but thee!" (Ben Jonson). (5) See kāryânantaraja; dūrakārya is evidently the same as asamgati alamkāra and virodha II, but a nuance deserves to be noted. If the "cause acts at a distance", then "'I blow my breath', said the old man, 'and the stream stands still ... '. 'I breathe', said the young man, 'and flowers spring up over the plain'" would serve as an example. But it lacks the "ap- pearance of contradiction" which Vāmana deems essential. nirvartya, `to be accomplished': (1) a type of kāraka hetu in which a transformation of quantity or intensity is reflected. (2) D 2.240. (3) (4) See the two subtypes, upabrhana and nivrtti. (5) Nirvartya is distinguished from vikārya hetu, where a qualitative transforma- tion is intended. nivrtti, 'cessation': (1) a type of hetu wherein the modality of the cause is diminution or cessation. (2) D 2.237 (238). (3) candanâranyam ādhūya sprstvā malayanirjharān | pathikānām abhāvāya pavano'yam upasthitah (Dandin: "The wind out of the south, touching springs and sandal forests in the southern mountains, is destined to relieve the weary wanderer"). (4) "The world is too much with us; late and soon, / Getting and spending, we lay waste our powers" (William Wordsworth). (5) See upabrhana; both are types of nirvartya hetu. prāpya, 'acquirable': (1) a type of kāraka hetu in which the cause and effect are simulated. (2) D 2.240 (243). (3) mānayogyām karomiti priyasthānasthitām sakhīm / bālā bhrūbhangajihmâkșī paśyati sphuri- tâdharā (Daņdin: "'I shall pretend I am angry', she said to her friend who stood before her in the place of the lover, and her brows arched, her eyes narrowed, her lovely lips trembled"). (4) "This was one of their old games, this elaborate envisaging of situation and outcome

GLOSSARY 331 'I should say, "Mr. Kitson, I think it better to be direct with you'." 'Good', said Humphrey, 'And next ...? ' 'Then he will say, "Excellent. Go ahead, Mrs. Pelham, go ahcad." I shall take my time: I shall give him a long look, then say ... "Why did you do this thing, Mr. Kitson?"'" (Pamela Hansford Johnson). (5) This is a kāraka hetu in which no transformation at all takes place, that is, in which simulation takes the place of action. Cf. nirvartya and vikārya. yukta, 'appropriate': (1) a type of hetu in which a cause is represented in figurative terms, but in such a way that, within the figure, the cause is appropriate to its effect. (2) D 2.253 (259). (3) pāņipadmāni bhūpānām samkocayitum īšate / tvatpādanakhacandrāņām arcișah kundanirmalāh (Dandin; here the [foot]-moon of the king closes the [hand]-lotuses of the retainers: in fact, the moon does cause this type of lotus to close; it blooms only during the day: "The jasmine- white rays from the ten moons of your toenails, O King, make the hand-lotuses of your courtiers to close"). (4) "Suddenly, as if the movement of his hand had released it, the load of her cumulated impressions of him tilted up, and down poured in a ponderous avalanche all she felt about him" (Virginia Woolf; outside the figure, the hand does, in fact, actuate the mechanism of dump trucks, etc.). (5) The contrasting term is ayukta, where the relation of cause and effect is, in literal terms, impossible. Cf. yukta and ayukta rūpaka, to which the present terms are but the addition of a causal relation. vikārya, 'to be transformed': (1) a type of kāraka hetu wherein the modality of the cause is alteration of quality. (2) D 2.240 (242). (3) utpravālāny araņyāni vāpyaḥ samphullapankajāḥ | candra pūrņaś ca kāmena pānthadrster vişam krtam (Daņdin; the traveller is separated from his lover: "The forests are sending forth new shoots, the tanks are full of lotuses, the moon is full; but love turns all this to poison in the eyes of the traveller"). (4) "All good Englishmen go to Oxford or Cambridge and they all write and publish books before their graduation, / And I often wondered how they did it until I realized that they have to do it because their genteel accents are so developed that they can no longer understand each other's spoken words so the written word is their only means of intercommunication" (Ogden Nash; the written word replaces the spoken word). (5) See hetu; vikārya is distinguished from nirvartya, where the modality is aug- mentation or diminution (change of quantity). Here the effect reflects a change of nature (quality as opposed to quantity).

GLOSSARY 332 sahaja, 'congenital': (1) a type of hetu in which the cause and effect appear simultaneously. (2) D 2.253 (256). (3) āvir bhavati narinām vayah paryastaśaiśavam / sahâiva vividhaih pumsām angajonmāda- vibhramaih (Dandin: "Maidenhood appears in girls, their childhood passed, and with it, in young men, the several passionate confusions of love"). (4) "All in green went my love riding / on a great horse of gold / into the silver dawn. / four lean hounds crouched low and smiling / my heart fell dead before" (e. e. cummings). (5) Sahaja forms the middle term in the triad dūrakarya ... kāryânantaraja. In all three, the cause and the effect are vyadhikarana, different as to substratum. The cause operates figuratively or through mental processes; hence Daņdin has included the three terms in jñāpaka hetu. They do not necessarily imply deformations of the relation of cause-effect, although other writers would say so (cf. notably the figure asamgati).

APPENDIX

Figures Defined for the First Time in the Encyclopedic Writers after Mammata

The number in parentheses refers to the serial order of the figure in the Kuvalayānanda of Appayyadīksita, from which we take the examples. atyukti (96): the representation of great nobility, compassion, etc. (variation on udātta); as: "by your gifts, King, beggars become horns of plenty!" anuguņa (78): a second quality augments or sets off the first (a variety of tadguna, the modality is association rather than imposition); as: "the blue lotuses appear a deeper blue in the presence of her sidelong glances". anujñā (71): an affected desire for a defect, setting off a quality (a virodha in which the contraries are specified as guna and dosa; the example is also an āśiș); as: "may our sufferings be prolonged, that we may praise Hari the more fervently". anupalabdhi (115): a poetic application of the mīmāmsaka pramāņa 'abhāva' (cf. pratyakșa (108)); as: "your delicate waist, clearly unable to support the weight of your full breasts, is deemed not to exist by all observers". arthāpatti (59, 114): the mīmāmsaka pramāna 'a fortiori (cf. pratyaksa (108)), the only one generally considered a figure before Appayya, and undoubtedly the analogical source for the others; as: "the moon himself has been conquered by your face; what hope is there for the lotuses?" alpa (42): the support exceeds the supported in minuteness (inversion of adhika); as: "the jeweled finger ring on your hand seems more like a rosary". avajñā (70): the description of an absent quality or defeet by means of

334 APPENDIX one present (a variety of ullāsa (69); apparently a vinok ti determined as guna or doșa); as: "let the lotuses close; what harm will there he for the moon?" (ex .: doșa). unmilita (81): a milita followed by the distinction reborn; as: "the God, seeing the Himälaya hidden behind your accumulation of glory, O King, gave it new life as ice" (i.e., indistinguishable in whiteness, the mountain is again made distinct as cold). upamāna (110): the mīmāmsaka pramāna 'comparison' (this one accepted also by other schools): in exemplification indeed difficult to distin- guish from the ordinary upama. Here the encyclopaedic spirit may be seen to have come full circle, with all sense lost for what in fact was being catalogued. ullāsa (69): description of a quality or defect by means of an associated one (cf. avajñā (70); a sahokti determined as guņa or doșa); as: "the Ganges, as it were, desires to be purified by your bath". ullekha (7): the description of one thing as many, in many contexts (in effect, a mālā rūpaka; in intent, perhaps a samuccaya); as: "he is indeed a Cupid with women, a wishing-tree to his suitors, death to his enemies". aitihya (117): the quasi-pramāna 'tradition' (allowed as authoritative in mīmāmsa and other systems when it does not conflict with Veda); as: "in the words of this old song, 'Joy accompanies the man living even for a hundred years'". güdhokti (87): veiled reference in the presence of the person really referred to (a possible specialization of several figures: paryāyokta personalized; or samāsokti with presence of implicit term specified; or leśa (I), again limited to persons. The figure, as defined, resembles the anyokti of Rudrata); as: "bull, leave this strange field! Here comes the gamekeeper" (reference is to the lover of an adulteress). nirukti (97): one word pun on a name; resembles tattva śleșa; as: "by your actions [inflaming the separated lover] you [Moon] are rightly called "doşākara" ['mine of defects, nightmaker']". Following atyukti, this figure would seem to have some relation to it, such as "ironic description of great iniquity" or the like; but none is apparent in the definition. parikarāūkura (25): accumulation of epithets distinguishing not the is subject (ef. parikara) but the object of the reference; as: "giver of the four goals of men, the God is four armed" (his four arms relate not to himself, but to what he does). pariņāma (6): a property of the subject of comparison is transferred to

APPENDIX 335 the object of comparison; in eflect, a vikriyā upama; as: "she of the intoxicated glance was looked at by the calm eyes of lotuses". pūrvarūpa (76): a tadguna followed by reappropriation of the lost quality; as: "though blue in reflection of Siva's neck, you regain whiteness from your own glory" (addressed to Pārvati). pratyaksa (108): the mīmāmsaka pramāņa 'perception' (Appayya appends to the Candrāloka of Jayadeva 17 figures, 10 of which are poetic applications of pramanas 'modes of true apprehension' or quasi- pramāņas such as aitihya, 'tradition', admitted in the mīmāmsā. This innovation probably dates from Bhojarāja; as: "the five senses attain perfection [in the presence of] this liquor, reflecting the beauteous faces of young women, smelling of fresh sahakāra blooms, sweet tasting, surrounded with buzzing bees, and cool". prastutāūkura (28): a relevant subject illuminates another relevant subject (instead of an apparently irrelevant subject; a variety of aprastuta- praśamsā); as: "you have the malatī flower, bee! What do you want with the ketaki, full of thorns?" This seems but a repeated aprastu- tapraśamsā, as the aprastuta-the girl he is wooing in the pleasure garden-is still obliquely referred to. praharșaņa (67): the attainment without effort of a desired goal (id., samādhi); as: "even as he thought of her, came a messenger suggest- ing rendez-vous". bhāvodaya, etc. (105-107): the dramatic moments of augmenting, diminish- ing, and muting an emotion. A late reaction showing an attempt to figurize the rasa theory. Cf. Ohv. 2.3. mithyādhyavasiti (65): an arthāntaranyāsa in which an impossible proposition is justified by another impossible proposition (cf. nidarśanā (II)); as: "he who would gain the fidelity of a prostitute does indeed wear the celestial lotus". ratnāvalī (74): the use of several descriptive epithets, each suggesting a different object of comparison (in effect, a mālā samāsokti): as: "O King, you are of clever mien [four-faced], the master of good fortune [husband of Lakşmī], knowledgeable [omniscient]" (refer- ences are to Brahmā, Vișņu, and Rudra). lalita (66): the description of a similar situation (id. anyokti): as: "now that the water has flowed away, she seeks to build a dam" (she seeks to restrain a lover already bound to another). vikalpa (54): a virodha determined as a choice; as: "bend, O Kings, either your heads or your bows!" vikasvara (62): an arthāntaranyāsa in which the justification is stated as a

336 APPENDIX

simile; as: "he is unconquerable, for the great are as difficult to tame as the sea". vicitra (40): the cause, though having a contrary attribute, yet produces the desired effect (cf. visama (III); as: "the pious bow down in order to obtain exaltation from on high". vidhi (99): pleonasm; the establishment of what is already established; as: "on hearing the fifth tone [a particularly sweet one], the kokila bird becomes truly a kokila bird!" Ironic contrary of mimāmsā usage? vivrtokti (88): a güdhokti (87) which the poet clarifies in other terms. vişādana (68): the attainment of a goal contradicting that desired (inverse of praharsana (67)); as: "as he was about to light the flame, the wind blew it out". śabdapramāņa (111): one of three figures representing the mīmāmsaka pramāna 'verbal authority' (cf. smrti (112), śruti (113)). It seems only a cover term for the latter two figures, q.v. śruti (113): the mīmāmsaka pramāņa 'verbal authority' (viz. śruti as opposed to smrti); as: "you are truly called 'Varada' ['giver of boons']" sabhava (116): the quasi-pramana 'possibility'; as: "some day I will find a congenial soul, for the world is wide and time is endless!" srti (112): the mīmāmsaka pramāņa 'verbal authority' (viz. smrti as opposed to śruti); as: "you are forced to do evil. Yet these deeds are not done, for Manu has said 'all things done from force are indeed not done'".

BIBLIOGRAPHY

Abhinavagupta. See Bharata. Agni Purāna, ed. G. B. Kale. (= Ānandāśrama Sanskrit Series, 41.) (1957). Anandavardhana. See Jacobi, H. and Dhvanikāra. Annambhatta. Tarkasamgraha, ed. Y. V. Athalye (= Bombay Sanskrit and Prakrit Series, 55.) 2nd ed. (Poona, 1963). Appayyadīkşita. "Kuvalayānanda", with the "Candrāloka" of Jayadeva. 10th ed.

Bādarāyaņa. The Vedāntasūtras of Bādarāyana, with the Commentary by Sankara, (Bombay, 1955).

trans. George Thibaut. (= Sacred Books of the East, 34.) (Oxford, 1890). Bhāmaha, Kāvyālamkāra. Appendix VIII to The Pratāparudrayaśobhūşaņa of Vidyā- nātha, ed. K. P. Trivedi. (= Bombay Sanskrit and Prakrit Series, 65.) (Bombay, 1909). -. Kāvyālamkāra; with the Udyāna Vrtti, ed. D. T. Tatacharya (Tiruvadi, 1934). Bharata Muni. Natyasastra; with the Commentary of Abhinavagupta, ed. M. Rama- krishna Kavi. 4 vols. (= Gaekwad Oriental Series, 36, 68.) (Baroda, 1926, 1934). Bhoja. Sarasvatikanthābharana; with the Commentary of Ratneśvara, ed. Jibananda- vidya (Calcutta, 1894). Böhtlingk, O., and R. von Roth. Sanskrit-Wörterbuch. 7 vols. (St. Petersburg, 1852- 1875). Croce, Benedetto. Breviario di Estetica. 12th ed. (Bari, 1954). Trans. D. Ainslie, "Breviary of Aesthetic", (= Rice Institute Pamphlet, Dec., 1915), pp. 223-310. Dandin. Dandin's Poetik (Kāvjādarša), ed. O. Böhtlingk (Leipzig, 1890). -, Kāvyādarśa; with the Commentaries of Vādijanghāladeva, Taruņavācaspati, and the anonymous Hrdayamgama. ed. D. T. Tatacharya (Tiruvadi, 1936). Dasgupta, S. N., and S. K. Dey [De]. History of Sanskrit Literature (Calcutta, 1962). De, Sushil Kumar. History of Sanskrit Poetics. 2 vols. 2nd ed., revised (Calcutta, 1960). Sanskrit Poetics as a Study of Aesthetic (Berkeley, 1963). Some Problems of Sanskrit Poetics (Calcutta, 1951). Dhvanikara and Anandavardhana. Dhvanyaloka; with the Locana of Abhinavagupta, ed. Durgāprasād. (= Kāvyamālā, 25.) (Bombay, 1891). Empson, William. Seven Types of Ambiguity (New York, Meridian Books, 1960). Gonda, Jan. "The Meaning of the Word Alamkara", A Volume of Eastern and Indian Studies in Honour of F. W. Thomas (Bombay, 1939). --. Remarks on Similes in Sanskrit Literature (Leiden, 1949). Gotama Muni. Nyāyadaršana. (= Kāšī Sanskrit Series, 43.) (Benares, 1942). Ingalls, D. H. H. See Vidyākāra. Jacobi, Hermann, trans. "Anandavardhana's Dhvanyāloka", Zeitschrift der Deutschen

338 BIBLIOGRAPHY Morgenländischen Gesellschaft, 56 (1902), pp. 392-410, 582-615, 760-789; 57 (1903) pp. 18-60, 311-343. . "Ruyyaka's Alamkarasarvasva", Zeitschrift der Deutschen Morgenländischen Gesellschaft, 62 (1908), pp. 289-336, 411-458, 597-628. "Über Begriff und Wesen der poetischen Figuren in der Indischen Poetik", Nachrichten von der Königlichen Gesellschaft der Wissenschaften zu Göttingen. Philologisch-Historische Klasse, 1908. Jagannātha. Rasagangādhara. (= Benares Sanskrit Series, 8.) (Benares, 1885-1903). Jaimini Muni. Mimāmsādaršana with Šābarabhāşya. (= Kāši Sanskrit Series, 42.) (Benares, 1942). Jayadeva. See Appayyadīkşita. Jhā, Gangānātha. Pūrva Mimāmsā in Its Sources (Benares, 1942). Kane. P. V. History of Sanskrit Poetics. 3rd revised ed. (Delhi, 1961). Keith. Arthur Berriedale. History of Sanskrit Literature (Oxford, 1928). The Karma Mimamsa (London, 1921). . Indian Logic and Atomism (Oxford, 1921). Kumar, Dharminder. "A Study of Dandin". Unpublished Ph.D. dissertation, Punjab University, Chandigarh, 1965. Kumārila Bhaļta. Mimāmsāślokavārtika. (= Chowkhamba Sanskrit Series, 3.) (Benares, 1898-1899). Kuntaka. Vakroktijivita; A Treatise on Sanskrit Poetics, ed. S. K. De. 3rd ed. revised Māgha. Śišupālavadha. 12th ed. (Bombay, 1957). (Original edition by Durgāprasād, (Calcutta, 1961).

1890.) Mammața. Kāvyaprakāśa; with the Commentary of Harišankaraśarman, ed. Dhun- dhirāja Šāstri. (= Kāši Sanskrit Series, 49.) (= Benares, vikrama samvat 2008 [1951]). Nobel, Johannes. Beiträge zur älteren Geschichte des Alamkāraśāstra. (Inaugural- Dissertation, Friedrich-Wilhelms-Universitat zu Berlin, 1911.) Rāghavan, Venkațarāma. Studies on Some Concepts of Alankāra Sāstra. (= Adyar Library Series, 33.) (Adyar, 1942). Rājaśekhara. Kāvyamīmāmsā. (= Gaekwad Oriental Series, 1.) (Baroda, 1916). Renou, Louis. L'Inde classique, II (Paris, 1953). -. "Sur la structure du Kāvya", Journal Asiatique, CCXLVII (1959), pp. 1 ff. Rudrața. Kāvyālamkāra; A Treatise on Rhetoric, with the Commentary of Namisādhu, ed. Durgāprasād and W. L. Paņaśīkar. (= Kāvyamālā, 2.) (Bombay, 1887). Ruyyaka. Alamkārasarvasva. (= Kāvyamālā, 35.) 2nd ed. (Bombay, 1939). -. See also Jacobi, H. Śālikanātha. Prakaraņapancikā. (= Chowkhamba Sanskrit Series, 17.) (Benares, 1903-1904). Śamkara. Brahmasūtraśāmkarabhāşya. 2nd ed. (Bombay, 1938). Tatacharya, D. T., ed. See Bhāmaha; Daņdin. Taylor, Warren. "Tudor Figures of Rhetoric". Unpublished Ph.D. dissertation, (University of Chicago, 1937). Udayanācārya. Nyāyakusumāñjali. (= Kāšī Sanskrit Series, 30.) (Benares, 1912). Udbhața. Kāvyālamkārasārasamgraha; with the Commentary, the Laghuvrtti of

(Poona, 1925). Indurāja, ed. N. D. Banhatti. (= Bombay Sanskrit and Prakrit Series, 79.)

Vācaspatimiśra. Tattvakaumudi. (= Poona Oriental Series, 10.) 2nd ed. (Poona, 1934). (Commentary on the Sāmkhyakārikā of Iśvarakrșņa). Vāmana. Lehrbuch der Poetik (Kāvyālamkāravrtti), ed. Carl Cappeller (Jena, 1875). Vāmana and Jayāditya. Kāšikā; A Commentary on the Sūtras of Pāņini, ed. Śobhita-

BIBLIOGRAPHY 339 miśra and Brahmadattajijňāsu. ( Kāši Sanskrit Series, 37.) (Benares, vikrama samvat 2009 [1952]). Van Buitenen, J. A. B., and Dimock, E. C. Jr., eds. Introduction to Indian Literature Vidyākāra. Subhāşitaratnakoșa, eds. D. D. Kosambi and V. V. Gokhale. ( Harvard (1967). Oriental Series, 42.) (Cambridge, Mass., 1957). Subhaşitaratnakoşa. Trans. D. H. H. Ingalls. (= Harvard Oriental Series, 44.) (Cambridge, Mass., 1965). Viśvanātha. Sāhityadarpaņa, ed. Jīvānandavidyāsāgara. 5th ed. (Calcutta, 1900). Wain, John. "In the Echo Chamber", The New Republic, Aug. 7, 1965, pp. 20ff. Wellek, René, and Austin Warren. Theory of Literature. 2nd ed. (New York, 1956). Wimsatt, W. K., Jr., and Cleanth Brooks. Literary Criticism: A Short History (New Winternitz, Moriz. Geschichte der Indischen Literatur, IlI (Leipzig, 1920). York, 1962). Yāska. Nirukta; With the Nighantu, ed. Sivadattaśarma. (Bombay, vikrama samvat 1969 [1912]).

INDEX

Entries in capitals are major figures; the others are subfigures. Figures followed by * appear only in the Appendix.

akşara yamaka, 225 apūrva samāsôkti, 318 angangi samsrsti. 308 aprastāvabhāj tulyayogitā, 192 acintyanimitta viśeșôkti, 273 APRASTUTAPRAŚAMSĀ (I), 111 ATADGUŅA, 97 APRASTUTAPRAŚAMSĀ (II), 116 ATIŚAYA, 97 APRASTUTAPRAŚAMSĀ (III), 116 atiśaya upamā, 147 aprākaraņika tulyayogitā, 192 ATIŚAYOKTI, 97 abhāva hetu, 327 ATYUKTI*, 333 abhinnakriyā ślea, 295 adbhuta upamā, 148 abhinnapada śleșa, 295 adhyavasāna atiśayôkti, 98 ABHIVYAKTI, 117 adhyāropa aprastutapraśamsā, 112 abhūta upamā, 149 ADHIKA (I), 102 ADHIKA (II), 102 ambuja citra, 178 ayukta rūpaka, 243 ananyatva atiśayôkti, 99 ananvaya upamā, 148 ayukta hetu, 328

aniyama upamā, 148 ayuktakāri arthântaranyāsa, 118

aniścaya samdeha, 313 ARTHA, 117

anuktanimitta viśeșôkti, 273 artha citra, 178

ANUGUŅA*, 333 artha vyatireka, 277

anugrāhyânugrāhaka samsrșți, 309 artha śleşa, 295

ANUJÑA*, 333 ARTHĀNTARANYĀSA, 118 ANUPALABDHI*, 333 ARTHĀPATTI*, 333 ANUPRĀSA, 102 arthâvrtti āvrtti, 128 ANUMĀNA, 108 ardhaparivrtti yamaka, 226 anulomaviloma citra, 178 ardhabhrama citra, 178 anta dipaka, 194 ALPA*, 333 anta yamaka, 225 AVAJÑĀ*, 333 antâdika yamaka, 225 AVAYAVA, 122 ANYÔKTI, 108 avayava (I) rūpaka, 244 ANYONYA, 109 avayava (II) rūpaka, 244 anyonya upamā, 148 avayavi rūpaka, 244 APAHNUTI (I), 109 AVASARA, 122 APAHNUTI (II), 109 APAHNUTI (III), 110 aviruddhakriyā śleșa, 296 avirodhin śleșa, 297 avivakșita drstânta, 200

INDEX AVIŚEȘA, 123 upanāgarika anuprāsa, 103 341 avyaktâmśa samsrsți, 309 upabrhaņa hetu, 328 avyapeta yamaka, 226 UPAMĀ, 140 aślişța rūpaka, 244 upamā rūpaka, 246 ASAMGATI, 123 upamā śleșa, 297 asat nidarśanā, 202 UPAMĀNA*, 334 asamasta rūpaka, 245 upamāna dharmadyotakalupta upamā, asamāsa rūpaka, 245 152 ASAMBHAVA, 123 upamānalupta upamā, 152 asambhava upamā, 149 UPAMĀRŪPAKA (I), 170 asambhāvita upamā, 149 UPAMĀRŪPAKA (II), 171 asādhāraņa upamā, 150 UPAMĀSAMUCCAYA, 171 AHETU, 124 upameyadyotakalupta upamā, 153 ahetu vyatireka, 277 upameyadharmadyotakalupta upamā, 153 ākşipta vyatireka, 277 UPAMEYOPAMĀ, 171 ĀKȘEPA, 124 ubhaya upamā, 153 ākşepa rūpaka, 245 ubhaya rūpaka, 247 ākşepa vyatireka, 277 ubhaya vyatireka, 278 ākşepa śleşa, 297 ubhayacchanna prahelikā, 210 ācāra upamā, 150 UBHAYANYĀSA, 171 ācikhyāsā upamā, 150 ubhayâvrtti āvrtti, 128 ādi dipaka, 195 ULLĀSA*, 334 ādi yamaka, 226 ULLEKHA*, 334 âdimadhya yamaka, 226 ādyanta yamaka, 226 ŪRJASVI, 171 ādhikya atiśayôkti, 99 ādhikya vyatireka, 278 eka vyatireka, 278 āmredita yamaka, 227 ekacchanna prahelikā, 211 ārthi apahnuti, 110 ekadeśavivarti rūpaka, 248 ārthī upamā, 151 ekadeśin upamā, 153 āvalī yamaka, 227 ekabahu upamā, 154 āvrti yamaka, 227 ekaśabdâbhidhāna samsrșți, 310 ĀVRTTI, 128 ekânga rūpaka, 248 ĀŚIȘ, 129 ekântarapāda yamaka, 228 āhārya rūpaka, 246 ekârtha dīpaka, 195 EKĀVALI, 172 ukta ākșepa, 125 uktanimitta viśeşôkti, 273 AITIHYA*, 334 UKTI, 129 AUCITYA, 172 UTTARA (I), 130 AUPAMYA, 173 UTTARA (II), 130 utpādya upamā, 151 kalpapprabhrti upamā, 154 UTPREKȘĀ (I), 131 kalpita upamā, 154 UTPREKȘĀ (II), 133 kāku vakrôkti, 261 UTPREKȘĀ (III), 134 kāñcī yamaka, 228 UTPREKȘĀ (IV), 135 KĀNTI, 173 UTPREKȘĀ (V), 136 kāraka dīpaka, 195 UTPREKȘĀ (VI), 137 kāraka hetu, 329 UTPREKŞĀVAYAVA, 138 kārakagūdha citra, 179

utprekșita upamā, 152 KĀRAŅAMĀLĀ, 173 UDĀTTA, 139 kāraņāntara vibhāvanā, 264 UNMĪLITA*, 334 kārya aprastutapraśamsā, 112

INDEX 342 jāti vyatircka, 279 kāryakāraņapaurvâparyaviparyaya jāti svabhāvôkti, 326 atiśayôkti, 100 jñāpaka hetu, 329 kāryânantaraja hetu, 329 KĀVYADRȘȚĀNTA, 174 KĀVYALIÑGA, 174 ņamul upamā, 155

KĀVYAHETU, 175 kiņcitsadṛśī upamā, 155 TATTVA, 190

komala anuprāsa, 103 tattvâkhyāna(I) upamā, 156

KRAMA, 175 tattvâkhyāna (II) upamā, 156

kriyā dīpaka, 196 tattvâpahnava rūpaka, 249

kriyā virodha, 266 TADGUŅA, 191

kriyā viśeşôkti, 271 taddhita upamā, 156

kriyā śleșa, 297 tadvācisamkșepa upamā, 156 kriyā svabhāvôktì, 325 turagapada citra, 182 kriyāgūdha citra, 179 tulya aprastutapraśamsā, 113 krīdā citra, 179 tulyayoga upamā, 156 TULYAYOGITĀ, 191 khadga citra, 180 tulyâkāraviśeșaņa samāsôkti, 318

gajapadapāțha citra, 180 daņda citra, 183 gamana upamā, 155 datta citra, 183 gamyamāna vyatireka, 279 DIPAKA, 193 garbha yamaka, 228 DUŞKARA, 199 guņa dīpaka, 196 duşkara citra, 183 guņa virodha, 266 duşkara yamaka, 229 guņa viśeșôkti, 271 dūrakārya hetu, 330 guņa śleșa, 298 DRȘȚĀNTA, 199 guņa svabhāvôkti, 325 drșțânta vyatireka, 279 guņakriyā virodha, 267 dyotakalupta upamā, 157 gupta citra, 181 GUMPHANĀ, 175 dravya dīpaka, 197

GŪŅHOKTI*, 334 dravya virodha, 267

gomūtrika citra, 181 dravya viśeôkti, 272

grāmya anuprāsa, 103 dravya svabhāvôkti, 326

cakra citra, 181 dhanu citra, 184

cakraka yamaka, 228 dharma upamā, 157

cakravāla yamaka, 229 dharmadyotakalupta upamā, 157

cakrâbjaka citra, 182 dharmalupta, 158

cațu upamā, 155 dharmôpamānalupta, 158

caturvyavasita yamaka, 229 CITRA, 175 nānātva atiśayôkti, 100

cyuta citra, 182 nāman śleșa, 298

cyutadatta citra, 182 nāmântarita prahelikā, 212 NIDARŚANĀ (I), 201

CHĀYA, 190 NIDARŚANĀ (II), 202

cheka anuprāsa, 103 nindā (I) upamā, 158 CHEKĀNUPRĀSA, 190 nindā (II) upamā, 158 nindā tulyayogitā, 192 JĀTI, 190 nibhrta prahelikā, 212

jāti dīpaka, 197 nimitta aprastutapraśamsā, 113

jāti virodha, 267 nimittadrsți višeșôkti, 273 jāti viśeșôkti, 271 nimittadrşți vyatireka, 280 nimittâdrsți viśeşôkti, 274

INDEX nimittâdrsti vyatireka, 280 prakrti śleșa, 300 343

pratiloma yamaka, 232 niyama upamā, 159 pratilomânuloma citra, 185 niyama citra, 184 niyamavat śleşa, 298 prativastu upamā, 160 niyamâkşeparūpa śleşa, 299 PRATIVASTŪPAMĀ, 208 niravayava rūpaka, 249 pratiședha upamā, 160 NIRUKTI*, 334 PRATĪPA, 208 nirņaya atiśayôkti, 100 pratīyamāna vyatireka, 280 nirņaya upamā, 159 PRATYAKȘA*, 335 nirvartya hetu, 330 PRATYANĪKA, 209 nivrtti hetu, 330 pratyaya upamā, 160 niścaya upamā, 159 pratyaya śleșa, 300 niścayagarbha samdeha, 314 pramușita prahelikā, 214 niścayânta samdeha, 314 praśamsā upamā, 160 nyūna parivṛtti, 204 PRAŚASTI, 209 PRAŚNA, 210 pañkti yamaka, 229 praśna citra, 185 pada śleșa, 299 prasiddha ākșepa, 126 padârthavrtti upamā, 159 prastāvabhāj tulyayogitā, 192 padâvṛtti āvītti, 129 PRASTUTĀÑKURA*, 335 padma citra, 184 PRAHARȘAŅA*, 335 paramparita rūpaka, 249 PRAHELIKĀ, 210 paraspara upamā, 159 prahelikā yamaka, 232 PARIKARA, 203 PARIKARĀÑKURA*, 334 prākaraņika tulyayogitā, 193 prāpya hetu, 330 PARIŅĀMA*, 334 PREYAS (I), 217 PARIVRTTI, 203 PREYAS (II), 218 parivṛtti yamaka, 229 PARISAŅKHYĀ, 205 praudha anuprāsa, 104

parihārikā prahelikā, 212 bandha citra, 186 parușa anuprāsa, 104 bahu upamā, 161 paruşa prahelikā, 213 PARYĀYA (I), 205 bahveka upamā, 161

PARYĀYA (II), 205 binducyuta citra, 186

PARYĀYOKTA, 205 pādamadhya yamaka, 230 bhaāga yamaka, 232 bhadra anuprāsa, 104 pādayor ... yamaka, 230 bhavişyat ākșepa, 126 pādasamudgaka yamaka, 230 pādâdi (I) yamaka, 231 BHĀVA (I), 218 BHĀVA (II), 219 pādâdi (II) yamaka, 231 BHĀVIKA, 220 pādânuprāsa anuprāsa, 104 BHĀVODAYA*, 335 pādânta (I) yamaka, 231 pādânta (II) yamaka, 231 bhāşā śleșa, 301 bhinnapada śleşa, 302 pādâbhyāsa yamaka, 231 pādaikadeśaja yamaka, 231 bhinnâbhinnaviśeşaņa samāsôkti, 319

PIHITA, 206 bhedabhāj rūpaka, 251

puccha yamaka, 232 bhedamātra vyatireka, 281

PUNARUKTĀBHĀSA, 207 bhedânukti samdeha, 314 bhedôkti samdeha, 314 pūrņa upamā, 159 PŪRVA (I), 207 BHRĀNTIMAT, 220

PŪRVA (II), 208 PŪRVARŪPA*, 335 MATA, 221

prakalpita prahelikā, 213 madhura anuprāsa, 105 madhya dīpaka, 197

344 INDEX vamśa yamaka, 234 madhya yamaka, 233 vaktra yamaka, 234 madhyânta yamaka, 233 VAKRA, 260 mahāyamaka (I) yamaka, 233 mahāyamaka (II) yamaka, 233 VAKROKTI (I), 260 VAKROKTI (II), 262 mātracyuta citra, 186 mālā (I) upamā, 161 vakşyamāna ākșepa, 126 mālā (II) upamā, 161 vacana śleşa, 302

mālā dīpaka, 198 vañcita prahelikā, 214 mālā nidarśanā, 203 vati upamā, 163

māla dipaka, 198 varņa śleșa, 303

mālā nidarśanā, 203 varņânuprāsa anuprāsa, 106 mālā (I) yamaka, 233 varņâvrtti anuprāsa, 106 mālā (II) yamaka, 233 vartamāna ākșepa, 127 mālā rūpaka, 251 vastu upamā, 164 mālā vyatireka, 281 VĀKOVĀKYA, 263 MITHYADHYAVASITI*, 335 vākya upamā, 164 MĪLITA, 221 vākyârthavrtti upamā, 164 mukha yamaka, 234 vâdilopa upamā, 165 MUDRĀ, 222 VĀSTAVA, 263 muraja citra, 187 VIKALPA*, 335 musala citra, 187 vikalpa citra, 188 moha upamă, 162 VIKASVARA*, 335 vikārya hetu, 331 YATHĀSAMKHYA, 222 vikrānta yamaka, 235 yathêvaśabha upamā, 162 vikriyā upamā, 165 YAMAKA, 223 VICITRA*, 336 YĀVADARTHATĀ, 238 VIDARŚANĀ, 263 yukta rūpaka, 251 VIDHI*, 336 yukta hetu, 331 VINOKTI, 264 yuktâtman arthântaranyāsa, 119 yuktâyukta arthântaranyāsa, 119 viparīta arthântaranyāsa, 119

YUKTI, 238 viparīta upamā, 165

yugmaka yamaka, 234 viparyaya (I) arthântaranyāsa, 119 viparyaya (II) arthântaranyāsa, 120

RATNĀVALI*, 335 viparyāsa upamā, 165

rathapada citra, 187 vibhakti śleșa, 304

raśanā upamā, 162 VIBHĀVANĀ, 264

raśanā rūpaka, 252 viruddha ākșepa, 127

RASAVAT, 239 viruddha rūpaka, 253

RŪPAKA, 239 viruddhakarman śleșa, 304

rūpaka rūpaka, 252 viruddhârtha dīpaka, 198

rūpaka śleșa, 302 VIRODHA (I), 265 VIRODHA (II), 268 LALITA*, 335 VIRODHA (III), 268

lalita anuprāsa, 105 virodha upamā, 165

lāța anuprāsa, 105 virodhavat arthântaranyāsa, 120 LAȚĀNUPRĀSA, 259 VIRODHĀBHĀSA, 268 liūga śleșa, 302 virodhin śleşa, 305

lupta upamā, 163 vivakșita drstânta, 200 LEŚA (I), 259 VIVRTOKTI*, 336 LEŚA (II), 260 viśișța parivrtti, 204 LEŚA (III), 260 VIŚEȘA (I), 269 VIŚEȘA (II), 269 VIŚEȘA (III), 270

INDEX viśeşa aprastutapraśamsā, 113 viśeşaņa rūpaka, 253 śrautī upamā, 166 345

viśeşastha arthântaranyāsa, 121 SLIȘTA, 288 viśvavyāpin arthântaranyāsa, 121 ślişta dīpaka, 198 VIŚEŞOKTI (I), 270 ślişta rūpaka, 255 VIŚEŞOKTI (II), 272 ŚLEȘA, 288 VIŚEŞOKTI (III), 274 śleșa aprastutapraśamsā, 114 VIŞAMA (I), 275 ślea, upamā, 166 VIŞAMA (II), 275 śleşa vakrôkti, 262 VIŞAMA (III), 275 śleşa vyatireka, 282 VIŞAMA (IV), 275 śleşâviddha arthântaranyāsa, 121 vişama rūpaka, 253 ślokâbhyāsa yamaka, 236 vişaya apahnuti, 110 SAMŚAYA (I), 306 VIȘĀDANA*, 336 SAMSAYA (II), 306 vrtta ākșepa, 128 samśaya atiśayôkti, 101 vȚtti anuprāsa, 107 samśaya upamā, 166 vaidharmya arthântaranyāsa, 121 SAMSRȘȚI, 307

vaidharmya drstânta, 200 sakala rūpaka, 255

vyaktâmśa samsrsți, 310 SAMKARA, 312

VYATIREKA (I), 276 samkara samsrsți, 311

VYATIREKA (II), 284 SAMKĪRŅA, 312

vyatireka upamā, 166 samkīrņa prahelikā, 215

vyatireka rūpaka, 254 saņkīrņa rūpaka, 255

vyatireka śleşa, 305 samkīrņa samsrsți, 311 SAMKȘEPA, 312 vyapeta yamaka, 235 vyapetâvyapeta yamaka, 235 samkşepa upamā, 167

vyasta yamaka, 235 samkhyāta prahelikā, 215

vyasta rūpaka, 254 sat nidarśanā, 202

vyasta vystireka, 282 sadrśa vyatireka, 283

vyutkrānta prahelikā, 215 sadrśapada upamā, 167

VYĀGHATA (I), 285 sadrśi upamā, 167

VYĀGHĀTA (II), 285 samdamśa yamaka, 236

VYĀJA, 285 samdașța (I) yamaka, 236

VYĀJASTUTI, 286 samdașța (II) yamaka, 236

VYĀJOKTI, 286 saņdaşțaka yamaka, 237 SAMDEHA, 312 vyāhrtârtha citra, 188 samdeha samsrsți, 311

śakti citra, 188 SAMA, 315

ŚABDA, 286 sama parivrtti, 204

śabda vyatireka, 282 samakakșatā samsrsți, 311

śabda śleșa, 305 samasta yamaka, 237

śabdacitra citra, 188 samasta rūpaka, 256

ŚABDAPRAMĀŅA*, 336 samasta vyatireka, 283

ŚABDĀRTHA, 287 samastapāda (I) yamaka, 237

śabdârthavarti samsrsti, 311 samastapāda (lI) yamaka, 237

śabdôpādāna vyatireka, 282 samastapādaja yamaka, 237

ŚAYYA, 288 samastavastuvişaya rūpaka, 257

śara citra, 188 samastavişaya upamd, 167

śābdī apahnuti, 111 samastavyasta rūpaka, 258

śikhā yamaka, 235 samasyă citra, 189

śuddha rūpaka, 254 samāgata prahelikā, 216 samādhāna rūpaka, 258 sūla citra, 188 SAMĀDHI, 315 ŚRUTI*, 336 samāna upamā, 168

346 INDEX

samānarūpa prahelikā, 216 sāmānyâbhāva upamā, 169 samānaśabda prahelikā, 216 SĀMYA (I), 322 samāsa (I) upamā, 168 SĀMYA (II), 322 samāsa (II) upamā, 168 sāmyatadvācisamkșepa upamā, 169 samāsa rūpaka, 259 sāmyavācakasamkșepa upamā, 169 SAMĀSOKTI, 316 sāmyôpameyatadvācisamkșepa upamā, samāsôkti aprastutapraśamsā, 114 169 SAMĀHITA (I), 319 SĀRA, 322 SAMĀHITA (II), 320 sâvayava rūpaka, 259 SAMUCCAYA, 320 SŪKȘMA (I), 323 samuccaya upamā, 169 SŪKȘMA (II), 323 samuccaya yamaka, 237 stuti upamā, 169 samudga yamaka, 238 SAMBHAVA*, 336 stuti tulyayogitā, 193 spașțapracchannârtha citra, 189 sambhavâsambhava atiśayôkti, 101 SMARAŅA, 323 saņbhāvyamānârtha atiśayôkti, 101 saņmūdha prahelikā, 217 SMRTI*, 336 SVABHĀVOKTI, 324 sarvatobhadra citra, 189 saviśeșaña rūpaka, 259 SVARŪPA, 327

sahaja hetu, 332 svarūpa apahnuti, 111

sahetu vyatireka, 283 svābhāvika vibhāvanā, 265

SAHOKTI, 320 sahôkti śleşa, 306 hala citra, 189

SĀDRŚYA, 321 HETU, 327

sādrśyamātra aprastutapraśaņsā, 115 hetu upamā, 170

sādharmya arthântaranyāsa, 122 hetu rūpaka, 259

sādharmya drsțânta, 201 hetu viśeşôkti, 272

SĀMĀNYA, 321 hetu vyatireka, 283

sāmānya aprastutapraśamsā, 115 hetu śleșa, 306