1. Kavya Praksa, Translation of Chaps 1-6 Mammatas Kavya Prakasa Alexander Havemeyer Catlin (Thesis)

Copyright

by

Alexander Havemeyer Catlin

2005

The Dissertation Committee for Alexander Havemeyer Catlin Certifies that this is the approved version of the following dissertation:

The Elucidation of Poetry:

A Translation of Chapters One through Six of

Mammața's Kāvyaprakāśa

With Comments and Notes

Committee:

Stephen Phillips, Supervisor

Nicholas Asher

Joel Brereton

Robert Hankinson

Kathleen Higgins

Robert Solomon

The Elucidation of Poetry:

A Translation of Chapters One through Six of

Mammața's Kāvyaprakāśa

With Comments and Notes

by

Alexander Havemeyer Catlin, D.E.U.G .; B.A .; M.A.

Dissertation

Presented to the Faculty of the Graduate School of

the University of Texas as Austin

in Partial Fulfillment

of the Requirements

for the Degree of

Doctor of Philosophy

The University of Texas at Austin

December 2005

Acknowledgements

Dr. Stephen Phillips has been a constant source of support and inspiration for this

work. He helped with literally every stage of the project, from the research that led to the

selection of the topic through the final day. During these long years he as always been

ready to inspire, direct, suggest improvements, and provide moral support. It is a rare

student who finds a mentor so generous with his time and energy. Thank you, Stephen,

you were all a supervisor should be, and much more.

I have also been fortunate to receive the support of my family, through good times

and bad. Thanks especially to my nieces, whose innocent love has been a great source of

hope and strength for me.

Conor Roddy offered support, camaraderie, and inspiration. I am lucky to have

such a friend. All the members of my committee offered assistance at key times. Many

other friends have helped throughout the years: Ed Allaire, Jill Glenn, Tami Wright, John

Garcia, George Dickey, and Ellen Briggs merit special thanks.

Thanks finally to Connie Rochelle, whose love gave me motivation and whose

support gave me the strength to finish. I know we will accomplish great things together. I

am looking forward to helping you with your dissertation!

iv

The Elucidation of Poetry:

A Translation of Chapters One through Six of

Mammața's Kāvyaprakāśa

With Comments and Notes

Publication No.

Alexander Havemeyer Catlin, Ph.D. The University of Texas ay Austin, 2005

Supervisor: Stephen Phillips

Mammața's Kāvyaprakāśa is an eleventh-century Sanskrit work on poetics that

synthesizes two earlier schools of poetics: the Alankāra School (the School of Poetic

Ornamentation) and the Dhvani School (the School of Suggestion). The first six chapters

establish the method of synthesis and explain the position of the Dhvani School. Jyatsna

Mohan's edition of the Kāvyaprakāśa (Nag Publishers, 1995) is transliterated, edited, and

translated from Sanskrit into English.

The dissertation's introduction places the work in conceptual, artistic, and

historical context. Here a new theory about the nature and importance of Mammata's

theoretic synthesis is advanced. It is argued that Mammata was aware of the tensions

between the two schools and sought to create a system of poetics that could incorporate

the theoretical and critical tools of both.

Explanatory comments are added to make the work more accessible and to clarify

philosophic difficulties. Additional comments seek to show how the poetic examples

serve to illustrate theses of the larger philosophic discussion.

vi

Contents

11 Introduction

1. Mammata and the Kāvyaprakāśa

3 2. Conceptual Introduction 3

3. Sanskrit Poetry 30

4. Mammata and the Early History of Sanskrit Poetics 36

5. Mammata's Legacy 41

6. Translation Method 44

7. Purpose and Content of Comments 49

Appendix to the Introduction 1 - Transliteration Details 51

Appendix to the Introduction 2 - Translation Details 53

Chapter One: Specification of the Purpose, Cause, and

Types of Poetry 54

1.1. Opening Auspicious Verse 54

1.2. The Purpose of Poetry 57

1.3. The Cause of Poetry 60

1.4. The Definition of Poetry 62

1.5. Three Types of Poetry Defined 66

vii

Chapter Two: Investigation of the Nature of Words and Meanings 73

2.1. Types of Words and Meanings Introduced 73

2.2. Definition of Literal Designation 84

2.3. Definition of Metaphoric Indication 91

2.4. Two Types of Pure Metaphors: Additive and Commutative 94

2.5. Impositional and Determining Metaphors Defined and Explained 98

2.6. Purposes of Metaphor Explained 107

2.7. Suggestion Based on a Single Word 109

2.7.1. Suggestion Based on a Metaphoric Word 109

2.7.2. Suggestion Based on a Literal Word 119

Chapter Three: Investigation of Suggestiveness Based on Meaning 126

3.1. Suggestion Based on the Character of the Speaker 128

3.2. Suggestion Based on the Character of the Person Addressed 131

3.3. Suggestion Based on Intonation 132

3.4. Suggestion Based on the Sentence 134

3.5. Suggestion Based on the Literal Words 135

3.6. Suggestion Based on the Presence of Another 136

3.7. Suggestion Based on Occasion 137

3.8. Suggestion Based on Location 138

viii

3.9. Suggestion Based on Time 139

3.10. Suggestion Based on Gesture 140

Chapter Four: The Exposition of Suggestive Poetry 144

4.1. Suggestive Poetry in which the Literal Meaning Is Unintended 148

4.2. Suggestive Poetry in which the Literal Meaning Is Intended 150

4.2.1. Suggestive Poetry Whose Psychological Sequence Is Unnoticed 152

4.2.1.1. Rasa Defined 155

4.2.1.2. A Four Stage History of Rasa Theory 159

4.2.1.3. The Eight Rasas of Bharata Listed, with Examples 179

4.2.1.4. The Abiding Emotions and Auxiliary States 197

4.2.1.5. The Ninth Rasa 201

4.2.1.6. The Emotions, Auxiliaries, Pseudo Rasas, Etc. 204

4.2.2. Suggestive Poetry in which the Psychological Sequence Noticed 217

4.2.2.1. Suggestive Poetry Based on the Power of the Word 218

4.2.2.2. Suggestive Poetry Based on the Power of the Meaning 227

4.2.2.2.1. Suggestive Poetry in which the Suggestor is Self-Existent 228

4.2.2.2.2. Suggestive Poetry in which the Suggestor is Invented by the Poet 233

4.2.2.2.3. Suggestive Poetry in which the Suggestor is Invented by a

Character in the Poem 238

4.3. Suggestive Poetry Based on Both the Words and the Meanings 243

ix

4.4. An Objection to the Classificatory System 245

4.5. Suggestive Poetry in which a Suggestive Word Adds to the Charm 249

4.5.1. Secondary Suggestion in which the Literal Meaning Is Unintended 249

4.5.2. Secondary Suggestion of Rasa 253

4.5.3. Secondary Suggestion Based on the Power of the Word 256

4.5.4. Secondary Suggestion Based on the Power of the Meaning 258

4.6. Suggestion Based on Context 276

4.7. Suggestion Based on Word Parts 27

4.8. Poems with More than One Suggestion 296

Chapter Five: The Investigation of the Types of Combinations that Exist

Between Suggestive Poetry and Poetry of Subordinate Suggestion 302

5.1. The Eight Types of Poetry of Subordinate Suggestion 302

5.1.1. Poetry of Non-Concealed Suggestion 304

5.1.2. Poetry of Dependant Suggestion 307

5.1.3. Poetry Whose Suggestion Depends on an Established Literal Meaning 321

5.1.4. Poetry Whose Suggestion is Unclear 323

5.1.5. Poetry Whose Suggestion is of Doubtful Prominence 324

5.1.6. Poetry Whose Suggestion and Literal Sense are of Equal Prominence 326

5.1.7. Poetry Whose Suggestion is Implied by Intonation 327

5.1.8. Poetry Whose Suggestion is Non-Striking 328

5.1.9. Further Varieties of Suggestive Poetry 331

5.2. Demonstration that Suggestion is Necessary 333

5.2.1. First Objection: A Single Semantic Power is Sufficient 343

5.2.2. Second Objection: Denotation is Sufficient 347

5.2.3. Third Objection: The Power of the Speaker's Intended Meaning

Is Sufficient 348

5.2.4. Fourth Objection: Metaphorical Indication is Sufficient 361

5.2.5. Fifth Objection: Denotation of the Sentence is Sufficient 366

5.2.6. Sixth Objection: The Nyāya View that Inference is Sufficient 368

Chapter Six: Definition of Poetry Based on Verbal or Semantic Display 375

6.1. Verbal Display 377

6.2. Semantic Display 379

Appendix - Correction to the Mohan Text 381

Bibliography 385

Vita 397

xi

Introduction

1. Mammața and the Elucidation of Poetry (Kāvyaprakāśa)

1.1 Mammata. We know little of Mammata's life. Three facts establish that he wrote

after 1000 CE. First, Mammata quotes a poem that eulogizes King Bhoja of Dhara,

whose reign extended from approximately 1005 to 1054. Second, he quotes

Abhinavagupta, whose dated works are from 990, 992, and 1012.1 Third, Mammata

draws four stanzas from Padmagupta's Navasāhasankacarita, which is commonly dated

circa 1020. Two further facts establish that he wrote before 1150. First, Mammata is

quoted in Ruyyaka's Alankārasarvasva, which is thought to have been written between

1135 and 1155.3 Second, there is a dated commentary on the Kāvyaprakāśa that dates

from 1159-1160.4 Arguing from these facts, the scholarly consensus dates the

Kāvyaprakāśa to the middle or end of the eleventh century.' The form of his name, his

reference to King Bhoja, and his familiarity with Abhinavagupta's work suggest that he

Ingalls, et al. (1990:32). 2 Gajendragadkar (1939:11). 3 Gajendragadkar claims this is the same Ruyyaka who wrote the Samketa commentary on the Kāvyaprakāśa (1939:11). 4 Mohan (1995:xxviii). 5 De (1960:145-148) and Kane (1961:273-275). Interested reader should refer to these works for the arguments supporting these dates.

hailed from Kashmir.6 Apart from the Kāvyaprakāśa, there is the only one extant work

attributed to Mammata, the Sabdavyaparavicara. However, there are references to a lost

work, the Samgitaratnamāli.7

Rājānakānanda, a 160 century commentator, claims that Mammata was a Saiva.8

Nothing in Mammata's extant work corroborates this claim. In his introductory verse,

Mammata praises Sarasvatī, the goddess of speech and poetry. This has been used to

suggest that Mammata was a Sārsvata Brāhmana.9

There is evidence, beginning with a manuscript dated 1159-1160, that part of the

Kāvyaprakāśa was written by a certain Alaka or Allata, about whom nothing else is

known.1 As there is no certainty about whether and in which parts Mammata was

"helped" (only the 10th chapter? also the 7th? the vrtti? the whole work?), no philological

or interpretative thesis can be reasonably based on double authorship. Like the scholarly

mainstream, I speak of Mammata as the sole author of the Kāvyaprakāśa.

1.2 The Text of the Kāvyaprakāśa. Like many late influential Sanskrit works, there is

little variation among the manuscripts. The earliest manuscript known dates from 1159-

1160.11 Unfortunately, a true critical edition of the text has not been made. However, the

Mohan edition, though lacking scholarly apparatus, includes seventeen commentaries and

6 Gajendragadkar (1939:9-10). 7 Gajendragadkar (1939:10). 8 Gajendragadkar (1939:9). 9 Gajendragadkar (1939:9). 10 Gajendragadkar gives the most extensive treatment of this issue (1939:18-24). He concludes that the theory of double authorship is based on a mistake. 11 De (1960:145) and Kane (1961:274).

2

is free from major errors. I have romanized this edition and made a list of corrections.

The errors in Mohan's edition are mostly typos. All the corrections were confirmed in

other editions and few of them seriously affect the meaning.

2. Conceptual Introduction

Before turning to Mammata's role in the history of Sanskrit criticism, I would like to

introduce some of his principle conceptual tools. I hope to provide primarily non-

specialists with an overview of the topics they will encounter in Sanskrit poetics, and,

more specifically, in Mammata. Many of the topics I introduce here are examined in

greater detail either in the following sections of the introduction or in my comments on

the individual theoretical verses (kārikās).

2.1 Definition of Poetry and Poetic Ornaments. I will often refer to two traditions of

poetics that Mammata attempts to synthesize. In the first chapter, he lays the conceptual

foundation for his synthesis while discussing his definition of poetry. Early Sanskrit

poetics conceives poetry in formal terms. Broadly speaking, the early critics strive to

define poetry in terms of syntactic or linguistic features. They agree that all poetry

contains some striking element that sets it off from non-poetic writing. Later, the Dhvani

School shifts the emphasis of poetics from form to content. Its proponents believe poetic

discourse to be characterized by a dominance of suggested meaning. They claim that

3

suggested meaning is the soul of poetry and strive to make the semantics of suggestion an

objective science.

Mammata's conception of poetry attempts to synthesize the two approaches. He

insists that poetry must be defined formally. He defines poetry is meaningful speech that

has poetic excellences, avoids poetic faults, and can be decorated by poetic ornaments.

To this definition he adds that the quality of a poem is determined by the intensity of its

suggested content; the more intense, the better the poem. By means of this integration, he

avoids the formalism of the early poetics on the one hand and the overly inclusive

emotive theory of the Dhvani School on the other. This conceptual synthesis is

Mammata's most original contribution to Sanskrit poetics. Indeed, I will attempt to show

that it structures his entire investigation and through his influence gives shape to late

Sanskrit poetics. Let us examine the terms of his definition.

Despite their differences, the early critics all realize that there is something

stylistically distinct about poetry. To emphasize its difference from normal discourse

(svabhāvokti), they speak of poetry as "deviant speech" (vakrokti), "excellent speech"

(atiśayokti), and "super-mundane speech" (lokottarokti). All these expressions point to

the idea that poetry is distinguished by its mode of expression, not its content. The most

important of these expressions, "deviant speech," suggests that the early critics believe

that the difficulty of poetry contributes to its aesthetic value.

Many of the poetic features the early poetics investigate are grouped together

under the heading "poetic ornaments" (alankāra). These are of such importance to the

4

early critics that the study of poetics (alankāraśāstra) came to be named after them.

Critics divide ornaments into verbal ornaments and semantic ornaments. Verbal

ornaments include difficult meters, rhyme, alliteration, onomatopoeia, etc. Mammata also

classifies punning (ślesa) as an ornament verbal ornament because the effect depends on

a particularity of the word (the fact that it has two meanings). Generally, any ornament

that loses its force when the word is replaced by a synonym is a considered a verbal

ornament. Sanskrit poetics also categorizes over one hundred semantic ornaments,

including many types of metaphor and simile, as well as hyperbole, synecdoche, zeugma,

and others. For the early theorist, these ornaments do not decorate poetry, they decorate

speech and thereby turn it into poetry.

The primacy of the structural elements is especially clear in the eighth century

critic Vamana, the first critic to speak not only of the body of poetry (word and sense),

but also of its soul. He claims that style (rīti) is the soul of poetry and that the styles are

grounded in particular arrangements of words (viśistapadaracanā).12 Styles are

differentiated by the presence or absence of poetic virtues or excellences (guna), such as

compactness of word structure (ojas), smooth word coalescence (ślesa), and absence of

long compounds (mādhurya).15 We have seen that Mammata integrates the concept of

poetic excellences in his definition.

The early critics also identified several poetic faults (dosa) that can mar or destroy

a poem. Mammata agrees that poetry should be free from faults. Faults include technical

12 De (1960:2:90). 13 De (1960:2:94-95).

5

errors in grammar, logic, and meter, as well as stylistic errors such as inappropriateness,

harshness in sound, and errors that unexpectedly arise when certain words are combined

(unintentional puns, etc.).

The early critics could not turn to meter as a distinguishing feature of poetry

because not all Indian poetry is metrical and metrical verse is not confined to poetry in

Sanskrit literature. Prominent religious, philosophical, and scientific works share this

form. Metrics is viewed as a separate science closely allied with grammar. However,

poetics does treat the artistic value of certain meters. Rhyme is rarely used in Sanskrit

poetry. It is discussed by the late critics, perhaps because of the increasing popularity of

rhymed verse in the vernacular languages after the tenth century. But it is never

considered a candidate for poetry's defining feature.

The Dhvani School claims that suggestion of rasa is the soul of poetry. We will

return to the very important concept of rasa presently. Here it is enough to notice that the

Dhvani School defines poetry as emotionally evocative speech. The theorists observe that

there are poems that contain no ornaments, while all good poetry moves its audience. The

Dhvani philosophers are particularly troubled by the many examples of pure technical

virtuosity in Sanskrit poetry that seem to involve no suggestion. They conclude that

despite appearances, even these poems must be suggestive to be considered poetry.

Mammata is the first critic to separate clearly the issue of poetic quality from that

of its essence. As we saw, he defines poetry as meaningful speech that contains formal

excellences and is free from faults. The Dhvani School raises strong objections to the

6

claim that the ornaments define poetry, and Mammata agrees that ornaments are optional.

With the formal definition in place, Mammata asserts that the prominence or strikingness

of the suggested content determines a poem's worth. We will return to this issue in our

discussion of rasa. But let us note here that Mammata attempts to resolve the problem

posed by poems of technical virtuosity. Formally, such poems are indeed poems.

However, their lack of emotive content makes them bad poems. In this regard,

Mammata's system captures the taste of his period perfectly.

The Indian critics, however, are not content with definitions and descriptive

classifications of the formal elements found in poetry; they also want to understand how

poems convey information. Let us turn to this issue.

2.2 Semantic Powers. All classical Indian theorists hold that words, whether singly or in

sentences, are capable of conveying meaning in a number of ways. While Indian

philosophy of language addresses a broad range of topics, poetic theorists focus primarily

on two: In how many ways can words convey meaning? And how does each way

function? The various ways a word conveys meaning are called "semantic powers," and

Mammata, like most late critics, believes there are three such powers.

A first semantic power is denotation (abhidha). Although almost all Indian

theorists accept denotation, how it functions is hotly debated.1 Perhaps because the

14 I say "almost all" to except the Buddhist "exclusion" (apoha) theory, which is important in classical Indian philosophy as a whole, but not in poetics.

emphasis in poetry is on suggestion, Mammata takes a fairly non-committal view of

denotation. In chapter two, he claims that by means of social conventions certain words

come to refer to certain types of objects. He leaves open the question whether words have

this power singly or only when combined into sentences. Furthermore, he makes no

explicit claims concerning the ontological status of the various types of referent, although

he seems to believe that denotative words primarily refer to physical objects and facts.

Finally, he seems to believe that truth lies in correspondence, but gives no account of the

relation between the truth-bearer and truth-maker.

Mammata deals with two Mīmāmsaka theories that purport to explain how

individual words combine to make a sentence, the Bhätta's "relation-of-the-designated"

theory and Prabhakara's "designation-of-the-already-related" theory. We will examine

the details of these views in chapters two (kārika 6) and five (in the defense of

suggestion). There is an important rival theory of sentence meaning, Bhartrhari's sphota

theory. Mammata, however, does not discuss it. Sentence meaning is important in

Mammata's theory of suggestion. However, Mammata feels that he theory works equally

well on either Mīmāmsaka theory.

There are three main factors govern the combining of words into sentences:

syntactic expectancy (ākānksā), semantic fittingness (yogyatā), and contiguous

pronunciation (samnidhi). We will examine these in the next section: metaphoric

indication.

8

A second semantic power is commonly called metaphoric indication (laksanā).

Like denotation, metaphoric indication is almost universally accepted in Indian

philosophy of language. This power comes into play only when the denoted meaning

does not mesh with the broader semantic context. This "blocking" of the literal meaning

allows the speaker to communicate secondary meaning related to the denoted meaning.

For example, if we say of a man that he spends all his time chasing skirt, the literal

meaning of the word "skirt" is blocked for two reasons. First, skirts, being inanimate, are

not the sort of thing that can literally be chased. Second, even if one could imagine a

context were skirts could be chased, such a context does not fit into our broader

knowledge of the man in question. Rather, the literal meaning is blocked and the word

comes to refer to those who commonly wear skirts, i.e., women (whether they are

actually wearing skirts or not). Again, when we say of a boxer that he is a real tiger, the

literal meaning is blocked by the fact that a human cannot literally be a tiger. In this case

we mean that the man shares qualities with the animal, e.g., speed, strength, and ferocity.

The most common Sanskrit example is the sentence, "The village is on the Ganges." The

use of the locative case in Sanskrit, more so than the preposition "on" in English, makes

it appears that the village is actually located on top of the river. As villages were not built

on top of rivers in ancient India, the literal meaning is blocked and a secondary meaning

is understood, namely that the village is on the bank of the Ganges. The relation between

the primary and secondary meaning in this case is said to be proximity.

9

Indian philosophy of language attempt to catalogue the different ways this

semantic power operates, focusing on how the literal meaning is blocked, how other

possible meanings are generated, and how one of the possible meanings is selected. Let

us take these up in turn.

There is blocking of the normal process of denotation when the individual words

that make up a sentence cannot be combined into a meaningful unit. Three main factors

govern the combining of words into sentences. First, the individual words of a sentence

are expected to combine into a syntactic whole (ākānksā). Minimally, a sentence must

contain a subject and a verb. Depending on verb and the context, other words may be

required. Second, the meanings of the words are expected to combine into a meaningful

whole according to the principle of semantic fittingness (yogyata). The classic example

of a sentence that fails this condition is "He wets with fire" (siñcati agnina), although

"The village is on the Ganges" is another. Third, the words have to be perceived

contiguously in a reasonable span of time (samnidhi). Only failure of the first two

conditions can be overcome by metaphoric indication. Words create both syntactic and

semantic expectations, which, when frustrated forcing the hearer to search for an alternate

meaning for one or more of them. The ways in which there can be such frustration are

numerous.

Two types of metaphor occur even when the denotations of the words do combine

into a meaningful whole. The first occurs when the resultant sentence does not fit the

larger semantic context. Here the concept of semantic fittingness (yogyatā) is expanded

10

to operate on semantic units above the sentence level. The second occurs when the

resultant sentence is ambiguous. In this case theorists resort to the speaker's intended

meaning (tātparya) to disambiguate. This is especially important when part of a group is

metaphorically used to indicate the entire group. For example, "The king is leaving,"

used to mean that the entire royal entourage is leaving.

When investigating how one generates possible alternative meanings, special

attention is paid to the types of relations that can exist between primary and secondary

meanings. Metaphors are divided into those based on similarity (gaunī laksanā) and those

based on other relations (śuddhā laksanā). The similarity relation is seen to be

particularly important and is analyzed into similarity of location, similarity of purpose,

similarity of appearance, and so on. The Nyāyasūtra lists ten other types of relations that

can ground metaphoric indication, including such relations as part-whole, cause-effect,

and means-end.15

If more than one alternative meaning is possible, the hearer must determine which

is intended by the speaker. This is done be attempting to determine what purpose the

metaphor serves. By examining details of the context, the speaker's goals, and, in poetry,

the constraints and expectations of the genre, the hearer eliminates unlikely options. If

there are still several options, especially in poetry, the reader should consider possible

relations between these and attempt to determine if the ambiguity is intentional.

15 Nyāyasūtra 2.2.62.

11

Mammata shows particular sensitivity to poems that use intentionally ambiguous

suggestive metaphors in chapter five.

Metaphoric indication is of special interest to critics of poetry for two reasons. On

the one hand, critics focus on the aesthetic value of this method of communicating. As we

saw, they classify many types of metaphor, synecdoche, hyperbole, etc. Even when the

same meaning can be communicated by denotation, they believe that these poetic

ornaments either comprise or augment the beauty of a poem. Mammata is also aware that

figurative speech slows the reader down and potentially making her more sensitive to the

texture of the words and the resonances of meaning.

On the other hand, critics notice that metaphoric indication not only serves to

decorate a verse, but almost always to suggest a further meaning.1 For example, the

synecdoche we just examined (chasing skirt) could be used to suggest that the man is

only interested in women sexually. Likewise, the metaphoric use of the locative in the

sentnece, "The village is on the Ganges," is said to suggest that the village shares the

coolness and purity of the holy river. This leads us to the third semantic power,

suggestion.

Suggestion (dhvani) is the most controversial semantic power. Although the term

has its roots in descriptions of ritual practice, critics of poetry, including Mammata, use it

to describe the semantic function that conveys a meaning, especially an emotive meaning,

which cannot be explained by means of the other two semantic powers. This negative

16 Indian philosophers are aware of dead metaphors. Mammata discusses them in chapter two.

12

definition leads many schools of Indian philosophy to deny the existence of suggestion as

a semantic power. To explain poetic suggestion, the critics attempt to adapt the

explanation of how a dramatic spectacle suggests its rasa or mood. The adapted theory,

however, does not prove adequate to the new medium and the critics do not provide an

alternative (more on this in the section on rasa). In line with Anandavardhana, Mammata

and the other defenders of suggestion admit that they cannot fully explain how it works.

Instead, the defenders of suggestion call attention to semantic phenomena that

cannot be explained by the other semantic powers. These range from the connotation of a

single word, thru the allusions of a sentence, to the evocation of an emotion or mood by

an artwork as a whole. We have already seen an example of the connotation of a single

word ("skirt" and "Ganges") in connection with metaphoric indication. Mammata, while

setting out his system, provides us with many more. Let us first examine Mammata's

presentation of suggestion and then turn to his defense of it against objections raised by

other schools.

In chapter three, Mammata lists various factors that can bring about suggestion of

additional meaning. He then clarifies the nature of suggestion by means of the

classificatory system and almost one hundred poetic examples that comprise chapter four.

For each poem he points out why denotation and metaphorical indication cannot account

for the total meaning. Although indirect in nature, Mammata's demonstration is

imposing. He makes the reader aware of vast quantities of semantic content any rival

eliminative theory would have to explain.

13

Mammata follows the tradition, which divides suggestion according to what is

suggested. The first level of division is comprised of three groups: suggestion of facts

(vastu), suggestion of poetic ornaments, and suggestion of emotive content (rasa and

emotions). It is striking how disparate these three types are. We will look at each in turn.

We have already encountered a metaphoric expression that suggests a fact: "The village

is on the Ganges." This sentence suggests that the village is cool and pure. In this case,

the additional semantic content communicated by this type of suggestion could be

communicated literally. In cases such as this, suggestion adds to the beauty of the

expression. However, sometimes a poem suggests several mutually incompatible facts. In

this case the resultant ambiguity can be an integral and non-reducible part of the poem's

aesthetic value (see poem 8, for example).

When the literal meaning suggests an ornament this also enriches the non-emotive

semantic content. Consider poem 57, for example:

Honor the trident god's artistic prowess! Even without color or canvas, He paints the wondrous world.

The poet uses a comparison to suggest the poetic ornament of "distinction" (vyatireka).

Śiva, the trident god, is compared to an artist, but the comparison does not hold: no artist

can paint without medium (colors) or support (canvas). The literal meaning makes a

comparison, but the suggestion is that Siva's mode of creation is incomparably superior.

Although in this case one can paraphrase the additional meaning, this is not always the

case. Several incompatible ornaments can be suggested, as in poem 65.

14

Mammata believes that literary works can be described objectively as erotic,

heroic, etc. He asserts that neither the denoted nor metaphorically indicated meaning can

convey the information necessary to ground such statements. Mammata gives three

reasons in support of the claim that a piece of literature does not communicate its rasa by

denotation. First, many pieces of literature do not state their rasa. Second, merely

denoting the rasa does not insure that the audience experiences it aesthetically. Finally, a

single rasa is often communicated by the piece as a whole, and denotation does not

normally work this way. Mammata assumes that if a piece of information is denoted, we

ought to be able to indicate where and explain how. Neither of these expectations holds

for suggestion of rasa. As we will see in the section on rasa, the suggested emotive

content of a poem is not limited to the rasas. This type of suggestion can also evoke the

full gamut of human emotions.

In the second half of chapter five, Mammata attempts to defend suggestion

against objections raised by various philosophic schools. The largely negative nature of

Mammata's justification of suggestion leaves it vulnerable to alternative theories. The

most powerful alternative comes from the Naiyayikas, who urge that the first two

semantic powers, along with world-knowledge and inference, are sufficient to explain

verbal knowledge. Although established before Anandavardhana, this theory is most

forcefully set out by Mammata's contemporary, Mahimabhatta.17

17 Despite the fact that they both probably hailed from Kashmir, there is no evidence that Mammata and Mahimabhatta knew each other.

15

Mahimabhatta argues that any additional meaning "suggested" by a poem can be

deduced from the literal meaning by means of the ordinary process of inference.

Inference is based on the inference grounding-pervasion (vyāpti) by the property to be

proved (sādhya) of the prover property (sādhana) and the presence of the prover in the

inferential subject (paksa).1 The inferential subject in the case of poetry is a certain

verse, the prover property is "having a certain literal meaning," the property to be proved

is, "having the desired additional meaning." Mahimabhatta argues that the pervasion can

be established by non-cognition (anupalabdhi) of the suggested meaning in the absence

of the literal meaning, identity (tādātmya) of the words that communicate the two

meanings, and causation (tadutpatti) of the additional meaning by the literal meaning.19

Mammata, in his reply, shows that the inferential process, as spelled out by the

Nyāya philosophy, cannot account for the suggested meaning. He does this by pointing

out that in many poems the "prover property" is not invariably concomitant with the

"property to be proved."2 He claims if you move a phrase from a poem into another

linguistic context, the same "prover property" (the literal meaning) might exist without

"the property to be proved" (the suggested meaning), or might be linked with a different

property (an alternative suggested meaning). Mammata does not avail himself of the fact

that many suggestions depend on the exact wording of the poem, not on its literal

meaning. These poems could be rewritten without changing their literal meaning in a

18 Phillips (2002:8-12). 19 De (1960:2:197). 20 Mammata is summarizing the arguments found in the Dhanyāloka 1.4b and Abhinava's comments thereon.

16

manner that would strip them of their suggestiveness. These poems clearly show that the

literal meaning does not always imply the suggested.

Mammata adds a second critique. Not only are the "prover quality" and "quality

to be proved" not invariably linked in poetry, they sometimes contradict each other.

Consider poem 2, for example;

O false messenger! You know not the pain you bring. The sandal paste is washed clean off your rounded breasts, And the rouge from your lips is completely rubbed off. Gone is the makeup from the corners of your eyes, And your slender body still shakes. So you went for a bath in the tank, And not to be with that wretch?

The poem affirms that the young woman has gone to bathe, but its suggested meaning is

that she has gone to make love and not to bathe. Here the supposed prover proves its own

negation; clearly not a desirable outcome. The same might be said for poem 57, which we

just examined, where the literal comparison suggests that the objects are incomparable.

Mammata also discusses poems that suggest several mutually incompatible

meanings and leave the reader in an intentional state of doubt as to which is intended (see

poem 111, for example). Mammata claims that part of the aesthetic pleasure of these

poems is the unresolved ambiguity. The Naiyāyikas would be hard pressed to explain

how this is possible within their system of inference.

Although the defenders of suggestion do not convince all Naiyāyikas to adopt it,

at least one famous late Naiyāyika, Jagannātha, did. Furthermore, they are generally

credited with causing other late Naiyāyikas to expand their conception of "the speakers

17

intended meaning" (tātparyajñāna), which, as we saw above, was introduced primarily to

resolve instances of ambiguity.2 Kunjunni Raja writes, "According to some of the later

Naiyāyikas a general knowledge of the meaning intended by the speaker is an essential

factor in all cases of verbal comprehension."2- Indeed, the theory of suggestion and this

broad theory of the speaker's intended meaning cover much of the same theoretical

territory.

2.3 Rasa. Mammata shows in chapters three and four that there are many types of

suggested meaning. Historically, however, suggestion of rasa is held to be the most

important. Broadly speaking, rasa is the mood, tone, or feeling evoked by a piece of

literature. Rasas are the aesthetic counterparts to certain basic human emotions, called

"abiding emotions" in the literature on aesthetics. The nine rasas generally accepted are

the erotic, the comic, the pathetic, the wrathful, the heroic, the frightful, the disgusting,

the wondrous, and the tranquil. The corresponding abiding emotions are sexual desire,

mirth, sorrow or grief, anger, heroic energy, fear, disgust, wonder, and world-weariness.

As we will see, the relation between the basic emotions and the rasas is explained

differently at different points in the history of the literature on aesthetics.

Bharata's Nātyaśāstra contains the oldest extant discussion of rasa as an aesthetic

concept. "Rasa" also means "flavor," and Bharata likens the rasa of a dramatic spectacle

21 Kunjunni Raja (1963:176-178 and 290-292). 22 Kunjunni Raja (1963:176).

18

to the flavor of a dish.23 Just as various foods and spices combine to create the flavor of a

certain dish, so the various elements of a drama come together to produce the mood.

These elements include the emotions, actions and reactions of the characters, the setting,

the plot development, etc. Neither Bharata nor his early commentators discuss the

aesthetic response that an artwork provokes. Although present in a theatrical

performance, rasa, for the early thinkers, is a quality either of the characters in the play

or of the actors. Let us look at these two theories.

In the Abhinavabhāratī, Lollata, the earliest commentator on Bharata whose work

survives, is reported as claiming that a rasa is an intensified abiding emotion.24 The

erotic, for example, is an intensified form of sexual desire. The abiding emotion is

intensified primarily by the artistic representation of the determinants, symptoms, and

auxiliaries (we will return to these below). The rasa is chiefly experienced by the

character represented, but is also tasted by the representing actor.

I believe that Lollata, like Bharata himself, writes primarily to instruct

playwrights and actors. Like Aristotle, his theory explains which elements need to be

included in a successful plot. Lollata discusses the story of Rāma's love for Sītā and

claims that more than mere sexual desire must be presented. Rāma's fervor needs to be

intensified by being developed through out the play. In a good romance, the lovers go

through trials, doubts, and frustrations that augment their ardor. The climax leaves the

23 Nātyaśāstra 6.31. I suspect that this passage is later than the sūtra it is attached to. Part of it is quoted by Lollața (early 9th century), so it must predate him. 24 Gnoli (1968:26).

19

viewers with the impression that the couple is deeply in love and that their love will last.

Rāma's love, lived over a period of several years, is focused and intensified by being

condensed into an artwork.

Rāma's love, however, not only changes in degree, it changes its nature. Lollata is

aware that an artistic presentation of a love story is a different type of thing that a real

love story. Here we see that from the beginning, rasa is understood as the aesthetic

counterpart to a real emotion. The actor who presents Rama's love is filled with the erotic

rasa, but is not himself in love with either Sīta or the actress playing that role. Some later

theorists misunderstand this and claim that the actor would forget his lines if Lollata's

theory was correct!

Bharata explains various factors that allow a spectacle to evoke rasa, including

costumes, sets, music, dance, plot developments and many others. The elements of plot

that most interest the poetry critics are the determinants, the symptoms, and the

auxiliaries of the emotion. These are best explained by example. Juliet is a "determinant"

of Romeo's sexual desire (other determinants include the proper setting and time, i.e., the

Capulet garden at night). The symptoms of Romeo's desire include the actions taken for

the sake of the beloved, such as climbing garden walls, writing letters, exchanging

smiles, etc. Also included are certain behavioral changes like his willingness to risk

danger and his desire to befriend Tybalt. Finally, the auxiliaries of Romeo's desire

include both secondary emotions, such as the jealously Romeo feels toward Paris, and

"involuntary states" that arise from being in love, such as perspiration and trembling.

20

Bharata sets out guidelines regulating the use of these factors. He explains which

symptoms and auxiliaries pertain to each rasa and which are antithetical, he advises

which factors can be mixed and the result of their mixing, he shows which rasas can be

introduced in digressions without diminishing the dominant rasa, and other such things.

Although Bharata's system can seem both mechanical and limiting, it does explain how a

dramatic spectacle successfully conveys meaning beyond the literal and metaphoric.

Bharata's is the only aesthetic system that survives from the early period of Indian

Aesthetics (pre-sixth century) and it has been adapted to many different art forms. This is

not surprising when one remembers how complex both Indian dramatic spectacle and

Bharata's text are. In so far as they are important to the production of a dramatic

spectacle, Bharata discusses poetry, music, painting, and dance. The more specific Indian

theories of both music and painting, as we understand them presently, are both rooted in

Bharata's system. When the Dhvani School critics looked for inspiration, it seems this

was the only place to find it. Rasa theory offered two distinct theoretical tools: the theory

explained the nature of dramatic emotional expression and how such expression is

achieved. The Dhvani School attempted to use both.

When rasa theory was adapted from dramaturgy to poetics by Anandavardhana,

several changes were necessary. As we will see in the following section, Sanskrit poetics

deals primarily with short, self-contained poems. The limited length of the poems forced

the critics to reconsider the list of elements that were said by Bharata to produce rasa.

25 Almost all modern rasa based theories show the strong influence of Abhinava's reworking of Bharata.

21

Although they retain Bharata's requirement that the determinants, symptoms, and

auxiliaries of the emotion must be present, the critics allow that some of the elements

may be inferred by the spectator rather than explicitly presented in the artwork. They also

realize that a short poem can present only a fraction of the elements that a play can.

More importantly, Bharata's explanation of the mechanics of evoking rasa proves

too cumbersome for the short verses. Almost without exception, the verses describe a

single moment or scene or observation. In this way they are more closely related to

painting than drama. There simply is no room in a couplet for the digressions and details

that serve to develop and reinforce the rasa in drama. When determinant, symptoms, and

auxiliaries are presented, they are used to fill in the details of the poetic image, not to

advance the plot.

The impossibility of applying the mechanics of theatrical spectacle to short verse

leaves the critics without a theoretical explanation for the functioning of suggestion. As

we saw in the section on suggestion, this pushes them to attempt a classification of the

types of suggestion and to show that the "suggested" meaning cannot be arrived at by

means of the traditional semantic powers. They fare better with the other half of rasa

theory: the explanation of the nature of aesthetically presented emotions.

The critics adopt the idea that the emotions, when evoked by an artwork, undergo a

transformation. They model their theory of the poetically evoked emotions on rasa

theory, but the difference of medium forces them to make important changes. While only

a limited number of themes were treated in classical Indian drama, Indian short poems

22

address almost every aspect of classical Indian life. Bharata's list of eight rasas was most

probably a list of the rasas actually embodied in the plays he knew. Later drama theorists

add "tranquility" (śanta) to the list, perhaps to account for the dominant mood of the

Mahābhārata and perhaps because of religious works such as Aśvaghosa's long poem

(mahākāvya), the Buddhacarita (first century). In either case, the art seems to have driven

the theory. And when confronting a different art form, even this expanded list proves

inadequate.

Rudrata, a ninth century poetician, argues that any emotion, if developed by

determinants, etc., can become a rasa.2 Coupled with the view that the emotions are

infinite in number, this leads to the view, attributed to Lollata, that the rasas are

infinite.7 Pratīharenduraja claims that the original list should be maintained for

tradition's sake and the other emotions dealt with under different titles.28 The final great

critic, Jagannātha, takes this view to the extreme by claiming that the original system

should be retained merely for the sake of convenience.29

Mammata seems to follow Pratīhārendurāja. He retains the nine rasa limit, but

blurs the distinction between the nine abiding emotions (those that can become rasas)

and other important emotions by grouping them together in one type of suggestion.

Several other emotions can be (and actually have been) developed into a full drama, and

as such could ground an additional rasa. Devotion, family love, and friendship are among

26 Rudrata, Kāvyālankāra 12:3-4, discussed in Raghavan (1967:127). 27 In the Abhinavabhāratī, vol. 1, 346. 28 Raghavan (1967:129). 29 Raghavan (1967:142).

23

the additional emotions that are considered as possible candidates for a corresponding

rasa. While they are ultimately denied rasa status, Mammata seems to consider them on

a par with the abiding emotions of sexual desire, etc. He lists them in the chapter on the

best type of poetry, giving excellent poems as examples.

Mammata also recognizes that many less robust emotions that may not be able to

ground a classical Indian drama can certainly be suggested in a poem. Indeed, the

Subhāsitaratnakosa, an anthology of short verse compiled in Mammata's time, shows a

great interest in verses that suggest emotions other than the nine that can become rasa.50

Mammata is well aware that the evocative powers of poetic suggestion extend far beyond

the nine rasas.

Although he includes poems that treat the traditional "pseudo-rasas" in the

chapter that treats the best type of poetry, it is difficult to determine Mammata's attitude

toward them. Indian philosophy resorts to the concept "pseudo" to describe a close

imitator of what is being investigated. Items so described ultimately fail to meet the

standard, giving the term a strong negative connotation. Thus in logic there are "provers"

(hetus) which are sufficient to establish the desired inference and "pseudo-provers" (hetu-

ābhāsas) which appear to be provers but fall short under close scrutiny. Abhinava uses

the common example of seeing pseudo-silver that turns out to be mother-of-pearl.31 This

30 Ingalls puts its date at "shortly before the year 1100." Mammata is believed to have written in the second half of the eleventh century (see Introduction). Vidyākara includes a whole section on adulterous women and myriad other poems that fail the standards of rasa. 31 Ingalls, et al. (1990:217).

24

would be an example of pseudo-perception. Likewise in Bharata's drama theory, there

are emotions that seem transformable into rasas, but fail to do so on closer inspection.

Every example of pseudo-rasa I have encountered fails to be a true rasa for moral

reasons. Raghavan suggests that the category of pseudo-rasa was invented to handle

Rāvana's illicit passion for Rama's wife Sīta.52 Indeed, Abhinava cites an earlier critic

who claimed that Ravana's lust (laulya) should be counted as a new rasa, with the

abiding emotion covetousness (gardha)." Abhinava rejects this view, claiming instead

that until we realize the inappropriateness of a pseudo-rasa, we experience it as rasa.

After the realization, however, our experience changes according to which emotion is

portrayed." Mammata's example concerns the pseudo-erotic rasa evoked by a woman

with several lovers.

Given the differences between a public spectacle and a court poem, it is hard to

understand how and why the moral constraints of drama theory should be adopted by

poetics. Mammata position on the issue is not clear. The main evidence that Mammata

feels that poetry should obey the same moral constraints as drama is his retention of the

term "pseudo-rasa," which we saw has strong negative connotations. However, he could

be using the name merely to identify certain poems. The main evidence for this view is

that he includes suggestion of pseudo-rasa in the group of poems described as the best

32 Raghavan (1969:125). 33 Abhinavabhāratī (1956:1:342). 34 Ingalls, et al., (1990:107).

25

poems. I would like to think that Mammata is aware that the immoral, like the horrible

and the frightful, can be aesthetically appreciated.

The difference between the abiding emotions and the other emotions is a key

element in Bharata's system. Only the abiding emotions are robust enough to be

developed throughout an entire dramatic spectacle. They alone can be elaborated through

the presentation of the determinants, symptoms, and auxiliaries. As Abhinava makes

clear when attempting to find the abiding emotion for the rasa of tranquility, the abiding

emotions are primary: the rasas are merely their aesthetic counterparts. When the brevity

of the Sanskrit poems made it impossible to use Bharata's mechanics of rasa in poetics,

the main justification for distinguishing the abiding emotions is lost.5 While drama

theorists need to be certain an emotion is stable enough to ground a play, poetics has no

such concern. No emotion is too fleeting to ground a poem. If it is admitted that the

distinction between the abiding emotions and the other emotions is not relevant to poetic

beauty, the elevated status of the nine rasas is seriously threatened.

The logical conclusion for poetics was drawn by Lollata: the number of rasas

should be considered infinite. In other words, each emotion should have an aesthetic

correlate. This move would have greatly facilitated criticism. However, the fact that the

distinction between the stable emotions and the other emotions is still important to drama

theory militates again this move. Unfortunately, no other way to express the fact that all

35 Abhinava puts forth two additional arguments for the distinction in the Abhinavabhāratī (1956:665-667). First, only the abiding emotions do not depend on the object that produces them, being innate, and second, all the other emotions presuppose at least one abiding emotion. Neither argument seem convincing to me.

26

emotions can be appreciated aesthetically is found. With the exception of the nine abiding

emotions, Indian theorists, Mammata included, simply call emotions suggested in poetry

and emotions lived in the world by the same name. While Mammata's broad appreciation

of poetry is shown throughout chapter four, he never discusses the relative poetic merit of

the stable emotions and the other emotions. If I am correct in thinking that the abiding

emotions lose their primacy, this is never explicitly stated.

The ninth and tenth century critics not only reworks and expands Bharata's theory

of the nine rasas to better analyze poetry, they also add a whole new orientation to

aesthetic theory. When the early dramaturges discuss rasa, they focus primarily on how

to create a play that evokes rasa. The poetic critics too are interested in showing how to

create rasa rich artwork. However, they increasing consider the psychology of the

aesthetic experience as well.

The Dhvani School, striving to understand the psychology of audience response,

develops a sophisticated theory that centers on the aesthetic universalisation of the

emotions. Rasa is no longer considered an object to be enjoyed, but as the ongoing

process of aesthetic enjoyment itself. This view, which can be usefully compared to T.S.

Eliot's theory of the universal correlate, became the orthodox for later writers. Let us

examine it in greater detail.

Several key elements of the Dhvani School theory were worked out by the ninth

century critic Sankuka. He claims that rasa is not a heightened form of an emotion, but

rather an artistic reproduction of the emotion. For Sankuka, sexual desire exists in the

27

world, while the erotic rasa arises from its reproduction in art. The aesthetic experience

arises in one who is trained to appreciate art by means of the perception of an artistic

imitation. The rasa is located in the play, and not, contra Lollata, in the actors. The actor

makes manifest the play's rasa through his skill in imitating the emotions of the

characters.

Śankuka notices that the audience perception of the rasa comes about by means

of a cognitive process that is unique to art. He says that in aesthetic experience, "there is

neither doubt, nor truth, nor error. The notion that appears is, 'This is that,' not, 'This is

really that.""36 In aesthetic perception, in other words, there is "willing suspension of

disbelief" that allows the audience to fully engage in the production in a manner distinct

from normal experience.

Another important pre-Abhinava critic, Bhattanāyaka, builds on Śankuka's

theory. He believes that the abiding emotions, determinants, symptoms, and auxiliary

states were universalized by a special power of words." The universalization process of

artistic production transforms a particular emotion and the events surrounding it into a

general emotion free of context. The audience can appreciate these general elements

without personal attachment. We can see that this view is much more radical than

Śankuka's. The audience is not "disinterested" simply because they understand that art,

as imitation, is different from reality and engage with it accordingly. Bhattanāyaka's view

36 Gnoli (1968:32). 37 Gnoli (1968:45).

28

undermines the very idea of art as imitation. He sees art as universalizing, not imitating,

reality. His view of what is enjoyed in an artist production is correspondingly novel.

Ingalls writes that Bhattanāyaka "sees rasa not as an object to be enjoyed, but as

the ongoing process of (aesthetic) enjoyment itself."38 Rasa is no longer thought of as in a

play, much less in the actors or characters, but in the experience of an artistic production.

Rasa, in other words, is a process, not an entity or quality.

The process of enjoyment is causally dependant on the artwork enjoyed and yet

experientially free of it. Abhinava represents him as saying, "is due to the emergent state

of goodness (sattva), is pervaded by beatitude (ānanda) and light (prakāśa), and is

similar to the tasting (asvada) of the supreme brahman.' ,,39 We can see that once the

artwork gives rise to the rasa experience, it is left behind. The experience itself is of bliss

and light alone unqualified by the details of the artwork. It is the transcendent nature of

the experience that invites comparison to mystical experience.

Abhinava refines Bhattanāyaka's theory and his version that becomes the

touchstone for the Indian tradition, a place it continues to hold today. He accepts the

theory of universalization, both of the determinants, etc., and of the abiding emotion. He

maintains that the universalized emotion must be perceived, not inferred, by the

individual spectator. He provides three conditions that the spectator must meet to be fit

for the rasa experience: he must be trained in the arts, he must be in a state of willing

suspense of disbelief, and he must be disinterested to the point of self-forgetting. These

38 Ingalls, et al. (1990:37). 39 Gnoli (1968:47-48).

29

conditions are necessary to enter the state of sympathetic union that Abhinava describes.

He further agrees that rasa is an experience, not an object, affirming that rasa exists only

in the spectator.

Abhinava also agrees that the rasa experience is akin to the mystic union with

Brahman. However, he differentiates the aesthetic from the yogic by insisting that the

aesthetic experience has an object, namely the artwork. This is very important, for it

provides Abhinava a way of avoiding Bhattanayaka's homogenizing of aesthetic

experience.

Abhinava's condition that the spectator be cultured underpins his claim that the

rasa experience produced by a certain work is always of the same type. While some

connoisseurs are more receptive than others, they are all receptive in the same way to the

same thing. As is clear from the disinterested requirement, if any personal eccentricities

enter into the aesthetic experience, they act to limit or destroy it. The rasa can only be

evoked in a learned connoisseur that is capable of aesthetic distance.

3. Sanskrit Poetry

The following brief introduction to Sanskrit poetry is specifically intended to help the

non-specialist understand Mammata's examples. The earliest identified poems

encountered in the Kāvyaprakāa are excerpted from the two great epics, the

Mahābhārata and the Rāmāyana. The Mahābhārata is, "gigantic, comprising a length of

about seven times that of the Iliad and the Odyssey combined ... [It] narrates the story of

30

two rival groups of cousins both laying claim to the succession in a region somewhere

near present-day Delhi, and the war that ensues." The core story dates from

approximately the fourth century BCE. Over the next eight centuries, many shorter tales

were added, giving the text is current volume. The Rāmāyana, the core of which dates

from approximately the second century BCE, recounts the story of the abduction of

Rama's wife Sīta by the demon king Ravana and the war that follows. Apart from the

verses Mammata excerpts directly from the epics, many of his other examples are based

on episodes found in the two epics.

Classical poetic literature (kāvya) is commonly held to have blossomed in the first

few centuries of the Common Era in the context of royal courts. Although many hold that

the tradition climaxed in the fourth to nonth centuries with such figures as Kālidāsa,

Bhavabhūti, and Sriharsa, there are major writers at least until the Mogul invasions of the

sixteenth and seventeenth centuries (e.g., Rājasekhara, Jayadeva, and Jagannātha).

Classical Sanskrit literature is highly ornate and displays great technical virtuosity. This

is partly due to the fact that Sanskrit was a second language for the artists, and partly to

its courtly nature. It is literature written by connoisseurs for connoisseurs and strives for

both emotional power and intellectual charm.

Sanskrit kāvya aims at beauty through the presentation of ideal types in tightly

crafted verse. Rarely does it touch on political matters, and never with the aim of

effecting regime change. Neither does it generally strive to be expressive of the poet's

40 Van Buitenen (1974:17).

31

personal emotions. Although the poetry is emotionally charged, and many writers

probably write from their own experience, neither the personality nor the experience of

the poet is normally the focus. Rather, idealized types - the lover, the beloved, the hero,

etc - are used to evoke universal human emotions. The great individuals, e.g. Rāma,

Sītā, and Chandragupta, are normally either mythical or historically distant from the

writer. Even the most common exceptions to this rule, panegyrics written for specific

royal patrons, portray the kings and princes in general terms, with very few personal

details. Furthermore, when the protagonists of kāvya are individuated, it is normally more

by their actions than by their character.

Three main forms of kavya are considered important by aesthetic theorists. These

are great poetry (mahākāvya), drama (nātya), and short poetry (subhāșita). The great

poetry presents mythic, royal, and religious themes in works that often run to several

cantos. Also popular in this genre are ideal descriptions of nature and the seasons. As the

plot is usually fairly simple, this form offers the poet great scope to show his control over

the technical aspects of poetry. Generally, each canto is written in a single meter and the

audience expects to find a great number of the poetic ornaments (alankāra) skillfully

woven into the story. While the "formalism" of Sanskrit poetry was often criticized in

early Western evaluations of Sanskrit literature, it is now beginning to be appreciated on

its own terms.41

41 Macdonell (1899:328) offers a good example of the first approach.

32

Rather than striving to portray character development or the internal logic of a

story, Sanskrit drama focuses primarily on the evocation of rasa. A simplified treatment

of character and the various elements of plot (see introduction 2.3) are used to build up

the spectators' appreciation of the rasa bit by bit. Ultimately, the evocation climaxes with

an appropriate ending that reinforces all that has come before. In a romance, for example,

the lover will be strong and noble, the beloved beautiful and pure. The lovers will meet in

an ideal setting, such as a forest glade, and fall instantly in love. Their love will be tried

and tested in various episodes that show the depth of their devotion. In the end they are

united in loving union that reinforces or completes the romantic theme of the entire piece.

The three most common themes in Sanskrit drama are the romantic, the heroic, and the

comic.

The dramatist can excel in two areas: plot structure and language. The detours and

obstacles the protagonists face must be carefully crafted to capture the audience's

attention and augment the dominant rasa. The poet must avoid the opposing faults of

monotony and incoherence. Although extensively discussed by the dramaturges, plot

holds little interest for the poetic critics.

Generally, Sanskrit plays are written partly in prose and partly in verse. At

climatic moments the main characters speak in polished verses that allow the dramatist to

display his poetic capacity. These poetic sections vary in length from a single couplet to a

hundred lines or more. These verses are often passionate, but also include descriptions of

33

natural beauty or general truths. Individual verses from these poetic passages are isolated

for study by the poetic critics.

The third form of literature, short verse (subhāsita), is of the greatest interest to

the poetic theorists. Indeed, almost all of Mammata's examples belong to this class.

Many of the short poems exalted in the Sanskrit tradition are extracted from the first two

types of literature (larger poems and plays). The idealized heroes and the simplified plots

we discussed above made it relatively easy to extract verses and treat them as

independent poems. Sanskrit poets also composed detached poems (muktaka) that are

free of context are often collected into anthologies according to their subject.42 As Ingalls

has pointed out, wherever these verses come from, they ideally should be able to stand on

their own, both in terms of meaning and mood." The critics single out verses either

because of their stylistic excellence or their emotional power. The early poetics addresses

the former in terms of poetic ornaments, while the Dhvani school addresses the later by

means of suggestion. Mammata attempts to explain both.

The short poems are broken into quarters, the length of which is determined by

meter. The common meters result in poems between forty and eighty syllables, although

meters that yield more than one hundred syllables exist." This places the subhāsita

midway between the Japanese haiku and the English sonnet in length.

42 Vasudeva (2005:15-16). 43 Ingalls (1965:33). We will see that several of Mammata's examples do not meet this ideal. 44 The eight syllable śloka of the Rāmāyana was not common in classical poetry.

34

These poems can be compared to Indian miniature paintings because of their vivid

visual imagery. With a remarkable economy of words, the poets evoke physical settings

through carefully chosen details. In these settings, the parting or reunion of lovers,

momentous decisions, victory in battle, and the like, are given emotional depth by

focusing on minutiae of behavior and appearance. The poets can make much of a

trembling lip, a sidelong glance, or fiery eyes. They are also very effective at letting their

characters speak, choosing short phrases that reveal much more than they say. Again and

again, the reader is surprised by the quantity of information the poets convey by the

accumulation of such details.

It is too early to determine the exact extent of Mammata's poetic culture. As of

now, less than half of the poems he quotes have been traced. From the approximately

two hundred poems that are identified, we know that Mammata was familiar with the two

great epics (the Mahābhārata and the Rāmāyana) and the major classical poems (both

mahākāvyas like the Meghadūta and the Raghuvamśa, and anthologies like the

Amaruśataka).46 However, from what we know, he quotes most frequently from the

classical dramas. Indeed, his quotes span almost all the extant plays, including works by

Kālidāsa, Bhavanhūti, Śrīharsa, Sūdraka, and others. I will discuss the great poems and

plays individually in the comments.

45 Dwivedi lists sources for almost three hundred of the six hundred poems (1970: appendix C). However, approximately one third of these references are other works of poetics that quote the poems as examples. This leaves only about two hundred poems fully identified. 46 Interestingly, no Bhartrhari poems have been identified.

35

The latest poet that Mammata quotes is Padmagupta, who either lived shortly

before him or was his contemporary. Further research into the sources of Mammata's

examples and into tenth and eleventh century literature, especially the short verse, may

reveal more connections between Mammata and his immediate intellectual context.

4. Mammata and the Early History of Sanskrit Poetics

The Kāvyaprakāśa occupies a unique position in the history of Sanskrit poetics. Written

in a period when several writers were trying to weave together the main threads of the

previous poetics, Mammata's text became the standard compendium of poetics.47 As

Mammata is more a genius of integration than creation, a brief look at the history of

Sanskrit poetics is necessary to understand his importance.48

Mammata shows extensive knowledge of earlier Sanskrit poetics." His work

draws from two very different traditions of poetics. The first chapter and chapters seven

to ten are drawn from the older Alankāra School writers (Bhāmaha, Dandin, Udbhata,

and Rudrata). Chapters two to six, on the other hand, are indebted to the Dhvani School

writers (mainly Anandavardhana and Abhinavagupta). This has led some to say that

47 For information on Mammata's competitors, like Bhoja, see Gerow (1977:271). Even Mammata biggest rival, Viśvanātha (14th c.), wrote a commentary on the Kāvyaprākaśa and uses its structure in his own work, enlarging it only to include a section (5) on plot in drama. On the standardization of Mammata see

48 The two standard histories are that of Kane (1961) and De (1960). While these works remain invaluable, De (1960:2:216).

the Gerow's history (1977) improves on them in many areas. It is, in my opinion the best introduction currently available. 49 Mammata quotes several writers by name (Bharata, Rudrata Anandavardhana, Abhinava, inter alia). Almost half of the identified poems that Mammata quotes had already been quoted in earlier works. In many cases Mammata uses poems in the same theoretical context as the earlier writers. His familiarity with the earlier writes is further shown by the inclusive nature of his list of alankāra.

36

Mammața is better termed a "synthetic" writer than a "systematic" one.50 While I agree

with this assessment, I believe the synthesis that Mammata achieves is both stronger and

of more importance than is generally acknowledged.

The earliest writers on poetics are inspired most directly by the Grammarians and

the Mīmāmsakas, not by the Nātyasāstra." This is only logical, considering their desire

to understand the "grammar" of poetic ornaments (alankara), poetic blemishes (dosa),

and several syntactic stylistic excellences (guna). In pursuing these issues, they seek to

differentiate poetic language from religious, moral, and scientific language. Poetry is held

to be decorated or ornamented speech, thus the name "science of ornamentation"

(alankaraśastra) for poetics." The main concerns of the early poetics are the definition

of, justification for, and teaching of the art of poetry, as well as a classification of its

various types. Its goal is to educate future poets. Drama theory (nātyaśāstra) is treated as

a separate science by these writers. This is not surprising when one considers the

enormous difference between the writing of an epic poem and the production of a

dramatic spectacle.53

Anandavardhana recreates poetics on an entirely new ground adapted from

dramaturgy and philosophy. None of the important aspects of earlier poetics (poetic

50 Gerow (1977:272). 51 Kane (1961:335ff), Gerow (1977:221) and De (1960:2:18). 52 Originally the poetic ornaments were thought to decorate speech, thereby transforming speech into poetry. Later the Dhvani school claimed that the ornaments decorated poetry, making them unessential. They introduced the famous analogy that the poetic ornaments are to poetry as jewelry is to a woman. 53 Gerow (1977:227). Gerow has argued throughout many works against the standard vision that Sanskrit poetics develops toward Abhinava and then, having climaxed, goes into decadence. De is the most highly regarded proponent of this view (1960 v2:213 to 217 gives a wonderful summary) and Coomaraswamy the most popular. Among contemporaries, M. Chari's Sanskrit Criticism as well as most "popular" works on Indian art carry forward this highly suspicious theory.

37

ornaments, blemishes, and excellences) are taken up seriously in his work. Instead, he

focuses on the concept of mood or flavor (rasa) as developed by Bharata in the

Nātyaśāstra (see introduction 2.3). Ānandavardhana claims that producing rasa is the

essential function of poetry. Furthermore, he modifies a key term from Mīmāsaka

philosophy of language, "suggestion" (dhvani), to explain how a poem communicates

rasa. This brilliant maneuver allows him to supplant the earlier writer-based poetics that

focuses on techniques for producing good poetry with a spectator-based poetics that

focuses on the reader/audience response. Gerow argues that the massive cultural changes

that followed the fall of the Gupta Empire brought drama and poetry together. Plays

became more read than preformed and devotional (bhakti) poetry was often sung, rather

than read, blurring the difference between the two arts. These changes may have acted

as an impetus for Anandavardhana's fusion of poetics and dramaturgy.

The devotional potential of poetry is not lost on Anandavardhana's commentator

Abhinavagupta. Anandavardhana begins analyzing the effect poetry has on its audience

and includes a ninth rasa "tranquility" (śanta) to extend his theory to the great religious

epics.5 Abhinava greatly refines the notion of rasa, increasing both the importance and

the scope of aesthetic experience.5 By closely examining the earlier commentators on the

Nātyaśāstra, Abhinava is able to develop a theory of aesthetic psychology that has often

54 Gerow (1977:252). 55 Ingalls, et al. (1990:814). 56 Ingalls, et al., (1990:35).

38

been compared to T.S. Eliot's theory of the universal correlate.57 From this he argues for

the religious value of artistic appreciation, claiming that pure rasa experience is

analogous to blissful union with Brahman.

Mammata bases his synthesis in the structure of the older poetics. The

Kāvyaprakāśa, like the early works on poetics, strives to explain what Gerow calls, "the

formal expressive repertoire of the poet, as a practitioner of a different and uniquely

intelligible kind of language." Kārikās 1 through 4b are entirely within the early

tradition. In 4ab, Mammata defines poetry using the three technical terms of that tradition

(excellences, blemishes, and poetic ornaments). The entire second half of the

Kāvyaprakāśa is devoted to an encyclopedic presentation of these three concepts.

However, in karika 4cd, he introduces Anandavardhana's grammatical tool, suggestion.

Having already defined the essence of poetry along traditional lines, he cannot follow

Anandavardhana in making suggestion its essence. Instead, he ingeniously uses

suggestion to establish a hierarchy of excellence among poems. By this move he manages

both to accept the post-Anandavardhana preference for suggestive poetry and resolve

Anandavardhana's problem of how to account for non-suggestive poetry and suggestive

literature that is not poetry.

Mammata reviews the grammar of suggestion, along with denotation and

indication, in the second and third chapters. In these sections, suggestion is treated as one

57 Perhaps the most striking example is Sahal, K.L., "Objective Correlative and the Theory of Rasa", Calcutta Review, N.S. 2, no. 2 p.237, but the comparison is common. 58 Gerow (1977:258).

39

of the three factors (or semantic powers) that comprise the meaning of a sentence (see

introduction 2.2). Included in suggestion are such linguistic phenomena as irony, pun,

and the intention of a metaphor, as will be explained in various comments on particular

poems.

Mammata gives little detailed attention to Abhinava's psychological theory of

rasa and less to his theologico-aesthetic theory. In kārikā 2, he repeats the claim that rasa

experience is analogous to experience of Brahman, but offers no defense or development

of the idea. In his commentary to kārikās 27 and 28, Mammata quotes the Nātyaśāstra

6.31 and summarizes Abhinava's history of rasa theory. After summarizing and

approving of Abhinava's theory, Mammata simply moves on, adding neither criticism nor

elaboration.

However, the central claim of Abhinava's Locana is presupposed in all of

Mammata's chapters four, five, and six. Abhinava establishes that the semantic power of

suggestion, prevalent in poetry, makes possible a particular aesthetic effect for a receptive

audience. The fact that Mammata has nothing theoretical to add to Abhinava's claim does

not reduce the importance of this theory in Mammata work. The three chapters in

question focus on illustrating the workings of poetry as an evoker of rasa." It should be

noted that suggestion has a much expanded role in these chapters. It is no longer only a

necessary tool used to understand the grammar of a sentence. Suggestion is also viewed

59 Chapter four treats poems in which suggestion is primary, chapter five poems in which it is secondary, and chapter six poems in which it is absent. Even chapter six uses rasa (or more exactly, absence of rasa) as the defining character for a class of poetry. Strictly speaking, these chapters also deal with the suggestion of emotions not eligible to become rasa.

40

as conveying both the intended meaning of an entire piece and its mood. If it is true that

Mammata is responsible for popularizing Abhinava, he did so through a streamlined

presentation of Abhinava's theories and more than a hundred examples that show the

theory in application.

The Kāvyaprakāśa can be divided into four parts. The first chapter presents the

creative synthesis that will be worked out in the other three parts. It begins along

traditional lines, but manages to integrate suggestion with the older aesthetics of

ornament in a new and ingenious manner. Chapters two and three provide a philosophy

of language inspired by some of Anandavardhana's innovations, most notably the concept

of suggestion as a third semantic power. Chapters four through six illustrate Abhinava's

refinement of rasa theory. Chapters seven through ten return to an encyclopedic

presentation of the traditional concepts of alankāraśāstra, namely excellences, blemishes,

and ornaments.

5. Mammața's Legacy

Mammata's integration of the Dhvani School writers into the older tradition proved

definitive for the Sanskrit tradition in that all later theorists that discuss poetics as a

whole do so within Mammata's framework. Mammata's solution has the additional

advantage of being inclusive. Every poem in the Sanskrit tradition can find a place in his

60 That Viśvanātha added a chapter on drama to Mammata's structure does nothing to undermine this claim. A house with an extra room is still the same house. Similarly, although there are differences of emphasis, Jaganātha's Rasagangādhara can be mapped onto the Kāvyaprakāśa with little difficulty.

41

synthesis in contrast with many of the earlier theories.61 His system has the opposite

advantage of not being too inclusive, in contrast with Anandavardhana's.62 We saw that

Mammata's synthesis was presented in his definition of poetry and his creation of a

hierarchy of quality among poems. Both of these were scrutinized by later theorists, who

tend to reject his definition and accept his hierarchy.63

Given the synthetic and encyclopedic nature of the Kāvyaprakāśa, one would

expect it to be very dense, and it is. What is surprising, however, is its appearance of

simplicity. In presenting his compilation of earlier theories, Mammata rarely presents the

arguments found in those theories (the largest exception being the verse on rasa).

Mammata also rarely develops the arguments for his own view. And when he does so,

they are given in skeletal form. Opposing positions are often only briefly sketched. With

the exception of the second half of chapter five, Mammata presents sparse counter

argument. Mammata's method of presentation makes the Kāvyaprakāśa seems more a

textbook than polemic or creative addition to an ongoing debate.

I would argue that the Kāvyaprakāśa was intended more to teach the "correct

view" than to defend this view against objections. This theory might explain its extreme

popularity among commentators, most of whom fill in the missing arguments to a greater

61 There exist poems with no ornaments, on the one hand, and poems with no rasa, on the other. Jagannātha claims that there are poems that are excluded by Mammata's definition, but his examples are too short to be convincing. See the Rasagangadhara 1.6. 62 There are types of writing that evoke rasa yet are not poetry. 63 Viśvanātha attacks it in his commentary on the Kāvyaprakāśa and in his Sāhityadarpaņa, Jagannātha in the Rasagangādhara 1.6, and, more recently, Gajendragadkar in his notes to the Kāvyaprakāśa (1939:151- 155).

42

or lesser extent.64 Indeed, I have found it necessary to do so in my commentary as well.

Mammata's text practically cries out for commentary, and if it was truly written to

present the views, Mammata himself probably expected it to be commented upon either

orally to students or by teachers writing out the fine points of the analysis.

Mammata's streamlining of the argument is counterbalanced by his extensive use

of examples to illustrate philosophic claims. Indeed, Mammata reverses the priorities of

his hero Abhinava in this regard. Instead of lengthy arguments enlightened by an

occasional example, Mammata gives hundreds of examples structured by the theoretical

framework. When the theoretical issues are largely resolved (and thus not in need of

further argument), it is reasonable to turn one's attention to the poems themselves.

Mammata's popularity may also be due to his preference for example over minutia of

argumentation.

Mammata manages to pull together most of the earlier poetic theory into one

theoretic whole. His text is both superficially attractive and rewarding of deeper study.

His framework is inclusive enough to capture all that had been admired in Sanskrit

poetry, yet also capable of distinguishing poetry from other forms of writing. The theory

itself is grounded in a deep philosophic tradition, but Mammata does not let the

philosophy overshadow the poetry it is supposed to explain. The Kāvyaprakāśa reveals

and tries to explain poetry's unique power and value. To those who are so inclined, it is

an invitation to investigate theoretical issues at greater depth, and an indication of how to

64 It is often claimed that only the Bhagavadgītā has more commentaries.

43

do so. To others it is an invitation to use the existing theory to enrich their appreciation of

the poems so nicely collected, pushing them again and again to see details that they might

otherwise miss. For all these reasons, Mammata well deserves his position as the most

popular Sanskrit poetic theorist.

6. Translation Method

The Kāvyaprakāśa is a theoretical work. It seeks to define poetry, classify its various

types, explain how they convey their meaning, and examine why some poems are better

than others. It consists of three distinct types of writing, each of which must be handled

differently in translation. The core text consists of theoretical verses in fixed meter. These

are explained in a prose commentary, also by Mammata. Finally there are illustrative

examples, which are drawn from the massive realm of Indian poetry. The purpose and

form of all three must be considered. One can ask, for example, why Mammata writes the

core text in meter. What is gained by using the verse and commentary form? Precisely

what role do the examples serve? Answers to these questions help shape the form of the

translation.

Before I can turn to these questions, I should make clear the desired audience and

the goal of the translation. This translation is targeted at three distinct audiences. The first

is philosophers and critics. Mammata's text is primarily theoretic and is philosophically

rich. It treats extensively questions of semantics, emotion, and aesthetics. It also touches

44

on questions of ontology, epistemology, religion, and ethics. Those interested in

comparative literature or comparative aesthetics might find the text especially interesting.

My second audience is readers of Indian poetry. One of Mammata's goals is to

teach his reader how to read Sanskrit poetry. Anyone interested in Sanskrit poetry, either

in the original or in translation, might profit from this work. Furthermore, because of

Sanskrit's historical influence, I imagine that readers and writers of poetry in the modern

languages of the sub-continent will also find this text rewarding.

My final audience is readers and writers of English poetry. Some of Mammata's

ideas can certainly travel outside the Sanskrit context. I also hope that readers of poetry

in translation will find some of these poems excellent. If not, the fault lies with me, not

with the original Sanskrit, for Mammata has chosen some great poems.

My goal has been to produce a text readable without any knowledge of Sanskrit.

To this end I have attempted to make the translation complete, which means that no

Sanskrit words that are not already current in English have been left untranslated. For

those acquainted with Sanskrit, I have tried to include the Sanskrit technical terms in

parentheses the first time I translate them.

I turn now to the structure of the Kāvyaprakāśa and how the various types of

writing there should be translated. First, why did Mammata write theoretic verses and a

self-commentary to explain them? The verse - commentary form is about as old as

Sanskrit philosophy, beginning with the grammarian Pānini, c.500 BCE, and the

Mīmāmsāsūtra, c.200 BCE. The form is used in hundreds of later philosophical and

45

scientific works. The verses were to be memorized by students. Like cue cards in drama,

they can be used to evoke the rest of the argument. The headings of Descartes' Principles

serve a similar function and were (and still are, in France) used in a similar way. To fulfil

their function, the verses must to be short and easy to remember. Again, they are cues,

not summaries. The commentary fleshes out the argument, providing much that is both

essential and absent from the verses. Neither was intended to be what we would call an

independent text.

Although the initial justification for composing theoretical verses was to aid

memorization, later in the tradition, verses became also a means of showing one's

technical virtuosity. How should these Sanskrit verses be translated? First, I think it is

safe to assume that nobody will memorize an English translation of the Kāvyaprakāśa,

thus the first possible motivation for a verse translation of these verses is removed.

Second, it is not part of Western traditions to philosophize in verse, and even less to show

off by doing so. Thus those reasons too are removed. Remembering that Mammata's

main goals were theoretical, and that in modern English theoretical goals are pursued in

prose, I argue that the theoretical verses are most fully translated when rendered in prose.

The next question is how to deal with Mammata's division of the theoretical text

into two layers, verse and commentary. Along the lines just set out, it is tempting to

assert that these two layers should be combined to conform with normal modern English

theoretical style. What would be lost if it were suppressed? First, we would lose the

implication that the verse material is primary, while the commentary is secondary. If this

46

were all, I would argue that this difference of importance would be better represented in

an established English form. But Mammata's commentators - and there were many -

would be difficult to understand if the structure of the original were altered. As of now,

none of these have been translated, but hopefully this situation will change in the future.

Finally, the poems. I assume that Mammata chooses the poems he does for two

reasons: they illustrate his point and he thinks them good poetry. Given that they are

examples, clearly the translations must function as examples to English readers. As

obvious as this seems, the existing translations fail on this score. Secondly, I feel that any

translator of this text should, to the best of his ability, render the poetic examples as

poetry. A failure to do so deeply falsifies the impression a reader will receive of the

original.

What is a poetic translation of poetry? I suggest that it is a translation that takes

seriously the aesthetic aspects of the original. These include qualities such as the tone, the

level of diction, the rhythm, the rhyme, the line length, the use of vowels and consonants,

the weight of the words, the images, the metaphors, etc. Another aspect is the resonant

semantic capacities of the words used. Most of all, for these poems one has to consider

the aesthetic emotion suggested or produced by the whole. Not all the aspects of Sanskrit

poetry can be rendered in modern English. Some metaphors and images are culturally

specific and the connotations of words shift from language to language. Not even all the

aspects that can be rendered can be rendered in a single translation: different translations

will capture different aspects. My method has been to first understand how poem is

47

functioning as an example. Then I have tried to isolate the most important aesthetic

aspects of each poem and capture as many of those as I can while retaining the poem's

function as an example. All translation is a compromise, the richness of poetry makes this

doubly true of poetic translation. Where other translations exist, I have occasionally

quoted them.

A word about puns, which form an important part of the Sanskrit poetic tradition.

Sanskrit punning poems are often highly complicated and contain many words with

double meaning. The poets often strive to present a pair of parallel meanings throughout

the entire poem. Here I have found no option but to give a pair of juxtaposed literal

translations for each poem, one reflecting each of the punning word's meanings. Where

many words have a single meaning shared by both "sides" of the pun, I have inserted a

center column. For example, poem 78:

Primary Meaning Common Meaning Secondary Meaning Who is not flooded with delight By the good scripture! By a lover's visit! Always intent on Moral advice Enjoyment And leading to final salvation And release from frustration In a heavenly place In a secluded place.

The way these poems are used in the text (along with the relative difficulty of punning in

English) has forced me to sacrifice any attempt to achieve poetic merit in English. As I

said above, my primary obligation is to make sure the English translations really do serve

as illustrations of what they are supposed to illustrate. I hope the parallel presentation is

more useful and enjoyable than a prose explanation of the puns.

48

As I hope I have made clear, my primary concern at all three levels of the text

(theoretical verses, commentary, and poetic examples) has been to make the translation

'work' as an English book on poetics. I hope I have chosen the correct method to make

Mammata's text as useful, accessible, and accurate as possible.

7. Purpose and Content of Comments

The main purpose of my comments is to provide easier access into intricacies of

Mammata's text. I do this in four ways. First, I explain particular cultural references, such

as the names of people and places. When a poem requires it, I sketch the plot of the larger

work from it is drawn. I also explain metaphors and images that might not be clear in

English and indicate when the connotations of a Sanskrit word are importantly different

than those of its English translation.

Second, I expand and clarify Mammata's terse presentation of philosophical

arguments. This often requires providing the context of the debate and the implications of

Mammata's views. Readers familiar with Indian philosophy will notice that Mammata

does not give the traditional exposition of his opponents' views (pūrvapaksa). I try to

supply this by making clear which positions Mammata is fighting against and the main

arguments in their favor.

Third, I explain how the poems function as examples. When necessary, I clarify

the suggestions and poetic ornaments. Then I tie the poem back into the philosophic

context by trying to show exactly what aspect of the poem serves to illustrate each

49

theoretical point. This is particularly challenging when the example turns on a detail of

the Sanskrit language, but I have tried to avoid being tedious.

Fourth, I give historical information pertaining to Mammata's views, filling in the

details of the historical outline of the introduction (sections 4 and 5). As Mammata rarely

mentions his sources, I try to identify them. I also try to note when Mammata either

introduces something new into the tradition or actually contradicts one of his respected

sources. Finally, I try to indicate which of Mammata's points are especially influential for

later criticism and poetry.

50

Appendix to the Introduction 1 - Transliteration Details.

Concerning the Sanskrit text presented:

1. The devanāgarī text has been transliterated into the Roman alphabet using the

standard modern method.

2. I have followed the sectioning of the text that the early commentators imposed (which

was also followed by the later commentators). The only exception to this is an

occasional transition word that has been moved to make the English more clear.

3. I have followed the Mohan text (1995) as my base text. I have checked it against the

Jha (1925) and the Dwivedi (1977) texts. Where there is a conflict, I have further

checked it against the text of the Anandāśrama Sanskrit Series No.66 (1929).

4. I have listed corrections to the Mohan text in an appendix. If the correction is clearly

a typo, not appearing in the other texts, the correction is not further justified. If the

correction is more substantial (a very rare occurrence), the alternatives and a

justification are given.

5. I have included the Mohan page number in parenthesis after each kārikā or section

thereof that appears independently in the text.

6. I have not transliterated Mohan's numbering of the kārikās or the poems as these

numbers do not exist in the earliest manuscripts. They are, however, represented in

the translation.

51

7. Mohan is not consistent with his use of bold text. I have used it only for the chapter

headings and the kārikās. Mohan decided to use commas in his text. I have either

transliterated these as dandas or dropped them, as seemed appropriate.

8. I have inserted an additional danda when one seemed desirable on grounds of

sentence unity if it exists in other versions of the text.

9. Mohan has added paragraph brakes in his text. Most of these have been suppressed.

10. Mohan has no consistent use of the double danda in prose. This is partially, although

not wholly, due to his addition of paragraph breaks. I have tried to use them

whenever there is a paragraph break in the English, which makes it slightly easier to

move back and forth between the texts.

11. I have not restored sandhi. However, I have tried to insert a space between words

where this is possible without restoring the sandhi. I hope this make the text more

useful to readers who want to identify a single word.

52

Appendix to the Introduction 2 - Translation Details

The following list gives some of the details of my translational strategy with reference to

specific Sanskrit constructions:

1. Parentheses are used to indicate meanings not on the surface of the Sanskrit text but

very closely implied. I supply them in the interest of readability.

2. Long Sanskrit sentences broken in translation often require repetition of the subject.

3. The antecedents of Sanskrit pronouns are often supplied.

4. Sometimes Sanskrit singulars are translated by English plurals in theoretical contexts.

5. The sense of adi (etc.) is often spelled out. Other times X-ādi is translated as "X and

the like." For example, gotvādi is sometimes translated, "universals such as

cowhood."

6. When a Sanskrit adjective is used to modify two or more nouns, two different English

adjectives are sometimes employed.

7. Following the normal Sanskrit convention, Mammata (or a later copyist) places

chapter names at the end of each chapter. These have been moved to the head of the

chapter in the translation. Similarly, the last sentence of a section sometimes

identifies its speaker. These sentences have been moved to the head of the section. I

have, however, indicated these relocations in the notes.

8. In Mammata's Sanskrit, the immediate future is usually expressed by the present and

the future tense used to refer to a more distant future. Thus I have sometimes inserted

'now' in the first case and 'later' in the latter.

53

9. The optative of Vas (to be) is often translated by the present indicative. Mammata's

use of the optative, in this case, is prescriptive and not potential.

10. When making a comparison, Mammata often only supplies what is different, leaving

the reader to fill in the rest. This is possible in Sanskrit because nouns are declined,

making the parallel obvious. In English, the whole comparison is often given without

parentheses.

11. Mammata uses the locative to indicate the group or class (C) to which an item (x)

belongs. Often there is no way to retain the locative in English and so I often render

this x is a case of C or x is a C.

54

kāvyaprakāśaḥ

atha pratham ullāsaḥ

ity kāvyaprakāśe kāvyasya prayojanakāranasvarūpaviśesanirņayo

nāma pratham ullāsaḥ II

The Elucidation of Poetry

Chapter One

Specification of the Purpose,

Cause, and Types of Poetry

granthārambhe vighnavighātāya samucitestadevatām granthakrt parāmrśati

niyatikrtaniyamarahitām hlādaikamayīm ananyaparatantrām I

navarasarucirām nirmitam ādadhatī bhāratī kaver jayati II (1)

niyataśaktyā niyatarūpā sukhadukhamohasvabhāvā

paramānvādyupādānakarmādisahakārikāranaparatantrā satrasā na ca hrdyaiva taih

55

tādrśī brahmaņo nirmitir nirmāņam I etad vilaksaņā tu kavivānnirmitiḥ I ata eva jayati I

jayaty arthena ca namaskāra āksipyate iti tām praty asmi pranata ity labhyate II

(1.1 Opening Auspicious Verse)

At the beginning of the book, the author focuses on the goddess properly preferred (by

poets), hoping that she will remove all obstacles.

1. The poet's goddess, Bharatī, excels by establishing a world composed of

pure delight, dependant on nothing beyond herself, resplendent with the nine

rasas, and free from the laws set forth by Fate.66

The creation of Brahman is necessarily ruled by the power of Fate. It consists of pleasure,

pain, and spiritual ignorance, it is dependant on material causes, i.e. atoms, etc., and

auxiliary causes, i.e. karma, etc, and finally it consists of (only) six rasa67, which are not

even always delightful. But the creation of the poet's speech is different from this, and for

precisely this reason it excels. And by using excels reverence is implied. Thus the verse

means, "I bow to her."

Comments

It is traditional to begin any classical Sanskrit work with a prayer or invocation of the god

that the author feel most relevant to his work. This practice is called mangala, "doing

65 The nine rasas are: erotic, comic, heroic, horrific, enraged, terrific, pitiful, wonderful, and the peaceful. 66 Niyati is the goddess Fate (the daughter of Meru). Niyati also means "necessity," "natural laws," etc. 67 Here rasa refers to the six gustatory flavors: sweet, salty, bitter, sour, pungent, and astringent.

56

something auspicious." Bhāratī is another name for Sarasvatī, the goddess of speech and

poetry. Mammata playfully compares her to Brahman, the Supreme Being, to vaunt the

advantages of poetry over reality. These include its independence from physical matter,

its freedom from the laws of providence, its diversity of subjects, and its pleasant nature.

ihābhidheyam saprayojanam ity āha

kāvyam yaśāse 'rthakrte vyavahāravide śivetarakșataye

sadyah paranirvṛtaye kāntāsammitayopadeśayuje II (41)

kālidāsānādīnām iva yaśah śrīharsāder dhāvakādīnām iva dhanam

rājādigatocitācāraparijñānam ādityāder mayūrādīnām ivānarthanivāranam I

sakalaprayojanamaulibhūtam samanantaram eva rasāsvādanasamudbhūtam

vigalitavedyāntaram ānandam prabhusammitaśabdapradhānavedādiśāstrebhyah

suhrtsammitārthatātparyavatpurāņādītihāsebhyaś ca śabdārthayor guņabhāvena

rasāngabhūtavyāpārapravaņatayā vilaksanam yat kāvyam

lokottaravarņanānipuņakavikarma tat kānteva sarasatāpādanenābhimukhīkrtya

rāmādivad vartitavyam na rāvanādivad ity upadeśam ca yathā yogam kaveh sahrdayasya

ca karotīti sarvatha tatra yatanīyam II

(1.2) The author now states the purpose of poetry;

2. Poetry brings fame, wealth, knowledge of the world, destruction of the

inauspicious, and instant, complete liberation. Poetry also serves to give

advice in the manner of a beloved woman. 57

Poetry brings fame, like that of Kālidāsa68 and others, wealth like that of (kings) like Srī

Harsa and (poets) like Bānao, knowledge of the proper conduct displayed by kings and

others, protection from adversity, like (in the case of) the poet Mayūra by (his poems for)

the sun", etc., and bliss, the highest of all goals, which, arising immediately from the

taste of a rasa, consumes all other knowledge. Poetry is the work of poets skilled in

descriptions of this world and others. Religious texts like the Vedas are composed

essentially of authoritative statements (śabda) and thus are like teachers. In the Puranas

and histories, what really matters is the point or moral (of the story), and they are thus

like friends. But poetry is different. In it, one strives to produce rasa through graceful

expression and a worthy subject. By turning the heads of the poet and the reader and

filling them with passion (sarasatā) like a beloved woman, poetry entices them to act

like Räma and other heroes and not like Rävana and other villains."For all these reasons, una and other he poetry is to be pursued.

Comments

Following tradition, Mammata first explains why the subject of his work, poetry in this

case, should be studied. This passage is close to Bhämaha verses two to eight. After

giving examples of famous and rich poets, and claiming that poetry can be used to

68 Kalidāsa was a Sanskrit poet and playwright of the early 5th century. 69 I follow Gajendragadkar in reading Bāņa for Dhāvaka. Bāņa was a poet active in the court of Srī Hārśa

70 Mayūra was another 7th century poet. Having contracted leprosy, he composed poems to the god of the (7th century).

sun and was cured of his disease as a reward. 71 Rāma is the hero of the epic poem the Rāmāyana. Rāvana is the antagonist. 58

petition the gods, he states that poetry can instruct. Poetry instructs not like a religious

text, nor like a history that aims to convey a moral, but rather like a lover that pleases and

thus improves."2 The goal of instruction made easy by pleasure is familiar from western

poetics (one thinks of Dryden, for example), but that is not quite the point here. The idea

is not to sugarcoat morality, but rather to spur one to greater actions through love of

exemplary people illustrated in poetry. Think of the Greek desire to emulate Achilles, or

the chivalrous emulation of Arthur. What is striking here, however, is that the portrayal

of human ideals is compared to a woman. Like a woman, a good poem can capture one's

heart and entice one to do its bidding. Would it be too much to see Freud's theory of the

sublimation of sexual desire prefigured? Probably. But in any case, sexual desire and

culture ideals are here seen to move people in analogous ways.73

The analogy seems to fall apart when one considers that a beloved woman only

evokes a single rasa (the erotic), while poetry can evoke any of the nine. Mammata picks

an example, however, in which a woman Sīta, inspires not only the erotic passion, but

many others as well in the course of the Rāmāyana (indeed, a clever critic could easily

show that she inspires all nine).

The real downside of this analogy is that there is no reason to believe that the

beloved woman will always influence their lovers for the better. Sīta, for example,

evoked not only the morally sanctioned passions of Räma, but also the morally

72 These metaphors date to Abhinava, at least, who uses them three times (Ingalls, et al., 1990:71, 437, 533). 73 This analogy has a long history in Sanskrit. The Rg Veda (10.71.4) claims that only to one who understands the hidden or inner significance of words does speech reveal herself completely. Note also that the goddesses of speech and learning are both female. 59

reproachable passions of Rāvana. Why, analogously, is there any reason to believe that

poetry will evoke only commendable passions? If there is something distinct in poetry

that makes its moral influence less ambiguous than that of a beloved, Mammata needs to

tell us what it is. If there is not, there is no basis for the claim that a poem will fill one

with the desire to act like a hero and not like a villain.

evam asya prayojanam uktvā kāraņam āha I

śaktir nipuņatā lokaśāstrakāvyādyavekșaņāt l

kāvyajñaśiksayābhyāsa iti hetus tadudbhave II (63)

śaktih kavitvabījarūpah samkāraviśesah I yām vinā kāvyam na prasaret prasṛtam vā

upahasanīyam syāt l lokasya sthāvarajangamātmakalokavrttasya | śāstrāņām

changovyākaranābhidhānakośakalācaturvargagajaturagakhadgādiloksanagranthānām I

kāvyānām ca mahakavisambandhinām l ādigrahaņād itihāsānām ca vimarśanād

vyutpattih I kāvyam kartum vicārayitum ca ye jānanti tadupadeśena karane yojane ca

paunah punyena pravṛttir iti trayah samuditāh na tu vyastās tasya kāvyasyodbhave

nirmāņe samullāse ca hetur na tu hetuvaḥ II

(1.3) Having explained the purpose of poetry, the author states its cause:

3. The cause of poetry consists of: capability, skill resulting from familiarity

with the world, the sciences, poetry, etc., and practice of the instruction given

by experts of poetry.

60

Capability is an individual propensity and is the seed of poetry. Without capability,

either poetry does not appear, or, if it does, it is laughable. Familiarity with the world

means study of the behavior of all things physical, animal and spiritual. Familiarity with

the sciences means study of technical works on meter, grammar, and semantics, as well

as works on the arts and the four goals of life", and those on elephants, horses, weapons,

etc. Familiarity with poetry means study of the works of the great poets. Finally, the

etc. implies the study of history. Practice proceeds by repeated creation, i.e., composition

guided by those who know how to create and critique poetry. Capability, skill, and

practice are not three separate causes, but rather the single collective cause of the

composition and brilliance of poetry.

Comments

This section is admirable in its acknowledgment of the complexity of human nature and

culture." It bares comparison with Horace's Ars Poetica, lines 268-69 and 408ff. The

topics Mammata raises received much greater attention in a slightly earlier work, The

Kāvyamīmāmsā. For example, it divides capacity into seven types. By claiming that a

poet needs to have studied the arts he mentions, Mammata touches on issues raised by

Plato in the Republic, et al., but these issues receive no extended attention in the Sanskrit

tradition to my knowledge.

74 The four goals of life: pleasure, wealth, fulfillment of duty, and liberation. 75 This section is very similar to Bhāmaha verses nine and ten.

61

When a single entity gives rise to another single entity, one speaks only of a

single cause. If the first entity is complex, as in this example, one can speak of different

parts of the entity combining to make a complex cause. The point here is that although all

the parts are necessary, none is sufficient for the creation of good poetry.

Despite the use of this technical vocabulary, Mammata does not rise to the

challenge of giving a precise description the causal process implicit in the creation of a

poem. But I doubt he is to be condemned for this; such "descriptions" tend to falsify by

oversimplification of the creative process. Jagannātha criticizes Mammata's definition in

the Rasagangādhara 8, where he defends intuition as the only cause of poetry.

evam asya kāraņam uktva svarūpam āha I

tad adoşau sabdārthau saguņāv analamkrtī punaḥ kvāpi | (81)

dosagunālamkārāh vaksyante | kvāpīty anenaitad āha yat sarvatra sālamkārau kvacit tu

sphutālamkāravirahe 'pi na kāvyatvahānih l yathā

yah kaumāraharah sa eva hi varas tā eva caitraksapās te conmīlitamālatīsurabhayah praudhāh kadambānilāh l sā caivāsmi tathāpi tatra suratavyāpāralīlāvidhau revārodhasi vetasītarutale cetah samutkanthate II

atra sphuto na kaścit alamkārah l rasasya ca prādhānyān nālamkāratā II

(1.4) Having stated the cause of poetry to be such, the author states its nature:

4ab. Poetry consists of words and meanings that possess excellences and are

free from blemishes. In some cases, it lacks poetic ornament.

62

Faults, excellences, and poetic ornament (all technical terms) will be described

later. By saying, in some cases, the author states that generally there is a poetic

ornament. But even when there is no striking poetic ornament, poetry is possible.

For example:

He who made love to me first loves me still. Ever the same are the nights of spring, As are the breezes from the flowering grove Heavy with the scent of jasmine blooms. I too remain the same. And yet - On the bank of the Reva river, In a secluded glade among the reeds, I long for those games That bring deep sexual rapture. (1)76

Here there is no striking poetic ornament. (Why not call the rasa an ornament?)

The rasa is the primary purpose of the poem, not a poetic ornament.

Comments

76 Attributed to Śīlābhattārikā in several later anthologies. Dwivedi says it is found in the Sūktimuktāvalī, a collection of poems the post-dates Mammata by at least a century. Gajendragadkar adds that it is found in the Sārngadharapaddhati as verse 3768. It is also found in the Subhāsitaratnakosa as verse 815, although with several important differences. See Kasambi and Gokhale (1957:150). 1a. yaḥ - who, kaumāraharāh - deprive of virginity, sa - he, eva - still, hi - indeed, varas - husband, tāḥ - the, eva - same, caitrakșapās - nights of Caitra month. 1b. te - the, ca - and, unmīlita - opened, mālatī - jasmine, surabhayah - perfumed, praudha - heavy/luxuriant, kadambānilāḥ - the breeze of/from/in the kadamba trees, which bloom in the spring. 2a. sā - it, ca - and, eva - same, asmi - I am, tathāpi - even so, tatra - there, surata - being deeply satisfied sexually, vyāpāra - coming together with/bringing about, līlā - play, vidhau- type 2b. revārodhasi - on the bank of the Revā, vetasī - cane/citrus, tarutale - clear ground under a tree, cetah - heart, samutkanthate - longs after. 63

Attempts to give a precise definition of poetry fill the Sanskrit tradition. Gajendragadkar

gives a detailed critique of Mammata's definition, comparing to the definitions given by

the other major writers of poetics." The debate over which definition is best, like so

many others in poetics, seems to me to turn on whether one takes the view of the poet or

the reader/audience. The early definitions, which Mammata echoes, are writer-based. For

example, Bhamaha, after defining the faults and ornaments, claims, "Poetry is a

combination of verbal and semantic ornaments." They tend to focus on what linguistic

elements need to be incorporated and avoided in composition. Later definitions, like that

of Anandavardhana, are audience-based and focus on the experience to be triggered by a

piece of writing. Thus he begins the Dhvanyaloka with the claim, "Suggestion is the soul

of poetry." One is not surprised when a twentieth century commentator like

Gajendragadkar, heavily influenced by Anandavardhana and Western romantic theories,

criticizes Mammata's definition for leaving out evocation of rasa. Such criticism,

however, tend to show little appreciation for the way Mammata integrates rasa into his

system.

In his attempt to write a comprehensive summary of poetic theory, Mammata is

faced with the task of trying to integrate Anandavardhana 's reader-based perspective

within the older writer-based tradition. I have shown that Mammata's definition of poetry

is traditional by relating his text to Bhāmaha's. In the next kārikā, he will try to use the

77 Gajendragadkar (1939:151-155). 78 Kāvyālankāra 1.16. I feel this verse, when read in context, has the meaning I have given. The common translation is non-sense: "poetry is a combination of words and meanings." For more on this issue, see the introduction. 79 Dhvanyāloka 1.1. 64

types of poetry (from Anandavardhana) to introduce the reader-based perspective. This

will lead him to a "digression" extending to the end of chapter six. In chapter seven, he

takes up the definitions of faults, excellences, and poetic ornaments, which, according to

tradition, should have been explained right after their introduction in this definition.80 I

think Mammata's awareness of the novelty of his method shows through in the

accompanying assurance that he will eventually explain these terms.

The poem treats a familiar theme in Sanskrit poetry. Compare: Near death/ by the

sacred Täpi River/ I will tell the truth/ even now my eyes/ return to the stand of reeds/ by

the water 81

tadbhedān kramenāha

idam uttamam atiśayani vyangye vācyād dhvanir budhaiḥ kathitaḥ II (112)

idam iti kāvyam I budhair vaiyākaranaih

pradhānabhūtasphoțarūpavyangyavyañjakasya śabdasya dhvanir iti vyavahāraḥ

krtaḥ I tatas tanmatānusāribhir anyair api

nyagbhāvitavācyavyangyavyañjanaksamasya śabdārthayugalasya l yathā

nihśesacyutacandanam stanatatam nirmrstarāgo 'dharo netre dūramanañjane pulakitā tanvī taveyam tanuh I mithyāvādini dūti bāndhavajanasyājñātapīdāgame vāpīm snātum ito gatāsi na punas tasyādhamasyāntikam II

atra tadantikam eva rantum gatāsīti prādhānyenādhamapadena vyajyate II

80 Indeed, if you modified chapter one and cut chapters two through six, Mammata's text could easily pass as pre-Abhinava. 81 Merwin and Masson (1977:43). 65

(1.5) The author now states the division of poetry in order (of merit):

4cd. That (poetry) is best in which the suggested meaning dominates the

literal meaning. It is called suggestive poetry by the learned.

That means poetry. The learned means the grammarians, who define suggestion

as: a word that suggests the suggested meaning by means of the eternal concept

(sphota) that is the primary meaning. Following this view, others also define

suggestion as: a connection of expression and meaning capable of suggesting a

suggested sense that subordinates the literal sense. For example, a woman says the

following to her go-between:82

O false messenger! You know not the pain you bring. The sandal paste is washed clean off your rounded breasts, And the rouge from your lips is completely rubbed off. Gone is the makeup from the corners of your eyes, And your slender body still shakes. So you went for a bath in the tank, And not to be with that wretch? (2)83

82 The information in this sentence is implicit in the use of the word dūti. 83 This verse is in many recensions of the Amarusataka, but falls outside the first hundred. I have found it as 103 and 105. Dwivedi and Gajendragadkar refer to it as 105. Frustratingly, it is excluded with no explanation from all existent English translations. la. niḥśesacyuta - completely, candanam - sandal ointment, stanatațam - rounded female breast, nirmrsta - wiped off, rägah - color/redness , adharaḥ - lower lip/lips. 1b. netre - in the eye, dūram - far edge, anañjane - free from collyrium, pulakitā - horripilated (a symptom of stimulation), tanvī - possessed of a body, tava - of yours, iyam - this, tanuh - slender. 2a. mithyāvādini - o speaker of lies, dūti - o messenger, bāndhavajanasya -, ajñāta - not knowing, pīda - the suffering, agame - that comes to me, 2b. vāpīm - tank, snātum - to bath, itah - from here, gatāsi - are gone, na - not, punar - on the contrary, tasyādhamasya - of that wretch, antikam - presence. 66

Here, "You have gone near him only to fool around (rantum)" is suggested by the

prominence of the word "wretch".

Comments

Mammața here leaves the early tradition and draws heavily from

Anandavardhana. Notice that he has been forced to use a whole new technical

vocabulary concerning suggested meaning. The explanation of this vocabulary

will occupy the next two chapters. Using rasa to impose a qualitative hierarchy on

poems is Mammata biggest innovation. Chapters four, five, and six are devoted to

explaining the three levels of quality.

It is well known that Mammata, while identifying the same three types of

poetry as Anandavardhana, introduces a hierarchy of quality, but nobody, to my

knowledge, has offered an explanation of Mammata's motivation. If my thesis is

correct, Mammata's reason for doing so is not hard to find. In so much as

Anandavardhana considered suggestion (of rasa) the essence of poetry, he had no

need to explicitly develop such a hierarchy. It is implicit in his definition.84

Mammata, by insisting on a formalist definition of poetry, has to assert what

Anandavardhana could assume, namely, that suggestive poetry is better. His

reward, however, is great. He successfully creates a theoretic structure that can

84 See his commentary on 3.36b (Ingalls, et al. 1990:611). Anandavardhana's problem was how to explain that last, non-suggestive, type could be considered poetry. See Anandavardhana 3.40 to 3.42 and Abhinava's comments (Ingalls, et al. 1990:632 to 648). 67

house both of the earlier traditions. But what is more, his definition of poetry also

better reflects the actual linguistic usage.

atādrśi guņībhūtavyangyam vyangye tu madhyamam | (136)

atādrśi vācyād anatiśāyini l yathā

grāmataruņam taruņyā navavañjulamañjarīsanāthakaram paśyantyā bhavati muhurnitarām malinā mukhacchāyā II

atra vyañjulalatāgrhe dattasamketā nāgateti vyangyam gunībhūtam tadapeksayā

vācyasyaiva camatkāritvāt II

5ab. A different type of poetry occurs when suggested meaning is not

dominant. It is called poetry of subordinate suggestion, and is of middle

rank.

A different type of poetry occurs means cases where the suggested meaning does not

dominate the literal meaning. For example:

On seeing the young man Holding a fresh ashoka bouquet, The young woman's face Suddenly flushes a deep red. (3)85

Here the suggested meaning, namely, (that the girl) did not go to the rendezvous

in the ashoka grove, is subordinate because the literal meaning is more striking.

85Ruhrata's Kāvyālankāra 7.39. Also discussed by Ingalls, et al. (1990:168). 1a. grāmataruņam - village youth, taruņyā - young woman, nava - fresh, vañjula - aśoka, mañjarī - bunch of flowers, sanätha - having a protector in, karam - hand, 1b. paśyantyā - having seen, bhavati - is/becomes, muhur - momentarily/repeatedly, nitarām - very much, malinā - dark color, mukhacchāyā - facial complexion. 68

Comments

The boy reminds the woman of their rendezvous in the ashoka grove and shows

he was there by bring a bouquet of ashoka flowers back to town. Upon seeing the

flowers, the girl blushes deeply, ashamed that she did not keep the appointment.

Mammata claims that one is tempted to dwell on the description of the flowers in

the hand of the youth and the girl's reaction, rather than on the suggested sense.

Because of the post-Anandavardhana focus on suggestion, this type of poem is

excluded from the highest category.

śabdacitram vācyacitram avyangyam tv avaram smrtam | (145)

citram iti guņālamkārayuktum l avyangyam iti sphutapratīyamānārtharahitam I

avaram adhamam l yathā

svacchandocchaladacchakacchakuharacchātetarāmbhucchatā mūrcchanmohamaharsiharsavihitasnānāhnikāhnāya vaḥ | bhidyād udyad udāradarduradarī dīrghādaridradruma drohodrekamahormimeduramadā mandākinī mandatām II

vinirgatam mānadamātmamandirādbhavatyupaśrutya yadrcchayāpi yam I sasambhramendradrutapātitārgalā nimīlitāksīva bhiyāmarāvatī II

5cd. Poetry based on semantic or verbal display is non-suggestive and is

known as the lowest type poetry.

69

Brilliant means possessing excellences and poetic ornaments. Non-suggestive

means having no striking suggested meaning. Lowest means inferior. For

example:

May your doubts dissolve into the gentle Ganges, Whose infinite waters absolve the great seers and daily wash away their ignorance, Whose sparkling clarity cuts through the fields, offering a home to boisterous bull-frogs, Whose pride flows out in mighty waves Pushed up by the falling of towering trees. May your doubts dissolve into these slow-flowing waters! (4)86

(Mammata's second example:)

Having heard that Destroyer of Pride is about, Even if just for a stroll, Indra, flustered, quickly bolts the gates, And the city of the gods seems to shut its eyes From fear. (5)87

o This poem is in śārdūlavikrīditam metre (m sj s ttg 12+7). Much more than Mammata's other examples, this seems to be an exercise in virtuosity. Many of the words are chosen largely for their sounds. This leaves the translator with the choice of trying to capture either the meaning or the play of sounds. I have chosen the meaning, and tried to give some indication of the auditiory effect in the comments. That said, this poem is more loosely translated than most. 1a. svacchanda - acting at pleasure (echo of svaccha - beautiful/transparent), ucchalad - shining/sparkling, accha - clear, kaccha - bank, kuhara - crevasse, chätetara - the other having been cut (chāta + itara), ambhu - water, chatā - lump/mass 1b. mūrcchan - dissolved, moha - spiritual ignorance, maharși - great sage, harșa - joy, vihita - accomplish, snāna - bathing/ablution, āhnika - daily rite, āhnāya - at once, vaḥ - your. 2a. bhidyād - may it destroy/brake/cleave or dissolve/untangle, udyad - rising/jumping (ud+Vi), udāra - excellent/loud, dardura - frog (I am aware that the bull-frog is native to North America), darī - cave, dīrghā - long, daridra - roving/mendicant, druma - tree. 2b. droha - injury, udreka - to grow high, mahormi - great waves, medura - thick/dense/great, madā - rapture/pride/intoxication, mandākinī - the slow streaming one, i.e., the Ganges, mandatām - indolence/ignorance/doubts. 87 Found in Mentha's Hayagrīvavadha according to Dwivedi. 1a. vinirgatam - gone out, mānadam - destroyer of pride - the demon king Hayagrīva, ātma - his, mandirād - palace. 1b. bhavati - is, upaśrutya - to be heard, yadrccha - spontaneous, accidental, yā - going, api - even, yam - him. 70

Comments

Mammata's first example has a difficult meter and a pronounced technical brilliance. Part

of its beauty comes from the remarkable alliterations. The first half-line uses "cch" (a

fairly rare group) six times. Likewise with "d-r" and "m-d" later on. There are also

myriad larger repetitions (moha/maha, harsi/harsa, darī/dari, mandā/manda, etc.). The

compounds are both long and complex. Finally, there is a striking parallelism between

the two lines, which both start with descriptions, move to the holy water, and finish with

the removal of doubts.

The second poem is built around the city walls - face simile. This

personification is decorated by a sustained juxtaposition of nasals and dentals.

It is striking that from my perspective, these poems are among the most

beautiful in the chapter. While these poems were greatly admired before the

theory of Anandavardhana, the fact that neither of these poems has a striking

suggestion mares them. One of the great merits of Mammata's synthesis is that

he can acknowledge the importance of suggestion and also save poems like these.

I would like to imagine that Mammata was reacting to a larger awareness among

2a. sasambhrama - quickly/angered, Indra - king of the gods, druta - without delay, pātita - made to fall, argalā - wooden bolt. 2b. nimīlita - having closed eyes, ākși - eyes, iva - like, bhiyā - fear, āmarāvatī - the city of the gods. 88 See The Kāvyamīmāmsā for an example of the importance of citra kāvya before Ānandavardhana (Parashar translation pg. 155). 71

pundits of the danger of prescriptive definitions. "In every man sleeps a zealot,

and when he awakes there is a bit more evil in the world."89

89 Cioran (1949:8), my translation. 72

atha dvitīya ullāsaḥ

iti kāvyaprakāśe śabdārthasvarūpanirņayo nāma dvitīyaullāsah I

Chapter Two

Investigation of the Nature of Words and Meanings

krameņa śabdārthayoh svarūpam āha l

syād vācako lākșanikaḥ śabdo 'tra vyañjakas tridhā | (160)

atreti kāvye l esām svarūpam vaksyate |

(2.1. Types of Words and Meanings Introduced.)

The nature of words and meanings is now explained in order.

6ab. Here words can be considered to be of three types: literal designators,

metaphoric designators, and suggestive designators.

Here means "in poetry." These types will be explained (in 7cd ff).

73

Comments

Mammata is not implying that the language of poetry requires a unique philosophy. He is

merely acknowledging that there are certain types of literature where suggestion is not

often employed (law texts, for example).

vācyādayas tadarthāḥ syuḥ I (166)

vācyalaksyavyangyāh I

tātparyārtho 'pi keșucit | (166)

ākānkșāyogyatāsamnidhivaśād vaksyamānasvarupānām padārthānām samanvaye

tātparyārtho viśesavapur apadārtho 'pi vākyārthah samullasatīty abhihitānvayavādinām

matam I vācya eva vākyārtha ity anvitābhidhānavādinaḥ II

6c. Literal meaning, etc., are their (respective types of) meanings.

(Here is meant) literal meaning, metaphoric meaning, and suggested meaning.

6d. According to some, there is also the speaker's intended meaning.

Those who hold the "relation-of-the-designated" theory (abhihitānvaya-vādin) claim that

the speaker's intended meaning, which is the meaning of the sentence and not of the

individual words, appears when there is a coming together of the word meanings (which

are going to be explained) according to syntactic expectation, semantic fittingness, and

pronunciation rules.

74

Those who hold the "designation-of-the-already-related" theory (anvitābhidhāna-

vādin) (on the other hand) claim that the literal meaning is just the sentence meaning.

Comments

All of mainstream Sanskrit philosophy of language assumes that words and meanings are

distinct. There is debate over what type of entity a meaning is, how each word is

connected to its respective meaning, both within a sentence and without, and how many

types of connection are possible. Mammata explains the three semantic capacities in this

and the next chapter. With regard to sentence meaning, he merely indicates that among

the philosophers that use this concept there are two opposing camps.

Both camps accept the premise that whole sentences convey more than their

individual words. Mammata calls this the intended meaning, others call it the sentence

meaning. Both schools are also committed to an ontology containing both particulars and

relations (among other types of entities). The first school claims that the words denote

only their corresponding ontologically independent entities. That is to say, each word is

saturated (in the Fregean sense). In a sentence, they claim, a new element is introduced

among the words. This element is a complex relation, the elements of which are listed in

the text. The sentence meaning is a complex fact, which is grasped from both the words

and the relation between the words.

The second school claims that the sentence is the basic linguistic unit - the

utterance of a single word cannot be meaningful (unless other, elided, words can be filled

75

in by the hearer). The individual words of a sentence do denote entities, but only as

existing in some possible relation. A dictionary entry must give not only the entity, but

also the types of relation the entity can enter into (i.e., words are divided into parts of

speech). Sentence meaning thus arises directly from the denotations of the words, as

combined in a fact. This view is committed to the claim that all words are unsaturated,

i.e., the meaning of the each word can change from sentence to sentence."

Mammata does not attempt to resolve the debate, which continues throughout

Sanskrit philosophy." Indeed, for the rest of the text he makes no use of this philosophic

tool (sentence meaning). We will see the ramifications of this in connection with his

explanation of meaning in chapter four.

sarvesām prāyaśo 'rthānām vyañjakatvam apīșyate I (176)

tatra vācyasya yathā

māe gharovaaranam ajja hu natthitti sāhiam tumae tā bhaņa karanijjam eme a ņa vāsaro thāi I192

90 Because Sanskrit nouns are declined, word order is relatively unimportant in Sanskrit. Thus many of the obvious English-based objections to this theory were not considered (i.e., in Sanskrit you do not need word order to know if John kissed Mary or vice versa). 91 I wanted to use Siderits's translation of the two theory names (1991:35 note 21). He calls the Bhātta position (abhihitānvayavāda) the "word-plus-relation theory" and the Prābhākara (anvitābhidhānavāda) the "related designation theory". In the context of his explanation this condensed translation is useful, but in isolation I feel it is quite confusing. Kunjunni Raja (1969) chapter 5 is a good place to start for anyone interested in the issue. Gopinath Bhattacharya's explanation in his commentary on the Tarkasamgraha- Dīpikā (p. 300ff) is also good. 92 All of the following translations from Prakrit into Sanskrit, unless stated otherwise, are from the Bālacittānurañjanī of Sarasvatītītha, which is included in Mohan. 1a. māe = mātar - o mother, gharovaaraņam = grhopakaranam - household provisions, ajja hu = atha khalu - now indeed, ņatthitti = nāsti + iti - is not, sāhiam = kathitam - discourse, tumae = tvayā - with/by you 76

atra svair avihārārthinīti vyajyate I

laksyasya yathā

sāhentī sahi suhaam kaņe kaņe dūmmiāsi majña kae l sabbhāvanehakaraņijjasarisaam dāva viraiam tumae I193

atra matpriyam ramayantyā tvayā śatrutvam ācaritam iti laksyam I tena ca

kāmukavisayam sāparādhatvaprakāśanam vyangyam

vyangyasya yathā

ua niccalaninpamdā bhisinīpattammi rehai balāā ņimmalamaragaanapariththiā samkhasutti vva 1194

atra nispandatvena āśvastatvam I tena ca janarahitatvam I atah samketasthānam etad iti

kayācit kimcit pratyucyate l athavā mithyā vadasi na tvam atrāgato 'bhūr iti vyajyate II

7ab. We accept that all the types of meaning are sometimes also suggestive.

There can be suggestiveness in literal meaning, as in:

Mother, you said that we are out of supplies.

1b. tā = tasmāt - thus, bhaņa = bhana - say, kim = kim - what, karanijjam = karanīyam - is to be done, eme a = evameva - in this way, ņa = na - not, vāsaro = vāsarah - day, țhāi = tisthati- stay. (Mohan:179) 95 1a. sāhentī = sādhayantī - win over/seduce, sahi = sakhi - confidante, suhaam = subhagam - a beloved, kaņe kaņe = kșaņe ksaņe - every instant, dūmmiāsi = dūnā asi are suffering pain, majña = mama - of me/mine, kae = krte - when doing, 1b. sabbhāvanehakaraņijjasarisaam = sadbhāva - true/good/faithful + sneha - affection + karaņīya - to be done/made/effected + sadrśam - resembling/having the appearance, dāva = tāvat - certainly, viraiam = viracitam - is performed, tumae = tvayā - with/by you. 94 Gāthāsaptaśatī Weber 4 (1881:4), Patwardhan 3 (1980:2). 1a. ua = paśya - see!/look! ņiccalaņiņpaņdā = niścalanișpandām - undisturbed and motionless, bhisiņīpattammi = bisinīpatre - on the lotus leaf, rehai = rajate - whitish silver color, balāā = balākā - crane. 1b. ņimmalamaragaaņaparițhțhiā = nirmala - flawless + marakata - emerald + bhājana - dish + pratisthitā - resting, samkhasutti vva = śankhaśuktir iva - like a conch-shell pearl, which is, of course, mother of pearl. However, as there are traditionally eight sources of pearls, the conch shell being one, I feel it was chosen in this case for stylistic, and not semantic, reasons. 77

Tell me what to do. The day will not stay light for long. (6)

Here it is suggested that the girl desires to go out for reasons of her own.

There can be suggestiveness in metaphoric meaning, as in the following verse

spoken by a woman to her female friend who is supposed to be acting as her go-between:

My dear friend! For my sake You've gone Time after time To seduce my beloved. Your efforts do seem to spring From honest affection. (7)95

The metaphoric meaning here is, "By giving pleasure to my beloved, you have acted like

an enemy." And by means of this (metaphoric meaning) the suggested meaning is that the

guiltiness of the lovers is known.

There can be suggestiveness in suggested meaning, as in:

Look! A silvery white crane Motionless and calm upon a lotus petal Like a pearl Poised on a flawless emerald platter. (8)

By the word motionless is suggested a state of confidence (in the crane), and by this is

suggested an absence of people. Therefore, the speaker of the poem is pointing out to her

lover, "This is a place for a rendezvous." Alternatively, she is suggesting, "You lie! You

did not come here (for our rendezvous)."

95 The second half of the introductory sentence communicates what a Sanskrit reader would understand from the gender of the words in the poem and from the word "sakhi," which means "friend" but also "confidante" and "romantic go-between." 78

Comments

The first poem presents a girl old enough to go out alone but young enough to be still

living with her mother (and thus unmarried). She reminds her mother that they are short

of some household supply (notice the past tense "said"). That she does so as evening

approaches suggests that she wants to go out for her own (i.e., romantic) purposes.

In the second poem the word "friend" and the words "seduce my beloved" create

a blocking of literal meaning, triggering the reader to search for an alternate meaning."0

One realizes that the word "friend" is being used ambiguously, if not outright ironically.

But the reader is pushed to ask why the speaker would use such irony. We conclude

(from the use of seduce in place of something like "help me seduce") that the speaker

must be aware of illicit relations between her erstwhile friend and her lover.

The third poem has no blocking of literal sense, it can be read as a simple

description of a nature scene. The crane's calm suggests, to an attentive reader, that there

are no people around." Given the context (that the speaker is a young girl speaking to

paramour in the presence of others), one can see that the speaker is suggesting the

absence of people would be meaningful to them (for one of the reason given above). The

ambiguity of the second level suggestion adds to the poem's charm, for both possible

suggestions evoke the erotic rasa.

96 Mammata explains how metaphoric indication works later in this chapter. 97 How it effects this suggestion is left unanswered. Some argue that all suggestion can be explained as an inference, others claim either that this cannot be done or that it is, in fact, not done in actual cases. 79

The poems reveal a problem that follows us though the next four chapters,

namely, how much context and imagination are needed to understand the verses that are

given as examples? All these verses originally were part of larger units (sometimes epics,

sometimes plays, etc.), and many of them, like the poem at hand, cannot be fully

understood out of context. The question I would like to pose is: how did Mammata intend

his readers to understand to his examples? There are several possible answers. First, he

believed that the larger context, despite what I have just claimed, is not necessary to

understand the poems. Second, he expected his readers to know the larger poems well

enough to provide the necessary context for themselves. Third, he expected his readers to

read a commentary that explained the context.

Clearly the first solution would be ideal, and Mammata's text seems to imply that

he held this view. Unfortunately, in the poem above and in countless others, I firmly

believe it is impossible to get the full meaning from the verse alone. For example, how

would you know that two lovers are involved? Or that they need to communicate

secretly? Worst of all, even Mammata himself is not sure of what the second suggestion

is! He claims it could be a reproach or an invitation. But could it not equally well be a

sweet reminder of a past rendezvous? Or an accusation that the lover has been there with

another girl? The possibilities seem endless.

When considering the second and third solutions, we should note that they are not

mutually exclusive. The second solution would apply to the ideal reader (and the pundits

are know for their elephantine memories), while the third solution would apply to the less

80

ideal reader (like myself). If a similar text in English were illustrated entirely with

Shakespeare quotes, an educated reader would not need the context explained. A fact that

mitigates against this view is that the original locations of most of the poems are not

identified in the commentaries, although usually a minimal context is given.98

With regard to the suggestions in the present poem, the situation is further

complicated by the fact that a character is speaking to another character, and not (only)

the poet to the audience (reader or listener). The poet has presented a scene in which a

woman suggests certain things to a man that we readers, as eavesdroppers, as it were, are

supposed to understand. Our case is thus quite different from that of the man being

spoken to, who is obviously familiar with the context (himself, the woman, and their

history together). The audience has to be brought into this fantasy by some means, for the

understanding of a whole scene is often integral to understanding the verbal power of

suggestion.

We shall return to these issues in the third chapter, and also in the fourth and fifth

chapters.

vācakādīnām krameņa svarūpam āha |

sākșāt samketitam yo 'rtham abhidhatte sa vācakaḥ | (198)

ihāgrhītasamketasya śabdasyārthapratīter abhāvāt samketasahāya eva śabdo

'rthavişeśam pratipādayatīti yasya yatrāvyavadhānena samketo grhyate sa tasya

vācakaḥ II

98 One wonders if the context is always given correctly, or if it is invented in the case of some poems. 81

(2.2.) The author explains the nature of literal designators, etc., taking them up in order:

7cd. A literal designator is a word that directly denotes a conventional meaning.

Since there is no knowledge of the meaning of a word whose convention is not grasped, a

word causes one to understand a particular meaning only in so far as it is joined with a

convention of use. When the convention of a word is grasped in relation to a meaning

without an intermediary, the word is the literal designator of the meaning.

Comments

Mammata uses a biconditional to show that the conventional meaning of a word is its

basic meaning: you understand a word if and only if you understand the convention(s)

governing its use. This view is traditionally associated with the Naiyāyikas, who argued

for it early on." The Grammarians and the Mimamsakas, on the other hand, both argue

for a natural or eternal connection between words and their senses.The Mīmāmsakas,

however, do recognize the role of convention in limiting the nature ability of a signifier to

a certain meaning. It is striking that Mammata, who normally sides with the

Grammarians and the Bhatta Mīmāmsakas, takes the Nyāya position here.

99 Nyāyasūtra 2.1.5. 100 See The Mīmāmsāsūtra 1.1.5. Also see Bhartrhari 1.23. 82

samketitaś caturbhedo jātyādir jātir eva vā l (208)

yady apy arthakriyākāritayā pravrttinivrttiyogyā vyaktir eva tathāpy ānantyād

vyabhicārāc ca tatra samketah kartum na yujyate iti gauh śuklaś calo dittha

ityādīnām visayavibhāgo na prāpnotīti ca l tadupādhāv eva samketah II

upādhiś ca dvividhah I vastudharmo vaktryadrcchāsamniveśitaś ca l vastudharmo 'pi

dvividhaḥ l siddhah sādhyas ca l siddho 'pi dvividhah I padārthasya prānaprado

viśesādhānahetuś ca l tatrādyo jātih uktam hi vākyapadīye na hi gauh svarūpeņa gaur

nāpy agauh gotvābhisambandhāt tu gauh iti l dvitīyo guņah I śuklādinā hi

labdhasattākam vastu viśisyate I sādhyah purvāparībhūtāvayavah kriyārūpah

ditthādiśabdānām antyabuddhinirgrāhyam samhrtakramam svarūpam vaktrā yadrcchayā

ditthādisv artheșūpādhitvena samniveśyata iti so 'yam samjñārūpo yadrcchātmaka iti I

gauh śuklaś calo dittha ityādau catustayī śabdānām pravrttiḥ iti mahābhāsyakāraḥ I

paramānvādīnām tu guņamadhyapāthāt pāribhāsikam guņatvam II

8ab. The conventionally designated is either fourfold consisting of natural kinds,

etc., or it consists of universals only.101

(Arguments for the first view:) Even though it is fitting to act or avoid action with regard

to particulars (and not universals) because particulars have causal efficiency, still, the

meaning convention cannot relate a word directly to individuals because of (the faults of)

infinite application (if the word is related to the entire class of individuals) or

101 I translate jāti first by "natural kind" then by "universal" because I believe Mammata is drawing from two different traditions, each of which use the word differently. Failure to recognize this makes this section impossible to understand. See the comments below. 83

misapplication (if it is based on a subclass). Furthermore, the individual entity does not

serve to differentiate (various aspects of) the cognitive object in a (complex) cognition

like, "This is a cow, white, walking, Dittha." Therefore, a linguistic convention connects

only to a characteristic of the individual (and not the individual itself).

Characteristics are twofold: (1, 2, 3) properties of the object and (4) those

imposed (on the object) by the will of the speaker. The properties of the object are also

twofold: (1, 2) the established properties and (3) the properties that are to be established.

The established properties are also twofold: (1) those which determine a kind (as

different from other kinds) and (2) those which impose differentiation (on members of a

single kind).

(1) The first of these are natural kinds. In support of this, there is a saying from

the Vākyapadīya: "A cow is a cow by its connection to cowhood. It is neither a cow nor a

non-cow by virtue of its own nature."102

(2) A second type of property is the quality (guna), because an object whose

natural kind is already ascertained is differentiated by qualities like white, etc., (from

others of the same kind).

(3) Properties that are to be established each take the form of an action or process,

whose parts are the proceeding and the subsequent moments.103

102 I have not been able to find this reference. 103 Processes are not often thought of as properties in Western ontology. When we say a man is walking, we mean that he has been walking and that we expect he will continue to do so for at least another moment. He is thus in the process of establishing the property "walked a certain distance or time." 84

(4) A word like "Dittha," etc., that by nature is assembled letter by letter and

grasped mentally at the end, is attached to an object, Dittha, in this case, by the will of the

language user (vaktr) as a characteristic (i.e., "being named Dittha"). Such a word is

called a proper name and has its essence in the voluntariness (of conventions made by

language users).

(To conclude the presentation of this view I quote) the author of the Mahābhāsya,

"The basis for the application of words is fourfold, as in a complex cognition like, "This

is a cow, white, walking, Dittha."104

Atoms, etc., are technically defined as qualities because they are listed among the

qualities (by the early realist philosophers).105

Comments

As I have shown in the introduction, Mammata is heavily indebted to both the

Grammarian and the Mīmāmsa traditions. As the quotes show, the view presented here is

from the Grammarians. The Mīmāmsa view is presented in the next section. Mammata

makes no attempt to determine which is correct, although he does present a couple

strands of argument.

104 I have not been able to find this reference. I follow Renou in translating pravrttih as "basis for the application" (1957:2:36). 105 According to Gajendragadkar (1939:175) Atoms means atomic dimension and etc. means qualities like all-pervadingness. Apparently these were controversially listed among the qualities in the oldest Vaiśesika ontology. See his discussion for the details of this debate. Mammata's point is that we do not need a fifth type of characteristic to account for them. 85

The first claim in the Grammarian argument is that only individuals enter into the

causal web of the world. A stock illustration claims that it is individual cows and not

cowness that one milks. The second claim is that the individuals cannot ground linguistic

usage. For one must either take the entire class of individuals or some subclass thereof as

the meaning of a particular word. If the first option is chosen, the meaning will never be

certain because such a class is, in principle, infinite (or at least without any known limit).

If the second option is chosen, further instances of the type not included in the defining

group will be, by the definition, not examples of the type. This would result is a

misapplication of the word as restricted to only the original group, whereas clearly new

tokens should be designated." The conclusion is that the linguistic conventions connect

words to universals as instantiated in particulars. Notice that an understanding of the

general concept of universal is implicit in this argument. Without such a concept, each of

the four cases would have to be argued separately.

The Sanskrit of the kārika at hand is confusing because Mammata uses the same

word, jati, to name one of the subclasses of universals as he uses to name the entire class.

This confusion evaporates when one understand the evolution of this word. In its earliest

uses in grammatical texts, jāti meant "natural kind" or "species." This use is preserved

in the Grammarian school of philosophy much later than in other schools. Most notably

for our purposes, Bhartrhari uses the word this way. There seems to be no word in the

106 Kumārila makes this argument in the Ślokavārtika 5:13:1. 107 See Renou (1957:1:148). 86

early grammarians that refers to universals in general.108 The word used here for an

instantiated universal is a general word for properties, which I translate as

"characteristic" (and which the Sanketah commentary glosses: tadupādhau iti

vyaktiviśesene)." In this section, the term "characteristic" is used to explain the

ontological status of word meanings. Each type of word (common noun, adjective, verb,

proper noun) is explained with reference to a subclass of characteristics.

The second view Mammata refers to is the Bhatta Mīmāmsaka view, and jāti here

refers to the general ontological category "universal." Mammata now proceeds to

explain this view.

guņākriyāyadrcchānām vastuta ekarūpāņām apy āśrayabhedād bheda iva laksyate I

yathaikasya mukhasya khangamukuratailādyālambanabhedāt l himapayah

śankādyāśrayesu paramārthato bhinnesu śuklādisu yadvaśena śuklaḥ śukla

ityādyābhinnābhidhānapratyayotpattis tat śuklatvādisāmānyam | gudatandulādipākādișv

evam eva pākatvādi l bālavrddhśukādyudīritesu ditthādiśabdesu ca pratiksanam

bhidyamānesu ditthādyarthesu vā ditthatvādy astīti sarvesām śabdānām jātir eva

pravṛttinimittam ity anye l tadvān apoho va śabdārthah kaiścid ukta iti

granthagauravabhayāt prakrtānupayogāc ca na darśitam II (234)

108 This should not be surprising. The grammarians worked to explain and distinguish nouns in general, verbs in general, adjectives in general, etc. They did not go beyond these to the more general concept of universals, which belongs more to ontology than grammar. See Renou (1957:1:78, 109, and 148). 109 See Renou (1957:1:109). 87

(Arguments for the second view:) Others hold that universals alone form the bases

of all words.1Just as a single face appears to be several when reflected in a sword, a

mirror, and a pool of oil, so each quality, action, and proper name has a single essence in

reality, even though they appear various due to the different substances (in which they are

instantiated). A universal, whiteness, for example, is that on account of which there are

identical expressions and thoughts, "(This) is white (and that) is white", for example, in

really distinct white substances: snow, milk, shells, etc. Likewise, there is a universal,

cookinghood, for example, in all the various distinct instances of cooking: cooking

jaggery, cooking rice, etc. Finally, there is a (word) universal, "Dittha"hood, for example,

(exemplified) in all the distinct utterances of the word "Dittha" by a child, by an old man,

by a parrot, etc., and in Dittha himself, who is changing from moment to moment. Thus

universals alone form the bases of all words.

Some (the Realists) hold that the substances which instantiate universals act as the

bases of all words, while others (the Buddhists) hold that (instances of) exclusion do so.

These views are not examined because they are not relevant and would, I fear, swell the

work (out of proportion).

110 This sentence is found at the end of the section in the Sanskrit (see translational note #6). 88

Comments

The Bhātta Mīmamsaka claims that universal are the literal meanings of words. As all of

Bhartrhari's "characteristics" can be considered universals, the difference between the

Grammarian and Mīmamsaka views is smaller than it first seems. The Grammarian view

places more emphasis on the four types of universals, the Bhatta view on the ontological

claim that all four types are subclasses of a single ontological category.

As we saw in the proceeding section, the personal identity of Dittha seems to

depend on the property "being called Dittha," which remains identical throughout

Dittha's life. Whether this claim concerning person identity is only supposed to explain

verbal usage, or whether it is an ontological claim, is not specified. In view of the

religious convictions that Mammata appears to hold, I would incline to the former

interpretation.

Mammata points out that there are other views concerning the ontological status

of the meaning element. I think he is right to claim that establishing which view is

correct is not important for his project. As long as all the schools active at this time agree

that there is such a thing (even the Buddhist exclusion class is an entity of sorts),

Mammata can move on to the semantic powers characteristic of poetry.

sa mukhyo 'rthas tatra mukhyo vyāpāro 'syābhidhocyate II (243)

sa iti sāksāt samketitaḥ I asyeti śabdasya II

89

8cd. That is the primary meaning, and with respect to it, the primary function of

this (a word) is called denotation.

That refers to directly denoted conventional meaning. Of this refers to a word.

Comments

All the philosophical schools admit there is such a thing as conventional meaning (with

the possible exception of Cārvāka).1Mammata's definition is thus vague enough to

avoid the dispute about the ontological status of meanings. What distinguishes poetry

from other uses of language lies in other semantic powers than denotation. To these he

now proceeds.

mukhyārthabādhe tadyoge rūdhito 'tha prayojanāt I

anyo 'rtho laksyate yat sā lakşaņāropitā kriyā Il (246)

karmaņi kuśalah ityādau darbhagrahanādyayogāt gangāyām ghosam ityādau ca

gangādīnām ghosādyādhāratvāsambhāvāt mukhyārthāsya bādhe vivecakatvādau

sāmīpye ca sambandhe rūdhitaḥ prasiddhe tathā gangātate ghosah ityādeh prayogāt

yesām na tathā pratipattih tesām pāvanatvādīnām dharmāņām tathā pratipādanātmanah

prayojanāt ca mukhyena amūkhyo 'rtho laksyate yat sa āropitah śabdavyāpāraḥ

sāntarārthanistho lakșanā II

111 Kunjunni Raja (1969:24). 90

(2.3. Definition of Metaphoric Indication.)

9. Metaphoric Indication is a superimposed semantic capacity of words (that occurs)

when the primary meaning is blocked and another meaning is metaphorically

indicated which is connected to the primary meaning either by a (second-order

linguistic) convention or by a special purpose.112

A non-literal meaning is metaphorically indicated by the literal meaning (when two

conditions are satisfied, the first of which is:) when the literal meaning is blocked. (This

can happen) either (1a) because of a (semantic) non-connection, as in expressions like

"karmaņi kuśalah" ("a discerning workman," originally, "a kuśa-grass worker") where

there is non-connection with gathering kuśa grass, or (1b) because of (semantic)

impossibility, as in expressions like, "The village is on the Ganges," where there is the

impossibility of a river like the Ganges being the site of a village. And (2) when a

(semantic) connection is established (between the two meanings), either (2a) by a second-

order linguistic convention, such as (the second order conventional meaning)

"discernment" (where the literal sense "sacred grass cutter" came to mean anyone who

works with the carefullness that is required of the kuśa cutter) or (2b) because of a special

purpose, such as (the indicated meaning) "proximity," which aims to bestow (upon the

village) such properties as purity, etc., (which are suggested properties and which) would

not be conveyed by such literal expressions as, "The village is on the bank of the

112 Kriyā is glossed as śakti in the Dīpikā. I follow this and translate it "semantic capacity" or "semantic power." Having semantic capacity means having the ability to convey meaning. Indicative words are said to be capable of conveying additional meaning (beyond the denoted meaning). Semantic capacities are that by which a word conveys meaning. Words can do this in three ways according to Mammata. Thus there are three semantic capacities or powers. 91

Ganges." A metaphor is a superimposed function of a word that is grounded an another

meaning (i.e., the literal).

Comments

The two examples point to two different linguistic phenomena that are covered by the

term "metaphoric indication." The first concerns the secondary or parasitic meanings of a

word that arise due to a semantic connection and/or addition. These become fixed by

convention and are available to all language users. Often they come in the form of faded

metaphors (nirūdhalaksaņā). Thus in English the word "maverick" first meant the cow of

a certain rancher named Maverick who did not brand his cows. It took on the meaning of

any cow that could not be identified and eventually any member of a group that does not

fit. Mammata would say that there are three conventions for the use of this word, and that

the later two are parasitic on the primary one. There is much discussion in the Sanskrit

tradition of whether metaphoric meaning can eventually become literal meanings (as with

"maverick" and also, apparently, in the case that Mammata discusses here)."5 This would

not be a large step for Mammata in that he holds that they are all equally conventional.

The winning argument appears to be that the "becoming literal" view can explain the

correct use of such words by people who are not aware that they are metaphors.

The second linguistic phenomenon we see here is closer to true metaphor. Here

the secondary use of a word is specific to the occasion of use and is created for a specific

13 One of the leading critics after Mammata, Viśvanātha, claims that they can become literal (see Kunjunni Raja 1969:264).

92

purpose. Thus when Shakespeare has Macbeth say, "Life is but a walking shadow ... " he

does so for a specific purpose (to suggest the fleeting nature of life, etc.). In this case,

there is no secondary linguistic convention that comes to include life as a meaning of

"shadow."

It is interesting to note that the specific purpose is said to be always a suggestion

(see kārikā 13, below). This had led some to claim that metaphor is always in the service

of suggestion. However, the first type Mammata cites seems to contradict this, as do part-

whole metaphors, which are often used simply for convenience (it is easier to say the

class is bored than to say that most of the students in the class are bored).

It is regrettable that Mammata does not tell us more about how we arrive at the

correct secondary meaning. It is clear from the tradition that the ability to do so is both a

result and a sign of good training. The training would thus have to teach both vast lexical

knowledge and subtle appreciation of literary devices and intentions. This linguistic

training would certainly compose part of the training described in chapter 1, kārikā 3.

svasiddhaye parākşepah parārtham svasamarpaņam I

upādānam laksaņam cety uktā śuddhaiva sā dvidhā II (264)

kuntāh praviśanti yastayah praviśanti ityādau kuntādibhir ātmanah praveśasiddhyartham

svasamyoginah purusā āksipyante I tata upādāneneyam laksaņā II

gaur anubandhyah ityādau śruticóditam anubandhanam katham me syād iti jātyā vyaktir

āksipyate na tu śabdenocyate viśesyam nābhidhā gacchet ksīņaśaktir viśesane iti nyāyād

93

ity upādānalaksanā tu nodāhartavyā l na hy ātra prayojanam asti na vā rūdhir iyam

vyaktyavinābhāvitvāt tu jātyā vyaktir āksipyate I yathā kriyatām ity atra kartā l kurv ity

atra karma I praviśa pindīm ityādau grham bhaksayetyādi ca l pīno devadatto divā na

bhunkte ity atra ca rātribhojanam na laksyate śrutārthāpatter arthāpatter vā tasya

vişayatvāt II

gangāyām ghosah ity atra tatasya ghosādhikaranatvasiddhiye gangāśabdah svārtham

apayati ity evamādau laksaņenaisā laksaņā II

ubhayrūpā ceyam śuddhā upacāreņāmiśritatvāt l anayor laksyasya laksakasya ca na

bhedarūpam tātasthyam I tatādīnām gangādisabdaih pratipādane tattvapratipattau hi

pratipipādayisitaprayojanasampratyayah I gangāsambandhamātrapratītau tu gangātate

ghoșa iti mukhyaśabdābhidhānāl laksanāyāh ko bhedah II

(2.4. Two Types of Pure Metaphors: Additive and Commutative.)

10. Pure metaphors are said to be twofold: the additive, which implies another

meaning along with the literal meaning, and the commutative (laksanam), which

surrenders the literal meaning for the other (the metaphoric meaning).

In an expressions like, "The lances enter," or, "The staves enter," in order to make the

entering of the lances possible, the men connected to them must be implied by the word

"lances," and likewise for "staves," etc. Therefore, this is a metaphor by means of

addition (of another sense, i.e., both the spears and the men bearing them are meant).114

114 Strictly speaking, of course, these are examples of metonymy, not metaphor. 94

In an expression like, "The ox should be sacrificed," the individual ox is not

mentioned by the word ("ox") according to the dictum: "The denotation (of a word) will

not go to the qualified (the individual) when it has already gone to the qualifier (the

universal)." (Some might claim that) the individual is implied by the universal, (because

otherwise it might be asked), "How is the sacrifice enjoined by the Veda possible for

me?" But this should not be cited as an additive metaphor, for here there is neither

convention nor a special purpose. Rather, the individual is implied by the universal

because of its invariable connection with the individual. Likewise in, "Do it!" an agent is

implied (because there is never an action without an agent). In the case of, "Do!" some

object is implied, (because there is never a transitive verb without an object). Likewise,

"Enter!" implies a building (to enter into) and "Cake!" implies that one should eat.115

With the expression, "Fat Devadatta does not eat during the day," nocturnal

feeding is not indicated metaphorically, because the sentence comes to have that meaning

by either verbal presumption or (unqualified) presumption.116

With the expression, "The village is on the Ganges," the word "Ganges" gives up

its literal meaning in order to establish that the village is on its bank. Expressions such as

this are metaphors by commutation.

115 In both cases, we might also cite the reason that a transitive verb and a direct object (an object in the accusative case) are syntactically dependant and thus invariably concomitant. 116 Kumārila Bhātta asserted the theory of verbal presumption (śrutārthāpatti) to explain this type of reasoning, while Prābhākara assert unqualified presumption (ārthāpatti). Other classical philosophers see all such presumption as a form of inference. Mammata apparently is not concerned to establish which view is right, but only to disregard this as a possible case of metaphoric indication. 95

These two types of metaphor are called "pure" because they do not involve the

assistance (of resemblance between the indicating (or literal) meaning and the indicated

(or metaphoric) meaning). However, the division between the indicated meaning and the

indicating meaning is not a complete separation. For when "the bank of the Ganges" is

understood by the word "Ganges," the correct interpretation is that a special purpose is

meant to be communicated. If only an external relation to the Ganges were to be

apprehended, "The village is on the bank of the Ganges," is what would have properly

been said. Why use metaphor if denotation communicates the same information?

Comments

The basic division of metaphoric indication is pure metaphor versus qualitative metaphor.

The latter type is based on similarity of meaning of the two terms, or more specifically on

their sharing some quality. The former is based on some other relation between the

meanings of the terms (e.g., cause - effect or part - whole). While early Mīmāmsakas

thought these were two different semantic powers, Mammata combines them and uses the

terminology to differentiate two subclasses of a single semantic power (metaphoric

indication).

Mammata begins his presentation of metaphor considering two types of pure

metaphor. The classification hinges on what part of literal meaning is brought into the

final understanding of the sentence. When one says, "The staves enter," the metaphor

7 Mammata combines several different (overlapping) ways of classifying metaphor into a single presentation. 96

supplies the people carrying the staves. The meaning of "staves" becomes both the staves

and the people carrying them: thus the term "additive." In, "The village is on the

Ganges," the word "Ganges" gives up its meaning "river," and takes on the meaning

"bank of the river." However, the loss of the original meaning is not complete, for certain

qualities inherent in the original object are available for communication by suggestion.118

The indication of a particular ox by the term "ox," whose denotation on the

Mīmāmsa view is the universal "oxhood," does not count as metaphoric indication

because the individual can be logically inferred from the universal concomitance between

the universal and a particular.1Similarly in the case of fat Devadatta, the knowledge is

gained by the means of knowledge known as presumption, which grounds negative

inferences. Fatness is universally non-concomitant with fasting. Although the literal

meaning seems "blocked," in both cases it is unblocked by a little logic. Such is not the

case, however, with, "The cribs cry," because there is no universal concomitance between

the cribs and babies. Mammata returns to this type of case presently (kārikā 12abc).

In the tradition there is a third type: a metaphor in which the literal term (or both

terms) gives up a part of its meaning, but retains another part. The most famous example

is from the Upanisads, "You are that (the Absolute)." The identification is not literally

true because there are differences between a human and Brahman, the Absolute. But, the

text claims, a part of what is referred to by "you" is identical to part of what is referred to

118 For additional details of this view, see Kunjunni Raja (1969:266). 19 No uninstantiated universals are admitted by the Mīmāmsakas. 97

by "that."120 Mammata does not treat the third type and I give a reason why he may have

chosen to ignore it in the comments to kārikā 12d.

sāropanyā tu yatroktau vișayī viayas tathā l (291)

āropayamānah āropavisayaś ca yatrānapahnutabhedau sāmānādhikaranyena nirdiśyete

sā laksaņā sāropā II

vişayyantah krte 'nyasmin sā syāt sādhyavasānikā Il (291)

vişayināropyamānenāntahkrte nigīrņe anyasmin nāropavisaye sati sādhyavasānā syāt l

(2.5. Impositional and Determining Metaphors Defined and Explained.)

11ab. There is another type of metaphor, the impositional, in which both the object

of comparison and the subject of comparison are stated.

When (1) the imposed upon and the imposed object of comparison are expressed with the

same case ending and (2) their difference is not denied, the metaphor is "impositional."

11cd. A determining metaphor occurs when the other (the subject of comparison) is

made interior to the object of comparison.

When the object of comparison swallows the subject of comparison, the metaphor is

"determining."

Comments

120 Appaya Dīkşita takes up this issue at length in the second part of his Vrttivārttika.

98

These types of metaphor become easier to understand with the examples, which are given

below.

bhedāvimau ca sādrśyāt sambandhāntaratas tathā l

gauņau śuddhau ca vijñeyau (300)

imāv āropādhyavasānarūpau sādrśyahetū bhedau gaur vāhīkah ity atra gaur ayam ity

atra ca l atra hi svārthasahacāriņo guņā jādyamāndyādayo laksyamānā api gośabdasya

parārthābhidhāne pravrttinimittatvam upayānti iti kecit l svārthasahacāriguņābhedena

padārthagatā guņā eva laksyante na parārtho 'bhidhiyate ity anye

sādhāraņaguņāśrayatvena parārtha eva laksyate ity apare l uktam cānyatra

abhidheyāvinābhūtapratītir laksaņocyate | laksyamānagunair yogād vrtter istā tu

gauņatā iti l avinābhāvo 'tra sambandhamātram na tu nāntarīyakatvam I tattve hi

mañcāh krośanti ityadau na laksanā syāt l avinābhāve cāksepeņaiva siddher laksaņāyā

nopayoga ity uktam II

āyur ghrtam āyur evedam ityādau ca sādrśyād anyat kāryakāranabhāvādi

sambandhāntaram l evamādau ca kāryakāranabhāvādilaksaņapūrve āropādhyavasāne

atra gaunabhedayor bhede 'pi tādrūpyapratītih sarvathaivābhedāvagamaś ca

prayojanam I śuddhabhedayos tv anyavailaksanyenāvyabhicāreņa ca kāryakāritvādi l1121

kvacit tādarthyād upacārah I yathā indrārthā sthūnā indrah | kvacit svasvāmibhāvāt

yathā rājakīyah purūso rājā l kvacit avayayavāyavibhāvāt l yathā agrahasta

ityatrāgramātre 'vayave hastah I kvacit tādkarmyāt l yathā ataksā takșā II

121 The preceding two sentences, while absent in Mohan, appear in Jha, Gajendragadkar, and Dwivedi. 99

12abc. Furthermore, these two divisions (of metaphor) are known as qualitative

when the metaphor is based on resemblance, and pure when it is based on another

relation.

Examples of the two divisions, impositional and determining metaphors, when the

metaphor is based on resemblance (and thus qualitative) are "The porter is an ox," and

"the ox" (said of a porter).1 (There are three prima facie explanations for this later type

of metaphor.) Some say that even thought the denotation of "ox" is another object (the

porter), qualities associated with the primary meaning, such as stupidity and slowness,

suffice to occasion the usage (of "ox" to mean the porter).123 Others say that only the

qualities of the object are indicated metaphorically, because they are identical with the

qualities associated with the primary meaning (the ox). Furthermore, the object (the

porter) is not denoted. Still others say that only the object (the porter) is indicated

metaphorically, because that is the substratum of the common qualities.

This is also said elsewhere, "A cognition of that which does not exist without the

object denoted is called a metaphor. This process should (be called) qualitative, because

122 Vāhika seems to be an alternate spelling of bālhika or bālhīka (Monier Williams 949). If vāhīka is a variation of the same word, this is a reference (and insult) to the Punjab. Gajendragadkar asserts this reading, but does not state his reasons. Dwivedi translates it as a personal name. Apte (1422) agrees that it is an alternate spelling of bālhīka but adds that it also means an irreligious person. Under bālhīka he adds "ox" as a second definition! I have taken it as a variation of vāhika and translated it "porter." Along similar lines, Jha's translates it "ploughman." The second example cannot be translated, "This (guy) is an ox." To do so would make this another example of impositional metaphor. An example of this expression in a full sentence is, "This ox is taking forever with my bags." 123 All the commentators agree that the porter is the referent of the phrase, "the other meaning". However, the passage makes more sense to me if the qualities of the porter (i.e., the individual instantiations of the universals) are what are referred to. 100

(the indicator - the ox) is connected to the indicated object (the porter) by the (common)

qualities." It has been said that by, "does not exist without," there is meant simply a

connection and not an invariable concomitance. If invariable concomitance were the

meant, no metaphor would occur in cases like, "The cribs cry," (or maybe, "The stands

roar.") Furthermore, if there were invariable concomitance, the indicated meaning would

be established by implication alone, and there would be no need of metaphor.

"Ghee is life," and "this true life" (Said of ghee) are examples (of metaphor)

based on a relation other than similarity - the causality relation.1 In such cases, the

superimposition or determining is based on a relation of the cause and effect relation or

the like.

The purpose (of the metaphor) in the two qualitative types of metaphor (e.g., "The

porter is an ox," and "the ox" used to refer to the porter) is cognition of identity despite

difference in substance (in the first type) and complete identity (in the second type). In

the pure types, on the other hand, the purpose is to show that ghee sustains life in a way

different from other (foodstuffs that also sustain life) or that ghee invariably sustains life.

(Four examples of pure metaphors based on other relations follow.) In some

cases, the metaphor indicates (a relation of) serving another's purpose. For example a

pillar dedicated to Indra is called Indra. In some cases, the metaphor indicates (a relation

124 While all agree that this quote is from Kumarila Batta, Jha (1925:24) locates it in the Ślokavārttika, while Dwivedi correctly gives Tantravārtika 1.4.22. See Tantravārttika (1937:354). 125 As in "the ox", the second example here cannot be translated as, "This (stuff) is life itself." To be an impositional metaphor, idam must be an article qualifying life, not a pronoun standing for ghee. An example of a full sentence containing this expression is, "Pour some of life itself into the fire." Compare the English expression, "He is all out of beans." 101

of) servant and master. For example an agent of the king is called the king's man. In some

cases, the metaphor indicates (a relation of) part to whole. For example, the foremost part

of the hand (i.e., the palm) is called the hand. In some cases, the metaphor indicates (a

relation of) of identity of activity. For example, a person who is not a carpenter (by trade)

is called a carpenter (when working wood).

Comments

There are three views of how the word "ox" can apply to the porter. All assert that the

universal ox-hood is denoted by the word "ox" and that this meaning is blocked by the

non-identity of the ox and the porter. The first view holds that metaphor works in three

stages. After the failed denotation, the qualities of the ox (stupidity and slowness) are

metaphorically indicated. Finally, the similar qualities ground the denotation of the

porter. The second view also asserts that ox-hood is denoted and blocked. Then the

qualities of the porter are metaphorically indicated. Finally, the porter is inferred as the

substratum of the properties. The third view asserts that the blocking of the literal sense,

the individual is metaphorically indicated on the basis of shared qualities.

A possible interpretation of this passage is that Mammata is presenting all the

possible views without taking sides (a common practice of his, as we have seen).

However, I believe that the first two views are "incorrect" (pūrvapaksa) and that for

Mammata, the third view is the "accepted and proven position" (siddhanta). Mammata

uses the Kumārila quote to give authority to his position. On this reading, Mammata is

102

offering a highly condensed summary of the arguments put forward by Kumārila in the

Tantravārttika.

The principle objection against the first view is the stability of meaning. Kumārila

argues that the meaning of a word does not change with the invention of every new

metaphor. Thus the porter cannot be part of the literal meaning of the word "ox" in the

expression, "the ox" (used to refer to the porter).120 A second objection is that denotation

can only function once. If the word "ox" successfully picks out a meaning, it will go no

further. 127

The objection Kumārila brings against the second view is that many qualities are

linked to the referent of the word ("ox" in this case) and only some of them are

metaphorically indicated in the expression, "the ox." There is no way to determine which

qualities are relevant without reference to the other object (the porter).128 Further, the use

of inference in the third step is criticized as cumbersome.129

The third view is the simplest of the three. The word "ox" cannot be understood

literally because the universal "ox-hood" is not instantiated in the porter. Objects that do

instantiate this universal, however, all share certain other qualities (like stupidity and

slowness). The hearer understands the word "ox" to metaphorically mean the porter on

the basis of shared qualities. The porter is thus not added to the literal meaning of the

word, nor is inference needed.

26 Tantravārttika (1937:356). 127 See Śabara on Mīmāmsāsūtra 1.1.25 for the earliest Mīmāmsā presentation of this common Bhațta belief. 128 Tantravārttika (1937:358). 129 Gajendragadkar (1939:202). 103

Kumarila's own view is that the meaning of the word "ox" is fixed to the universal

ox-hood. When the primary meaning of the word is blocked, the hearer searches among

the qualities associated with the primary meaning for qualities shared by the primary

meaning and the object at hand (ox-hood and the porter). The literal meaning of the word

"ox" is then provisionally limited to those qualities. This allows Kumārila to claim that

metaphor adds no semantic content to the word. It is a process by which the total

meaning of a term is provisional limited and applied to another entity for a specific

purpose.130 In the present case, the metaphor servers to ridicule the porter by

exaggerating some of his qualities. Mammata will discuss the special purposes of

metaphor in chapter three.

Before discussing the pure metaphors, I would like to finish with the qualitative

metaphor by explaining Mammata's discussion of their purpose. He claims that the

metaphor is used to cause a cognition of identity, which must be an identity of qualities.

In an impositional metaphor, the difference of substance is still cognized, whereas in a

determinative metaphor, this difference is denied. The identity is between shared

qualities, not all qualities (in calling someone an ox, nobody means that he has hooves,

etc.). In claiming that the porter and the ox are identical in intelligence, one indicates how

exceptionally stupid the porter is by human standards.

Mammata asserts that pure metaphors function by means of a relation other than

resemblance. His list of types of possible relations is not exhaustive. Such list (along with

130 Tantravārttika (1937:354). On the basis of Kumārila's more detail exposition, I feel Mammata is using "denotation" in a more general sense in which a word can "denote" other meaning by means of metaphor. 104

the topic of metaphor generally) appear very early in Indian philosophy.131 However, no

general principles can be formulated for pure metaphors because the relationships are too

diverse. Instead, Mammata shows how one relationship (causality) can ground

metaphoric use. The others could be likewise demonstrated.

Ghee (clarified butter) was considered especially capable of sustaining life. One

can lay stress on its causal capacity by meanings of the metaphor, "Ghee is life." This

metaphor implies that ghee is more able to sustain life than other foodstuffs. One can also

use the word "life" metaphorically to refer to ghee as in, "Pour life into the fire."

Mammata claims that in doing so, one establishes an invariable relation between life and

ghee. Think of the expression, "He gave up the ghost." As the ghost (or spirit) was

thought of as the cause of life, this expression metaphorically means, "He died."

Mammata rounds out this section by explaining how several other relations

ground metaphoric usage. An intelligent reader could imagine motives for the metaphors

that Mammata gives and construct sentences with them.

laksaņā tena sadvidhā II (322)

ādyabhedābhyām saha II

12d. Metaphor is therefore of six kinds.

(The four just described) together with the two earlier types (described in kārikā 10).

131 The Nyāyasūtra 2.2.62 gives ten, the Mīmāmsāsūtra 1.4.23 gives six, etc. Kunjunni Raja has a detailed discussion of these lists. See Kunjunni Raja (1969:233 ff). 105

Comments

The six are: the additive, commutative, pure superimposition, qualitative

superimposition, pure determining, and qualitative determining. Thus, four pure types

and two qualitative types of metaphor. The first two discussed belong to a different

principle of classification than the remaining four. In particular, the commutative and the

qualitative superimposition overlap to some extent. I think Mammata left out the third

type mentioned in the comment to kārikā 10 (the jahadajahallaksanā) because it clearly

overlaps with the qualitative superimpositional type. Mammata turns now to the

suggestive possibilities of metaphor.

sa ca

vyangyena rahitā rūdhau sahitā tu prayojane I (325)

prayojanam hi vyañjanavyāpāragamyam eva II

tac ca gūdham agūdham vā l (325)

tac ceti vyangyam I gūdham yathā

mukham vikasitasmitam vaśitavakrimapreksitam samucchalitavibhramā gatir apāstasamsthā mati I uro mukulitastanam jaghanamamsabandhoddharam batenduvadanātanau taruņimodgamo modate II

agūdham yathā

106

śriparicayāj jadā api bhavanty abhijñā vidagdhacaritānām upadiśati kāminīnām yauvanamada eva lalitāni II

atropadiśatīti II

tad eșā kathitā tridhā Il

avyangyā gūdhavyangyā agūdhavyangyā ca II

(2.6. Purposes of Metaphor Explained.)

Furthermore,

13ab. When based on a second-order linguistic convention, a metaphor is without

suggestive meaning. When based on a special purpose, it has suggested meaning.

For the special purpose is inferred only through the operation of suggestion.

13c. Furthermore, it is either oblique or obvious.

It means suggested meaning. Here is an example of oblique suggested meaning:

O moon-faced one, blooming smile, seductive sidelong glance, mind spinning, breasts budding, and full thighs that excite enticing frolics, In you the fullness of youth bursts forth rejoicing. (9)132

Here is an example of obvious suggested meaning:

By piling up treasure, even the stupid grow clever. So youth's intoxication alone teaches women their wiles. (10)133

132 Attributed to Ruma by Suktimuktāvalī according to Dwivedi. 133 Found in Ravigupta's Subhāsita according to Dwivedi. 107

Here the word "teaches" (carries the suggestion)

13d. Thus, metaphor is declared to be three-fold.

(Namely, metaphor that is) without suggestion, with an oblique suggestion, and with an

obvious suggestion.

Comments

There are many words used metaphorically in the first poem. In each case, the reader has

to stop and think about what is suggested. For example, the word "blooming" in

"blooming smile" suggests freshness, pleasant smell, delicateness, etc. Intoxication

cannot literally teach, so "teach" is used metaphorically to mean "imparts to". The

suggestion is that all alone young women learn to be charming as easily as if they had an

expert instructor.

tadbhūr lākşaņikaḥ I (339)

śabda ity sambadhyate l tadbhūs tadāśrayah II

tatra vyāpāro vyañjanātmakaḥ I (339)

kuta ity āha

yasya pratītim adhātum lakșaņā samupāsyate II

phale śabdaikagamye 'tra vyañjanān nāparā kriyā l (339)

prayojanapratipipādayisayā yatra laksanayā śabdaprayogas tatra nānyatas tatpratītir

api tu tasmād eva śabdāt I na cātra vyañjanād te 'nyo vyāpārah II

108

14a. That which supports the metaphor is the metaphoric (word).

Word is to be connected (to the word metaphoric). That which supports the metaphor

means that which is the basis of the metaphor.

(2.7. Suggestion Based on a Word.)

(2.7.1. Suggestion Based on a Metaphoric Word.)

14b. There (concerning the metaphoric word) the function (that achieves the special

purpose) has the nature of suggestion.

Objection: Why?

14cd - 15ab. Reply: Metaphor is used to bring about an idea connected with that

(suggestion). The purpose (of the metaphor) is inferred from the (metaphoric) word

alone. There is no other semantic capacity (that could achieve this) apart from

suggestion.

When a word is used metaphorically in order to set forth a special purpose, the cognition

of that (purpose) does not come from another (word), but rather from that (metaphoric)

word itself. And in this (type of situation), there is no (semantic) function other than

suggestion (that can accomplish this). this).

Comments

Types of suggestion divide into two groups: those based on words and those based on

meaning. Because Mammata is treating the nature of words in this chapter, he begins

109

with the suggestions based on words. Those also divide into two groups: those based on

metaphoric words and those based on literal words. Mammata begins with suggestion

based on metaphoric words and will treat suggestion based on literal words at the end of

this chapter (kārikā 19). In chapter three he will take up suggestions based on meaning.

We have already seen the stock example, "The village is on the Ganges."

Mammata has shown that "Ganges" must be understood metaphorically to mean "bank of

the Ganges." The purpose of this metaphor is traditionally understood to be

communication of certain qualities, like purity and coolness. Mammata will try to prove

in this section that the only semantic operation that can communicate the desired meaning

is suggestion. He will use an eliminative argument with the three types of meaning as

candidates. By ruling out denotation and metaphor, he can conclude that suggestion is the

semantic power used.

tathāhi

nābhidhā samayābhāvāt | (345)

gangāyām ghosa ityādau ye pāvanatvādayo dharmās tatādau pratīyante na tatra

gangādiśabdāh samketitāh II

The author explains the above (claim that suggestion is needed to explain the cognition of

the special purpose of a metaphor).

110

15c. (The required semantic capacity) is not denotation because the conditions (for

denotation) are not present.

In sentences like, "The village is on the Ganges," the properties of purifying, etc., are

cognized as belonging to the bank. There is no linguistic convention for the word

"Ganges" (that can account for this).

Comments

Mammața presents the example of, "The village on the Ganges." The statement causes

cognition of certain qualities associated with the river, namely coolness and purity.

Mammata claims that denotation cannot account for such cognition because "the purity of

the Ganges" is not part of the literal meaning of the word "Ganges."

The claim is supported by Kumārila's argument from the Tantravārttika that

Mammata used earlier concerning metaphor: a word cannot change its meaning every

time it is used in a new metaphor. In this example, "Ganges" cannot be said to come to

mean "the purity of the Ganges." Even if it becomes a faded metaphor (as in "London is

on the Thames"), the meaning "purity" would not be included in the literal meaning of

"Ganges." In the case of faded metaphors, the new meaning is conventionally attached to

the word and thus no special purpose is communicated. For example, by calling someone

a maverick, you do not suggest qualities associated with the original rancher (Maverick)

or his cattle.

111

Some believe, contra Mammata, that the statement, "The village is on the bank of

the Ganges," even understood only in its primary, literal sense, suggests the properties of

purity, etc. The metaphorical formulation is used to emphasize the suggested meaning on

this view. It does this by creating an obstacle that slows the reader down and demands an

explanation (why "Ganges" and not "bank of the Ganges"). The metaphor highlights the

word "Ganges," and invites the reader to reflect on its connotations. However, the

suggested meaning, even on this view, could not be communicated by the literal meaning

alone. 134

hetvabhāvān na lakșaņā | (347)

mukhyārthabādhāditrayam hetuḥ II

tathā ca

lakşyam na mukhyam nāpy atra bādho yogah phalena no I

na prayojanam etasmin na ca śabdah skhaladgatiḥ II (347)

yathā gangāśabdah strotasi śabādha ity tatam laksayati tadvat yadi tate 'pi sabādhaḥ

syāt tat prayojanam laksayet I na ca tatam mukhyo 'rthah I nāpy atra bādhah I na ca

gangāśabdārthasy tatasya pāvanatvādyair laksanīyaih sambandhah I nāpi prayojane

laksye kimcit prayojanam I nāpi gangāśabdas tațam iva prayojanam pratipādayitum

asamarthaḥ II

134 See Abhinava's Locana on 1.4b for a more complete presentation of the argument Mammata advances.

112

15d. (The required semantic capacity) is not metaphor because the necessary

conditions are lacking.

The necessary conditions refers to the three (given in kārika 9), i.e., the blocking of the

literal meaning, and the rest.

16. Specifically, the metaphoric meaning is not a literal meaning. There is no

blocking. There is no connection (between the metaphoric meaning) and the

purpose (of the utterance). There is no (further) special purpose in this (second

metaphor). Finally, the word is not powerless (to convey the purpose itself).

As the word "Ganges" indicates the bank because its (literal meaning) "river" is blocked,

similarly, if the (meaning) "bank" were blocked, the word "Ganges" could indicate the

special purpose (the qualities purity, etc., by a second order metaphor). But "bank" is not

a literal meaning, and, further, it is not blocked. Furthermore, there is no connection

between the "metaphorically indicated properties," like purifying, and the bank as the

(metaphoric) meaning of the word "Ganges." Additionally, there is no (additional)

special purpose for the metaphorical indication of the (first) special purpose. Finally, the

word "Ganges" is not unable to set forth the special purpose, as it was unable to set forth

the meaning "bank" (without metaphor).

Comments

In this nice example of reductio ad absurdum, Mammata shows the unacceptable

consequences that follow from assuming that the special purpose of a metaphor can be

113

communicated by a second metaphor. He does this by recalling the necessary conditions

for metaphor set out in kārikā 9.

evam apy anavasthā syāt yā mūlakșayakāriņī l

evam api prayojanam cel laksyate tat prayojanāntareņeti tad api proyojanāntareņety

prakṛtāpratītikrt anavasthā bhavet I nanu pāvanatvādidharmayuktam eva tatam laksyate

I gangāyās tate ghoșah ityato 'dhikasyārthasya pratītiś ca proyojanam iti viśiste lakșanā l

tat kim vyañjanayety āha II

prayojanena sahitam laksaņīyam na yujyate II (355)

17ab. Furthermore, this (the purpose of a metaphor being made known by a further

metaphor) would lead to infinite regress, which would destroy the original

(metaphor).

Furthermore, this means that if the purpose (of the original metaphor) had to be itself

indicated metaphorically there would be another purpose of this (second metaphor) and

another purpose (for the third metaphor). Thus there would be an infinite regress and no

cognition of the matter at hand (i.e., the purpose of the first metaphor). (And as this

contradicts the data of experience, the premise that the goal of metaphor is metaphoric

must be false.)

Objection: The bank joined with the qualities like purity, etc., is metaphorically

indicated. The purpose is cognition of additional meaning beyond (the meaning of) the

114

sentence, "The village is on the bank of the Ganges." Thus metaphoric indication

indicates an object qualified (by certain qualities such as purity). What is the use of

suggestion?

17cd. Reply: (The theory that) metaphorically indicated meaning is communicated

together with the purpose (of the metaphor) is not correct.

Comments

Mammata argues that if a first metaphor required a second metaphor to be

understandable, the second would require a third, and so on. This regress would be

vicious because it would block any understanding of the original metaphor. However, we

do understand the original, so the theory is false.

The objection tries to counter by making the purpose communicated by the

original metaphor along with the original metaphoric meaning. In this way, the purpose

would be evident in the original metaphor without recourse to another metaphor,

avoiding the regress. Mammata responds in the next kārikā.

kuta ity āha

jñānasya vişayo hy anyaḥ phalam anyad udāhrtam I (361)

pratyaksāder nīlādir visayah phalam tu prakatatā samvittir vā |

viśişte laksaņā naivam

vyākhyātam I

115

viśeşāh syus tu lakșite II (361)

tatādau ye viśesāh pāvanatvādayas te cābhidhātātparyalaksaņābhyo vyāpārāntareņa

gamyāh l tac ca vyañjanadhvananadyotanādiśabdavācyam avaśyam eșitavyam |

Objector: Why (is my theory not correct)?

18ab. Reply: Because the intentional object of cognition is one thing, while its result

(phala) is said to be another.

The intentional object of perception, etc., is a blue thing, for example. But the result (of

the cognition) is either "being-manifest" (i.e., the quality "being known" now belongs to

the blue thing) or a cognition (i.e., "This is something blue" in the self).135

18c. Therefore, metaphor does not indicate a qualified object.

This has been explained.

18d. The properties would be (cognized) in the indicated (object).

For example, in the case of the bank, the qualities of purifying, etc., are understood by a

semantic function different from denotation, speaker's intended meaning, and metaphor.

And this function has to be the one called suggestion, semantic reverberation, and

illumination, etc.

Comments

35 The extra information in my translation of this sentence comes from the Sanketa commentary. 116

The section starts with an argument from analogy in reply to the objection given in the

previous kārikā. Mammata states a general rule of epistemology (the intentional object of

a cognition is different from its result) and claims it applies analogously to semantics (the

meaning of a metaphor is different from it purpose). The analogy is as follows: A

cognition stands in relation to its intentional object as a word does to its meaning.

Furthermore, they both have a phala. "Phala" means "fruit," and by extension both

"result" and "purpose." I am using the Sanskrit "phala" because the argument turns on

the ambiguity of this term. The relation between the intentional object and the phala

(result) of the cognition should be (by analogy) the same as the relation between the

meaning and the phala (purpose) of the word. The argument fails, it seems to me,

because the causal result of a perceptual cognition is not analogous to the human purpose

motivating a metaphor in any obvious way.

Mammata concludes the larger argument, begun in kārikā 14cd., by claiming that

the special purpose for using a metaphor must be communicated by suggestion. He

assumes that the purpose must be communicated by one of the semantic powers; it is not

communicated by denotation; it is not communicated by speaker's intended meaning; it is

not communicated by metaphor; therefore, it must be communicated by suggestion.

Mammata does not argue against speaker's intended meaning for two reasons.

First, speaker's intended meaning is used to establish the literal meaning of a sentence in

cases of ambiguity, and thus falls under denotation. Second, this section only concerns

suggestions based on a single word. Clearly, a semantic power based on the entire

117

sentence, like speaker's intended meaning, cannot account for the semantic capacity of a

single word.

evam laksanāmūlam vyañjakatva muktam II abhidhāmūlam tv āha

anekārthasya śabdasya vācakatve niyantrite I

samyogādyair avācyārthadhīkrdvyāprtir añjanam II (372)

samyogo viprayogaś ca sāhacaryam virodhitā l arthah prakaranam lingam śabdasyānyasya samnidhih II sāmarthyamaucitī deśah kālo vyaktiḥ svarādayah | śabdārthasyānavacchede viśesasmrtihetavah II

ity uktadiśā

saśankhacakro hariḥ aśankhacakro harir ity acyate l rāmalaksmaņāv iti dāśarathau I

rāmārjunagatis tayor iti bhārgavakārtavīryayoh I sthānum bhaja bhavacchide iti hare I

sarvam jānāti deva iti yusmadarthe l kupito makaradhvaja iti kāme l desasya purārāter

iti śambhau I madhunā mattah kokila iti vasante | pātu vo dayitāmukham iti sāmmukhye I

bhāty atra parameśvara iti rājadhānīrūpāt deśād rājani | citrabhānur vibhātīti dine

ravau rātrau vahnau I mitram bhātīti suhrdi mitro bhātīti ravau l indraśatrur ityādau

vede eva na kāvye svaro viśeșapratītikrt l ādigrahaņāt

eddahamettatthaniā eddahamettehi acchivattehim I eddahamettāvatthā eddahamettehim diaehim |1136

ityādāv abhinayādayah I

136 1. eddahamettatthaņiā = etāvanmātrasthanī eddahamettehi acchivattehim = etāvanmātrābhyām akșipātrābhyām upalakșita 2. eddahamettāvatthā = etāvanmātrāvasthā eddahamettehiņ diaehiņ = etāvanmātrair divasaiḥ 118

(2.7.1.) Suggestion based on metaphor is explained thus. But suggestion based on

denotation is explained as follows:

19. When the referential capacity of a word with several meanings is limited by a

contextual element, such as a relation, the semantic operation that creates the

cognition of the non-denoted meaning is suggestion.

(Now follows a partial list of the possible contextual elements:)137

12. Conjunction, (2) disjunction, (3) association, (4) enmity, (5) motive, (6) situation,

(7) particular quality, (8) proximity of another word, (9) capacity, (10) fitness, (11) place,

(12) time, (13) gender, (14) accent. And others.

According to the direction just stated (there are these examples):

2. "Hari" with the conch and disc means "Acyate."138

3. "Hari" without the conch and disk means "Acyate."

4. "Rāma and Laksmaņa" mean "the sons of Daśaratha."

5. "They act like Rāma and Arjuna" mean "the son of Bhrgu and the son of

Krtavīrya."139

6. "Worship the immovable for release from the world." ("The immovable") means

"Hara."140

7. In "Lord knows all," ("lord") refers to the person addressed.

137 From Bhartrhari's Vākyapadīya 2.317 - 318. This is not Bhartrhari's own view according to his commentator Puņyarāja, but this is disputed by others. See Kunjunni Rāja (1969:50). 138 "Acyate" is a name for Visnu. 139 Rāma and Arjuna are famous enemies. 140 "Hara" is a name for Siva. 119

8. In "The one with the shark banner is angry," ("the one with the shark banner") means

"the god of love."141

9. In "It belong to the god, the enemy of the cities," (by proximity to the word "enemy-

of-cities," "god") means "Sambhu."142

10. "The cuckoo enraptured with (honey/wine/spring)" (the word that can mean

"honey/wine/spring" is limited to) "spring," (because spring alone has the power to

enrapture the cuckoo).

11. In "May the face of the beloved drink/protect/attend to you," ("drink/protect/attend")

means "observing" (because this meaning alone fits).143

12. In "The highest lord shines here," ("the highest lord") means the "king," because of

the place, which is the king's residence.

13. In "The bright one is shining," ("the bright one") means "the sun" in the day and

"fire" at night.

14. In "Mitram shines, " ("Mitram" means) "a friend." In "Mitram shines," ("Mitram"

means "the sun."144

15. In words like "Indra-killed" the accent does not cause the cognition of a particular

meaning in poetry, as it does in the Vedas (where the meaning of the compound is

determined by accent to be either "killer of Indra" or "killed by Indra").

141 The other bearer of a shark banner - the ocean - cannot have this particular quality. 142 "Sambhu" is a name for Siva. 143 The commentators disagree about the exact meaning of this example. 144 In the first sentence, "mitram" is masculine, in the second, it is neuter. 120

By and others other contextual elements, such as gestures, are understood (to play a role

in delimiting meaning. The following is an example of a gesture resolving the ambiguity

of a term:)

Her breasts have reduced to just this, Her petal-eyes have shrunk to merely this, She has dropped as far as this, In only this many days. (11)

Comments

Mammața takes up the suggestive possibilities of literal usage. The goal of this section is

to review the many ways a single ambiguous word can come to denote one of its

meanings and not another. This list of delimiting factors comes out of the Grammarian

tradition, but the use of suggestion (explained in the next section) comes from

Anandavardhana.14> The list contains three distinct types of factor. First, there are

grammatical (syntactical) factors such as gender and accent (13 and 14). Second, there

are inter-textual factors that draw on the larger semantic context (1, 2, 3, 7, 8, 9, 10).

Third, there are extra-textual factors that draw on a larger pragmatic context (6, 11, 12,

and the poem). The Grammarians investigate the role and types of contextual figures at

great length. 146

In the poem, the word "this" is ambiguous; it could refer to many sizes. An

accompanying gesture would show how the word is to be understood. That Mammata

145 Bhartrhari's Vākyapadīya 2.317 - 318. Ānandavardhana (2.21). 146 Kunjunni Raja (1969:58). 121

uses an example from drama adds credibility to Edwin Gerow's claim that the distinction

between drama and poetry has been obscured by this point in the Sanskrit tradition.147

ittham samyogādibhir arthāntarābhidhāyakatve nivārite 'py anekārthasya śabdasya yat

kvacid arthāntarapratipādānam tatra nābhidhā niyamanāt tasyāh I na ca laksaņā

mukhyārthabādhādyabhāvāt l api tv añjanam eva vyāpārah l yathā

bhadrātmano duradhirohatanor viśālavamśonnateh krtaśilīmukhasamgrahasya I yasyānupaplutagateh paravāraņasya dānāmbusekasubhagaḥ satatam karo 'bhūt II (395)

A word with more than one sense is prevented from denoting another meaning by a

contextual element like a relation, as was just explained. However, sometimes such a

word does convey another meaning. The way the word conveys this is not denotation,

because it has been delimited (to denoting a single meaning). Further, it is not metaphor,

because the conditions, such as the blocking of the primary sense are not present. Rather,

the semantic operation is suggestion itself. For example (the following punning poem):

Literal meaning Suggested meaning

The hand The trunk Of the noble-souled one Of the one who belongs to a good species Whose family is great Who is tall as bamboo Whose body is unassailable Who is difficult to mount Who has gathered a stock of arrows Who holds a swarm of bees Whose knowledge is unclouded Whose gait is unimpeded Who wards off enemies Who is an excellent elephant That hand was ever graceful by the That trunk was ever graceful in the sprinkling of the water of gifts sprinkling of the water of oblations. (12)

147 Gerow (1977:252). 122

Comments

For lack of a better method, I have used two columns to render the apparent and the

hidden meanings for each phrase in the Sanskrit original of this punning poem. One

should read the left poem first, then the right, and then try to understand how they can be

translations of the same text. Although there is little poetic merit in this method, at least

the reader should be able to understand the example.

In the case of puns, ambiguous words convey more than one meaning. Punning

received much attention in the early poetics.148 I should note that puns play a much larger

role in Sanskrit poetry that they do in English poetry, by and large. For several reasons

(limited number of roots, compounding, freedom of word order, etc.), it is easier to pun in

Sanskrit than English. The possibility of sustained puns (such as the poem above)

enlarges the poetic scope of punning. While in English puns are mostly used for comic

effect, in Sanskrit the uses are many. Mammata will return to the subject several times in

chapter four (poems 54 and 78) and at length in the section on ornaments (chapters nine

and ten). In this section, he is presenting only the argument that the second meaning in a

pun is conveyed by suggestion.

Mammata use once again an eliminative argument.The additional meaning

cannot be conveyed by denotation because this semantic power exhausts itself in

148 See Bhamaha 3.17 and Dandin 2.333, for example. 149 Mammata condenses a rather large argument in Anandavardhana (Ingalls, et al. pp. 291-311), losing much of the subtlety in the process. 123

denoting the primary meaning.150 It cannot be conveyed by metaphor because none of the

necessary conditions for metaphor are present. Therefore, it must be conveyed by

suggestion. Mammata will return to other parts of Ānandavardhana's discussion in

chapters three, four, and nine.

tadyukto vyañjakaḥ śabdaḥ

tadyukto vyañjanayuktah I

yat so 'rthāntarayuk tathā I

artho 'pi vyañjakas tatra sahakāritayā mataḥ II (404)

tatheti vyañjakah II

20a. The suggestive word is joined to that.

Joined to that means joined to suggestion.

20bcd. Since a word is so (i.e., suggestive) when joined with another meaning, that

meaning is also thought of as suggestive, because it is an auxiliary.

So means suggestive.

150 This argument dates back to Sabara's commentary on Jaimini's Mīmāmsāsūtra 1.1.25. It is only contradicted (as far as I know) by the Prabhākara school of Mīmāmsā. See Kunjunni Raja (1969:300). 124

Comments

Mammata's topic in this chapter has been the suggestiveness of words. But one could ask,

"How are words suggestive, but not their meanings?" To this question kārikā 20 answers

that when a word used metaphorically (like "Ganges") is joined with an additional

meaning ("the bank of the Ganges"), this additional meaning can be called suggestive as

long as it remains joined to the suggestive designator. This qualification avoids the

objection that the meaning "bank" is not suggestive in many other case (i.e., all the cases

where the word "bank" is used only literally). In the case at hand, the suggestiveness of

the meaning "bank" could be called secondary suggestiveness because it is dependant on

the primary suggestiveness of the word "Ganges."

Likewise, the homonymous word "karas" that can mean either "hand" or "trunk"

in the punning example above. It suggests "trunk" only when it means "hand". Thus the

meaning "hand" is called suggestive in this context. If the same word occurred in a poem

with the primary meaning "trunk," that meaning could suggest "hand." As we saw above,

a word can only have one literal meaning in any given usage.

Mammata will consider suggestions based on meaning in the next chapter. Here it

will be clear that when a meaning is primarily suggestive, the words that convey the

meaning are also suggestive in a secondary way (see kārikā 23). Thus the general truth:

neither words nor meanings are ever suggestive in isolation.

125

atha trtīya ullāsaḥ

iti śrikāvyaprākaśo 'rthavyañjakatānirņayo nāma trtīya ullāsah

Chapter Three

Investigation of Suggestiveness Based on Meaning

arthāḥ proktāḥ purā teșām (409)

arthāh vācyalaksyavyangyāh I tesām vācakalāksanikavyañjakānām II

arthavyañjakatocyate l (409)

21a. The meanings of these have been explained above (in chapter two).

The meanings are literal, indicative, and suggestive. Of these refers to literal designators,

indicative designators, and suggestive designators.

21b. Suggestiveness based on meaning will now be explained.

126

Comments

In the previous chapter, Mammata explained suggestions based on homonymous words.

Each instance of such a word was said to have a single literal meaning, with the other

meanings "resonating" in the mind of the reader. A poet can use such resonating

meanings suggestively.

In this chapter, Mammata examines poems in which the words have a single,

specific meaning. Furthermore, the sentences formed from such words have a single,

literal meaning. However, Mammata claims that the literal meaning (both of a word and

sentence), in combination with some other factor, can call forth other meanings. This

process is called suggestion based on meaning, and it is elucidated by means of ten

examples.

kīdṛśīty āha

vaktrboddhavyakākūnām vākyavācyānyasamnidheḥ II

prastāvadeśakālāder vaiśistayāt pratibhājușām I

yo 'rthasyānyārthadhīhetur vyāpāro vyaktir eva sā Il (413)

boddhavyah pratipādyaḥ I kākur dhvaner vikāraḥ I prastāvah prakaraņam l arthasya

vācyalaksyavyangyātmanah I

Objection: "What is it like?"

127

21cd - 22. Reply: Suggestiveness is exactly that semantic function of the meaning

which causes cognition of an additional meaning on the part of those keen on

subtleties by means of particularities of speaker, person addressed, intonation,

sentence, literal words, presence of another, occasion, location, time, etc.

The person addressed is the one communicated with. Intonation is variation of the

voice. Occasion means the context (of present activities). Of the meaning means the

literal meaning, the indicated meaning, or the suggested meaning.

krameņodāharaņāni I

aipihulam jalakumbham ghettūņa samāgagahmi sahi turiam I samaseasalilaņīsāsanīsahā vīsamāmi khanam ||151

atra cauryaratagopanam gamyate | (419)

(3.1.) Here are the examples in order (starting with a suggestion based on the character of

the speaker, who we know to be a girl by the gender of the word "I" and its predicated

adjectives):

O friend, having taken a full jug of water, I have come swiftly. I am tired, sweating, and breathless. I will rest a moment. (13)

151 Gāthāsaptaśati Weber 881. Weber's version show numerous differences (1881:478). 1. sahi = he sakhi - o friend, aipihulam = atiprthulam - full, jalakumbham = jalakumbham - a jug of water, ghettuņa = grhītvā - having taken, turiam = tvaritam - swiftly, samāgagahmi = samāgatāsmi - I have come, 2. samaseasalilaņīsāsaņīsahā = śrama-sveda-samlilaņiḥ śvāsanih sahā - weary, sweat-water panting with, khaņam = kșaņam - for a moment, vīsamāmi = viśrāmyāmi - will rest. 128

Here concealment of a secret love affair is known (through suggestion).

Comments

The verse is an example of a suggestion that functions by means of particularities of the

speaker. Gajendragadkar claims that the friend knows the speaker to be a "woman of

disreputable character."12 Jha also claims that "the character of the speaker is known to

be that of a woman with loose morals."153

Gajendragadkar's claim cannot be correct, for it is not the friend, but rather the

reader, that needs knowledge of the woman's character in order to understand the poem.

If the reader knew that the friend considered the speaker to be a woman of loose

character, the suggestiveness of the self-description would be reinforced. If the reader, in

other words, knew that the friend took the sweet, etc., to be signs of a love affair, he

could understand how the word "sweet" was being used suggestively. But there are two

problems with this approach. First, the poem contains no information concerning the

friend's opinions or beliefs. Second, it confuses suggestive literature with suggestive

discourse inside literature. Is the poem suggesting additional meaning to the reader, or is

the woman suggesting additional meaning to her friend? If the first, the opinion of the

friend is only relevant in so far as it is adopted by the reader. To a reader knows the

friend to be mistaken in her assessment of the woman's character, the poem will not be

suggestive. If the second, the reader would need to understand the friend's reaction in

152 Gajendragadkar (1939:237). 153 Jha (1925:41). 129

order to understand that the woman was speaking suggestively. We can imagine that if

the friends needed to communicate in the presence of others, the poet could show them

speaking in code (compare poem number eight). If this were the case, the reader would

need much more information than the poem conveys.

Jha does not say who knows about the woman's character. If he means the reader,

he fails to explain how the reader comes to know anything about the woman's character

that is not explicitly stated in the poem. The suggestion of a tryst, it should be noted, is

not supposed to depend on a larger textual context. Ingalls claims that all the verses

collected in the volumes of short verse (subhāsita) do not depend on other verses within a

drama or poem, but rather stand alone.1 Unfortunately, some of Mammata's examples

do seem to depend on a larger textual environment (see, e.g., poem eighteen below).

So how does the speaker's character inform the suggestion? I speculate that this

poem exemplifies a certain literary topos, poems about woman of loose morals. In a

poem such as this, the genre provides the necessary background information. This

solution remains within the parameters set out by Mammata - word, meaning, and a

cultured reader. Such a reader would recognize the genre, make the correct assumptions

about the woman's character, and thus understand the suggested meaning.

Unfortunately, as I mentioned, some of Mammata's examples do not work this

way. The inclusion of examples that require knowledge of the text from which the verse

is drawn, weakens the explanatory value of the theory of suggestion. In such cases it is

difficult to rule out inference or one of the other semantic powers as the means by which

154 Ingalls (1965:33). 130

the information is grasped. Furthermore, the larger textual context is not included in the

parameters Mammata gives as necessary to understanding suggestion (word, meaning,

and a cultured reader), unless, of course, knowledge of all the major dramas and the

ability to place any given verse is assumed for an educated reader.

onniddam dobballam cimtā alasattanam sanīsasiam I mahamamdabhāiņīe keram sahi tuha vi ahaha parihavai I1155

atra dūtyās tatkāmukopabhogo vyajyate | (424)

(3.2. Next an example of a suggestion based on the character of the person addressed.

The characters are two women:)

O friend! In the service of my wretched self You too are overcome with Sleeplessness, weakness, anxiety, fatigue, and sighs. (14)

Here it is suggested that the speaker's lover has enjoyed sex with the messenger.

Comments

In this genre of poetry, the go-between (friend, messenger, etc.) seems prone to affairs

with the lover of the woman she is "helping". Knowing this, an educated reader will read

155 1. oņņiddam = aun+nidryam - sleepelessness, dobballam = daurbalyam - weakness, ciņtā = cintā - anxiety, alasattaņam = alasattvam - tiredness, saņīsasiam = saniśvasitam - having sighs. 2. maha = mama - of me, mamdabhāiņīe = mandabhāgyāyāh - of the unlucky one, keram = krte, sahi = sakhi - o friend, tuha vi = tvām api - you too, ahaha = aha ha - certainly indeed, parihavai = paribhavati - comes over.

131

beyond the literal meaning (i.e., that the friend tired herself working to bring the two

lovers together) and understand the suggested meaning (i.e., she tired herself having sex).

See poem two in the first chapter for another example of this genre.

tathabhūtām drstvā nrpasadasi pāñcālatanayām vane vyādhaih sārdham suciram uşitam valkaladhaih I virātasyavāse sthitam anucitārambhanibhrtam guruh khedam khinne mayi bhajati nādyāpi kurușu II

atra mayi na yogyah khedah kurusu tu yogya iti kākvā prakāśyate I na ca vācyasidd

hyańgam atra kākur iti guņībhūtavyangyatvam śankyam I praśnamatreņāpi kākor

viśrānteh II (429)

(3.3. Next an example of a suggestion based on intonation:)

Having witnessed the Panchala princess laid low, We left for the forest, dressed in bark, and lived among hunters. We then went to Virata's palace, doing dirty work to remain unknown. My brother is angry with me, thus afflicted, And not with the Kurus? (15)156

The suggested meaning in this verse, i.e., "Anger towards me is not proper, but towards

the Kurus it is proper," is revealed by intonation. And it should not be suspected that the

intonation only serves to establish the literal meaning and that thus this is an example of

156 Venīsamhāra 1.11. The English convention of question marks makes the assertion/question ambiguity impossible to reproduce in writing. This convention marks that fact that these statements are disambiguated by intonation in English, just as Mammata claims for Sanskrit. 132

poetry of subordinate suggestion. The intonation can stop with (the suggestion of) a mere

question.

Comments

The worry is that if intonation merely serves to suggest that the last line is a question, the

suggestion will be subservient to the literal meaning, rendering this an example of

"poetry of subordinate suggestion" (see kārikā 5ab). In reply, Mammața agrees that the

suggestion needed for completing the literal meaning is merely that the last line is a

question. He also agrees that this is accomplished by proper intonation. However, the

resultant meaning suggests further that the brother's anger is improper, and that he is

being reprimanded. This suggestion of reprimand is the central meaning of the poem,

guaranteeing that it is truly an example of suggestive poetry.

taiā maha gamdatthalanimiam ditthim na nosi annatto I eņhim saccea aham te a kavolā ņa sā ditthī ||157

atra matsakhīm kapolapratibimbitām pasyatas te drstir anyaivābhūt calitāyām tu tasyām

anyaiva jātety aho pracchannakāmukatvam te iti vyajyate II

157 1. Taiā = tadā - then, maha = mama - my, gamdatthala = gandasthala - region of the cheek, ņimiam = nimagnām - sunken, immersed, fixed, ditthim = drstim - glance, na nosi = na nayasi - you do not lead, aņnatto = anyataḥ - elsewhere. 2. eņhiņ = idānīm - same, saccea = saiva - it too, aham = aham - I, te a = tāveva - they too, kavolā = kapolau - cheek, ņa sā dițthī = na sā drstih - not the glance. 133

(3.4. Next an example of a suggestion based on the sentence:)

(A woman to her lover upon the departure of her female friend)

Then your glance was fixed, Immobile upon my cheek. I am still the same, My cheeks are the same, But your glance is not. (16)

Here the suggestion is, "Your glance was one way when you were looking at my friend,

who was reflected in my cheek. But when she moved away, it completely changed. Thus

you have a hidden desire (for my friend)."

Comments

The lover's fixed glance loses its fire without any change in the woman he is supposedly

admiring. This suggests that something beyond the woman was responsible for the loving

glance. From the implication of the genre an educated reader might be expected to know

that another woman was responsible.158 So far, so good, but it is a big step to the

knowledge that the woman's friend was reflected in her cheek and then left, occasioning

the drop in the lover's passion. One could also imagine that the lover was daydreaming

passionately of another woman and then stopped, occasioning the change. Or one could

imagine that the lover really was looking passionately at the woman, but then

remembered his vow to chastity, etc.

158 Although this information seems more likely to be gained by inference. 134

Regardless of these problems, Mammata is right to point out that the

suggestiveness of the poem derives from the mysterious change in the lover's glance that

is communicated very economically by the sentence as a whole. The poet first explains

that the focus of the glance has not changed, then that the object of the glance has not

changed, and finally that the quality of the glance has changed. The reader is left

wondering why, and this mystery invites the reader to search the real cause.

uddeśo 'yam sarasakadalīśreniśobhātiśāyī kuñjutkarșānkuritaramaņīvibhramo narmadāyāh kim caitasmin suratasuhrdas tanvi te vānti vātā yeşām agre sarati kalitākāndakopo manobhūh II

atra ratārtham praviśeti vyangyam II (443)

(3.5. Next an example of a suggestion based on the literal words:)

The banks of the Narmada are stunning, and this place surpasses all! Dense alluring blooms among lush banana trees entice a charming woman to wander. O slender one! Feel the breezes blow - they are such good friends to hot lovers. Isn't that Love himself racing before them, sowing passions impetuous and single-minded? (17)159

The suggestion here is, "Enter here for sexual enjoyment."

59 Attributed to Battendurāja by Sūktimuktāvalī according to Dwivedi. 135

Comments

The suggestion is based on the various words that point out the suitableness of the place

for a sexual encounter. It differs from the previous type in that the romantic words work

independently, each serving to further the suggested meaning. Cumulatively, they leave

little doubt as to the desires of the speaker.

ņollei anaddamaņā attā mam gharabharammi saalammi khaņamettam jai samjñāi hoi ņa va hoi vīsāmo II160

atra samdhyā samketakāla iti tatastham prati kayācid dyotyate II (448)

(3.6. Next an example of a suggestion based on the presence of another:)

My tough-minded mother-in-law pushes all the housework my way. Only in the evening, for a moment, will there perchance be rest, or there will not be any. (18)

Here the meaning, "The evening is the time for a rendezvous," is made clear by a woman

to someone standing nearby.

Comments

160 1. ņollei = prerayati(?) nudati - push on aņaddamaņā = anārdramanā - whose mind is not soft, attā = svaśrh - mother-in-law mam = mām - me, gharabharammi = grhabhare -to housework, saalammi = sakale - all, 2. khaņamettaņ = kșaņamatram, jai = yadi - if, saņjñāi = sandhyāyām - in the morning, noon, or evening (one of the three transitions of the day), hoi = bhavati - be, ņa va hoi = na vā bhavati- or not be, vīsāmo = viśrānaḥ - rest. 136

The woman is pretending to be speaking to herself, and the literal meaning of the words

is clear. But her words also serve to suggest an additional meaning to a man standing

nearby. This is a case of reading between the lines that can only be achieved by knowing

the context. In other words, this verse, at least, cannot stand by itself. So either Mammata

has picked a poor example, or Ingalls is incorrect in his claim that the short verses

admired by the critics (subhasita) do not depend on other verses within a drama or poem,

but rather stand alone.11 For the reasons given in the beginning of the chapter, the theory

of suggestion would have greater explanatory power if the verses could be read

independently of their original context.

suvvaha samāgamissadi tuj pio ajja paharamettena I eme a kitti citthasi tā sahi sajjesu karanijjam II162

atropapatim pratyabhisartum prastutā na yuktam iti kayācin nivāryate II (452)

(3.7. Next an example of a suggestion based on occasion:)

Your husband is supposed to come before the next gong. Why stand about, my friend? Prepare for his arrival. (19)

Here a woman beginning to move towards her lover is being restrained by a

woman (friend with the suggestion), "It is not proper."

161 Ingalls (1965:33). 162 1. suvvaha = evam śrūyate - is heard thus, samāgamissadi = samāgamisyati - coming, tuj = tava - your, pio = priyaḥ - lover/husband, ajja = adya - today, paharametteņa = praharamātreņa - only three hours (one strike of the gong) 2. eme a = evam eva, like this, kitti = kimiti - why, therefore, citthasi = tisthasi - you stand, tā = tat - this, sahi = sakhi - friend, sajjesu = sajjaya - make ready, karaņijjam = karaņīyam - what is to be done.

137

Comments

The friend has learned that the woman's husband is coming, and wants to tell her that she

does not have the time to go dally. Again, this verse requires a context, namely, one has

to know that the woman intends to go meet her lover. I have not read enough Sanskrit

love poetry to know whether the genre can be said to supply the premise that wives left

alone will invariable cheat on the husbands.

anyatra yūyam kusumāvacāyam kurudhvam atrāsmi karomi sakhyah nāham hi dūram bhramitum samarthā prasīdatāyam racito 'ñjalir vah II

atra vivikto 'yam deśa iti pracchannakāmukas tvayābhisāryatām iti āśvastām prati

kayācin nivedyate II (455)

(3.8. Next an example of a suggestion based on location:)

Collect your flowers elsewhere my friend. I'll collect them here, for I cannot wander far. Cupping my palms together, I salute you. Be pleased! (20)

A woman makes known (the following suggested meaning) to her confidant, "This place

is solitary; therefore send my secret lover here."

138

Comments

This verse is recited in a secluded, romantic, forest grove. The location leads the reader to

guess the real intention of the woman. Again, I do not know if the genre can be said to

supply the premise that young women always hope for the arrival of their lovers

whenever they are in a solitary place. 163

guruanaparavasa pia kim bhanāmi tuhi mamdabhāiņī ahakam ajja pavāsam vaccasi vacca saam jevva suņasi karanijam I1164

atrādya madhusamaye yadi vrajasi tadāham tāvat na bhavāmi tava tu na jānāmi gatim iti

vyañjyate I (459)

(3.9. Next an example of a suggestion based on time:)

The elders have decided for you, o loved one, What can I say? My luck is lost. Today you are going abroad. Go! Only you know what must be done. (21)

Here the woman suggests, "If you go today, in the spring time, and I know not your fate,

I will die."

163 I cannot think of a Sanskrit poem where a young maid and her friend remain alone in the woods without the wanted or unwanted arrival of the lover. 164 1. guruaņaparavasa = gurujanaparavaśa - dependant on the will of the elders, pia = he priya - o loved one, kiņ bhaņāmi = kim bhaņāmi, what say I say, tuhi = tvām - to you, mamdabhāiņī = mandabhāgyā - who am unfortunate ahakam = aham - I, 2. ajja = adya - today, pavāsam = pravāsam - abroad, vaccasi = vrajasi - you are going, vacca = vraja - you go!, saam jevva = svayam eva - just yourself, suņasi = jānāsi - you know karaņijam = karaņīyam - what is to be done.

139

Comments

This verse is recited in spring, so the rainy season is about to arrive. In the Sanskrit poetic

tradition, the worst fate for lovers is to be kept apart in this season of love. If the lover

leaves and there is no hope of his returning soon, the woman suggests she will not

survive. This is yet another poem whose suggestion cannot be understood without

understanding the larger context, for the season is not mentioned in the verse itself.

ādigrahaņāc cestādeh I tatra cestāyā yathā

dvāropāntanirantare mayi tayā saundaryasārāśriyā prollāsyoruyugam parasparasamāsaktam samāsāditam ānītam puratah śiromśukamadhah ksipte cale locane vācas tatra nivāritam prasaranam samkocite dorlate II

atra cestayā pracchannakāntavisaya ākūtavieso dhvanyate II

nirākānksapratipattaye prāptāvasaratayā ca punar punar udāhriyate l vaktrādīnām

mithaḥ samyoge dvikādibhedena l anena krameņa laksyavyangyayoś ca vyañjakatvam

udāhāryam II (463)

(3.10.) From the use of etc. (in kārikā 22, above), gestures (are to be included in the list

of possible contextual elements). An example of (suggestion based on) gesture:

She shines with the very essence of beauty! When I come near her door, She rubs her thighs, Bringing them together, One upon the other. Through her veil 140

She casts stray looks, Says little, And folds her creeper-arms. (22)

Here, by means of gesture, a particular desire or plan is suggested for the sake of the

secret lover.

Again and again a fitting example is cited to achieve understanding without

expectation (of something more to come). The factors such as who is speaker and so on

can be grouped into pairs with differences (in the way suggestion is achieved) according

to which two, or three, etc., factors are involved. Though such a sequence (pairs of

factors, triads, and so on), suggestiveness of combinations of metaphoric meaning and of

suggested meaning could also be illustrated.

Comments

Mammata claims that the gestures described in the poem are used to communicate

information to a lover and thus to suggest a course of action. On this reading, the reader

would need to understand the coded gestures to understand the suggested action.

Alternatively, the woman of the poem could merely be suggesting that she desires the

man who is speaking. On this simpler reading, the verse can stand alone.

Each of the factors could be combined with another factor and each of these

different pairs of factors (or triplets, etc.) could be called different types of suggestion.

Furthermore, examples could be given of each accordingly. Mammata asserts that

141

examples could be similarly constructed wherein the suggestiveness of metaphoric and

suggestive sentences were contained. However, he provides no examples here.

śabdapramāņavedyo 'tho vyanakty arthāntaram yataḥ I

arthasya vyañjakatve tat śabdasya sahakāritā II (477)

śabdeti nahi pramāņāntaravedyo 'rtho vyañjakah II

23. Since the meaning that suggests another meaning is known by way of words (i.e.,

testimony), in the suggestive function of a meaning, the word plays a role.

The suggested meaning is known by way of testimony, i.e., words, because a meaning

known by another means of knowledge would not be suggestive.

Comments

Mammata repeats the claim that both word and meaning are involved in every instance

of suggestion, linking this kārikā to kārikā 20. "Other means of knowledge" refers to

means of knowledge like perception, inference, etc.163 In poetry, no means of knowledge

other than verbal knowledge is necessary, according to Mammata. He does not address

the epistemological problems that arise from the combination of present textual

information with memories of earlier textual information.

165 Of course, there is active debate over how many means of knowledge there are. Based on his use of five means in this text and respect for the Mīmāmsaka theories, Mammata probably follows the Bhatta view that there are five. 142

Furthermore, it is very hard to imagine understanding poem eleven in chapter two

without perceptual information. "Her breasts have reduced to just this, / Her petal-eyes

have shrunk to merely this, / She has dropped as far as this, / In only this many days."

Technically, of course, the meaning of each of uses of "this" could be understood by

previous textual information, but this poem is given as an example of disambiguation

through gesture.

143

atha caturtha ullāsaḥ I

iti kāvyaprakāśo dhvaninirņayo nāma caturtha ullāsah l1166

Chapter Four

The Exposition of Suggestive Poetry

yady api śabdārthayor nirņaye krte dosagunālankārānām svarūpam abhidhānīyam

tathāpi dharmiņi pradarśite dharmāņām heyopādeyatā jñāyata iti prathamam

kāvyabhedān āha l

(According to the definition of poetry given above167,) now that the exposition of

word and meaning is done, the nature of blemishes, excellences, and poetic

ornaments should be explained. However, the author will first state the divisions

of poetry because only when the property-bearer is known can one determine

which properties are to be rejected and which accepted.

166 Taken from end of chapter (Mohan 1129). 167 "Poetry consists of words and meanings that possess excellences and are free from blemishes. In some cases it lacks poetic ornaments." (kārikā 4ab). 144

Comments

Normally, in Indian philosophy, one defines the object of study and then explain each

term of the definition. Mammata will not return to this task until chapter seven.

Nowhere is the tension in Mammata's attempt to combine the pre-dhvani and

dhvani schools of poetry more clearly seen than here, where Mammata reverses the

essential and accidental qualities of a poem. The definition claims that qualities and faults

comprise the essence of poetry, thus these terms should be explained first. Poetic

ornaments should be explained next. Even though these are declared unessential in the

definition, each technical term in the definition should be explained. Only then should

accidental qualities be addressed.

An adequate definition applies to all sub-species of the definiendum, including

sub-species of value (the three types Mammata turns to now are those of best, middle,

and worst). Clearly the merit of a poem is an accidental quality and, as such, does not

effect the fact that it is a poem. After all, a bad poem is still a poem. If Mammata had

wished to follow the traditional order of investigation, he would have proceeded with

chapters seven through ten, and then returned to the material of chapters four to six.

On the principle of determining what a thing is before determining its properties

compare Socrates' question, "How can I know a property of something when I don't even

know what it is?" (Meno, 71b). Many thinkers in the Indian tradition emphasized this

principle. Kumārila Bhatta is particularly famous for his strict adherence to it.

145

The first half of this chapter (though kārikā 42ab) explains eighteen different

types of suggestive poetry. In all these types, it the sentence as a whole, and not an

individual word that is suggestive. In the second half of the chapter Mammata goes

through the types again with examples in which an individual word is suggestive. As it is

easy to miss the forest for the trees in this chapter, I offer the following outline of the

types:

A. Literal meaning unintended (kārikā 24).

1. Transformed into another meaning (poem 23).

2. Entirely set aside (poem 24).

B. Literal meaning intended.

B1. Psychological sequence unnoticed.

3a. Suggests rasa (poems 25-26 and 30-44).

3b. Suggests pseudo-rasas, emotions, etc. (poems 45-53).

B2. Psychological sequence noticed.

B2.1. Based on the power of the word.

4. Suggesting an ornament (poems 54-57).

5. Suggesting a fact (poems 58-59).

B2.2. Based on the power of the meaning.

B2.2.1. The suggestor is self-existent.

6. A fact suggests a fact (poem 60).

7. A fact suggests an ornament (poem 61).

146

8. An ornament suggests a fact (poem 62).

9. An ornament suggests another ornament (poem 63).

B2.2.2. The suggestor is invented by the poet.

10. A fact suggests a fact (poem 64).

11. A fact suggests an ornament (poem 65).

12. An ornament suggests a fact (poem 66).

13. An ornament suggests another ornament (poem 67).

B2.2.3. The suggestor is invented by a character in the poem.

14. A fact suggests a fact (poem 68).

15. A fact suggests an ornament (poem 69).

16. An ornament suggests a fact (poem 70).

17. An ornament suggests another ornament (poem 71).

B2.3. Based on the power of both the words and the meanings.

18. Suggests an ornament (poem 72).

āvivakșitavācyo yas tatra vācyam bhaved dhvanau I

arthāntare samkramitam atyantam vā tiraskratam II (481)

laksaņāmūlagūdhavyangyaprādhānye saty eva avivaksitam vācyam yatra sa dhvanau iti

anuvādāt dhvanir iti jñeyah I tatra ca vācyam kvacid anupayujyamānatvād arthāntare

pariņamitam l yathā

tvām asmi vācmi vidusām samavāyo 'tra tisthati I ātmīyām matimāsthāya sthitim atra vidhehi tat II

147

atra vacanādi upadeśādirūpatayā pariņamati II

kvacid anupapadyamānatayā atyantam tiraskṛtam l yathā

upakrtam bahu tatra kim ucyate sujanatā prathā bhavatā param I vidadhadīdrśam eva sadā sakhe sukhitam āssva tatah śaradām śatam II

etad apakāriņam prati viparītalaksaņayā kaścid vakti II

(4.1. Suggestive Poetry in which the Literal Meaning Is Unintended.)

24. A literal meaning that is not intended and is either shifted to another

meaning or entirely set aside should be considered a type of suggestive poetry

(dhvani).168

A semantic unit (vācyam) wherein a hidden suggested meaning based on metaphorical

indication is predominant, and the literal meaning (vācyam) is not intended, should be

known as suggestive poetry because of the use of this word in the kārikā.

Sometimes the literal meaning turns into another meaning because it is not

proper in the context. For example:

I say to you, "Here stands a group of the learned So remain here with your mind composed." (23)

Here "saying, etc." takes the form of "advising."169

168 For Mammata's sourse for this definition, see Anandavardhana 2.1 and Abhinava's commentary in Ingalls, et al. (1990:200-212). 169 The Sanketa and Sampradāyaprakāśinī commentaries explain this suggestion by claiming that mere saying is not proper for the context, and that the speaker is hoping to help his listener with this information. The suggestion might be something like, "Take advantage and be attentive!" 170 Viparītalaksana, which I translate here as "irony," is literally "metaphorical indication by means of reversed or contrary expression." Note that this is a type of metaphoric indication, not an ornament. 148

Sometimes the literal meaning, being not suitable for the context, is entirely set

aside. For example:

What a great help! What can I say? Your kindness is widely renowned. Behaving ever like this, my friend, May you enjoy a hundred autumns. (24)

Someone says this to an enemy with irony.170

Comments

In suggestive poetry, the suggested meaning stands in a relation to the literal meaning

that is different from the relation of metaphoric meaning. As Mammata will explain,

there are many types of suggestive relations. He begins with a poem in which the

suggested meaning extends the literal meaning ever so slightly. The speaker is not only

trying to communicate information, but to suggest a course of action. The suggested

meaning refines or colors the literal meaning.

In the second poem, the metaphorically indicated meaning is the opposite of the

literal. Unlike pure metaphoric usage, the meaning arrived at by the blocking of the literal

meaning is not the main point of the verse. The verse communicates a third meaning

beyond the literal and metaphorically indicated. It suggests a certain attitude in the

speaker and a certain relation between the speaker and the hearer.

149

vivaksitam cānyaparam vācyam yatrāparas tu saḥ | (502)

anyaparam vyangyanistham I eșa ca II

(4.2. Suggestive Poetry in which the Literal Meaning Is Intended.)

25ab. Next, when the literal meaning and another meaning are both intended, there

is another type of suggestive poetry.

Another meaning refers to a suggested meaning. And this (type of suggestive poetry is

divided into two based on the way the suggestion functions):

Comments

Mammata begins his discussion of the most important type of suggestive poetry. One of

Anandavardhana's central theses was that good poetry communicates both a description

of an event or scene and an emotional tone or rasa. The description is communicated, at

least in part, literally. The rasa is suggested by the details of the literal. Mammata will

examine many different subtypes of this kind of suggestion.

ko 'py alaksakramavyangyo laksyavyangyakramah parah II (505)

alaksyeti na khalu vibhāvānubhāvavyabhicāriņa eva rasah I api tu rasas taih ity asti

kramaḥ I sa tu lāghavān na laksyate II

150

25cd. Suggestion whose psychological sequence (krama) is unnoticed (alakșa) and

suggestion whose psychological sequence is noticed (laksa).171

By "unnoticed" we do not mean that rasa is the same as the determinants, the symptoms,

and the auxiliaries. Rather, rasa exists because of them. Thus there is a psychological

sequence. But it is called "unnoticed" because of the quickness (by which the hearer

picks up the suggestion).172

Comments

Mammata explains that in some types of suggestion one is aware of the process leading

up to the suggestion while in others one is not. A complex literary figure, for example,

has to be deciphered. When this is done, we still have to figure out the artist's intention.

Mammata, following Anandavardhana, believes that by using a complex figure the artist

can suggest additional meaning. In this type of suggestion, we are aware of the sequence

that leads up to our understanding of it. On the other hand, when a poem suggests a

mood, we often are not aware of when or how it does so. We might say, for instance, that

a poem is melancholy without being able to point to any concrete part of the poem that

grounds this emotion. The first half of the remainder of this chapter addresses types of

171 Mammata is following Anandavardhana in this section. See Anandavardhana 2.2 and Abhinava's commentary in Ingalls, et al. (1990:212-213). Ingalls, Masson, and Patwardhan translate alaksakrama as "without apparent sequence." The choice of words in Sanskrit is a bit odd, but the meaning is clear. In this type of suggestion the person does not usually notice the sequential process leading to the suggested meaning. 172 Alternatively, the last line could read, "The sequence is called 'unnoticed' because it is simpler (than saying, 'usually unnoticed')." However, experiencing the sequence is elsewhere likened to piercing a hundred leaves with a single arrow. The point seems to be that it happens so quickly it is hard to notice which leaf gets pierced first. 151

suggestions whose psychological sequences are unnoticed. At kārikā 37, Mammața turns

to suggestions whose psychological sequences are noticed.

To understand the shift from suggestion to rasa, keep in mind that the paradigm

suggestion whose sequence is unnoticed is suggestion of rasa. Let us remember that rasa

was originally the mood of a play. It is notoriously difficult to say exactly why the

Ramayana fills a responsive reader with pathos or why Romeo and Juliet is romantic.

In his self-commentary, Mammata introduces technical vocabulary from drama

theory that he will explain in kārikā 27 and 28.

atra

rasabhāvatadābhāsabhāvaśāntyādir akramaḥ I

bhinno rasādyalamkārad alamkāryatayā sthitaḥ II (513)

ādigrahaņād bhāvodayabhāvasamdhibhāvaśabalatvāni | pradhānatayā yatra sthito

rasādis tatrālamkāryah yathodāharisyate l anyatra tu pradhāne vākyārthe yatrāngabhūto

rasādis tatra [guņībhūtavyangye rasavatpreyaūrjasvinsamāhitādayo 'lankārāh I te ca]173

gunībhūtavyangyābhidhāne udāharișyante II

(4.2.1 Suggestive Poetry in which the Psychological Sequence Is Unnoticed.)

Of these

173 This text is not in Mohan. Frustratingly, Mohan does not say if he has reason to cut this line.

152

26. The type of suggestion whose sequence is unnoticed consists of rasas, emotions,

pseudo (rasas and emotions), cessation of emotions, and the like. Such a suggestion is to

be embellished by poetic ornaments and is different from (a poem in which) rasa, etc. are

poetic ornaments.

By mention of "and the rest" is meant the arising of emotion, the co-existence of two

emotions, and the admixture of many emotions. Where rasa, etc. are the main purpose of

a suggestion, they are to be embellished with poetic ornaments, as will be explained and

illustrated presently. But in other poems, the literal meaning is the main purpose (of the

poem) and rasa, etc., are secondary. [In these poems there are poetic ornaments such as

rasavat, preyas, ūrjasvin, and samāhita.] This type of poem will be explained and

illustrated under the heading "Poems of Subordinated Suggestion" (chapter five).

Comments

Mammata focuses now on those types of suggestion whose psychological sequence is

unnoticed, starting with suggestion of rasa and emotion. He will turn to suggestion of

pseudo-rasa and emotions in kārikā 36b, and to the suggestion of the cessation, arising,

and mixing of emotions in kārika 36cd. Finally, he will consider suggestions whose

psychological sequence is noticed beginning with kārikā 37cd. It is important to note that

all of these types of suggestion are valued both by Mammata and the Sanskrit tradition

generally. Many of the greatest poems suggest emotions that do not correspond to the

nine rasas.

153

Some of the earlier critics (Lollata, for example) dealt with this situation by

claiming that there were an infinite number of rasas. Mammata retains the limit of nine

rasas, but accepts the poetic value of the other emotions under other titles. I will return to

this point when considering suggestion of pseudo-rasas, etc.

The pre-Anandavardhana writers on poetics were familiar with the concept of

rasa but did not consider it to be the soul of poetry. De observes that the early theorists,

"include it [rasa] in poetic ornaments or by allowing it to form an element of one of the

excellences of diction." Anandavardhana recognized that rasa can be used this way,

but claims that in the best poetry it becomes the primary focus of the poem. He writes,

"When it [suggested sense] predominates, that is the soul of suggestion."> Mammata

accepts Anandavardhana's distinction, treating poems in which the suggested sense

dominates in this chapter, and poems of subordinate suggestion in the next.

The four poetic ornaments Mammata mentions are found in Bhamaha's

Kāvyālankāra 3:5 - 10. From his examples, one can surmise that they refer to poems of

erotic love, devotion, misplaced passion, and happy coincidence. Udbhata organized

these into rasa poems, poems of other affections (devotion, friendship, filial love, etc.),

pseudo-emotions, and cessation of passion. 1 As we will see, Mammata organizes the

section on suggestive poems of unnoticed psychological sequence according to Udbhata's

system. In poems, however, in which the suggested emotion is secondary to the literal

174 De (1960:2:117) 175 Ingalls, et al. (1990:214). 176 Ingalls seems to confuse Bhämaha and Dandin on this point in his otherwise useful note on the history of these poetic ornaments in Ingalls et al. (1990:243n6). 154

meaning or to the structure of the poem, Mammata calls the suggestion an ornament and

retains the original names. Thus, if the most striking element of a devotional poem is it

form, for example, Mammata will say it has the poetic ornament of devotion, not the

suggested emotion of devotion (see chapter ten).

Mammata effectively double each of Udbhata's types by considering whether the

suggestion dominates the poem or not. This reduplication of the original system was

necessitated by Anandavardhana's division of poems into suggestive poems and poems

of secondary suggestion.

tatra rasasvarūpam āha

kāraņāny atha kāryāņi sahakārīņi yāni ca

ratyādeḥ sthāyino loke tāni cen nāțyakāvyayo II

vibhāvā anubhāvās tat kathyante vyabhicāriņaḥ I

vyaktaḥ sa tair vibhāvādyaiḥ sthāyī bhāvo rasah smrtaḥ Il (521)

uktam hi bharatena vibhāvānubhāvavyabhicārisamyogād rasanispattih iti II

(4.2.1.1.) Now, the nature of rasa is stated.

27 - 28. The causes, effects, and concurrent elements of a lasting emotion like love in

the world are called "determinants," "symptoms," and "auxiliaries" in drama and

poetry. An abiding emotion made manifest by the determinants, symptoms, and

auxiliary states is known as rasa.

155

For it is said by Bharata, "Rasa arises out of the combination of the determinants, the

symptoms, and the auxiliary states."

Comments

Mammata introduces five important technical terms of poetics. He quotes the

Nātyaśāstra, 6.31, which, however, does not mention abiding emotions. In English we do

not call Romeo's love for Juliet by a different name than Abélard's love for Héloïse. But

in Bharata's aesthetics, the first is "rasa" and the second "abiding emotion". Likewise,

Héloïse is a "cause" of Abélard's love, whereas Juliet is a "determinant" of Romeo's

(other determinants include the proper setting and time, i.e., the Capulet party or the

garden at night). The effects of love on Abélard's behavior are call "symptoms" when

found in Romeo. These are the actions undertaken for the sake of love, such as climbing

garden walls, writing letters, exchanging smiles, etc. Finally, the concurrent elements in

Abélard are called auxiliaries when found in Romeo. These include both secondary

emotions, like the jealously Romeo feels toward Paris because of his love, and

"involuntary states" that arise from being in love, like perspiration and trembling.

Bharata sets out guidelines regulating the use of these factors. For example, he

explains which symptoms and auxiliaries are appropriate to each rasa and which are

antithetical, he advises which factors can be mixed and the result of their mixing, and he

shows which rasas can be introduced in digressions without diminishing the dominant n scem both mechani rasa. Although Bharata's system can seem both mechanical and limiting, it does go a

156

long way towards explaining how a dramatic spectacle successful conveys meaning

beyond the literal and metaphoric.

"Abiding emotion" is a technical term used to designate the nine emotions (listed

below) that can be central to a poetic work. The tradition deriving from Bharata thinks

that only these nine emotions are capable of transformation into rasa. There are other

poetically viable emotions not included in this group. Poetic theorists struggle to

understand the relation between the abiding emotions and the additional emotions

capable of artist expression. Mammata will return to this topic in kārikā 35cd -36ab.

The abiding emotions are sometime called the basic human emotions. Gnoli says

of the abiding emotions, "They permanently exist in the mind of every man, in the form

of latent impression (vāsanā) derived from actual experience in the present life or from

inherited instinct, and, as such, they are ready to emerge into his consciousness on any

occasion." Aesthetics uses the supposed universal nature of these emotions to explain

how a person can sympathize with the representation of emotions he has never personally

experienced.

Mammata will examine the technical terms one by one. To treat rasa, he

summarizes the histories of rasa theory, culminating with Abhinava. Then he turns to the

"determinants," etc. Before continuing, however, it is worth quoting more fully the

Nātyaśāstra section excerpted by Mammata. It is the root text on this topic and provides

the context for the following discussion.

177 Gnoli (1968:xvi). 157

"In that connexion I shall first explain the rasas. No [poetic] meaning proceeds

[from speech] without [any kind of] rasa. Now rasa is produced from a combination of

determinants, symptoms, and auxiliary states. Is there any instance [parallel to it]? [Yes,]

it is said that as flavor (rasa) results from a combination of various spices, vegetables,

and other articles, and as the six flavors are produced by articles such as raw sugar or

spices or vegetables, so the abiding emotions, when they come together with various

other psychological states attain the quality of rasas ... It is apparent that the rasas arise

from the psychological states and not the psychological states from the rasas. For [on this

point] there are [traditional verses such as]: The psychological states are so called by the

experts in drama for they make one feel the rasas in connection with various modes of

dramatic representation. Just as by many articles of various kinds auxiliary cooked foods

are brought forth, so the psychological states along with different kinds of histrionic

representation will cause the rasas to be felt."178

We will see that the key concepts introduced by Bharata are capable of divergent

interpretations. Mammata now examines four.

etad vivrņvate I vibhāvair lalanodyānādibhir ālambanoddīpanakāranaih ratyādiko bhāvo

janitah anubhāvaih katāksabhujāksepaprabhrtibhih kāryai pratītiyogyah krtah

vyabhicāribhir nirvedādibhih sahakāribhir upacito mukhyayā vrttyā rāmādāv anukārye

tadrūpasamdhānān nartake 'pi pratīyamāno rasah iti bhattalollataprabhrtayah II (534)

178 Nātyaśāstra 6.31-35 in M. Ghosh (1967:105-107). I have modified his translation slightly. 158

(4.2.1.2. A History of Rasa Theory)

(1.) Bhattalollata and his followers explain the sūtra thus:19 Rasa is an emotion like love

that is (1) born of the determinants, i.e., a basic cause like a playful woman and

enflaming causes like gardens, that is (2) made observable by the symptoms, i.e., the

effects (of being in love) like sideways glances and embraces, and that is (3) further

augmented by the auxiliary states, i.e., concurrent elements like distractedness. Rasa is

cognized primarily in respect to someone being imitated, like Rāma, but also is also

cognized by an actor when he affects the appearance of Rāma.

Comments

Mammata begins his history, following Abhinavagupta's Abhinavabhāratī, with Bhatta

Lollata. Lollata and the second theorist Mammata presents, Srī Sankuka, are thought to

have lived in the ninth century, directly before, or contemporaneous with,

Anandavardhana.Like Abhinava, Mammata does not present Anandavardhana's

position, but rather skips to one of Anandavardhana's tenth century critics, Bhattanāyaka.

Finally, Mammata presents Abhinava's late tenth century "defense" of Anandavardhana's

theory, which, as Ingalls has argued, is substantially different from Anandavardhana's

own statement of his theory.18 Therefore, it should be clear that Mammata presents not a

history of rasa theory generally, but a partial history of the ninth and tenth century debate

179 This is a translation of "iti battalollataprabhrtayah" taken from the end of this presentation of his views. 180 Gnoli (1968:xvii) and Chari (1990:xi). 181 Ingalls, et al. (1990:18-35). 159

that culminated in Abhinava's theory.12 It should be noted that the only extant sources

for this debate are the Abhinavabharatī, and later texts based on it. In other words, we

only know the thoughts of the first three theorists through a hostile critic. When

Abhinava quotes these writers we have no way of knowing whether he is really quoting

their works or speaking for them. Although the later practice is common in Indian

philosophy, such glosses are usually very accurate. Mammata is almost undoubtedly

glossing Abhinava, although he may have additional sources for the four views he

presents.

In the Abhinavabhāratī, Lollata is reported as claiming that a rasa is an

intensified abiding emotion.1' The erotic, for example, is an intensified form of sexual

desire. The abiding emotion is intensified primarily by the artist representation of the

determinants, symptoms, and auxiliaries. The rasa is chiefly experienced by the character

represented, but is also tasted by the representing actor.

I believe that Lollata, like Bharata, writes primarily to instruct playwrights and

actors. Like Aristotle, his theory explains which elements need to be included in a

successful plot. Lollata discusses the story of Rāma's love for Sīta and claims that more

than mere sexual desire must be presented. Rāma's fervor needs to be intensified by

being developed through out the play. In a good romance, the lovers go through trials,

182 The best general introduction to this debate is Gnoli (1968). His introduction sketches the basic positions and arguments and is followed by the text and a translation of the Abhinavabhāratī, the main existent source for this subject. For a more complete history that takes into account comments by Kālidāsa and the early poetic theorists, see De (1960:2:108-138). However, De's account is limited by the fact that he, like Abhinava and Mammata, considers Abhinava's position to be the pinnacle of aesthetic thought. Like them, he tends to judge other theories by Abhinava's criteria, regardless of their stated goals. 183 Gnoli (1968:26). 160

doubts, and frustrations that augment their ardor. The climax leaves the viewers with the

impression that the couple is deeply in love and that their love will last.

Lollata's theory shows how to represent a rasa in a manner that will convince and

entertain an audience. Lollata's theory strives to teach playwrights and actors how to win

an audience. Lollata considers a play and, more specifically, the elements of a dramatic

production as causes and the audience response an effect. To achieve the desired effect, it

is best to put in place the necessary causal factors.

Lollata's view is criticized for leaving the audience out of the description of the

aesthetic experience and for claiming that the actors feel the emotion of the characters

they are representing. I feel both these criticisms miss the point of his theory. It is true

that Lollata ignores two possible topics of aesthetic theory. First, he has nothing to say

about the psychology of audience response. Why, one can ask, is an audience moved by

an artwork? While this is an interesting question, answering it is not vital to the writing of

a good play.14 Lollata shows what works, without worrying about why it works. Second,

one may investigate the psychological states typical of an actor. This too is an interesting

topic, and a complex one. Lollata claims that the actor feels the rasa, "by virtue of the

power of realization (amusamdhāna)."1 According to Gnoli, realization allows the actor

to feel the emotion without losing his self-awareness.186 Certainly more needs to be said,

but it is an acceptable start that refutes the common criticism of Lollata's theory, namely,

that the actor, besotted by love, would forget his lines.

184 Gnoli (1968:xix). 185 Gnoli (1968:26). 186 Gnoli (1968:xviii). 161

rāma evāyam ayam eva rāma iti na rāmo' yam ityauttarakālike bādhe rāmo 'yam iti

rāmaḥ syād vā na vāyam iti rāmasadrśo 'yam iti ca

samyanmithyāsamśayasādrśyapratītibhyo vilaksanayā citraturagādinyāyena rāmo 'yam

iti pratipattyā grāhye nate I

seyam mamāngesu sudhārasacchatā supūrakarpūraśalākikā drśoh I manorathaśrīr manasah śarīriņī prāneśvarī locanagocaram gatā Il

daivād aham adya tayā capalāyatanetrayā viyuktaś ca I aviralavilolajaladah kālah samupāgataś cāyam II

ityādikāvyānusamdhānabalāc chiksābhyāsanirvartitasvakāryaprakatanena ca națenaiva

prakāśitaih kāranakāryasahakāribhih krtrimair api tathānabhimanyamānair

vibhāvādiśabdavyapadeśyaih samyogat gamyagamakabhāvarūpāt anumīyamāno 'pi

vastusaundaryabalād rasanīyatvenānyānumīyamānavilaksaņah sthāyitvena

sambhāvyamāno ratyādir bhāvas tatrāsann api sāmājikkānām vāsanayā carvyamāno

rasa iti śrīśankukah II (543)

(2.) Śrī Śankuka explains Bharata's sūtra as follows:18 In the same way a painted horse

(is recognized as a horse), the actor is grasped by the cognition, "This is Rāma," which is

different from (1) a factual cognition, "Rāma is he, he is Rāma," (2) a false cognition,

"He is Rāma," suppressed in the next moment by, "He is not Rāma," (3) a cognition of

doubt, "He may be Rāma, or not," and (4) a cognition of similarity, "He resembles

Rāma.”

187 This is a translation of "itiśrīśankuka" taken from the end of the section (Mohan 543). 162

Rasa is an emotion, like love, which is being cognized as enduring. It is relished

by the trained188 a audience and is absent from the actors. The cause, effect, and concurrent

elements are made manifest by the actor by force of close study of poems like the

following and by learning and practicing his role. These three, even though artificial, are

not to be thought of as such and are known by the words, determinants, symptoms, and

auxiliary states. (By saying that rasa arises) out of the combination (of these three,

Bharata means a relation), which essentially is that of the inferred to its inferential marks.

The inferring of rasa, however, is different from the inferring of other things, because

rasa is relishable on the strength of its inherent beauty.

(Now the examples:)

She herself - the nectar in my limbs, The healing camphor ointment for my eyes, The embodiment of the glorious desires of my heart, The ruler of my life - has come into sight! (25)

Because of the gods, I am now separated from her Of trembling broad eyes. And the season of the thick, rolling, rain clouds has arrived. (26)189

Comments

Unlike Lollata, Śankuka is interested in how a play produces pleasure in the audience.

We thus move from drama theory to aesthetics. He claims that rasa is not a heightened

form of an emotion, but rather an artistic reproduction of the emotion. Abhinava

188 "Vāsanā" is translated here as 'trained' because the idea is that the spectator must arrive at the work with the proper experiences in place. I believe Sankuka simply meant the proper cultural education, but experience from past lives could also be meant. 189 Also quoted in Rudrata's Kāvyālankāra 7:29 (D). 163

represents him as claiming, "Rasa is simply a permanent state of mind, and, more

precisely, the reproduction (anukarana) of the permanent state of mind proper to the

person produced - Rāma, etc." For Sankuka, sexual desire exists in the world, while

the erotic rasa arises from its reproduction in art. The aesthetic experience arises in one

who is trained to appreciate art by means of the perception of an artistic imitation. The

rasa is located in the play, and not, contra Lollata, in the actors. The actor makes

manifest the play's rasa through his skill in imitating the emotions of the characters.

The audience perception of the rasa comes about by means of a type of inference

that is unique to art.According to Abhinava, Sankuka says, "Representation

(abhinayana), indeed, is nothing but a power of communication (avagamanaśakti) - this

power differing from the one of verbal expression."192 Śankuka comes close here to

Anandavardhana's theory of suggestion and closer still to Bhattanayaka's theory of

enlivening or revelation (bhāvanā). But Sankuka, focusing on the power of a production,

not the power of words, turns to inference as the key cognitive function.195 Abhinava

quotes Sankuka as saying that in aesthetic experience, "there is neither doubt, nor truth,

nor error. The notion that appears is, 'This is that,' not, 'This is really that.'"4 Mammata

expands this quote, giving examples of each type of cognition. In aesthetic perception, in

190 Gnoli (1968:29). 91 Gnoli (1968:xix). 92 Gnoli (1968:31). 93 On Abhinava's debt to this theory, see Masson and Patwardhan (1970:1:10). Clearly perception, the other main non-verbal means of knowledge, is not available for a general theory of poetics, for while reading a poem or play, you see words, not emotions. 194 Gnoli (1968:32). 164

other words, there is "willing suspension of disbelief" that allows the audience to fully

engage in the production in a manner distinct from normal experience.

na tātasthyena nātmagatatvena rasah pratīyate notpadyate nābhivyajyate api tu kāvye

nātye cābhidhāto dvitīyena vibhāvādisādhāranīkaranātmanā bhāvakatvavyāpāreņa

bhāvyamānah sthāyī sattvodrekaprakāśānandamayasamvidviśrāntisattvena bhogena

bhujyate iti bhattanāyakah II (560)

(3.) Bhattanāyaka explains Bharata's sūtra as follows:1 Rasa is not cognized, produced,

or manifest in either a manner unrelated to (the spectator, as in (1) above) or as a property

of oneself (i.e., the spectator)190. Rather, in poetry and drama, rasa, which is lasting, is

actualized through a special function of words (bhāvakatva) that is different from

denotation (and indication) and essentially universalizes the determinants, etc. It is

relished by a delight that is essentially repose and a conjunction of light and bliss

resulting from the dominance of the quality of goodness (sattvodreka).

Comments

Bhattanayaka was an important theorist who prefigures many of Abhinava's most

important ideas. Ingalls writes, "Most of the components of Abhinava's new theory are

195 This is a translation of "iti Bhattanāyaka" taken from the end of the section (Mohan 560). 196 This is commonly taken to refer to Sankuka, but could also refer to Anandavardhana, or an unknown critic. The view presented in neither Sankuka's nor Anandavardhana's, but does resemble Abhinava's.

165

borrowed, strange to say, from Ānanda's main critic (Bhattanāyaka)."197 Most

importantly, Bhattanāyaka believes that the emotions, determinants, symptoms, and

auxiliary states were universalized by a special power of words called bhāvakatva.198

The universalization process transforms a particular emotion and the events surrounding

it into a general emotion free of context that can be shared without personal attachment.

Not only is the audience "disinterested," but the plot is made generic, as well. Romeo's

love for Juliet, correctly appreciated, becomes generic erotic experience. Ingalls writes

that Bhattanāyaka "sees rasa not as an object to be enjoyed, but as the ongoing process of

(aesthetic) enjoyment itself." This enjoyment, Abhinava represents him as saying, "is

due to the emergent state of goodness (sattva), is pervaded by beatitude (ānanda) and

light (prakāśa), and is similar to the tasting (āsvāda) of the supreme brahman."201 As

akin to supreme mystical union, Bhattanāyaka argues that the rasa experience is not in

any particular subject (neither the character, the actor, the audience, nor some unrelated

person).

In later Indian theorizing on rasa, all rasa experience is often considered

identical. - Given his vague, mystical description of rasa experience, it is difficult to see

how Bhattanāyaka would avoid this homogenizing of aesthetic experience. As we will

197 Ingalls, et al. (1990:35). 98 Gnoli (1968:45). 199 Unfortunately, laking Bhattanāyaka's description of the generic experience, we cannot know he meant by a generic erotic experience. 200 Ingalls, et al. (1990:37). 201 Gnoli (1968:47-48). 202 For example, B.N. Goswami (1986:21) writes that the rasas, "are separately listed because even though rasa is defined as one and undivided, it is one or the other of these nine rasas through which an aesthetic experience takes place." As he makes clear on p.24, the "aesthetic experience" is precisely the experience of this single, undivided rasa. 166

see, Abhinava struggles to avoid reducing the varieties aesthetic experience to a single

type.

Bhattanāyaka claims, like Anandavardhana before him, that a poem is able to

produce rasa experience in a qualified audience by means of a special semantic power,

which he calls "enlivening" or "revelation" (bhāvanā). Abhinava argues that this power

is not different in any important way from the semantic power that Ananda calls

suggestion. With only Abhinava's summary to go on, this seems a valid point. But as

Ingalls observes, there was probably more to Bhattanayaka's theory than Abhinava

reports.203

loke pramadābhiḥ sthāyy anumāne 'bhāsa pātavavatām kāvye nātye ca tair eva

kāraņatvādiparihāreņa vibhāvanādivyāpāravattvād alaukika vibhāvādi śabda

vyavahāryair mamaivaite śatror evaite tatasthasyaivaite na mamaivaite na śatrorevaite

na tatasthasyaivaite iti sambandhaviśesasvīkāraparihāraniyamānadhyavasāyāt

sādhāraņyena praśatrorevaite na tatasthasyaivaite iti sambandhaviśeșasvīkāraparihāra

niyamānadhyavasāyāt sādhāranyena pratītair abhivyaktah sāmājikānām vāsanātmatayā

sthitah sthāyī ratyādiko niyatapramātrgatatvena sthito 'pi sādhāranopāyabalāt

tatkālavigalitaparimitapramātrbhāvavaonmișitavedyāntara-

samparkaśūnyāparimitabhāvena pramātrā sakalasahrdayasamvādabhājā sādhāraņyena

svākāra ivābhinno 'pi gocarīkrtaś carvyamānataikaprāno vibhāvādijīvitāvadhiḥ

pānakarasanyāyena caryamāņah pura iva parisphuran hrdayam iva praviśan

203 Ingalls, et al. (1990:36). 167

sarvāngīņam ivālingan anyat sarvam iva tirodadhat brahmāsvādam ivānubhāvayan

alaukikacamatkārakārī śrngarādiko rasah I sa ca na kāryaḥ l vibhāvādi vināe 'pi tasya

sambhavaprasangāt l nāpi jñāpyah siddhasya tasyāsambhavāt l api tu vibhāvādibhir

vyañjitaś carvanīyah | kārakajñāpakābhyām anyat kva drstam iti cet na kvacit drstam ity

alaukikasiddher bhūsaņam etan na dūsaņam I carvaņāniśpattyā tasya niśpattir

ūpacariteti kāryo 'py ucyatām | laukikapratyaksādipramānatātasthyāvaboddhaśali-

mitayogijñanavedyāntarasamsparśarahitasvātmamātraparyavasitaparimitetarayogi-

samvedanavilaksanalokottarasvasamvedanagocara iti pratyeyo 'py abhidhīyatām |

tadgrahakam ca na nirvakalpakam vibhāvādi parāmarśapradhānatvāt l nāpi savikalpam

carvyamāņasyālaukikānandamayasya svasamvedanasiddhatvāt

ubhayābhāvasvarūpasya cobhayātmakatvam api pūrvavallokottaratām eva gamayati na

tu virodham iti śrīmadācāryābhinavaguptapādāh II (574)

168

(4.) The great teacher Abhinavagupta explains Bharata's sūtra as follows:204

An abiding emotion like love exists in a cultured spectator as a latent trace of

previous experience (vāsanā). In common life, clever people repeatedly infer that

women, and the like, cause an abiding emotion. In poetry and drama these causes are

referred to with technical terms, such as the determinants, because they can bring about

the effects without the normal causal conditions. The determinates, etc., are not cognized

as bound by a principle of acceptance or rejection of a specific relation as in: "These are

mine," "These are my enemy's," "These are belonging to a neutral person," or "These are

not mine," "These are not my enemy's," "These are not belonging to a neutral person."

Thus the abiding emotion is manifested by the cognitions (of the determinants, etc.) as

universal. Though the emotion is restricted to the individual spectators (niyatapramātr),

by the power of the universalizing process it exists as universal in a spectator who (1) has

knowledge of poetics, (2) has his limited perceiverhood swallowed up for a period of

time, (3) and whose emotions are not delimited, being free from connection with

anything other than what is to be known and manifest (in the artwork/drama).

(The universalized emotion of sexual desire becomes) the erotic rasa, which has

its own form, even though it is not distinct from the spectator. The life of (a rasa

experience) is one with (the experience of artistic) relishing and thus its life span is

determined by (the existence) of the determinants, etc. (In this) it is analogous to the

flavor of a beverage (which lasts only as long as the beverage itself). (The rasa appears)

204 This is a translation of "iti śrīmadācāryābhinavaguptādāh" at the end of the section (Mohan 574).

169

as if sparkling before the eyes, entering the heart, embracing the whole person, and

eclipsing everything else. (Rasa experience is) akin to the experience of the enjoyment of

Brahman and causes extraordinary wonder. Rasa is not an effect, because that would

entail the false proposition that its existence would be possible after the determinants,

etc., disappear. Nor is it cognized, because it does not exist as an independent entity.

Rather it is relished as manifested by the determinants, etc.

Objection: "Where does one see that which is different from the material causes

and (the ideas that) cause knowledge?" Reply: "It is found nowhere." Given the

extraordinary nature (of rasa), this is a merit, not a fault.

By the production of the relishing, there is manifestation of the production of the

rasa, so it can be figuratively called an effect. Rasa can even be called an object of

cognition because it is within the scope of a paranormal consciousness which is distinct

from (1) the normal means of knowledge, such as perception, etc., (2) the knowledge of

the limited yogin, which is indifferent (to the normal objects of cognition), (3) and the

self-contained knowledge of the unlimited yogin, which is free from contact with any

other object of knowledge.

Furthermore the grasping of rasa is not a case of non-conceptual perception

because of the necessity of first assimilating the determinants, etc. Neither is it a case of

conceptual perception because the rasa experience is a self-consciousness of the

extraordinary bliss of relishing (the rasa). Furthermore, from its nature as negation of

170

both and yet having the character of both, we infer its extraordinary nature, as above, and

not that the claims are inconsistent.

Comments

Abhinavagupta is a complex and fascinating thinker whose thought is only touched upon

here. A full presentation of his aesthetic theory can be found in the Abhinavabhāratī and

in his commentary on Anandavardhana, which both have excellent English

translations.205

Abhinava accepts Bhattanāyaka's theory of universalization, both of the

determinants, etc., and of the abiding emotion. He maintains that the universalized

emotion must be perceived, not inferred, by the individual spectator. He provides three

conditions that the spectator must meet to be fit for the rasa experience: he must be

trained in the arts, he must be in a state of willing suspense of disbelief and self-

forgetting, and he must be disinterested. These conditions are necessary to enter the state

of sympathetic union that Abhinava describes. He further agrees that rasa is an

experience, not an object, affirming that rasa exists only in the spectator.

Abhinava also agrees that the rasa experience is akin to the mystic union with

Brahman. However, he differentiates the aesthetic from the yogic by insisting that the

aesthetic experience has an object, namely the artwork. This is very important, for it

provides Abhinava a way of avoiding Bhattanāyaka's homogenizing of aesthetic

205 Gnoli (1968) and Ingalls, et al. (1990). Masson and Patwardhan (1969 and 1970) are both invaluable. The secondary literature on rasa far outweighs that on any other topic in Sanskrit poetics. 171

experience. Unfortunately, modern theorists tend to ignore, even to misrepresent, this

merit of Abhinava's theory.

When one tries to understand just what type of experience aesthetic experience is,

the complications begin. Abhinava carefully distinguishes it from other forms of

experience. He insists that it is neither a perceptual experience nor a purely cognitive

experience, although it shares qualities with both of these. Furthermore, although it is a

paranormal experience, it is not a yogic experience. Abhinava further claims that the rasa

experience is different from non-conceptual and conceptual perceptual experience. He

may be referring to the experience of raw sense data (nirvikalpaka pratyaksa) and

concept-laden experience (savikalpaka pratyaksa). If so, Abhinava has argued that all

aesthetic experience occurs by means of a psychological sequence, so its content cannot

be momentary sense data. On the other hand, its nature as self-consciousness of bliss

precludes conceptual perception. Ultimately Abhinava's description has the effect of

urging the reader to attend to his own aesthetic experience to appreciate its unique

character.

As should be clear from Mammata's gloss, this failure of reason to explain rasa

does not bother Abhinava. Indeed, he considers it a good thing that art is "beyond" all

rational explanation. However, upon reading his works, one discovers that Abhinava is

not an irrationalist. He seeks to establish rasa experience as a separate form of experience

that has qualities distinct from the forms of experience identified by the philosophers. His

position is finely argued with great attention to minute distinctions. Furthermore, he

172

points out many difficulties facing anyone desiring an explanation of aesthetic experience

that reduces it to one of the other types.26

Mammata, who seems ever in awe of Abhinava, is content to allow him the final

word on rasa. Rather than attempting to reproduce the details of Abhinava's lengthy

arguments, Mammata focuses on Abhinava's description of the experience (rasa appears,

"as if sparkling before the eyes, entering the heart, embracing the whole person, and

eclipsing everything else"). This slightly erotic description ties in nicely with the earlier

claim that poetry teaches like a lover. Mammata neither explains Abhinava nor

attempts to improve upon him; instead, he passes on to other technical terms within

Bharata's sūtra.

vyāghrādayo vibhāvā bhayānakasyeva vīrādbhutaraudrāņām aśrupātādayo 'nubhāvāḥ

śrngārasyeva karunabhayānakayoh cintādaye vyabhicāriņah śrgārasyeva

vīrakaruņakānām iti prthag anaikāntikatvāt sutre militā nirdistāh II

Such things as tigers are determinants, as it were, with respect to the heroic, the

marvelous, and the furious rasas, as well as to the frightful. Shedding tears and such are

symptoms of the pathetic and frightful rasas, as well as the erotic. Reflection is an

auxiliary state connected with the heroic, the pathetic, and the frightful rasas, as well as

206 There were writers who did, mostly from the Logicians. Chari (1963) tries to give a rational explanation of rasa in modern analytic terms. 207 Chapter one, kārikā 2. 173

the erotic. These three are pointed out together in the sūtra because separately they are

not connected to one (rasa).

Comments

Having treated rasa, Mammata explains the other technical terms of kārikās 27-28 by

means of examples. The crux is the phrase, "because separately they are not connected to

one" (prthag anaikāntikatvāt). In the context of this passage, it seems to mean that a

determinant is not limited to being a determinant for only one rasa. The same is true for a

symptom and an auxiliary state. In the context of the following passage, it seems to mean

that no determinant, symptom, or auxiliary state can bring about a rasa experience

without the other two. Both claims are part of Mammata's poetics.

viyadalimalināmbugarbhameghyam madhukarakokilakūjitair dirśām śrīh l dharaņir abhinavānkurānkatankā pranatipare dayite prasīda mugdhe II

ity ādau

(Objection: Not all verses contain all three elements stated in the sūtra - the

determinants, etc. For example,) the following verse and others like it:

The swarming swollen monsoon clouds are beautiful! The coos of the cuckoos too are beautiful! Beautiful also is the buzz of the honeybees. Beauty herself has come - sashed with the earth's tender shoots. Come along too, sweet girl, be pleased with your adoring lover. (27)208

208 Indian vegetation is reborn each year with the monsoon. Before it arrives, all is parched and brown. The rainy season is also the traditional season of romance. 174

Comments

An objector points out that not all rasa poems contain the three elements supposed

necessary to rasa experience. Indeed, the objector gives examples of good poems in

which only one of the three factors is present. The first example shows only the

determinants of the erotic, i.e., a pair of lovers, a beautiful setting in nature, and the

season of spring. No symptoms, such as romantic interaction (sidelong glances, holding

hands, etc.) are described.Neither are any auxiliary states mentioned. This seems to

prove that the explanation in kārikā 27-28 is incorrect insofar as it states that rasa arises

from all three factors together.

parimrditamraņālīmlānam angam pravrttiḥ katham api parivāraprārthanābhi kriyāsu I kalayati ca himāmśorniskalankasya lakșmīm abhinavakaridantacchedakānta kapolaḥ I

ity ādau

The following verse and others like it (state only the symptoms. They mention neither the

determinants nor the auxiliaries.)

Her weakened limbs are threshed lotus stems. Barely, by the urging of friends, she manages to move, Her cheeks gleaming like spotless ivory, Lit only by the cold-rayed moon. (28)210

209 One might object that the lover's invitation to his beloved is a symptom. 210 Mālatīmādhava 1.25 (D). 175

Comments

This poem focuses on the symptoms of separation. The woman has grown thin, weak,

and pale. No determinants are described (except, of course, the woman herself). Neither

is any emphasis placed on the auxiliaries. One could argue that several auxiliaries, as

depression, exhaustion, and lifelessness, are described. However, when these states arise

as a direct result of the abiding emotion, they are included as symptoms. A proper

auxiliary in this case would be, for example, anxiously looking about, or fits of anger. In

so far as the poem focuses only on the symptoms, it acts as a second counter-example to

the claim that all three factors must be present.

dūrād utsukam āgate vivalitam sambhāsiņi sphāritam samślisyaty arunam grhītavasane kimcāñcitabhrūlatam māninyāś caraņānativyatikare bāspāmbupūrņeksanam caksurjātam aho prapañcacaturam jātāg asi preyasi II

ity ādau ca

The following verse and others like it (state only the auxiliaries.)

How the eyes of a jealous woman work to capture a lover who has strayed! They search about when he's far, but look elsewhere when he draws near, Grow wide when he speaks, and redden when he reaches out for her. They arch their creeper-brows as he touches her clothes, But flood with tears as he falls at her feet. (29)211

211 Amaruśataka 49. Also found in the Subhāșitaratnakoșa 641, where it is attributed to Ratipāla. Ingalls interprets this poem differently (1965:513). 176

Comments

Finally, a third counter example that presents only the auxiliaries, and neither the

determinants nor the symptoms. Because a woman is a determinant for the erotic, the

example is not perfect (although only her eyes appear in the poem, not the woman

herself). Among the auxiliaries shown are suspicion, envy, pride, joy, and confusion. The

complete list of auxiliaries is given below in kārikās 31-34.

yady api vibhāvānām anubhāvānām autsukyavrīdāharsakopāsūyāprasādānām ca

vyabhicāriņām kevalānām atra sthitih tathāpy etesām asādhāranatvam ity

anyatamadvayāksepakatve sati nānaikāntikatvam iti II

Reply: Even though the determinants, the symptoms, and the auxiliary states - such as

eagerness, embarrassment, joy, anger, carping, reconciliation - stand alone in each of

these examples, because of the interdependence of their natures, the other two of the three

are inferred. Thus our definition does not succumb to counterexamples.

Comments

Mammata nicely shows how to adapt the traditional drama theory to poetry. In drama, of

course, all three factors are easily presented, often more than once. In a short poem, the

available space can make the presentation of all three factors problematic. Furthermore, a

poem may gain in power by deliberately focusing on only a single element. Mammata

177

overcomes these problems by explaining that while only a single factor may be

mentioned, the other two are invariably inferred from what is directly communicated,

taken together, we might point out, with considerable practical and literary background

knowledge. Only when the necessary inference has occurred, can one experience rasa.

From the literal content of poem 27, for example, we infer symptoms, such as the lovers

looking at one another, conversing romantically, the man's urging, and the woman's shy

reluctance. We also infer auxiliary states, such as rashness, joy, excitement, pride, and

longing. Indeed, Mammata would claim that if we fail to infer this we miss the sense of

the poem and it seems unlikely that it will produce a rasa experience.

Mammata's adaptation, however, weakens considerably the explanatory power of

the theory. By no longer pointing exclusively to factors objectively present in the

artwork, Mammata opens his theory to the threat of subjectivism. As we have seen, the

critics believe that the aesthetic response to a given artwork is uniform across cultured

audiences. This allows them to maintain that the factors that produce the suggestion are

objective despite the fact that some of them are inferred. Surprisingly, no major figure

objects to the claim of uniform aesthetic response.

It is remarkable how differently the theory functions in the two art forms. As we

saw in the introduction (section 3), the rasa is developed in a play by means of a series of

reinforcing scenes. Layer upon layer, it gains intensity by showing the heroes'

consistency in different situations. The poem, on the other hand, focuses on a single

178

moment, concentrating all its force to create an image striking enough to suggest the rasa

all at once.

tad viśeşān āha l

śrngārahāsyakaruņaraudravīrabhayānakāḥ I

bībhatsādbhutasamjñau cety astau nātye rasāḥ smtā Il (641)

(4.2.1.3.) The types of rasa are now given:

29. The erotic, the comic, the pathetic, the wrathful, the heroic, the frightful, the

disgusting, and the wondrous are called the eight rasas of drama.212

Comments

This list of eight comes from Bharata's Nātyaśāstra 6.15. The ninth rasa, added to the

system early in the tradition, is discussed in kārikā 35ab. The list purports to contain all

the emotions that can be the central theme of a play or an epic poem. We have already

seen that when this theory is applied to all poetry, including isolated verses, a great

number more emotions have to be admitted. Rudrata argued that any emotion, if

developed by determinants, etc., can become a rasa." g 213 Coupled with the view that the

emotions are infinite in number, this leads to the view, attributed to Lollata by Abhinava,

212 Ingalls translates the eight rasas as: the erotic, the comic, the tragic, the furious or cruel, the heroic, the fearsome or timorous, the gruesome or loathsome, and the wondrous (Ingalls, et al. 1990:16). Gnoli translates: the erotic, the comic, the pathetic, the furious, the heroic, the terrible, the odious, and the marvelous (Gnoli 1968:xvi). 213 Rudrața, Kāvyālankāra 12:3-4, discussed in Raghavan (1967:127). 179

that the rasas are infinite.214 Mammata seems to take the line put forward by

Pratīhārenduraja that the original list should be maintained for tradition's sake and the

other emotions dealt with under different catagories.215 Jagannatha takes this view to the

extreme by claiming that the original system should be retain for convenience's sake.216

As the nature of the individual rasas and the additional artistic emotions only become

clear with illustration, I will discuss them separately below.

tatra śrngārasya dvau bhedau I sambhogo vipralambhaś ca l tatrādyah

parasparāvalokanālinganādharapānaparicumbanādy anantatvād aparicchedya eka eva

ganyate l yathā

śūnyam vāsagrham vilokya śayanād utthāya kimcicchair nidrāvyājam upāgatasya suciram nirvarņya patyur mukham | vistrabdham paricumbya jātapulakām ālokya gandasthalīm lajjānamramukhī priyeņa lasatā bālā cire cumbitā II (653)

Here the erotic is divided in two: love-in-enjoyment and love-in-separation. The first of

these is counted as one only and is not to be broken into types because (the possibilities),

such as exchanging glances, embracing, drinking the lips, kissing the entire body, are

unlimited. For example:

Seeing the inner rooms empty, She sits up slowly and carefully in the bed To study the face of her husband. Believing him asleep, she kisses his cheek.

214 In the Abhinavabhāratī, vol. 1, 346. 215 Raghavan (1967:129). 216 Raghavan (1967:142). 180

Seeing the kissed hair quiver, She hangs her head, abashed. After only a moment, her smiling lover pulls her near. (30)

181

Comments

Poems producing the erotic rasa are divided in two: those in which the lovers are able to

realize their love and those in which they are frustrated by some obstacle. Mammata

declines to introduce further subcategories of love-in-enjoyment, claiming they would be

endless. He will give the subcategories of love-in-separation below.

The poems of love-in-enjoyment tend to be charming in their innocent affirmation

of love in all its forms. Always sexual, but rarely overtly descriptive of sexual acts, the

poems are celebrations of this essential activity. The poem here shows a shy young wife

attempting to express her love without waking her husband. He pretends to be asleep to

enjoy her affection, but the hairs on his cheek rise in excitement, giving him away. He

allows her a moment of shame for her forwardness, but this only increases his

excitement. So he takes her in his arms.

The poem includes the three factors that evoke rasa. The determinants are the

lovers and the bedchamber. The symptoms include such acts as the woman's kissing her

lovers cheek and the lovers pulling her close. Shyness is an auxiliary state when present

in a woman and the hair standing on end is a classic involuntary state for both the erotic

and the frightful rasas. These details combine together, like the details of a painting, to

present a powerful image of a moment of realized love.

tathā

tvam mugdhāksi vinaiva kañculikayā dhatse manohārinīm lakşmīm ity abhidhāyini priyatame tadvīțikāsamsprśi |

182

śayyupāntanivistasasmitasakhīnetrotsavāndito niryātah śanakair alīkavacanopanyāsamālījanāh II (662)

"O sweet-eyed girl! You would be ravishing Even without this tight corset." So saying, the lover starts toying with its knot. Their friends, thrilled with the passion in the eyes Of the smiling couple edging toward the bed, Slip out, giving welcome excuses. (31)217

Comments

Mammata gives a second example of the innocent pleasures of erotic love. The poet

cleverly mixes in determinants (the busty woman and the bedroom), symptoms (amorous

speech), and auxiliaries (anxiousness). For more examples of this genre in translation, see

Merwin and Masson's Sanskrit Love Poetry, Bailey and Gombrich's Love Lyrics, and

Martha Selby's Grow Long, Blessed Night, are all excellent introductions.218

aparas tu abhilāsavirahersyāpravāsaśāpahetuka iti pañcavidhih | krameņodāharanam I

premārdrāh pranayasprśah paricayād udgādharagodayāstās tā mugdhadrśo nisargamadhurā cestā bhaveyur mayi I yāsv antahkaraņasya bāhyakaraņavyāpātarodhi ksanādāśamsāparikalpitāsv api bhavaty ānandasāndro layah II (667)

The other (love-in-separation) has five types: longing, estrangement, jealousy, dwelling

apart, and love blocked by a curse. An example of each in order:

217 Amaru 27. 218 Merwin and Masson (1977); Bailey and Gombrich (2005); Selby (2000). 183

Her soft eyes light up with love, Fanned by intimacy, they smolder with desire. Immediately my senses fail, My mind is consumed with bliss, Simply dreaming that she is here, loving me. (32)219

Comments

The critics are divided about the sub-types of love-in-separation. Most list four, but argue

over their natures. Mammata is alone in including "separation caused by jealousy" in the

list.220 It is also notable that Mammata does not include the sorrow of separation caused

by death (karunavipralambha). Many include separation caused by a curse under the

category separation caused by dwelling apart, for this is the normal result of the curse

(but consider Sakuntalā). Regardless of the sub-types, the separation must include a

chance of reunion to be included in the erotic rasa. Permanent separation evokes the

pathetic rasa.

In the poem, a man dreams of the woman he deeply desires. By exalting the

pleasure of imagining, instead of the realization, the poet suggests a far greater pleasure

in realization. It is interesting that this genre is not limited to male longing. Consider the

following poem, "When he comes back / to my arms / I'll make him feel / what nobody

ever felt / everywhere / me / vanishing into him / like water / into the clay of a new

jar."221 This contrasts nicely with the young bride's shyness in poem 30.

219 Found in the Mālatīmādhava 5.7 (D). The Sanskrit poem focuses on the actions of the fair-eyed girl, which I could not successfully bring into English. 220 Raghavan (1967:171). 221 Translated by Merwin and Masson (1977:107). 184

anyatra vrajatīti kā khalu kathā nāpy asya tādrk suhrd yo mām necchati nāgataś ca hahahā ko 'yam vidheh prakramaḥ | ity alpetarakalpaņākavalitasvāntā niśāntāntare bālā vrttavivartanavyatikarā nāpnoti nidrām niśi II

esā virahotkanthitā II (677)

(Next an example of estrangement:)

"He's gone off somewhere. What's the story? All his friends are on my side. Yet still, he hasn't come. O damn it! What's happening?" Sleepless, alone in her bed, the girl feels abandoned. And little by little, Her tormenting doubts Devour her. (33)222

This woman is longing for an estranged lover.

Comments

Unlike the first type of longing, this genre treats estranged lovers. The main character is

always a woman who has quarreled with her lover and isolated herself. The poems

describe both the anger of the woman and her longing for reconciliation.225 Most often,

the focus is on the physical suffering the woman feels because of the separation from her

lover. Here is a second example from the Subhāsitaratnakosa: "Ah, cursed moon, touch

not even in jest / my wasted body, burning with the heat / of separation from my dear

222 Some compromises have been made to make the girl's words to herself sound believable to modern ears. Literally the third line reads, "Alas! What new move of destiny is this?" 223 Ingalls et al. (1990:264n2). 185

one, / for when your rays fall on me, / though fair as ripened lotus stalks, / they hurt me

like burning brands."24 Here the moon reminds the woman of the man she longs for.

sā patyuh prathamāparādhasamye sakhopadeśam vinā no jānāti savibramāngavalanāvakroktam sasūcanam | svacchair icchakapolamulagalitaih paryasta netrotpalā bālā kevalam eva rodhiti luthal lolālakair aśrubhih II (681)

(Next an example of jealousy:)

When her husband first strays, She has no friend to teach her The armaments and insults of betrayal.

She merely weeps, Her lotus eyes filling with crystal tears Which trickle down her bright cheeks And roll into her flowing curls. (34)225

Comments

This genre expresses the feelings of a woman confronting competition. The poems focus

either on her rage and sorrow of a betrayed woman (as in this poem) or on the beauty her

anger lends. Here is a charming example of the later from the Subhasitaratnakosa: "The

same side-stepping of her glance and unclear words, / the same shrugging me away when

224 Translated by Ingalls (1965:234 poem 714). 225 Amaruśataka 29. 186

I embrace her body; / again the contradicting of every word and stubborn shaking of her

head: / my wife by anger has become a bride again."226

Male jealousy is not proper, so it is relegated to either pseudo-rasa, or the comic.

The point is not that men do not feel jealousy, rather, it is that such jealousy is not a

proper theme for erotic art. Men were expected to have affairs and doing so did not lessen

their worth. Unfaithful women, on the other hand, were scorned. One can feel jealous

when another has what one wants, and nobody wants an unfaithful wife. One who does

is ridiculous, not erotic, as in Molière's Tartuffe.

Many have commented on the essentially optimistic nature of Indian poetry and

its lack of tragedy. But their poetics is not without resources for analyzing such plays.

Certainly, the male ego is little comforted by scene like Othello, Act 3, Scene 3, when

Iago is first seeding the grains of doubt, "O curse of marriage, / that we can call these

delicate creatures ours, / and not their appetites." The Indian theorist could treat this play

as evoking the frightful rasa, from the perspective of Desdemona, or the pseudo-heroic,

from the perspective of Othello. In either case, it would be seen less as the downfall of a

great man and more as showing human emotion gone frightfully wrong.

prasthānam valayaih krtam priyasakhair asrair ajasram gatam dhrtyā na ksanam asitam vyavasitam cittena gantum pura yātum niścitacetasi priyatame sarve samam prasthitā gantavye sati jīvita priyasuhrtsārthah kim u tyajyate II (686)

226 Ingalls (1965:218 poem 637). 227 Think of Rama's concerns over Sīta's purity, for example. 228 J.A.B. van Buitenen (1974:82-83). 187

(Next an example of travel:)

My bracelets are gone, Tears, my sweet friends, have freely fallen, My courage stayed not an instant, And my heart went first of all. When my love decided to leave, Everything left together. O my life, if you must go to hurry! Why fall behind your dear friends? (35)229

Comments

The man must leave his beloved, but neither wants to part. The woman is so distraught

that she want to die.

tvām ālikhya pranayakupitām dhāturāgaih śilāyām ātmānam te caranapatitam yāvad icchāmi kartum asrais tāvan muhur upacitair drśtir ālupyate me krūras tasminn api na sahate samgamam nau krtāntaḥ (690)

(Next an example of love blocked by a curse:)

On a stone I painted you Pretending to be angry. I wanted to put myself, too, Falling at your feet. But I could no longer see through The flowing of my tears. Even here pitiless fate Will not let us be together. (36)230

229 Amaruśataka 35. 188

Comments

This genre depicts "a pair of star-cross'd lovers" held apart by a curse. This example

comes from near the end of Kālidāsa's long poem (mahākāvya), The Cloud Messenger.

The lover, "who had grossly swerved from his duty and was deprived of his greatness by

his lord's curse ... " sends a cloud home to his lonely wife bearing a message.251 The curse

keeps the lovers apart for a whole year. Part of the message the lover sends home is this

description of his sorry state and the power of the curse to separate them even in painting.

This theme is very popular in Sanskrit literature, as it is in world literature

generally. The curse is a good literary method of separating two lovers without either of

them incurring any blame. Cursed lovers are also found in the epic poems (e.g., Rāma

and Sīta) and in the classical drama (e.g., Duhsanta and Śakuntala).

hāsyādīnām krameņodāharanam

ākuñcya pāņim asucim mama mūrdhniveśyā mantrāmbhasām pratipadam prśataih pavitre | tārasvanam prathitathūtkamadāt prahāram hāhāhato 'ham iti rodhiti viśņuśarmā II

Now for examples of the comic, etc., in order:

Drops of water made holy by sacred verses May purify my head,

230 Meghadūta 2.45. 231 Meghadūta 1.1, translated by M.R. Kale (1969:2).

189

But that whore clenched her dirty hand And give it a loud smack anyway. How wretched am I, Visnu the Protector! (37)

Comments

The speaker of the poem is named "Visnu," people being commonly named after gods in

India. There is a humorous contrast of the speaker's namesake and situation in the poem.

Pretending to actually be the god Visnu, the speaker mocks his own "sanctity" and feigns

helplessness to the onslaught of a prostitute.

Bharata claimed the comic arises from mimicry of the erotic.232 It is hard to

believe that this could have been the sole source of comedy for classical drama. Indeed,

Bharata was criticized for this view before Mammata's time.253 In poetry, as we can see

in Mammata's example, there is no such restriction. Three centuries after Mammata,

Viśvanātha claims that the comic is, "whatever a person laughs at, when he beholds it

distorted in respect of form, speech, or gesture."234

hā mātas tvaritāsi kutra kim idam hā devatāh kvāśisah dhik prāņān patito 'śanir hutavahas te angesu dagdhe drsau ittham ghargharamadhyaruddhakarunāh paurānganānām giraś citrasthān api rodayanti śatadhā kurvanti bhittīr api II (701)

(Next an example of the pathetic:)

232 Nātyaśāstra 6.40. 233 By Bhoja (11th century) according to Ghosh (1967:107). 234 Śahityadarpaņa 228, translated in Ballantyne and Dāsa Mitra (1875:122). 190

Mom! Where are you running? What is going on? Life is hell! Lightning strikes. Your arms are on fire, your eyes scorched. O god, where is thy protection? These lamenting cries of the village women, Half muffled by their moans, Brake the walls and make the painted statues weep. (38)235

Comments

Poems that express the pathetic describe situations of extreme sorrow, usually due to

physical injury or death. Viśvanātha says the pathetic arises from the loss of what is

loved.236 He adds that the corpse of the loved one is a good determinant for this rasa. In

the poem the speaker describes her dying mother. Poem 116 in chapter five is another

good example. Modern usage would call these situations tragic, and they do arouse

Aristotelian "pity and fear." However, they lack some of the elements of tragedy that

Aristotle discusses in the Poetics, most notably the plot development from a state of

ignorance to recognition. Furthermore, there is no "tragic flaw" in the characters of

Sanskrit poems - a concept that came to be essential to later western tragedy

(Shakespeare and Racine, inter alios). For these reasons, I find it more accurate to call

235 By Narayana Bhatta (D). But is he confusing this with the following poem? 236 Śahityadarpana 230, translated in Ballantyne and Dāsa Mitra (1875:123). 191

this genre, following Gnoli, "the pathetic."237 The poems certainly do arouse sympathy

and pity.

krtam anumatam drstam vā yair idam gurupātakam manujapaśubhir nirmaryādair bhavadbhar udāyudhaih I narakaripuņā sardham tesām sabhīmakirītinām ayam aham asṛnmedomāmsaih karomi diśām balim II (705)

(Next an example of the wrathful:)

Krsna, Bhīma, and Arjuna! With weapons drawn you violated the bounds of morality. I will offer to the four quarters of the earth Your blood, fat, and flesh - all of you Who committed, accepted, or witnessed this outrage! (39)238

Comments

The wrath of the speaker shines forth in his intention to revenge the death of Drona,

wrongfully killed by those who owed him respect. The story of Battanārāyana's play

comes from the Mahābhārata. Bhīma has sworn to revenge the outrage done to Draupadi

by killing Duhsāsana. In the family feud that follows, Bhima and his brothers are forced

to kill their teacher, Drona, provoking the wrath expressed in the poem.

ksudrāh samtrāmete vijahata harayah ksunnaśakrebhakumbhā

237 Gnoli (1967:xvi). Remembering that the rasa experience, although it is enjoyed by the spectator, still has as its object the artistic expression of the abiding emotion of the characters in the artwork, one has to reject a spectator based translation like, "the sympathetic." 238 Veņīsamhāra 3.24. 192

yuşmaddehesu lajjām dadhati paramamī sāyakā nispatantah | saumitre tista pātram tvam asi na hi rupām nanv aham meghanādah kimcid bhrūbhangalīlāniyamitajaladhim rāmam anveșayāmi II (709)

(Next an example of the heroic:)

O you useless monkeys, control your fear! My arrows have pierced the forehead of Indra's elephant, They would be ashamed to end up in you. Son of Sumitra, stay where you are, You are not worth my wraith. Rāma subdues the very ocean with a mere play of his brows, I, Meghanāda, seek only him. (40)239

Comments

Rāma has led an army of monkeys and bears across the ocean to Sri Lanka to fight the

demon army of Rāvaņa. Here Rāvaņa's son Meghanāda seeks a worthy foe and scorns

those unequal to him. We also find this genre displayed in Henry IV, part I (act 5, scene

4), where the prince challenges Hotspur to single combat ("Two stars keep not their

motion in one sphere"). Not only the foe but also the occasion must be correct for the

heroic. Compare Achilles' unwillingness to fight Hector at the beginning of the Iliad.

grīvābhangābhirāmam muhur anupatati syandane baddhadrstiḥ paścārdhena pravistah śarapatanabhayād bhūyasā pūrvakāyam darbhair ardhāvalīdaih śramavivrtamukhabhramśibhih kīrnavartmā paśyod agraplutatvād viyati bahutaram stokam urvyā prayāti II (716)

239 Dwivedi writes, "Hanumannātaka?" 193

(Next an example of the frightful:)

The antelope turns slightly in a graceful manner Glancing back at our speeding chariot, Then it tucks its haunches in and bounds off for fear of a falling arrow. Its panting mouth scatters half-eaten grass onto the track. Look! Jumping so high, it moves more through the sky than on the earth. (41)

Comments

Viśvanātha writes that this genre displays a woman or a mean person in fright or panic.240

As such it is close to the pathetic, but here there is no lasting damage. From the present

example, we see that portraying the fear of an animal is also possible. This verse from the

Abhijñānaśākuntala portrays King Duhsanta out hunting at the beginning of the play. The

king thrills in the chase, clearly enjoying the fear of his prey.

The emphasis of the verse is clearly on the king's pleasure at seeing the terrorized

animal, not on the animal itself. Viśvanātha's description and his example also indicate

that pleasure in another's fear is the object of this genre. He quotes the following verse

from the Ratnävali that mocks the fear of base people: "The eunuchs fled, having

abandoned shame, because of their not being counted among men; the dwarf, in terror,

ensconces himself within the loose and wide trousers of the chamberlain; the

mountaineers, the guardians of the bounds, act in a style accordant with their name (i.e.,

the scatter); while the hump-backs, fearing they may be seen by the monkey, cowering

240 Śahitadarpaņa 235, translated in Ballantyne and Dāsa Mitra (1875:126). 194

down, slink quietly off."241 The reader is supposed to share this mixture of mockery and

scorn for the fear caused by a mere monkey.

Given that it is much more natural to identify with the king, if the reader actually

does so, the poem is not really frightful, for the king is not in a state of fright. The abiding

emotion coinciding with the frightful rasa is supposed to be fear, not pleasure in the fear

of others. This rasa serves two distinct ends in Sanskrit literature, as far as I know. First,

the fear of the base acts as a foil to the valor of the great and is thus used to develop the

heroic rasa. Second, the fear of women is thought to increase their beauty and is thus

used to develop the erotic rasa. These are roughly parallel to the fear of common people

and in early westerns, "The Outlaw," for example. Unfortunately we do not have a whole

drama devoted to this rasa that could help illuminate Bharata's original idea. Ironically,

the modern suspense novel, in which the reader shares the protagonist's fear, fits the

definition better than the verses given here.

utkrtyotkrtya krttim prathamam atha prthūtsedhabhūyāmsi māmsānyamsasphikprsthapindyādyavayavasulabhānyagrapūtīni jagdhvā ārttah paryastanetrah prakatitadaśanah pretarankaḥ karankādankasthādasthisamstham sthaputagatamapi kravyam avyagramatti II (720)

(Next an example of the disgusting:)

Having butchered and stripped the skin, And devoured the swollen and putrid lumps of flesh Easily torn from the shoulders, hips, and buttocks, The hungry ghost bares its teeth and looks about,

241 Ballantyne and Dāsa Mitra (1875:127). 195

Drawing out the tendons, guts, and eyes, Then deliberately picking out the meat that remains In the joints and bones of the corpse in his lap. (42)242

Comments

This genre is a component in the modern genre, "horror." These poems focus on the

decaying human corpses and the ghosts and vultures that feed on them. They are intended

to shock the reader and fill him with disgust. Viśvanātha list the determinants as,

"stinking flesh and fibre and fat."23 Death is among the auxiliaries, but note that death

itself is not disgusting, but rather the consumption of a decaying corpse. Death figures in

many other genres (the wrathful, heroic, and pathetic). Vidyākara has a small section

devoted to burial grounds in his Subhāsitaratnakosa that contains this verse and thirteen

other examples of this rasa.

citram mahān esa batāvatārah kva kāntir esābhinavaiva bhangih I lokottaram dhairyamaho prabhāvaḥ kāpy ākrtir nūtana eșa sargah II

(Next an example of the wondrous:)

How marvelous! Superb! O, what a divine incarnation! How lovely indeed, this novel sight! What extraordinary poise! What splendor! How shapely indeed is this new creature! (43)

242 Mālatatīmādhava 5.16. 243 Śahitadarpana 236, translated in Ballantyne and Dāsa Mitra (1875:127). 196

Comments

The unexpected, irrational, or unexplainable can evoke a sense of wonder. Viśvanātha

claims that anything supernatural can cause wonder.44 However, here and in the

Nāgānanda (1.13 +1) the wonder is inspired by artistic creation.24 Abhinava provides

one of his own verses as an example in which the poet wonders at an old man's stupidity:

"This is not mere age; the angry spitting snake of Death sprinkles his white venom on this

man's head. He sees this and still seem happy of heart and does not seek refuge in Śiva:

ah, how brave he is!"20 The old man, being so close to death, should have thoughts of the

afterlife. The poet is amazed that the man is oblivious to his fate.

esām sthāyibhāvān āha |

ratir hāsaś ca śokaś ca krodhotsāhau bhayam tathā l

jugupsā vismayaś ceti sthāyibhāvāḥ prakīrtitāḥ Il (728)

spastam |

(4.2.1.4) The abiding emotion of each rasa is stated:

30. The abiding emotions are as follows: sexual desire, mirth, sorrow or grief, anger,

heroic energy, fear, disgust, and wonder.

The meaning of the verse is clear.

244 Śahitadarpana 237, translated in Ballantyne and Dāsa Mitra (1875:128). 245 Ingalls, et al. (1990:519-520). 246 Ingalls, et al. (1990:700). I have modified the translation slightly.

197

Comments

Mammata continues by quoting the Nātyaśāstra 6:17, which follows the rasa verse

Mammața quoted in the previous kārikā (29). Having given examples of each rasa that

make clear the nature of the abiding emotions, he apparently feels that additional

comment is unnecessary. The list, however, is not without problems.

The first problem is that often the characters in the poems do not feel the emotion

that grounds the rasa. For example, we saw that a jealous woman can ground the erotic

even though she herself feels no sexual desire. Likewise the characters in humorous

poems are rarely themselves amused. Unlike descriptions of love or heroism, the origin

of the emotion in these poems is in the external reaction to the scene. The poet judges the

jealous woman sexy or the clumsy person funny and the spectators agree. In wonder we

saw one case (Mammata's example) where we share the wonder of the character and one

case (Abhinava's example) where we wonder at the character's stupidity.

In the case of fear the same problem of audience reaction occurs. However, we

also saw that with fear the audience is not supposed to enjoy a universalized essence of

fear, but rather the spectacle of another's fear.

Unfortunately the original eight rasas and their abiding emotions have received

very little attention in modern scholarship. Most writers merely list the rasas, provide a

few examples, and then move on to other points. Raghavan provides a starting point the

198

investigation by listing the subcategories suggested for each rasa.24 A full historic

investigation of each rasa with many examples is sorely needed.

nirvedaglāniśankākhyās tathāsūyā madaśramāḥ I

ālasyam caiva dainyam ca cintā mohah smrtir dhrtih II

vrīdā capalatā harșa āvego jadatā tathā I

garvo vişāda autsukyam nidrāpasmāra eva ca II

suptam prabodho 'marś cāpy avahittham athogratā l

matir vyādhis tathonmādas tathā maraņam eva ca II

trāsas caiva vitarkaś ca vijñeyā vyabhicārinah I

trayamstraśadamī bhāvāḥ samākhyātās tu nāmataḥ II (738)

nirvedasyāmangalaprāyasya pathamam anupādeyatve 'py upādānam vyabhicāritve 'pi

sthāyitābhidhānārtham I

31 - 34. The thirty-three auxiliary states are known by name as: world-weariness,

exhaustion, suspicion, envy, intoxication, fatigue, indolence, depression, worry,

confusion, remembrance, peace of mind, shame, rashness, joy, excitement,

lifelessness, pride, dejection, anxiety, sleeping, apoplexy, dreaming, awakening,

resentment, dissimulation, violence, attentiveness, sickness, insanity, dying, fright,

and perplexity.

247 Raghavan (1967:chap. 7). 199

Of world-weariness, which is generally inauspicious and thus not to be mentioned first, it

is listed (first) in order to denote that it is an abiding emotion even though it is listed in

the auxiliary states.

Comments

If you read this kārikā directly after the proceeding one, you see that this list begins

where the other leaves off. It is the same in the Nātyaśāstra, where these verses occur as

6:19-22. We will discuss the ninth rasa and its abiding emotion at greater length below.

Here I would like to focus on the argument for world-weariness being the abiding

emotion for tranquility. This argument predates Abhinava, who discusses it in his

commentary on Bharata." The problem confronting anyone who argued for an

additional rasa was finding textual evidence for the abiding emotion in the Nātyaśāstra.

There is a seemly corrupt version of the Nātyaśastra that lists the ninth rasa and an

abiding emotion (śama), but many critics either did not know of this version or did not

accept it. The argument Mammata repeats tries to locate world-weariness in both lists: by

means of its location it can be read as the final item on the proceeding list and the first

item on the following. The argument claims that this explains why world-weariness is

listed first among the auxiliaries, which would otherwise be inauspicious. Raghavan

248 Abhinavabhāratī (1956:1:334).

200

objects that world-weariness was not considered inauspicious and that this reveals the ad

hoc nature of the argument. Abhinava also rejects this argument.250

Pundits are very reluctant to put themselves above their teachers or the sages of

earlier generations. Probably out of deference to Bharata, Mammata accepts the attempt

to explain away the difference. It should be considered more a sign of respect than a flaw

in reasoning. Concerning the ninth rasa, Mammata appears unaware of Abhinava's

theories. This could lead one to speculate that Mammata knew Abhinava's commentary

on Anandavardhana but not his commentary on Bharata (in which his theories of the

ninth rasa are developed). For whatever reason, this is the only subject about which

Mammata does not follow or even discuss Abhinava's theories.

tena l

nirvedasthāyibhāvo 'sti śānto 'pi navamo rasaḥ | (778)

yathā l

ahau vā hāre vā kūsumśayane vā drsadi vā manau vā loste vā balavati ripau vā suhrdi vā l trne vā straine vā mama samadrśo yānti divasāh kvacitpunyāranye śiva śiva śiveti pralopatah II

(4.2.1.5. The Ninth Rasa.) Accordingly,

35ab. The ninth rasa is tranquility, whose abiding emotion is world-weariness.

249 Raghavan (1967:79). 250 Translated by Gerow (1994:195-196).

201

For example:

A snake or pearls around my neck, Flowers or rock slab for my bed, A jewel or a stone, A deadly foe or a friend, A woman or a mere blade of grass: They are all the same for me. In a sacred grove I pass my days Chanting and chanting softly, "Siva, Śiva, Śiva" (44)251

Comments

Mammata adds the ninth rasa (and it corresponding abiding emotion) to Bharata's list of

eight (given above in kārikā 29). It was probably introduce soon after Bharata, perhaps to

account for the dominant mood of the Mahābhārata, and perhaps because of religious

works such as Aśvaghosa's long poem (mahākāvya), the Buddhacarita (first century).

Made famous by Anandavardhana, the ninth rasa was already accepted by many

writers.252

The poem shows an ascetic who has no concerns for material good. He has left

the world and enjoys the equanimity of liberated experience. His peace is perfect; nothing

alters his internal state. The crux of this poem is the "kvacit" in the final line. I have taken

it to mean that the speaker moves about, from grove to grove, thinking that his snake, etc.

are just as good as a string of pearls, etc. However, it could also be taken to suggest that

the speaker lives in the world (with pearls, etc.), but in his indifference he is like an

251 Dwivedi writes, "By Utpala, Vairāgyaśataka?" 252 Raghavan explores the origin and reception of the ninth rasa at length (1967:1 -103). Bhattacarya's more recent book on this subject adds many datail and examples (1976). 202

ascetic in some grove. Jha resolves this by making the whole poem a wish on the part of

the speaker, but I do not see how this reading is possible without an optative verb.

Abhinava disagrees with the theory put forth here in his commentary on the

Nātyaāstra. He argues at length that world-weariness cannot be the abiding emotion of

tranquility and proposes the essential self (atman) instead. As can easily be imagined, this

gives the rasa of tranquility a privileged place in his system. For Abhinava it grounds and

makes possible all the other rasas. As we have seen, he takes the parallel between the

experience of the rasa of tranquility and the experience of Brahman very seriously.

Indeed, Masson and Patwardhan claim that Abhinava based his aesthetics on the

philosophy of his Kāśmīri Saiva religious tradition.25 Gerow argues convincingly that

this view needs to be questioned. He proposes that Abhinava's aesthetics predates and

informs his metaphysical writings, not the reverse.4 He shows that Abhinava's

conception of the self as active self-awareness allows both involvement with the artwork

and detachment from individual emotions.255

ratir devādi vișayā vyabhicārī tathāñjitaḥ II

bhāvaḥ proktaḥ (789)

ādiabdān munigurunrpaputrādivisayā l kāntāvisayā tu vyaktā śrngārah l udāharaņam |

kanthakonavinivistam īśa te kālakūtam api mahāmrtam I

253 Masson and Patwardhan (1970:32-33). 254 Gerow (1994:186-188). 255 Gerow (1994:199-201, especially the notes).

203

apy upāttam amrtam bhavad vapur bhedavrtti yadi me na rocate II

haraty agham samprati hetur esyatah ubhasya pūrvācaritaih krtam śubhaih I śarīrabhājām bhavadīyadarśanam vyanakti kālatritaye 'pi yogyatām II

evam anyad apy udāhāryam I

(4.2.1.6. The Emotions, Auxiliaries, Pseudo-Rasas, Etc.)

35cd - 36a. Love, when directed towards a god, etc., and the auxiliary states are

called "emotions" when suggested.

By etc. is meant (love) directed toward a sage, a teacher, a king, a son, etc. Love of a

woman, on the other hand, evokes the erotic (rasa). Examples:

O Lord! When sheltered in the curve of you neck, Even poison is the nectar of eternal life. But eternal life apart from your heavenly body, Has for me no taste. (45)

Your mere existence, O sage Narada, Destroys present evil, Brings about future rectitude, And testifies to virtues past. Your very presence Proves the worthiness of earthly beings In all the three times. (46)256

The others (i.e., love for a teacher, etc.) could similarly be exemplified.

Comments

Mammata address the other affections (preyas) here. Traditionally they are held to be

four: love of friends (rudrata), love of family (vātsalya), love between a social superior

256 I have followed Jha in identifying the great sage Nārada as the object of this poem. 204

and inferior (prīti) and love of a god (bhakti).257 Each of these affections can be (and

actually had been) developed into a full drama, and as such each could ground a tenth

rasa. Mammata here denies that this should be done, claiming instead that the rasas

should be limited to nine. Mammata gives no justification for the seemingly arbitrary

decision to include one new rasa (tranquility) and exclude the others.258 258 Clearly the

criterion cannot be duration, as the technical term "abiding emotion" suggests, for major

works of poetry had been written evoking these emotions (e.g., the Gītagovinda). As we

saw above, there was a movement in Sanskrit poetics to open the list of rasas to all

poetically developable emotions that was resisted in favor of the tradition.

Mammata, like all major late critics, restricts the number of emotions that can

become rasas and includes suggestion of the other emotions under other headings. So

whereas Mammata's system is usually descriptive and inclusive, here it is prescriptive

and exclusive. However, Mammata does not deny the value of the poems. They remain in

the highest category - poems in which suggestion is dominant. Indeed, one can see that

the concept of rasa is not absolutely central in the late poetics.

Although I cannot defend my theory here, I believe that rasa theory, when

incorporated into poetics, helped develop the theory of suggestion, and that suggestion,

being the more useful concept to criticism, soon outgrew rasa. Of the seventy-nine

suggestive poems that Mammata gives in this chapter, only eighteen are cited as

257 I do not understand why Mammata uses rati and not preya to discuss these affections. While literally rati can mean affections of the type discussed, its sexual connotations and its use as the abiding emotion for the erotic rasa make it seem a poor choice. 258 Raghavan discusses this issue at length (1967:119 - 162).

205

examples of rasa. The others are used to exemplify other types of suggestion. Indeed,

Mammata will tell us at the end of this chapter that there are 10,455 types of suggestion.

After Anandavardhana, suggestion, not rasa, is central to the discussion of the nature of

poetry and the quality of individual poems.

Chari presents the opposite view. He writes, "Rasa is the most important concept

in Sanskrit criticism and one that is central to all discourse about literature ... rasa is the

relishable quality inherent in an artistic work - which ... is its emotive content."259 Chari

express the common view that the concept of rasa covers all that can be suggested in an

artwork. We see here that this claim is historically false. The concept of rasa failed to

grow into a comprehensive theory of poetic emotive content.0 Mammata gives many

examples of emotionally rich poems that do not fall under of the rubric of rasa. Let us

return to these.

añjitavyabhicārī yathā l

jāne kopaparānmukhī priyatamā svapne 'dya drstā mayā mā mām samsprśa pānineti rudatī gantum pravrttā purah I no yāvat parirabhya cāțuśatakair āśvāsayāmi priyām bhrātas tāvad aham śathena vidhinā nidrādaridrīkrtah II

atra vidhim pratyasūyā l

An example of a suggested auxiliary state (becoming the dominant element):

O my brother! Today I saw in dreams

259 Chari (1990:9). 260 Rasa did become central to Indian aesthetics outside of poetics (in music, for example). 206

The face of my best beloved averted in anger. "Don't touch me" she wept and turned to go. I went to comfort her with endless entreaties and embraces, But just then sly Fate robbed me of my sleep. (47)

Here grumbling bitterly against fate is suggested.

Comments

The auxiliary state of frustration in the poem dominates the rasa (the frustrated erotic) by

means of the striking striking personification of fate in the last line. The trope of the first

four lines portrays a lover pretending to reunite with his beloved in a dream. Although it

is farely cliché in Sanskri poetry, it can be used to good effect. Take the following

poem for example: "Have mercy, sleep, / and show me once again / my darling though it

be but for a moment; / for when I see her / I shall hold her in my arms so tightly / she

shall not go, or if she goes / she must take me too."202 In this poem the focus falls

squarely on the lover's passion for his beloved. In the poem Mammata gives, however,

the image of the sly noctural robber holds more interest than the beloved. One is

distracted from the lover's sentiment towards his beloved by such questions as, "What

would Fate look like and how does he steal the lover's sleep?" Mammata correctly points

out that this poem suggests more about the character of the lover than his love.

tadābhāsā anaucityapravartitā | (806)

261 See poems 762, 763, and 779 in the Subhāsitaratnakosa for more instance of this trope. 262 Subhāsitaratnakoșa 806. Translated by Ingalls (1965:251). I added "she" to the final line, being unable to read it otherwise. 207

tadābhāsā rasābhāsā bhāvābhāsāś ca l tatra rasābhāso yathā l

stumah kim vāmāksi ksanam api vinā yam na ramase vilebhe kah praņān ranamakhamukhe yam mrgayase | sulagne ko jātah śaśimukhi yam ālingasi balāt tapahśrīh kasyaisā madananagari dhyāyasi tu yam II

atrānekakāmukavisayam abhilāsam tasyāh stuma ityādyanugatam bahuvyāparopādānam

vyanakti lI

36b. Manifestations (of rasas and emotions) that are deficient are pseudo-

occurrences of them.

By pseudo-occurrences of them is meant pseudo-occurrences of rasa and pseudo

occurrences of emotion. Here is an example of a pseudo-occurrence of rasa:

Whom should we praise, O fair-eyed one, such that without him you are not happy even for a moment? Who readily offered his life on the altar of war, such that you seek him now? Who was born at an auspicious moment, O moon-faced one, such that you embrace him so strongly? Who has such a treasure of religious merit, O incarnation of Love, such that you meditate upon him so profoundly? (48)

Here the many activities of the woman are inferred by the phrases, "Whom should we

praise," etc. These activities suggest a desire for more than one lover.

Comments

Indian philosophy resorts to the concept "pseudo" to describe a close imitator of what is

being investigated. Thus in logic there are "provers" (hetus) which are sufficient to

208

establish the desired inference and "pseudo-provers" (hetu-ābhāsas) which appear to be

provers, but fall short under close scrutiny. Abhinava uses the common example of

seeing pseudo-silver that turns out to be mother-of-pearl.263 Likewise in poetry, there are

emotions that seem transformable into rasas, but fail to do so on closer inspection. To

take a contemporary example, we often read of the "courage" of people who blindly enter

danger. This would count as pseudo-courage, because although the action would be

courageous if the agent were aware, the necessary awareness is absent.

Raghavan suggests that the category of pseudo-rasa may have been invented to

handle Ravana's illicit passion for Rama's wife Sīta.264 Abhinava cites an earlier critic

who claimed that Ravana's lust should be counted as a new rasa (laulya), with the

abiding emotion covetousness (gardha).2 Abhinava rejects this view, claiming instead

that until we realize the inappropriateness of an artistically represented emotion, we

experience it as rasa, after the realization, however, our experience changes according to

which emotion is portrayed.200 Abhinava cites Bharata's view that the erotic leads to the

comic in cases like these. However, it is hard to understand what is funny about the

lust that leads to the abduction of another's wife.

The rasa in the poem is not genuine because the man's love for a woman of loose

character is improper and thus not erotic (according to classical sensibilities). Whatever

one thinks of this moral judgment, the general point that the perverse is no longer erotic

263 Ingalls, et al. (1990:217). 264 Raghavan (1969:125). 265 Abhinavabhāratī (1956:1:342). 266 Ingalls, et al., (1990:107). 267 Nātyaśāstra 6.40. Abhinava seems to be forcing Bharata a bit here. 209

seems correct. But then again, what is the dominant mood of the opening parts of Lolita if

not the erotic?

bhāvābhāso yathā l

rākāsudhākaramukhī taralāyatāksī sā smerayauvanatarangitavibhramāngī tat kim karomi vidadhe katham atra maitrīm tatsvīkṛtivyatikare ka ivābhyupāyah II

atra cintā anaucityapravartitā l evam anye 'py udāhāryāh II

Here is an example of a pseudo occurrence of an emotion:

Her face resembles the full moon, her eyes are large and trembling, Her body quivers with budding youth. What should I do? How can I win her affection? (49)

Here "worry" is improperly employed. Similarly the other emotions could be illustrated.

Comments

The auxiliary emotion of worry (number nine on the list from kārikā 31 - 34) is not

proper for a man to feel towards a woman until a relationship has been established, and

thus this is not a genuine case of worry.208 While Abhinava specifically denies that

auxiliary emotional states can be suggested, Mammta chooses several examples of

suggested auxiliaries (e.g., poems 47, 50, and 51).269

268 Jha (1925:73). 269 Ingalls, et al. (1990:217). 210

It is worth repeating that while Mammata is denying that these poems actually

suggest the emotion or rasa they seem to suggest, he is not denying that they are

suggestive poems of the highest order. Indeed, the Subhāsitaratnakosa, an anthology of

short verse compiled in Mammata's time, shows a great interest in verses that suggest

emotions other than the nine that can become rasa.Mammata might be reflecting a

larger awareness of the broad powers of suggestion.

bhāvasya śāntir udayah samdhiḥ śabalatā tathā Il (818)

krameņodāharaņam I

tasyāh sāndravilepanastanatatapraśleșamudrānkatam kim vaksaś caranānativyatikaravyājena gopāyyate l ity ukte kva tad ity udīrya sahasā tat sampramārstum mayā sāślistā rabhasena tatsukhavaśāt tanvyā ca tadvismrtam II

atra kopasya II

(Additional ways a poem can be suggestive.)

36cd. There is pacification, arising, co-existing, and mixing of emotions.

Examples will be given in order (starting with pacification):

"Are you falling at my feet just to hide your chest? Did the pressure of some girl's rounded breast, thick with sandal-paste, leave marks?" "Do you see any?" I ask and quickly embrace her, vigorously rubbing the traces away.

270 Ingalls puts its date at "shortly before the year 1100." Mammata is believed to have written in the second half of the eleventh century (see Introduction). Vidyākara includes a whole section on adulterous women and myriad other poems that fail the standards of rasa. 211

Enraptured, the slender one forgets all about them. (50)271

Here (there is pacification) of anger.272

Comments

Mammata lists more elements that can be suggested in a poem. He moves from a static

suggestion of a rasa or an emotion to the suggestion of emotive processes. We can see

here a subtle understanding of both human emotions and how they can be suggested.

In the poem, anger subsides to sexual desire. According to V.K. Chari, this can

only be the case if the anger is weaker than the desire, for if they were equal, anger would

win out.273 This verse could easily be worked into a larger framework and used to help

suggest the erotic rasa. However when it is considered in isolation, it is the calming of

the woman's anger that dominates.

ekasmiñ śayane vipaksaramanīnām agrahe mugdhayā sadyo mānaparigrahaglapitayā cāțūni kurvann api I āvegād avadhīritah priyatamas tūsņīm sthitas tatksanam mā bhūt supta ivety amandavalitagrīvam punar vīksitaḥ II

atrautsukyasya II (823)

(Next an example of the arising of emotion:)

When he whispers the name of her rival in bed,

271 Amaruśataka 26. 272 I am reading kopa as a synonym of krodha. 273 Chari (1990:68).

212

The sweet girl withers and turns away. Furious, she ignores her lover's coaxing. But when he stops, she thinks, "He mustn't fall asleep." And quickly turns back to face him. (51)274

Here (the arising) of anxiety.

Comments

"Anxiety" is one of the auxiliary states (number twenty on the list from kārikā 31 - 34).

Here the girl is anxious that if her lover falls asleep without reconciliation she will lose

him. Despite the righteous anger of the woman, in a larger context the dominant rasa

would be the erotic. There are so many Sanskrit poems about lovers slipping up and

mentioning other women in bed that a separate section in often given to them in the

anthologies!275

utsiktasya tapahparākramanidher abhyāgamād ekatah satsamgapriyatā ca vīrarabhasotphālaś ca mām karșatah I vaidehīparirambha eșa ca muhus caitanyam āmīlayan ānandī haricandanendusisir asnigdho rūnaddhy anyatah II

atrāvegaharsayoḥ I (826)

(Next an example of co-existing of two emotions:)

The arrival of one so rich in austerity and penance Urges me on to a flowering of heroism And to the joy of good company. But the blissful embrace of the Videha princess,

274 Amaruśataka 23. 275 Ingalls (1965:217). 213

Sticky as a drop of yellow sandal paste in early spring, Envelopes and captures my very soul. (52)276

Here (the co-existence) of enthusiasm and joy.

Comments

As Masson points out, "vacillation between love and asceticism" is a very common

theme in Sanskrit poetry. He quotes Jagannātha's: "I have drunk deep of the Upanisads

and the Bhagavadgītā has traveled the highway of my mind. And yet, from out of the

house of my heart, she will not go."2/8 Mammata seems right that this type of poem

possesses a charm distinct from the charm of poems expressing a single rasa.

kvākāryam śaśalaksmaņa kva ca kulam bhūyo 'pi drśyeta sā doşāņām praśamāya na śrutam aho kope 'pi kāntam mukham kim vaksyanty apakalmasāh krtadhiyah svapne 'pi sā durlabhā cetah svāsthyam upaihi kah khalu yuvā dhanyo 'dharam dhāsyati II

atra vitarkautsukyam atismaranaśankādainyadhrticintānām śabalatā l bhāvasthitis tūktā

udāhṛtā ca II (830)

(Next an example of the mixing of several emotions)

What a shameful act for my lunar race! But to see her again! I learned the scriptures to extinguish all faults,

276 Bhavabhūti, Mahāvīracarita 2.22. The princess of Videha is, of course, Sītā. 277 Merwin and Masson (1977:4). 278 Merwin and Masson (1990:4), slightly altered. 214

Yet even in anger her face was beautiful! What would my faultless teachers say? But she is so rare, even in dreams! Be calm, my heart. (Try not to wonder) Which happy youth will drink her lips? (53)279

Here are mixed together deliberation, longing, resolve, recollection, doubt, dejection,

self-control, and anxiety.

Existence of an emotion has already been discussed and illustrated (in kārikā 35c-

36a).

Comments

This category is an extension of the previous one. Notice that the theme is similar: a

battle between the "higher" and "lower" emotions. The man knows he should leave the

woman, but is too besotted to do so. Rather than simple vacillation, the poet evokes

several allies on each side. The auxiliaries of the erotic - longing, recollection,

dejection, and anxiety counter the auxiliaries of the tranquil - deliberation, shame,

doubt, and resolve. The poet skillfully leaves the result unstated, focusing instead on the

man's torment. Ingalls reports that the poem is about a king who falls in love with a

brahmin woman despite the caste restrictions placed on such a relationship.280

mukhe rase 'pi te 'ngitvam prāpnuvanti kadācana | (835)

te bhāvaśāntyādayah I angitvam rājānugatavivāhapravrttabhrtyavat II

279 Quoted in the Dhvanyāloka 2.3. 280 Ingalls, et al. (1990:219).

215

37ab. Even while rasa is primary, these sometimes gain predominance.

The word These means the pacification of emotion, etc. Predominance means like that

of a royal subject who, during his marriage, is followed by the king.

Comments

Mammata very briefly explains the relation between the suggestion of rasa and

suggestion of non-rasa producing emotive states by analogy. Just as the king retains his

authority even when allowing his subject to occupy the place of honor, so the rasa is

primary, even when the suggestion of one of the other emotive states dominates a poem.

The crux is understanding what is meant by "primary." While Abhinava had stretched

Anandavardhana's system to acknowledge these types of suggestion, he still considers

them subservient to the development of rasa in a larger context.8 Although Mammata

follows Abhinava's ranking, his point seems to be that while suggestions whose

sequences are unnoticed are paradigmatically suggestions of rasa, many other such

suggestions are possible. Rasa suggestion dominates the field, even though there are

times when it cedes power to some other type of suggestion.

Here ends the lengthy discussion of suggestion whose psychological sequence is

not noticed. Mammata has considered many types of suggestion that arise without the

reader being able to pin down exactly why. They could be called suggestions of mood.

281 Ingalls, et al. (1990:117 and especially 215-217). 216

He will now move on to suggestion that require a deliberate mental process on the part of

the reader. Roughly put, in these suggests the reader becomes aware that the poet has

created some semantic or stylistic effect and when he figures out why, he has understood

the suggestion. Hopefully this will be made clearer and more precise by the discussion

and examples that follow.

anusvānābhasamlakşyakramavyangyasthitis tu yah II

śabdārthobhayaśaktyuthas tridhā sa kathito dhvaniḥ | (839)

śabdaśaktimūlānuraņanarūpavyangyah arthaśaktimūlānurananarūpavyangyah

ubhayaśaktimūlānurananarūpavyangyaś ceti trividhah I

(4.2.2. Suggestive Poetry in which the Psychological Sequence Noticeable.)

37cd - 38ab. Suggestions wherein the psychological sequence (connecting the

suggested meaning with the literal meaning) is noticeable - like an echo - are said to

be of three types: those arising from the power of the word, the meaning, or both.

The three types are echo-like suggestions based on the power of the word, echo-like

suggestions based on the power of the meaning, and echo-like suggestions based on the

power of both.

Comments

217

These suggestions are called echo-like because one understands the literal meaning first

and only afterwards the suggested meaning, hearing, as it where, the same thing

differently. An example (in English) would be the pun, "He gave me a laurel and hardy

handshake." You understand the literal meaning first, and then the punning meaning hits

you. The sounds remain the same (as in an echo), but the meaning changes. The

independence of the two levels of meaning accounts for the sequence in understanding.

You can flip back and forth between them, but you cannot read both at once (like

Wittgenstein's "rabbit-duck").

tatra

alamkāro 'tha vastv eva śabdād yatrāvabhāsate II

pradhānatvena sa jñeyaḥ śabdaśaktyubhdāvo dvidhā l (844)

vastv eveti analamkāram vastumātram l ādyo yathā

ullāśya kālakaravālamahāmbuvāham devena yena jarathor jitagarjitena I nirvāpitah sakala eva rane ripūņām dhārājalais trijagati jvalitah pratāpah II

atra vākyasyāsambaddhārthābhidhāyaktvam mā prasānkșīd iti

prākaranikāprākaranikayor upamānopameyabhāvah kalpanīya ity atropamālankāro

vyangyah II

(4.2.2.1. Suggestive Poetry Based on the Power of the Word.)

Of these (three types of sequential suggestion),

218

38cd - 39ab. Suggestions based on the power of the word are known to be of two

types, based on whether a poetic ornament or a mere fact is made manifest by the

word principally.

Mere fact means the fact only, without a poetic ornament. To illustrate the first (i.e.,

where a poetic ornament is made manifest):

Apparent Meaning Hidden Meaning The king The king Brings up the large cloud raises his sword Dark and fresh with a fierce roar And extinguishes and destroys With torrents of rain with its flickering edge The heat of destructive fires the glory of enemies Pervading the three worlds. extending over the three worlds. (54)282

Here a relation of simile is understood between the subject of the sentence (Indra) and

another object (the king) in order to avoid an irrelevant sentence meaning. Thus the

suggested meaning is a poetic ornament, in this case a simile.

Comments

Please recall that I have used two columns to render the apparent and the hidden

meanings for each phrase in the Sanskrit original of punning poems. One should read the

left poem first, then the right, and then try to understand how they can be translations of

the same text. Although this method has no poetic merit, at least the reader should be able

to understand the examples.

282 I am indebted to Jha (1925:77-78) for this reading of this, and the following, punning poems. 219

This and the following five poems can be read literally, so there is no reason to

invoke metaphoric indication. But the literal meaning does not mesh with the context,

which pushes the attentive reader to search for another meaning. The first four poems

suggest poetic ornaments, while the final two suggest facts. These suggestions arise from

the words, because in each case the effect would be ruined if synonyms were substituted.

The word "fact" is being used by Mammata to indicate any state-of affair

described in a poem. The opposite category is not that of fiction, but that of ornament.

Thus the heat of the sun and Romeo and Juliet's love are "facts." On the other hand, the

claim that the Pope is the shepherd of the faithful is an ornament (a metaphor), not a

"fact." The metaphor suggests the "fact" that he leads and protects the faithful. It is

important to keep this in mind for the rest of this chapter.

The current example describes Indra, the Vedic god of rain and king of the

heaven. Indra is a symbol of generous heroism for several reasons. First, his rain both

extinguishes forest fires and allows the earth to produce food. Second, Indra uses his

thunderbolt to fight the demons of darkness, thus protecting mankind. The poem ties in a

third aspect of Indra in the final line. In Vedanta Indra is equated with the Supreme

Being.23 Thus, in Dvaita Vedanta at least, he could be called on for aid in quenching the

fires of desire that keep one from realizing his identity with Brahman.

Although the apparent reading is coherent, it fails to fit the context. The poem is

about a king in battle, and his army raising huge rain-clouds, etc., is not literally possible.

283 Monier Williams (1899:166).

220

But when one understands that the king's conduct in battle is being compared to Indra's

conduct in fighting the harmful fires, the relevance becomes clear. The comparison is

never, however, directly stated. Rather, it is suggested by the attributing actions to the

king that are usually connected with Indra.

tigmaruchirapratāpo vidhuraniśākrd vibho madhuralīlaḥ I matimān atattvavrtti pratipadapaksāgraņīr vibhāti bhavān II

atraikaikasya padasya dvipadatve virodhābhāsah II (857)

Apparent Meaning Hidden Meaning284 O lord! Your splendor is fierce and sweet, Fierce splendor without splendor. You destroy your enemies, The moon and not the night maker Your actions are pleasing, The spring and devoid of beauty You act with intelligence and magnanimity, Intelligent and acting blankly, You are brilliant at the front of your army. The first day of the fortnight and not the first day. (55)

By taking each compound in this poem as two words, there is "pseudo- paradox."

Comments

Each line in the above literal translation translates one Sanskrit compound by reading the

compound as a single phrase. However, if you read the compound as two phrases, they

284 The compounds should be divided as follows for the apparent contradiction: tigmaruchirapratāpo becomes tigmaruchir + apratāpo; vidhuraniśākrd becomes vidhur + aniākrd; madhuralīlah becomes madhur + alīlaḥ; matimānatattvavrttih becomes matimān + atattvavrttih; pratipadapaksāgraņīr becomes pratipad + apaksāgraņīr. 221

contradict one another in each case. "Pseud-paradox" is a poetic ornament in which the

apparent paradox is dissolved once one understands the pun. This example is odd in that

the paradoxical reading is more difficult to arrive at than the non-paradoxical one. The

following poem has two more orthodox "pseudo-paradoxes." Although the degree of

virtuosity evidenced here is rarely appreciated in European criticism, this sort of poem

has an honored place in Sanskrit poetry.

amitah samitah prāptair utkarșair harsada prabho I ahitaḥ sahitaḥ sādhuyasobhir asatām asi II

atrāpi virodhābhāsaḥ | (862)

In wars for peace my lord excelled, Honorably renowned for kindly killing the wicked. (56)

Here too there is "pseudo-paradox."

Comments

I have incorporated two pseudo-contradictions into this loose translation (war/peace and

kindly/killing). The fact that the first two words of each line of the original seem to

contradict each other is quite apparent ("a" is a privative prefix, whereas "sa" is a

possessive prefix). However, as one continues each line, one quickly realizes that the

contradiction is only apparent. Gerow, in his discussion of this ornament, quotes Wilde:

"Until yesterday I had no idea there were any families or persons whose origin was a

222

Terminus."28> The pseudo-paradox dissolves when one understands that it is a railroad

terminus that is meant.

nirupādānasambhāram abhittāv eva tanvate | jagac citram namas tasmai kalāślāghyāya śūline II

atra vyatireka I alamkāryasyāpi brāhmaņaśramaņanyāyenālamkāratā II (865)

Honor the trident god's artistic prowess! Even without color or canvas, He paints the wondrous world. (57)286

Here the poetic ornament of "distinction" (vyatireka) (is suggested).

Even that which is to be ornamented is (here called) an ornament, by the maxim

of the Brahman-Buddhist monk.

Comments

Gerow defines the ornament "distinction" as follows: "a figure in which 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 comparability; hence

an inverted simile."28/ In the inverted simile of the poem Siva is compared to an artist,

only to show that the comparision cannot hold. No artist can paint without medium

(colors) or support (canvas), so Siva's creation excels. In is interesting to note that

285 Gerow (1971:269). 286 Nārāyana Bhatta, Stavacintāmaņi 9 (D). 287 Gerow (1971:276). 223

comparing the artists and the divine creator, which is often considered revolution of

romantic criticism in the West, was a commonplace in Sanskrit criticism.288

Although a person loses caste in becoming a Buddhist monk, the caste had

previously can still be used to distinguish him from other monks. Thus calling him a

Brahman, while literally false, is possible in an extended sense: thus the maxim. Here in

chapter four generally, and in the previous four poem specifically, the topic of discussion

is suggestion, not ornaments. Mammata believes that suggestion is the primary goal of

these poems, and as such, is to be established first and then decorated by poetic

ornaments. Dhvani School critics often use the analogy of a woman (the essence) and

jewelry (the decoration). Naming a type of suggestion after the ornament it suggests, like

naming a woman after the jewelry she wears, is correct only in the extended sense

illustrated by the maxim.

An Alankāra School critic would claim that the utterance becomes a poem

through the presence of the ornament, which is thus essential. They admit that the

ornament is suggestive, but insist that this feature remains secondary. In the given poem,

both types of analysis seem equally applicable.

vastumātram yathā

pamthia na ettha sattharamattha manam pattharatthale gāme annaapaoharam pekhkhiūņa jai vasasi tā vasasu I1289

288 See Abram (1983), for example. 289 The Sanketah commentary gives the following Sanskrit translation: pathika nātra samstaro 'sti manāk prastarasthale grāme I 224

atra yady upabhogaksamo 'si tadā āssveti vyajyate II (871)

Now an example (of a suggestions based on the power of the word in which a) fact (is

principally made manifest by the word):

O traveler, this barren plateau has no inn. Feel the charged atmosphere! If you want to stay with me, then please do. (58)290

Here, "If you want to make love, stay" is suggested (by a certain word used by the

woman).

Comments

The word in question could be translated, "large clouds," but it can also mean, "swelling

breasts." I have tried to capture the ambiguity with the expression "charged atmosphere,"

which could be charged emotionally as well as meteorologically.

śanir aśaniś ca tam uccair nihanti kupyasi narendra yasmai tvam I yatra prasīdasi punah sa bhāty udhāro 'nudāraś ca II

aviruddhāv api tvadanuvartanārtham ekam kārya kuruta iti dhvanyate II (876)

O lord of men! The evil stars and thunderbolts Strike fiercely he who raises your wrath. While he who pleases you gains both Brilliant fame and a loyal wife. (59)

unnatapayodharam prekșya yadi vasasi tad vasa II 290 Dwivedi writes, "Gāthāsaptaśatī?" 225

Here (the speaker) suggests that even opposed forces work together to please the king.

Comments

The opposed forces are expressed in two pairs of apparently contradictory words in

Sanskrit (śani - aśani and udāra - anudāra). Thus this could seem to be a case of

"pseudo-paradox." However, the opposed terms refer to the allies of the king and the

rewards of the faithful. There is no paradox in having two allies that oppose one another

(the British and the Soviets in WWII), nor in receiving two things that seem contrary (a

water heater and a freezer). Rather, the poem is another example of suggestion of a fact,

in this case the king's power. The sensitive reader understands that forces normally at

odds are working together, and then that this suggests that the king's power is so great it

can force these enemies' coöperation.

This ends the section on suggestions based on the words whose sequences are

noticed. The poems given in this section all relied on the double meaning or antonym of a

particular word. Changing the word for a synonym would ruin the effect. To return to the

example from Wilde ("Until yesterday I had no idea there were any families or persons

whose origin was a Terminus."), its effect would be lost if one replaced "terminus" with

"train station." Mammata now turns to poems in which the suggestion in based on the

meaning of the words, not the words themselves.

226

arthaśaktyudbhavo 'py artho vyañjakah sambhavī svataḥ II

praudhoktimātrāt siddho vā kaves tenombhitasya vā

vastu vālamkrtir veti sadbhedo 'sau vyanakti yat Il

vastvalamkāram atha vā tenāyam dvādaśātmakaḥ | (880)

svatahsambhavī na kevalam bhaņitimātranispanno yāvad bahir apy aucityena

sambhāvyamānah I kavinā pratibhāmātreņa bahir asann api nirmitah kavinibaddhena

vaktreti vā dvividho 'para iti trividhaḥ I vastu vālamkāro vāsāv iti śodhā vyañjakah I

tasya vastu vālamkāro vā vyangya iti dvādaśabhedo 'rthaśaktyubhdavo dhvaniḥ II

(4.2.2.2. Suggestive Poetry Based on the Power of the Meaning.)

39c - 41b. Suggestive meaning based on semantic power is self-existent, established

by a creative expression of the poet, or established by an expression of some poetic

character. Each of these (suggested meanings) is either a mere fact or a poetic

ornament; thus there are six types. Each of these six further suggests either a mere

fact or a poetic ornament, making for twelve kinds.

Self-existent means having an externality (objectivity) not established by the mere words

(of the poet). The other, which does not exist externally, is two-fold: it is either a creation

of the poet's imagination, or of a character portrayed by him. Thus we arrive at three

types. These three being either factual description or poetic ornament, we arrive at six

types of suggestor. The suggested meaning of each of these being either factual

227

description or poetic ornament, we arrive at twelve types of suggestive poetry based on

semantic power.

Comments

To discern the logic of this twelve-fold division we need to look closely at the twelve

examples that follow. In the first four poems, the suggestor is self-existent, in the second

quatrad it is a creation of the poet, and in the final group it is based on a character of the

poem. We shall see that in each group of four, the first two examples illustrate suggestive

facts, the second two suggestive poetic ornaments. Finally, with each of these pairs, the

first example suggests a fact, the second, an ornament.

krameņodāharaņam

alasaśiromani dhuttānam aggimo putti dhanasamiddhimao I ia bhanieņa naangī papphullaviloanā jāā 11291

atra mamaivopabhogya iti vastunā vastu vyañjate | (889)

(4.2.2.2.1. Suggestive Poetry in which the Suggestor is Self-Existent.)

Examples are now given in order:

My daughter, here is a leader of the virtuous, a paragon of taste, And a man of exceptional riches - Hearing this, the eyes of the shapely girl bloomed. (60)

291 The Sanketah commentary gives the following Sanskrit translation: arasaśiromaņir dhūrttānāmagrayaḥ yutri dhanasamrddhimayaḥ I iti bhaņitena natāngī praphullavilocanā jātā l

228

In this example, the fact, "(This man) is just right for my love", is suggested by the fact

(of her eyes blooming).

Comments

The category "self-existent" includes all objective things or events in the world. It

contrasts with the category of fantastical things invented either directly by the poet, or

indirectly by a character in the poem. People do reacted by widening their eyes, and this

reaction suggests excitement or fear universally. In the context of the poem, it suggests

the young woman's enthusiasm for the man described.

The crux is how a man who is "arasa" (flavorless, weak, without culture) can be a

perfect lover. The Sanketah glosses "arasa" with "arocakinah," which can mean

someone with fastidious or cultivated tastes. Alternatively, the man could be a rich old

saint who has lost his taste for the world. In this case the father might be suggesting a

marriage of interest to his greedy daughter.

dhanyāsi yā kathayasi priyasamgame 'pi visrabdhacātukaśatāni ratāntaresu I nīvīm prati pranihite tu kare priyeņa sakhyah śapāmi yadi kimcid api smarāmi II

atra tvam adhanyā aham tu dhanyeti vyatirekālamkārah l (893)

How fortunate you are to be able to recall All the sweet seductions

292 Monier-Williams (1899:89). 229

Whispered in the height of passion by your lover. I swear, my friend, that after my lover Starts to fondle the knot of my sash, I remember nothing. (61)293

In this example the poetic ornament of "distinction" is suggested (by the self-existent fact

described). The suggestion is, "It is I, not you, who is fortunate."

Comments

As we saw with poem 57, the ornament "distinction" is inverted simile. Two similar

things or experiences are brought together to point out the dissimilarity. Usually the

object compared is shown to be superior in some way to the object of comparison. Thus

Dwivedi is justified in calling this ornament "comparative excellence."24 Here the sexual

pleasures of the two women are compared. The ecstasy of the speaker, whose intensity

precludes any memory of it, is clearly superior to the pleasure of the second woman who

can remember all the details.

darpāndhagandhagajakumbhakapāțakūța - samkrāntinighnaghanaśonitaśonaśocih I vīrair vyaloki yudhi kopakaşāyakāntiḥ kālīkațāksa iva yasya kare krpānaḥ II

atropamālamkāreņa sakalaripubalaksayaḥ kśanāt karisyate iti vastu II (897)

293 Dwivedi writes, "Vijjākā, according to Śëktimuktāvalī." 294 Dwevedi (1967:103). 230

Having split the heads of elephants raging, rutting, and blind with lust, His sword did shine a thick blood red. He searches out his foes upon the field of war, Like Kālī with burning, wraith-red eyes. (62)295

Here the fact that he will quickly kill the entire enemy army is suggested by the (self-

existent) poetic ornament of simile.

Comments

Kālī (or Durgā, wife of Siva) is a most potent force of destruction. If the king is like her,

surely he can kill the whole enemy army, as Mammata explains.

The simile is "self-existent" because a real king is being compared to a real

goddess (or so it was believed). The two terms of the simile have objective existence and

the king really is like Kālī. The poet has invented nothing, he is merely making artisitic

use of an objective state of affairs.

gādhakāntadaśanaksatavyathāsamkatāt arivadhūjanasya yaḥ I austhavidrumadalānyamocayan nirdaśan yudhi ruşā nijādharam II

atra virodholamkāreņādharanirdaśanasamakālam eva atravo vyāpāditā iti tulyayogitā

mama ksatyāpy anyasya ksatir nirvar tām iti tadbuddhir utpreksyate ity utprekșā ca l

esūdāharanesu svatah sambhavī vyañjakah II (902)

Biting down on his lip in battle rage, He frees the coral lips of his enemies' wives From their lovers' teeth,

295 Dwivedi writes, "Vijjākā, acc. to Śuktimuktāvalī" (1967:102). 231

And any wounds they might inflict. (63)

The (self-existent) poetic ornament in the poem is a "contradiction," i.e., "the enemies are

killed as soon as the lip is bitten." (The ornament) suggests the poetic ornaments of

"equal joining" and "fancy." The suggestion by "equal joining" is, "May my wound

remove the wounds of others." The suggestion by "fancy" is the king's thought, since it is

fanciful.

In each of the preceding four examples, the suggestor was self-existent.

Comments

"Contradiction" is an ornament in which the poet includes an apparent contradiction.

Here the king is both wounding himself and wounding others. The contradiction is only

apparent because the king bites his lip while engaged in battle. The death of the enemies

comes about in fighting that is not explicitly described.

"Equal joining" is an ornament usually akin to the Greek zeugma. The poem gives

an inverted example, in which the king's biting stops the enemies from biting their wives

in love-making. There is a parallelism here, but not full zeugma. A charming and simple

example of the ornament is found in the verse, "The young men enjoyed the shady

terraces along with their wives."20 One can interpret the verse to mean that the men and

their wives enjoyed the terraces. But by "equal pairing," one can also interpret it to mean

that the men enjoyed both the terraces and their wives.

296 Māgha, Śiśupālavadha 3.35. Quoted in Ingalls, et al. (1990:345). 232

The king hopes his wound, which is strictly incapable of activity, will remove the

wounds of others. The ornament "fancy" pertains to imaginary ideas such as this,

especially to the personification of inanimate or abstract things.

kailāsasya prathamaśikhare veņusammūrchanābhih srutvā kīrtim vibudharamaņīgīyamānām yadīyām | srastāpāngāh sarasabisinīkāndasamjātaśankā dinmātangāḥ śravanapuline hastam āvartayanti II

atra vastunā yesām apy arthādhigamo nāsti tesām apy evamādibuddhijananena

camatkāram karoti tvatkīrtir iti vastu dhanyate | (908)

(4.2.2.2.2. Suggestive Poetry in which the Suggestor is Invented by the Poet.)

All the earth's elephants, Hearing tender flutes and sweet goddesses Sing your praise on Mount Kailasa, Mistake the sounds for juicy lotus stems And search their ears with hungry trunks. (64)

"Your fame is especially wondrous because it creates such ideas even in those creatures

who do not understand (the words sung)." This (fact) is suggested by the "fact" (created

by the poet).

Comments

The suggesting fact - that the sound of the song is so sweet that the elephants raise their

trunks to try to eat it - does not exist in the real world. It is a creation of the poet's

imagination. Mammata claims that it suggests that the fame of the king is especially

233

wondrous. The primary suggestion, it seems to me, is that the music is especially sweet.

Only this accounts for why the elephants want to eat it. The most obvious reason for the

sweetness is that the music is played by goddesses. The glory of goddesses is thus

suggested. Mammata would have to argue that part of the music's sweetness is due to the

content of the song in order for the fame of the king to be suggested.

kesesu balāmodia teņa a samarambhi jaasirī gahia jaha kandarāhim vihurā tassa dadham kamthaammi samțhaviām l1297

atra keśagrahanāvalokanod dīpitamadanā iva kandarās tadvidhurān kanthe grhnanti ity

utpreksā l ekatra samgrāme vijayadarśanāt tasyārayah palāyya guhāsu tisthantīti

kāvyahetur alamkārah I na palāyya gatās tadvairiņo 'pi tu tatah parābhavam sambhāvya

tān kandarā na tyajantīti apahnutiś ca II (912)

In the battle, the King forcibly took The goddess Victory by the hair. So the caves passionately embraced The necks of his enemies. (65)

Here (the imagined fact suggests) the ornament "fancy." The suggestion is, "The caves

embrace his enemies by the neck as if their lust were aroused by the seizure of Victory's

hair." (The imagined fact also suggests) the ornament "poetic cause," with the suggestion

as, "His enemies ran away and remained in a hidden place because they saw him

297 The Sanketah commentary gives the following Sanskrit translation: keśesu balāt kāreņa tena ca samare jayaśrīrgrhītā I yathā kandarābhir vidhurāstasya drdham kanthe samsthāpitāḥ Il

234

victorious in an engagement." (The imagined fact also suggests) the ornament "denial."

Here the suggestion is, "The enemies did not run away; rather, the caves, fearing the king

would defeat them, did not let them out."

Comments

In the real world, caves do not embrace men, so the "fact" described in an invention of

the poet. The poet gently mocks the cowardly behavior of the king's enemies. As

Mammata explains, the poet's use of this image suggests one of three ornaments. As we

saw with poem 63, "fancy" often involves the personification of inanimate objects, in this

case the caves. Caves, like victory, are often personified as women. Here they are both

seen as enamored woman. The desire of the caves is aroused by the king's passion for

Victory. I have followed this interpretation in my translation.

The second option merely establishes a "poetic cause." The enemies run away and

hide because they are afraid of the king's passion for, and control over, Victory. This

option, while adding an interesting twist, does not do justice to the image of the caves

embracing the enemies.

The third option sees a "denial" that the enemies ran away. Rather, they are

already in the caves and the caves embrace them out of fear, stopping them from leaving.

While this option seems better than the second, it does leave unexplained why the

enemies are in the caves to begin with.

235

Mammata does not directly discuss the relation of these three mutually

contradictory suggestions. By allowing that there are three he acknowledges that the

poem cannot be paraphrased without deflating the interpretive tension. He neither affirms

nor denies that part of the beauty of the poem arises from the tension. However, he does

make this claim with regards to poem 111, below.

gādhālimganarahasujjuammi daie lahum samosarai I mānamsiņīņa māņo pīlaņabhīa vva hiaāhim 11298

atrotpreksayā pratyālinganādi tatra vijrmbhate iti vastu | (918)

When the lover moves in impetuously For an earnest embrace, Pride quickly flees, as if afraid to suffer, From the hearts of the cheerful girls. (66)

Here the poetic ornament "fancy" suggests the fact that there was a swelling of reciprocal

embraces, etc.

Comments

The "fancy" here is the personification of pride and its imaginary fleeing. The ornament

298 The Darpana commentary has the following Sanskrit translation: gādhālinganarabhasodyate dayīte laghusamapasarati manasvinīnām mānaḥ pīdanabhīta iva hrdayāt l| The Sampradāyakāśinī commentary replaces "pīdanabhīta" with "preranabhīta," which seems inferior to me.

236

is created by the imagination of the poet and suggests that the lovers really do embrace. I

have not tried to capture the sexual punning of the first compound, which sounds like

gādha + linga + nara, which could be rendered, "the man whose penis is large or deeply

entered." The punning accounts for part of why pride is so ready to flee.

jā theram va hasantā kaivaanamburuhabaddhavinivesā l dāvei bhuaņamamdalamannam via jaai sā vānī l1299

atrotpreksayā camatkāraikakāranam navam navam jagat ajadāsanasthā nirmimīte iti

vyatirekah I eşu kavipraudhoktimātranispanno vyañjaka | (921)

Seated in the lotus-like mouths of poets, Speech is victorious, Portraying wonderful worlds to mock the Creator. (67)

In this example, "fancy" suggests the ornament "distinction." The suggestion is, "Speech

has a sentient seat and creates worlds ever new and of unmixed delight."

In each of the preceding four examples, the suggestor is based only in the poet's

imaginative assertion.

Comments

Speech here is both a goddess and the human faculty. Through "fancy" it is personified

and enthroned on a living throne. The Creator (Brahma) sits on a lotus and creates but

299 The Sanketah commentary has the following Sanskrit translation: yā brahmāņam iva hasanti kavivadanāmburuhabaddhaviniveśā l darśayati bhuvanamaņdalam anyad eva jayati sā vāņī ll

237

one world that is both pleasant and painful. Thus the implicit comparison is drawn to

illustrate the comparative excellence of Speech through an implicit inverted simile.

This poem recalls poem 57, in which the painter is shown to be inferior to the

divine creater because of his dependance on materials. Here, the poet is shown to be

superior to the divine creator, because his creation is more abundant (many worlds) and

pleasant. In poems the comparision is not fully spelled out, which suggests to me that it

was a common trope.

je lamkāgirimehalāsu khaliā sambhogakhiņnorari phārupphullaphaņāvalīkavalaņe pattā dariddataņam I te emhiņa malaānilā virahiņīņīsāsasampakkiņo jādā itti sisuttaņe vi bahalā tāruņņapunnā via I1300

atra nihśvāsaih prāptaiś varyā vāyavah kim kim na kurvantīti vastunā vastu

vyajyate | (925)

(4.2.2.2.3. Suggestive Poetry in which the Suggestor is Invented by a Character in the

Poem.)

"Gaunt the is wind off the Western Ghats. Upon the gold-crowned mountain slopes, Open-hooded cobras drank it down, Exhausted in ecstasy." "And yet it is reborn,

300 The Sanketah commentary has the following Sanskrit translation: ye lankāgirimekhalāyāścalitāh sambhogakhinnoragī sphārotphullaphaņāvalīkavalane prāptā daridratvam ta idānīņ malayānimā virahiņīniḥśvāsasamparkiņo jātā igiti śiśutve 'pi bahalāstāruņyapūrņā iva II Other commentaries (e.g., the Sārabodhinī) substitute ițiti for igiti in the final line.

238

Regaining its full strength From the suffering sighs Of women separated from their lovers." (68)301

Here the fact suggests another fact, namely, "Made bold by sighs, the wind will do

anything."

Comments

Following the tradition, Mammata separates poems in which the poet speaks in first

person from those in which he speaks through a character. In this and the follow three

examples, the suggestion arises from the words of a character in the poem. The first

example comes from a drama. The suggesting "fact" is the fanciful description of the

wind given by a character of the drama. The suggested "fact" is that the wind is now very

strong.

sahi viraiūna mānassa majjha dhīrattanena āsāsam piamdasanavihamlakhalakhanammi sahasatti tena osariam ||302

atra vastunākrte 'pi prārthane prasanneti vibhāvanā priyadarśanasya saubhāgyabalam

dhairyena sodhum na śakyate ity utpreksā vā II (930)

My pride swelled strong, Pushing me to stand up for myself. But it abandoned me, o friend, The moment I actually saw my lover. (69)303

301 Karpūramañjarī 1.20 (D). 302 The Sanketah commentary has the following Sanskrit translation: sakhi viracayya mānasya mama dhīratveņāsvāsam I priyadarśanavisrnkhalakșaņe sahaseti tenāpasrtam Il 239

In this example a fact suggests either an "unusual cause" or "fancy." In the first case the

suggestion is, "She is appeased even without any entreaty by her lover." In the second

case it is, "Pride cannot withstand the powerful pleasure of seeing one's lover."

Comments

Again, the suggesting fact is given in the voice of a woman, not in that of the poet. It

suggests one of the two ornaments. "The woman betrayed" (mānini) is such a common

theme in the erotic poetry that it became a sub-genre. Normally the lover needs to entreat

the woman he has betrayed before she will be appeased. The first ornament, "unusual

cause," occurs when the effect is present without its usual cause. A lover's entreats,

explanation, and begging are usually needed to appease a betrayed woman, but here the

woman is reconciled by the mere sight of her lover. This case of "fancy" differs little

from those we have already seen. The woman's pride and her pleasure are both

personified and pleasure proves itself stronger.

ollollakaraaraakhkhaehi tuha loanesu maha dinnam I ramttasuam paāo koveņa puņo ime na akkamiā ||304

atra kim iti locane kupite vahasi iti uttarālamkāreņa na kevalamārdranakhaksatāni

gopāyasi yāvat tesām aham prasādapātram jāteti vastu II (934)

303 Literally: The production of my pride was encouraged by my suppression of jealousy, But the moment I really saw my love, I was suddenly abandoned by it. 304 The Sanketah commentary has the following Sanskrit translation: ārdrārdrakarajadaśanakșataistava locanayor mama dattam I raktāmsukam prasādam kopenā punarime nākrānte II 240

My eyes are not eclipsed with anger. Their red clothing is a generous gift Given by your body's fresh scratches and bites. (70)

Here the ornament "reply" implies the question, "Why is there anger in your eyes?" This

suggests the fact that, "Not only are you trying to hide the fresh nail and teeth marks, but

I am grateful for the gift they give."

Comments

The ornament "reply" urges the reader to reconstruct a previous question or statement.

The woman in the poem explains that her eyes are red with anger, which leads the reader

to suppose that the lover asked something like, "Why are your eyes red with anger?"

The scene implied by the ornament suggests the fact that the woman is grateful to the

marks because they reveal her lover's misconduct. We see here that the "fact" suggested

can be as fanciful as some of the suggestor "facts" we have seen.

mahilāsahassabhaie tuha hiae suhaa sā amāanti l aņudiņamanannakammā angam tanuam vi tanuei I1305

atra hetvalamkāreņa tano stanūkarane 'pi tava hrdaye na vartate iti viśeșokti l eșu

kavinibaddhavaktrpraudhoktimātranispannaśarīro vyañjakah I evam dvādaśa bhedāḥ II

(939)

305 The Sanketah commentary has the following Sanskrit translation: mahilāsahasrabharite tava hrdaye subhaga sā āmāntī l anudinamananyakarmā anga tanu kamapi tanayati II 241

O lucky one! Your heart is so Crowded with women That she cannot find a place. So giving up all else She spends her days getting thinner and thinner. But still she cannot squeeze in! (71)306

In this example the ornament of "cause" suggests the ornament "remarked difference,"

namely, "Even though she becomes skinny, she find no place in your heart."

In each of the proceeding four examples, the suggestor was based on the words of

one of the characters in the poem.

These are the twelve types (of suggestion based on the power of the meaning).

Comments

"Cause" is the ornament that describes a cause for the described action or state-of-affairs.

There is usually something striking about the cause given. The girl is getting thinner

and thinner. The real reason is that her anxiety and dejection over not being able to win

the heart of her beloved. However, the character in the poem invents a different cause,

namely that the woman is getting thin to fit more easily in her beloved's heart.

"Remarked difference" is an ornament in which either a cause is present without

the expected effect (as here) or the effect without the cause. According to the character's

first trope, if the girl gets thin enough, she ought to fit anywhere, even in her beloved's

306 Gāthāsaptaśati 2.82. 307 This ornament is controversial. Gerow reports that Mammata rejects it, probably because he does not include it in the chapters on ornaments (1971:327).

242

crowded heart. However, the causal relation has failed. The girl has become thin, but still

finds no place in his heart. Thus the character uses an ornament to suggest another

ornament.

Mammata finishes the example of poems in which the suggestor is invented by a

character in the poem. These poems contrast with the preceding group in which the

suggestor is invented by the poet. Both these groups contrast with the first group in which

the suggestor is not invented, but naturally occurring. Poems of the first group can be

either narrative or put in the mouth of a character. As we saw, each of the group has four

types, giving a total of twelve.

śabdārthobhayabhūr ekaḥ | (944)

yathā

atandracandrābharaņā samuddīpitamanmathā l tārakātaralā śyāmā sānandam na karoti kam II

atropamā vyangyā II

(4.3. Suggestive Poetry Based on Both the Words and the Meanings.)

41c. There is a (type of suggestion) based on both word and meaning.

For example:

First Meaning Shared Meaning Second meaning

Who is not delighted by the beautiful girl dark night 243

Adorned with the bright head-jewel The clear moon

Who kindles love with twinkling eyes? stars? (72)

Here a simile is suggested.

Comments

The simile in the poem compares the girl with the moon. It turns on three expressions

with double sense. Thus it is a suggestion based on the power of the words. But

understanding not only the punning expressions, but also the meaning of the other words

is necessary to understand the simile. Thus it is based on both word and meaning. The

punning poems we have seen up to this point were comprised of long Sanskrit compound

words that could be broken in two ways, thus every semantic unit in the poem had double

meanings. This poem resembles more closely an English punning poem in that only three

of the words have double meanings while the structure remains the same on both readings

("Who is not delighted by an X, adorned with a Y, which kindles love with Z").

bhedā astādaśāsya tat II

asyeti dhvaneḥ I

nanu rasādinām bahubhedatvena katham astādaśety ata āha I

rasādīnām anantatvād bheda eko hi ganyate | (950)

244

anantatvād iti l tathā hi nava rasā l tatra śrngārasya dvau bhedau I sambhogo

vipralambhaś ca l sambhogasyāpi parasparāvalokanālinganaparicumbanādi-

kusumoccayajalakelisūryāstamayacandrodayasadrtuvarņanādayo bahavo bhedāḥ

vipralambhasyābhilāsādaya uktāḥ l tayor api vibhāvānubhāvavyabhicārivaicitryam I

tatrāpi nāyakayor uttamamadhyamādhamaprakrtitvam tatrāpi deśakālāvasthādibhedā

ity ekasyaiva rasasyānantyam | kā gananā tv anyesām l asamlaksyakramatvam tu

sāmānyam āśritya rasādidhvanibheda eka eva ganyate II

41d. Thus there are eighteen type of it.

It means suggestive poetry.

(4.4.) Objection: With so many types of rasas, etc., how can you claim there are (only)

eighteen (types of suggestive poetry)?

42ab. Reply: That there is a single type is acknowledged because of the infinite

number of rasas and the rest.

Infinite is explained as follows. There are nine rasas. The erotic has two types:

consummated and frustrated. Love-in-enjoyment has many types such as mutual glances,

embraces, kissing, flower-picking, amorous water sports, descriptions of sunsets, moon

rises, the six seasons, etc. Love-in-separation has already been divided into longing, etc.

There is also the variety of the determinants of the emotion, the symptoms of the

245

emotion, and the auxiliary states. Furthermore, the couple may be of upper, middle, or

lower class. Finally, there are variations of place, time, and circumstance. Thus even the

erotic rasa alone is infinite. How could one count the others? But the suggestive poetry of

rasa, etc. is counted as a single type, because (all these poems) share the common feature

of having the psychological sequence (of suggestion) unnoticeable.

Comments

The objector questions the logic of the classificatory system regarding rasa. Mammata

discusses suggestion of rasa under the third type of suggestion: poetry whose suggestion

is unnoticed. As we saw this category contains other types of poem, e.g., poems that

suggest emotions not capable of become rasa. If poetry that suggests rasa is the most

important type, it should at least have its own category, and not be confined to a single

type among eighteen. In fact, the objector urges that there should be a category for each

type of rasa. The objector is correct that system conceals the relative importance of rasa

poetry.

Mammata reminds his objector that suggestions were first divided according to

whether the literal meaning was set aside and then according to whether the suggestive

sequence was usually noticed. In the resulting classificatory system rasa suggestions are

all of a kind. Mammata urges that this system must be maintained, for if the subtypes of

rasa were allowed separate categories the system would be endless. Mammata assumes

246

that if you allow a first level of sub-types (the nine rasas), you must allow further levels

of sub-types (e.g., the two sub-types of the erotic and their sub-types).

The reply ignores the middle ground. Ten sub-types could have been given to the present

third category, one for each of the rasas, and one for unnoticed suggestions that suggest

something other than rasa. In this manner the importance of rasa could be shown without

destroying the system.

It is worth reviewing the list (at the beginning of the chapter) of the eighteen basic types

of suggestion, for their exposition has been spread out over the entire chapter to this point

and the rest of the chapter consists in further permutations of each type. The next section

gives poems in which the suggestion is based on a single word or compound.

A. Literal meaning unintended.

1. Transformed into another meaning (poem 73).

2. Entirely set aside (poem 74).

B. Literal meaning intended.

B1. Psychological sequence unnoticed.

3a. Suggests rasa (poems 75-76).

3b. Suggests pseudo-rasas, emotions, etc. (not much charm added, no examples).

B2. Psychological sequence noticed.

B2.1. Based on the power of the word.

4. Suggesting an ornament (poem 77).

5. Suggesting a fact (poem 78).

247

B2.2. Based on the power of the meaning.

B2.2.1. Where the suggestor is self-existent.

6. A fact suggests a fact (poem 79).

7.A fact suggests an ornament (poem 80).

8. An ornament suggests a fact (poems 81-82).

9. An ornament suggests another ornament (poem 83).

B2.2.2. Where the suggestor is invented by the poet.

10. A fact suggests a fact (poem 84).

11. A fact suggests an ornament (poem 85).

12. An ornament suggests a fact (poem 86).

13. An ornament suggests another ornament (poem 87).

B2.2.3. Where the suggestor is invented by a character in the poem.

14. A fact suggests a fact (poem 88).

15. A fact suggests an ornament (poem 89).

16. An ornament suggests a fact (poem 90-91).

17. An ornament suggests another ornament (poem 92).

B2.3. Based on the power of both the words and the meanings.

18. Suggests an ornament (not possible).

248

vākye dvyutthaḥ

dvyuttha iti śabdārthobhayaśaktimūlah |

pade 'py anye

api śabdād vākye 'pi l ekāvayavasthitena bhūsaņena kāminīva padadyotyena vyangyena

vākyavyangyāpi bhāratī bhāsate l tatra padaprakāśyatve krameņodāharaņāni I

yasya mitrāņi mitrāņi śatravah śatravas tathā | anukampyo 'nukampyaś ca sa jātah sa ca jīvati II

atra dvitīyamitrādiśabdā āsvastatvaniyantraņīyatvasnehapātratvādisamkramitavācyāh II

(957)

(4.5. Suggestive Poetry in which a Suggestive Word Adds to the Charm.)

42c. Suggestion based on both is found in a sentence.

On both means on both words and meanings.

42c. The others are found in a word also.

Also means, "also in a sentence." Just as a charming woman sparkles on account of a

gem worn on her finger, so a statement, although suggested by the sentence, sparkles on

account of individually suggestive words. The following, in order, are examples in which

there is also suggestiveness of a word.

(4.5.1. Secondary Suggestion in which the Literal Meaning Is Unintended.)

He is born and lives Whose friends are friends, Whose enemies are enemies, And whose loved ones are loved. (73)

249

In the poem the literal meaning of the second occurrence of the word friend is changed

into "encouraging," that of enemy into "reprehensible," and that of loved into "lovable."

Comments

Mammata gave an example of suggestion based on both word and sentence meaning in

poem 72. It comprised the eighteenth type of suggestion. This combination can

obviously be found only at the level of the sentence, not in an individual word. The other

seventeen types, however, are found in individual words. Mammata thus begins to gives

examples of types of suggestion that are parallel to the first seventeen types, but function

at the level of individual words instead of at the level of the sentence.

Mammata express the theory of primary and decorative suggestion by means of

an analogy with a bejeweled woman. As is so often the case, Anandavardhana uses the

same analogy to sum up his detailed presentation of this theory."8 The analogy functions

on two levels of beauty. The beauty of the woman is analogous with the beauty of the

sentence-level suggestion, while the beauty of the gem is analogous with the suggestion

based on a single word. Anandavardhana explains that, "there is beauty in all those

varieties of suggestion which appear in a single word, even though a single word serves

only as a reminder."The assumption is that a decoration serves to point out and

augment the beauty of the thing it decorates, not the other way around. If a decoration

becomes prominent, it is not longer a decoration, but rather a rival to the thing it is

308 Dhvanyāloka 3.1j. See Ingalls, et al. (1990:387). 309 Ingalls, et al. (1990:387).

250

supposed to decorate. In the poems that follow, the second suggestion remains

subservient to the primary. Thus these poems are properly included in this chapter, which

treats poems in which sentence-level suggestion is primary.310

The poem given here can be read literally as a series of tautologies. However, if

the literal meanings of three words are shifted, but not entirely set aside, the poem takes

on more interest. Each takes on a metaphoric meaning that intensifies the original

meaning. A paraphrase of the poem would then read: He is born and lives a worthy life,

whose friends are truly friends, whose enemies deserve enmity, and whose loved ones

both merit and reciprocate his love.

It is a bit difficult to see the difference between the suggestion of this poem and

that of poem 23 ("I say to you, 'Here stands a group of the learned / So remain here with

your mind composed.'"'). Supposedly, the sentence meaning (semantic content +

speaker's intention) of the earlier poem changed: a declarative sentence became an

advisory sentence. In the poem given here, on the other hand, the meanings of three

words change without radically altering the sentence meaning. But there are two

difficulties in this explanation. First, by pointing out that the meaning of the word "say"

shifts to "advise," Mammata gives us the impression that the first example also showed

the alteration of individual words, not of the whole sentence. Second, changing the

meanings of three words radically alters the sentence meaning in the poem given here.

Fortunately, the differences are much clearer in the other sixteen cases.

310 If the suggestion of a word becomes the prominent element of a poem, the suggestion is called a poetic ornament, and the poem becomes a poem of subordinate suggestion. Mammata discusses these issues in chapters five and ten. 251

khalavavahārā dīsanti dāruņā jahavi tahavi dhīranam I hiaavaassavahumaa ņa hu vavasāā vimujjhanti l1311

atra vimuhyantīti II (965)

Just as the conduct of a villain is seen to be vile, So the heart-felt efforts of the wise are never stupefied. (74)

Here stupefied (has its literal meaning completely replaced by a metaphoric meaning).

Comments

Actions, unlike people, cannot literally be stupefied. The suggestion is that the efforts of

the wise are never blocked or in need of changing. We need to compare this suggestion to

the irony of poem 24 ("What a great help! What can I say? / Your kindness is widely

renowned. / Behaving ever like this, my friend, / May you enjoy a hundred autumns."). In

poem 24, the literal sentence meaning is blocked by the context and is inverted, while

here only the meaning of a word is blocked and changed. Of course, this change does

alter the sentence meaning, but not to the same degree as in poem 24. If you rewrote the

two poems without metaphor, the first poem would lose its irony, which is essential to its

force. The poem here would only lose a decorative detail; the meaning of the whole

would remain the same.

311 The Sanketah commentary has the following Sanskrit translation: khalavyavahārā drśyante dāruņā yady api tathāpi dhīraņām I hrdayavayasya bahumatā na khalu vyavasāyā vimuhyanti Il 252

There is also a play on the word for villain, which derives from a word for

"threshing floor" and the word for vile, which derives from a word for "wood." Thus one

can to read the first line, "Just as the material of the threshing floor is seen to be wood."

This reinforces image of steadfast efforts of the honest.

lāvaņyam tad asau kāntis tad rūpam sa vacahkramah I tadā sudhāspadam abhūd adhunā tu jvaro mahān II

atra tadādipadair anubhavaikagocarā arthāh prakāśyante II (969)

(4.5.2. Secondary Suggestion of Rasa.)

That zest, that flare, that beauty, that voice, All - like ambrosia at the time - Are now some deadly disease. (75)

Words such as that suggests that the objects belong to the realm of experience.

Comments

The primary suggestion here is of the erotic rasa, more specifically the type of love-in-

separation in which the lover longs for his beloved. Sometimes these poem lack visual

detail, but here the repetition of word "that" makes the lover's words concrete, adding an

extra charm to the poem. The phrasing suggests that he has sharp memories of his

beloved's specific qualities. This lend force to the reader's conviction that the love is

both real and intense. Of a similar poem Abhinava says, "the word "those" serves as a

253

special cause of the [erotic] rasa by suggesting various memory-pictures of the

qualities ... which are indescribable and which can only be felt by the speaker."312

yathā vā

mugdhe mugdhatayaiva netum akhilah kālah kim ārabhyate mānam dhatsva dhrtim badhāna rjutām dūre kuru preyasi sakhyaivam pratibodhitā prativacastām āha bhītānanā nīcai śamsa hrdi sthito hi nanu me praneśvarah śrosyati II

atra bhītānaneti l etena hi nīcaihśamsanavidhānasya yuktatā gamyate bhāvādīnām

padaprakaśyatve 'dhikam na vaicitryam iti na tad udahriyate II (972)

Another example (of the third type of suggestion):

"O sweet one! Why do you cling so to mere sweetness? Be proud! Be courageous! Cast off your coyness with your lover!" To her friend's prodding she replied, fear-faced, "Whisper! My best-beloved may hear you from within my heart!" (76)313

Here fear-faced suggests the appropriateness of the request to whisper.

In the suggestion by a word of the emotions, the pseudo-emotions, etc., no beauty

is gained. Thus no examples of this are given.

Comments

312 Ingalls, et al. (1990:393). 313 Amaruśataka 70.

254

The primary suggestion in this poem is again of the erotic rasa. The suggestion is

ornamented by the suggestive phrase, "fear-faced." It suggests that the young woman is

so deeply in love that she feels her lover abides in her heart and thus might overhear the

friend's council. This detail adds credibility to the woman's love and thus charm to the

poem.

Mammata gives no justification of the claim that no beauty is gained by additional

suggestion of the emotions, etc., by a word. Although he is probably following

Ānandavardhana or Abhinava, I have been unable to locate a likely passage. They discuss

this issue in detail at 3.4a, where they address only suggestion of rasa by a single word.

Perhaps Mammata is making explicit a point the earlier writers made by omission.

In any case, the claim makes sense in Anadavardhana's system, where all genuine

suggestion is suggestion of rasa. If one believes this, one could argue that secondary

suggestions of other elements will not contribute to the dominant suggestion. However, if

one admits that the dominant suggestion can suggest emotions other than those that can

become rasas, the current claim is difficult to justify.

None of Mammata's commentators included in Mohan considered this claim

worthy of serious attention. They are content to repeat that these other suggestions based

on a word are possible in poetry, but unlike rasa suggestions, they add no charm.

This sentence could be used as evidence against the claim that Mammata no

longer thought of rasa as the sole essence of suggestion. If Mammata's notion of

suggestion were indeed broader than that of earlier critics, we would expect him to

255

continue. For example, a poem in which a single word suggestion of piety adds charm to

its devotional mood would be appropriate.

rudhira visaraprasādhitakaravālakarālarucirabhujaparighih | ititi bhrukutivitankitalalātapaddo vibhāsi nrpa bhīma II

atra bhīşanīyasya bhīmasena upamānam II (977)

(4.5.3. Secondary Suggestion Based on the Power of the Word.)

How you shine, my dread king! Your bent brows furrow your broad forehead, Your frightful iron arms beautifully Wielding the reddened, blood-spilling sword! (77)

In this example, the dreadfulness (of the king) is compared to that of Bhīmasena.

Comments

Mammata moves on the fourth type of suggestion (of the eighteen given at the beginning

of the chapter). This type uses a homonymic word (or words) to suggest a poetic

ornament. In the description of the king, the word "bhīma" (dread) reminds the reader of

Bhīmasena and thus suggests the poetic ornament of simile, "You are mighty like

Bhīmasena, the legendary warrior of the Mahābharata." The suggestion of this simile

adds charm to the main suggestion of the heroic rasa.

bhuktimuktikrd ekāntasamādeśanatatparah I

256

kasya nānandanisyandam vidadhāti sadāgamah II

kācit samketadāyinam evam mukhyayā vrttyā śamsati II (980)

Primary Meaning Common Meaning Secondary Meaning

Who is not flooded with delight By the good scripture! By a lover's visit!

Always intent on Moral advice Enjoyment And leading to final salvation And release from frustration In a heavenly place. In a secluded place. (78)

In this poem the woman approves of the proposed rendezvous by a homonymic

expression.

Comments

Mammata illustrates the fifth type of suggestion, in which homonymic words are used to

suggest a fact. Here a woman uses the extended pun to secretly communicate the fact

that she agrees to the lover's proposed tryst. The main suggestion on the first reading is

devotional enthusiasm, on the second it is the erotic rasa.

Next are the twelve suggestions based on meaning, beginning with a self-existent

fact that suggests another fact.

257

sāyam snānam upāsitam malayajenāngam samālepitam yāto 'stācalamaulim ambaramaņir visrabdham atrāgatih | āścaryam tava saukumāryam abhitah klāntāsi yenādhunā netradvandvamamīlanavyatikaram śaknoti tenāsitum II

atra vastunā krtapurușaparicayā klāntāsīti vastu adhunāpadadyotyam vyajyate | (984)

(4.5.4. Secondary Suggestion Based on the Power of the Meaning.)

A hot bath and a sandal-oil massage, Then you came here languidly, While the sky-jewel set over the Western Ghats. Yet still you are completely exhausted. How amazingly delicate you must be! Your eyes will not stay open for long. (79)

Here the fact that his fatigue is due to his having gone to another woman is suggested by

the facts (described in the poem) and made manifest by the word still.

Comments

The woman in the poem mentions various reasons why the man should not be tired: the

bath, the massage, the slow travel, and the coolness of the evening. The woman

highlights the incongruity between the man's story and his fatigue by using the word

"still." It shows that the woman suspects the man's other activities and pushes the reader

to look deeper. The fact that he is still tired after such a calm day suggests that he must

have done more than he admits, i.e., gone to see another woman, taken a bath and put oil

to hide her smell, and then traveled quickly to make up time.

258

Next Mammata gives a pair of verses that form an example of a self-existent fact

suggesting an ornament.

tadaprāptimahāduhkhavilīnāśesapātakā l taccintāvipulāhlādaksīņapuņyacayā tathā II cintayantī jagatsūtim parabrasvarūpiņam I nirucchvāsatayā muktim gatānyā gopakanyakā II

atra janmasahasrair upabhoktavyāni dukrtasukrtaphalāni

viyogaduhkhacintanāhlādābhyām anubhūtānīty uktam I evam cāśesacayapadadyotye

atiśayoktī II (988)

All her sin dissolves through the immense pain of not having him. All her merit evaporates into ecstatic dreams of him. Thus another milkmaid is released into salvation By contemplating the source of the world - Krsna! (80 & 81)314

Here the fruit of good and bad actions, which should take thousands of lives to

experience, is used up in the pain of separation and the joy of contemplation. This

hyperbole is made apparent by the word all (aśesacayapada).

Comments

The scene of the poem comes from one of the favorite parts of the Krsna myth. While the

avatāra is young, he dallies with milkmaids, who fall deeply in love with him. The poet

314 From the Visnupurāna, according to Dwivedi.

259

uses the love to suggest the ornament of hyperbole. The use of the word "all" adds force

to the ornament.

The poem is based on the belief that one will suffer for one's bad deeds and be

rewarded for the good ones, either in this life or a future one. During the process of

atonement one commits new deeds that cause the karmic process to continue. Escape is

only possible if one can atone for all one's past without accumulating new karma. Here

the girl suffers so intensely and experiences so much joy that she burns all her karma

instantly, thus being released from the world. The similarity between erotic love and love

of a god is a theme familiar to Western poetry as well (e.g., St. John of the Cross and St.

Theresa). If the two types of love were not so closely associated in the Indian tradition,

the poem would lose much of its force.

Next a poem in which a self-existent ornament suggests a fact.

kśanadāsāv aksanadā vanam avanam vyasanam asyasanam | bata vīra tava dvișatām parānmukhe tvayi parānmukham sarvam |I

atra śabdaśaktimūlavirodhāngenārthāntaranyāsena vidhir api tvām anuvartate iti

sarvapadadyotyam vastu Il (993)

The pleasure-givers (i.e., the nights) give them no pleasure, The shelterless (i.e., the forest) is their shelter, Their occupation is doing nothing (or watching the sheep). O king! When you are hostile to your adversaries, Everything is adverse to them. (82)

260

Here the fact, "Destiny itself follows your lead, (o king!)" is suggested by the ornament

apodixis and also by the subservient ornament "contradiction" taken with respect to the

primary sense of the words. The suggestion is made apparent by the word everything.

Comments

The poem primarily suggests that destiny itself follows the king's will by means of

apodixis: a figure in which the concluding line(s) of a poem or stanza explains the

proceeding ones. Here the new facts introduced in the concluding two lines of the

translation explain and justify the events of the first three lines. The ornament is

augmented by another ornament, namely the ornament "contradiction." A casual

inspection of the first line of the original will reveal the three "contradictions" that are the

basis for the poetic ornament ("a-" is a privative prefix). One member of each pair,

however, has a second meaning, which I have given in brackets. Thus the contradictions

are only apparent, which is key for this ornament. Compare, "The cat wolfed down the

jumbo shrimp on the mini grand piano." If the contradiction were not only apparent there

would be a blocking of the literal meaning that might trigger metaphor. The

"contradictions" add to the apodixis by suggesting that the king's power is so great that it

brings about the impossible.

The use of the word "everything" (itself a hyperbole) suggests that all the world

and destiny itself side with the king. The apodixis does not depend on this word because

261

the final line could be written with a different word without losing the ornament.

However, it does bring the ornament to light.

Next comes an example of a self-existent ornament suggesting another ornament.

tuha vallahassa gosammi āsi aharo milāņakamaladala I ia navavahuā souna kunai vaanam mahisammuham |1315

atra rūpakeņa tvayāsya muhur muhuḥ paricumbanam tathā krtam yena mlānatvam iti

milāņādipadadyotyam kāvyalingam I esu svatahsambhavī vyañjakah II (997)

"This morning your lover's lips are faded lotus petals." At this, the new bride let fall her face, blushing. (83)

Here the metaphor, based on the words, faded lotus petals, suggests a "poetic cause,"

namely, "Your lovers lips are faded because you kissed them continually."

In these (four previous examples) the suggestor is self-existent.

Comments

This charming example of love-in-enjoyment uses a metaphor in which the two terms

(lips and lotus petals) are objects one finds in the world, thus for Mammata, this is a

"self-existent" metaphor. The metaphor leaves the reader wondering in what why the

315 The Darpanah commentary has the following Sanskrit translation: tava vallabhasya prabhāte āsīdadharo mlānakamaladalam iti navavadhūḥ śrutvā karoti vadanam mahīsammukham II

262

lover's lips are like faded lotus petals. The woman's bashfulness in the final line suggests

that the cause is the young woman's intense kissing.

Gerow explains that "poetic cause" is an ornament only found in Mammata and

opines that Mammata invented it to remedy problems with the more common ornament

"cause" (hetu).316 If this is the case, Mammata does so, in my opinion, to indicate that the

causal relation in a poem should hold some special interest. Here it suggests the passion

of the lovers.

The metaphor is dependant only on the object of comparison. The poem could be

rewritten: "'This morning your lover's lips are worn out.' At this, the new bride let fall

her face, blushing." This would not effect the second ornament and shows that the first

ornament in based on a single compound word (faded-lotus-petals), not on the sentence

as a whole.

In the four previous examples the suggestor was self-existent fact, i.e., a

description of an actual or possible state of affairs. Now Mammata turns to poems in

which the suggestor is invented by the poet, starting with an invented fact that suggests a

further fact. The "facts" invented by the poet do not exist in the world, but are described

as if they did.

rāirsu camdadhavalāsu laliamapphāliuņa jo cāvam | ekacchattam via kuņai bhuanarajjam vijambhamto |1317

316 Gerow (1971:174-175). But, contra Gerow, Mammata does use the common ornament "hetu" (e.g., in poems 86 and 90 below). 317 The Sankerah commentary has the following Sanskrit translation: 263

atra vastunā yesām kāminām asau rājā smaras tebhyo na kiścid api

tadādeśaparānmukha iti jāgradbhir upabhogaparair eva tair niśātivāhyate iti

bhuanarajjapadadyotyam vastu prakāsyate II (1001)

Love, wantonly strumming his bow, so spreads his kingdom That the moonlit nights become his royal parasol. (84)

Here the fact that, "Those ruled by love stay awake enjoying sensual pleasure at

night, because none contradict his command (to do so)," is suggested by the fact

(invented by the poet) expressed in the word kingdom.

Comments

In this erotic poem, the poet imagines a fictitious kingdom ruled by the god of love. By

means of the night sky-parasol metaphor, he implies that Love's sovereignty extends over

the whole earth at night. The description of Love's dominion suggests that those under

the influence of love must obey devote themselves to love-making all night long.

niśitaśaradhiyārpayaty anango drśi sudrśah svabalam vayasyarāle I diśi nipatati yatra sā ca tatra vyatikarametyasamundmisanty avasthāh II

rātrișu candradhavalāsu lalitamāsphālya yaścāpam I ekacchatram iva karoti bhuvanarājyam vijrmbhamāņaḥ ǁ

264

atra vastunā yugapadavasthāḥ parasparaviruddhā api prabhavantīti

vyatikarapadadyotyo virodhah II (1006)

The glances of a nubile girl are Love's whimsical arrows. Mixed feelings spring up wherever her sexy eyes take aim. (85)

Here the fact suggests, by means of the word mixed, the ornament "apparent

contradiction," i.e., that even contradictory emotions appear simultaneously (in those

circumstances).

Comments

Mammata chooses another erotic poem. The poet invents the "fact" that the woman's

glances are Love's arrows. Note that Mammata does not consider this to be a metaphor:

the poet imagines that the god of love really shoots through such eyes. The poet claims

that the arrows arouse mixed feelings in the men they strike. This suggests a poetic

ornament, namely, "apparent contradiction." The expression "mixed feelings," usually

means "opposed feelings" (i.e., "I had mixed feeling about his coming"). But the

contradiction is only apparent, because what arises from love's arrows is a mixture of

various feelings linked both to the erotic and to a sense of dharma (ethical duty) as well.

vārijjamto vi puņo samdāvakadatthieņa hiaena I thanaharavaassaeņa visuddhajāir ņa calai se hāro ||318

318 The Sanketah commentary has the following Sanskrit translation: vāryamāno 'pi punaḥ santāpakadarthitena hrdayena | 265

atra viśuddhajātitvalaksaņahetvalamkāreņa hāro 'navaratam kampamāna evāste iti ņa

calaipadadyotyam vastu II (1009)

Even though repudiated by her enflamed heart, The noble necklace will not abandon Its closest friends - her swelling breasts. (86)

By the use of the words will not abandon in this poem, the ornament "cause" - which

is the nobility (of the necklace) - suggests the fact that the necklace holds fast, even

though trembling (from her heart's suffering).

Comments

The basic idea is that the necklace blocks a potential embrace, for it would cause pain if

crushed against the chest. The poet uses the ornament "cause" to convey the fanciful idea

that the nobility of the necklace causes it to guard the young girl's chastity. In so doing,

the necklace foils the heart, which suffers a burning longing for the embrace. The

necklace trembles as it reveals the tension.319

so muddhasāmalamgo ammillo kalialalianiadeho I

stanabharavayasyena viśuddhajātirna calatyasyā hāraḥ II 319 I have followed the Ādarśah commentary here. It says, "hārakampakatāpena," thus resolving the ambiguity as to what is trembling and why Alternately, the necklace serves the suffering heart by warding off an unwanted embrace (this is what Jha seems to think (1925:99)). And even though constantly scintillating, it does not move. I can see no way to incorporate vāryamāno into this reading. Jha translates "Even though forbidden by the heart ... " This is a possible translation, but nonsense in the context. 266

tīe khamdhāhi balam gahia saro suraasamgare jaai I|320

atra rupakeņa muhurmuhurākarsaņenā tathā keśapāśah skandhayor prāpto yathā

rativiratāv apy anivrttābhilāsah kāmuko 'bhūd iti khamdhapadyotyā vibhāvanā II

eșu kavipraudhoktimātraniśpannaśarīrah II (1013)

Her black braid swirls seductively, becoming the home of Love. Seconded by her shoulders, Fearless it faces passion's duel. (87)321

In this poem, the word reinforced is metaphorically employed to suggest the ornament

"unusual cause," namely, that the lover has repeatedly pulled her braid such that it has

reached her shoulder and that (her shoulders are so sexy that) the lover is not sated, even

after sex.

In the four preceding examples the suggestion is based on the imagination of the

poet.

Comments

The erotic rasa, which is primary, is enhanced with the metaphoric use of "reinforced,"

whereby a military theme is introduced into the poem. The woman's shoulders provide

320 The Sanketah commentary has the following Sanskrit translation: sa mugdhaśyāmalāngo dhammillaḥ kalitalalitanijadehaḥ I tasyāḥ skandenā balam labdhvā smaraḥ suratasamgare jayati 321 The Sanskrit uses the word for shoulder, which also denotes a division of an army, to ground a erotic/military metaphor. "Seconded" is the best I could find in English, altering the metaphor to one of dueling. Some metaphor must be present, of course, for the example to work. 267

extra sexual artillery to her already powerful hair, enticing the man to desire more

lovemaking. The ornament of "unusual cause" occurs when an effect is present without

its normal cause. The poem's context is that the couple has been making love and

furthermore that normally the man should be sated. The standard conditions for sexual

desire are thus absent, and yet desire is present. Thus the reader searches for the unusual

cause, i.e., the extraordinary attractiveness of the woman's hair and shoulders.

Mammata turns now to poems in which words spoken by characters suggest

further information (either facts or ornaments).

ņavapunņimāmiamkassa suhaa ko ttam si bhanasu maha saccam I kā sohaggasamaggā paosaraaņi vva tuha ajja I|322

atra vastunā mayīvānyasyām api prathamam anuraktastvam na tata iti ņavetyādi

paosetyādipadadyotyam vastu vyajyate II

O happy one! Tell me truly, are you the full moon, newly risen? And who, like the early evening sky, do you now love? (88)

Here the fact (that the lover has a new beloved) suggests, "You will love her only for a

while, like you did me." This is done by the use of words like newly and early.

Comments

322 The Sanketah commentary has the following Sanskrit translation: navapūrņimā śaśānkasya subhagaḥ kas tvam asi bhaņa mama satyam | kā saubhāgyasamagrā pradoșarajanīva tavādya II 268

Mammata separates poems in which the poet speaks in the first person from those in

which he speaks through his characters. In the following four poems the poet speaks in

the third person. Also, the "facts" discussed here are descriptions of possible states of

affairs, not necessarily actual states of affairs. They contrast with ornaments, not with

fictions. Fictitious descriptions of states of affairs are considered facts, being, so to say,

of the same genus as facts.

Mammata begins with a poem in which a woman chides a man who was her

lover. Her fictional description suggests his fickle nature. Jha explains as follows;

"During the early part of the evening the full moon appears 'red'; this redness is regarded

figuratively as representing the moon's love toward the evening; it is fleeting,

disappearing with the advent of night; and the indignant wife insinuates that her

husband's love for this new found lady will be as fleeting."323 It is hard to know whether

"happy one" is used ironically or not.

sahi ņavaņihuvaņasamarammi amkavālīsahīe ņibidie I hāro nivārio via uccheranto tado kaham ramiam ||324

atra vastunā hāraccedānantaram anyadeva ratam avaśyam abhūt tat kathaya kīdrg iti

vyatirekah kahampadagamyah II (1022)

323 Jha (1925:100). 324 The Darpanah commentary has the following Sanskrit translation: sakhi navanidhuvanasamare aņkapālisakhyā nibidayā | hāro nivārita evocchriyamānastatah katham ramitam II (My copy of the Sanketah drops the second word of the first line. Given the sense and the meter, this must be a copying error.) The Sāhityacūdāmanih changes ucchriyamanah to udvartamanah, which may better capture the sense. 269

Early in love's struggle Your necklace snapped Under your man's tight embrace! What happened next? (89)

In this poem the fact (described by the female friend suggests the poetic ornament)

"dissimilitude." From the term what one understands, "after the snapping of the

necklace, the love-play must surely have been different. What was it like?"

Comments

Here the fiction described by the woman suggests the poetic ornament "dissimilitude."

The woman suggests an inverted simile: the passion of the couple after the incident

described was unlike the passion before it occurred. The point of this ornament is to

convey the comparative excellence of the later passion.

pravisamtī gharavāram vivaliaanā viloiuņa paham I khamdhe ghettūna ghami hā hā naththotti ruasi sahi kim ti I1325

atra hetvalamkāreņa samketaniketanam gacchantam drstvā yadi tatra gantum icchasi

tadā aparam ghatam grhitvā gaccheti vastu kimitipadadyotyam II (1026)

As you came in the door, Your eyes turned,

325 The Sampradāyaprakāśinī commentary has the following Sanskrit translation: praviśāntī grhadvāram vivalitanayanā vilokya panthānam skandhād grhītvā kuțam hāhā nașta iti rodiși sakhi kim iti I|

270

Glancing back at the path Having shouldered a water pot, You wail, "Damn, it's broken!" Why do you weep, my friend? (90)

By the use of the words why do you weep in this poem the ornament "cause" suggests

the fact, "Knowing that he has gone to the rendezvous, if you want to go there, grab

another pot and go!"

Comments

When the girl's friend asks why she weeps, the implication is both that she seems to do so

because her water pot is broken and that there has to be a second reason. A condition

(noticing the broken pot) and its effect (weeping) described together constitute the

ornament known as "cause." Often there is something striking or irregular about such a

cause.320 The friend uses the ornament to suggest that the woman is really weeping

because she wants to go to her rendezvous at the riverside. She has just seen her lover on

his way and the water pot that was supposed to justify her trip is broken. However, she

can still go by taking another pot. This example shows, to my mind, that the "fact"

suggested can be rather complex.

yathā vā

vihatamkhalam tumam sahi dattūna kudena taralataraditthim vārappham samiseņa ā appā guruotti pādia vihino ||327

326 Gerow (1971:327). 327 The Sanketah commentary has the following Sanskrit translation: 271

atra nadīkūle latāgahane krtasamketam aprāptam grhapraveśāvasare paścād āgatam

drstvā punar nadīgamanāya dvāropaghātavyājena buddhipūrvam vyākulatayā tvayā

ghatah sphoțita iti mayā cintitam tatkim iti nāśvasiși tatsamīhitasiddhaye vraja aham te

śvaśrūnikate sarvam samarthayisye iti dvārasparśanavyājenety apahnutyā vastu II (1030)

Or this example (of an ornament suggesting a fact):

Seeing your exasperation, My trembling-eyed friend, Your water pot, thinking itself too heavy, Pushed off from the door to fall, smashing. (91)

Here the fact is: "You made a rendezvous in the canes by the riverbank (and so went to

get water). Just when you were coming back into the house, you saw him finally coming.

Wanting to return to the river, you have broken the water pot by pretending to hit it

against the door. All this I understand. Why are you not confident? Go fulfill your

desires! I will explain everything to your mother-in-law." (All of) this is suggested by the

ornament "denial" in the phrase, Pushed off from the door.

viśrkhalām tvā sakhi drstvā kuțena taralataradrsțim | dvāraparśamișeņa cātmā guruka iti pātayitvā bhinnaḥ Il The Sampradāyaprakāśinī commentary uses vihvalām ksanam in place of viśrnkhalām, which tones down the poem considerably. It also replaces pātayitvā bhinnah with pātitavibhinnah, which only seems to obscure the image. 272

Comments

This poem is a second example of a fact being suggested by an ornament. It gains charm

in Sanskrit by the ambiguity of the first line. Is the girl tired from carrying the pot and

thus exasperated by her clumsiness, or is she exasperated because she wants to go back to

the river? The key word (if we can trust the commentators) is "unchained," which can

mean "undone" or "uninhibited." The first meaning is more natural, but the second is

needed to explain the feeling of the water pot. Thus this word sets in the reader's mind the

idea that the girl is in a state of heightened passion. Unfortunately, I fear the reverse is

true for "exasperate:" one thinks first of the meaning "taxed to the limit" and only when

pushed by context of the meaning "to increase the intensity (of a passion or pain)."

The ornament "denial" is defined according to Gerow as "a figure in which the

object of comparison is affirmed in place of the subject of comparison."328 He quotes the

following verse from Bhamaha as an example: "It is not a swarm of bees, humming

incessantly of honey; it is the sound of the Love-hunter's bow being drawn." Here the

sound of the bees is likened to that of Love's bow, but the poet adds the trope of denying

the subject (the bees) and affirming the object (the bow).

The poem offers a rather tenuous example of this ornament, for there is no real

comparison between the young wife (subject) and the pot (object). Rather the action of

dropping the pot through knocking it against the door is likened to the pot knocking itself

against the door in order to smash. Of course, the pot, being inanimate cannot knock itself

against the door. Thus by assigning the action to the pot rather than to the woman, the

328 Gerow (1971:110). 273

speaker suggests that she knows the woman's secret motive for breaking the pot.

Furthermore, she suggests that she will play along.

johaņāi mahuraseņa a viiņnatāruņnautsuamanā sā l buhua vi ņavodhavvia parahuā ahaha harai tuha hiaam ||329

atra kāvyalingena vrddhām paravadhūm tvam asmānujjhitvābhilașasīti tvadīyam

ācaritam vaktum na śakyam ity āksepah paravahūpadaprakāśyah II

esu kavinibaddhavaktrpraudhoktimātranispannasarīrah I vākyaprakāśye tu pūrvam

udāhrtam śabdārthobhayaśaktyudbhavas tu padaprakāśyo na bhavatīti

pañcatrimśadbhedāh II (1034)

The moonbeams and wine Lend to her a newlywed's youthful lust Even though she is so very old. Alas! Another's wife has won your heart. (92)

The ornament "poetic reason" suggests the ornament "hinting" by the words another's

wife. The hint is, "You desire another's old wife and abandon us (young women); your

action is unexplainable."

In these last five examples, the suggestion is established by an expression of a

poetic character. Suggestion arising from a sentence has already been illustrated, and that

329 The Sanketah commentary has the following Sanskrit translation: jyotsnayā madhurasenāvatīrna tārunyotsukamanāḥ sā l vrddha 'pi navodheva paravadhūr ahaha harati tava hṛdayam Il

274

arising from the power of both the word and the meaning is never found in the word

(alone). Thus there are thirty-five types.

Comments

The ornament "poetic reason" implies that the man has fallen for the woman only

because she is another man's wife, and thus taboo.50 This is done by denying her the

qualities that usually captivate men: youth and beauty. The hint is that such a reason is

not the proper one, and therefore incomprehensible to the young beauties.

Mammata concludes the presentation of suggestion by a single word (or Sanskrit

compound). Of the original eighteen types of suggestion by phrases or sentences, only

seventeen can be brought about by a single word. Together they make the thirty-five

types Mammata mentions.

prabandhe 'py arthasaktibhūḥ II (1039)

yathā gṛdhragomāyusamvādādau I

alam sthitvā śmaśāne 'smin grdhragomāyusamkule I kankālabahale ghore sarvaprāņibhayamkare II na ceha jīvitah kaścit kāladharmam upagatah I priyo vā yadi vā dvesyah prāninām gatir īdrśī I

iti divā prabhavato grdhrasya purusavisarjanaparam idam vacanam

ādityo 'yam sthito mūdhāh sneham kuruta sāmpratam I bahuvighno muhūrto 'yam jīved api kadācana II

330 The other possible interpretation, that the man seeks to hurt his enemy by seducing his wife, is made improbable by two details: a seducer's heart would not be "won" and the other woman might well understand such plotting. 275

amum kanakavarņābham bālam aprāptayauvanam I gṛdhravākyāt katham mūdhāsty ajadhvam aviśankitāh II

iti niśi vijrmbhamāņasya gomāyor janavyāvartananistham ca vacanam iti prabandha eva

prathate l anye tv ekādaśa bhedā granthavistarabhayān nodāhrtāh svayam tu laksanato

'nusartavyāḥ l apiśabdāt padavākyayoḥ II (1039)

(4.6. Suggestion Based on Context.)

42d. Suggestion based on meaning exists in context also.

For example, in the following dialogue of a vulture and a jackal:

Skeletons crowd these burial grounds, Vultures and jackals lurk everywhere. Everyone is scared here - you shouldn't stay! Neither the loved nor the hated Come back from death Such is the fate of mortals. (93&94)331

These words aim to make people leave. They are spoken by a vulture, who is diurnal.

Fools! The sun holds fast. Now is the time for love! The present moment is inauspicious, But later the dead might be revived. How could you rashly leave this boy, This golden skinned bud of youth, On the words of a mere vulture? (95&96)332

These words aim to make the people stay. They are spoken by a jackal, who is nocturnal.

The suggestions appear only in the context.

331 Mahābhārata, Śāntiparvan, according to Dwivedi, but I have not found this dialogue. 332 Mahābhārata, Śāntiparvan. Ditto previous note. 276

The eleven further kinds (of suggestion by context) are not illustrated for fear

prolixity. However, they can easily be produced from the definition itself.

The word also means, "(in addition to being found) in the sentence and the word."

Comments

There are only twelve types, and not eighteen, because only the twelve types of

suggestion whose sequence is noticed and which arise from meaning can arise from the

meaning of the context. It might prove helpful to keep the list of the eighteen types ready

for the following discussion. It is found at the very beginning of this chapter.

The two types of suggestion based on metaphor (types 1 & 2) are excluded by the

very nature of metaphoric indication. Context is considered part of the literal meaning of

a text. Metaphoric indication comes into play when the literal meaning is blocked. When

the context is blocked, it cannot also be suggestive. The fact that the contextual

information is often used to arrive at the metaphoric meaning (as in irony) does not affect

this point. These two types of suggestion treat the suggestiveness of the metaphoric

meaning. That such poems might be otherwise suggestive and thus capable of serving as

examples of other types of suggestion is not relevant. The metaphoric meaning, strictly

speaking, is not contextual.

We saw that suggestions of the third type, suggestions whose sequences are not

noticed, are akin to moods that permeate a poem. Mammata probably excludes this type

because Anandavardhana had argued that all contextual understanding has a noticed

277

sequence.333 For example, in the poems, one first understands the words of the vulture

and jackal, and then, with the help of the context, the suggestion that the people should

leave or stay. This marked sequence is absent in suggestions of rasa, etc.

Types four, five, and eighteen depend of punning words, so they are excluded as

well. Once again, the fact that contextual information might be necessary if one is to

understand the puns does not vitiate Mammata's point. It is the punning word itself that is

suggestive, not the various factors that are necessary to understanding it.

This leaves types six to seventeen, which were illustrated above both for sentence

meaning and word meaning. Both the vulture and the jackal describe "facts" (situations)

that suggest "facts" (courses of action). Thus these are examples of the fourteenth type on

the list at the beginning of the chapter. Mammata claims that given the definition (and the

poems already given), it would be easy for an enthusiastic reader to find examples of the

other eleven types.

Suggestion is not easily mapped onto the syntactic/semantic/pragmatic division of

Western linguistics. As Mammata's mention of context shows, it seems to include parts

of both semantics and pragmatics. Both Bharata and Anandavardhana go into great detail

about the suggestiveness of larger semantic units. Indeed, in its earliest recorded aesthetic

uses, rasa refers to the mood of an entire play. Bharata often explains how details can

contribute to the overall rasa. Anandavardhana also addresses these issues.334 However,

reflecting his interest in short poetry, Anandavardhana treats primarily of suggestive

333 Ānandavardhana, Dhvanyāloka 3.33b, translated in Ingalls, et al. (1990:542). 334 Dhvanyāloka 3.10-14. 278

sentences or verses. However, he devotes several long sections to the suggestive

possibilities of texture (sanghatanā) and style or structure (racanā).335 Maybe because

shorter verse had gained prominence in poetics, Mammata omits these topics and turns

instead to even smaller semantic units: parts of words. In chapter eight, he continues the

discussion of the suggestiveness of certain sound combinations (alliteration, etc.). Finally,

in chapter nine he focuses on the relative merit of different regional dialects and their

suggestive power.

padaikadeśaracanāvarņesv api rasādayaḥ II (1046)

tatra prakṛtyā yathā

raikelihianiasanakarakisalaaruddhanaanajualassa ! ruddassa taianaanam parvvaiparicumbiam jaai |1336

atha jayatīti na tu śobhate ityādi I samāne 'pi hi sthaganavyāpāre lokottareņaiva

vyāpāreņāsya pidhānam iti tad evotkrstam II

(4.7. Suggestion Based on Word Parts.)

43ab. Rasa, etc. are (found) in parts of words, in style, in letters, etc.

Here is an example (of suggestion found) in the (verbal) stem:

Stripped of her clothes during sex, Pārvatī covers Siva's eyes with her lotus like hands. But his third eye, blinded with passionate kisses,

335 E.g., Dhvanyāloka 3.5-8 and 3.33. 336 The Sanketah commentary has the following Sanskrit translation: ratikelihrtanivasanakarakisalayaruddhanayugalasya | rudrasya trtīyanayanam pārvatīparicumbitam jayati Il 279

Triumphs. (97)337

Here triumphs is used instead of "shines" or some other synonym. Even though (all

three eyes) are similarly closed, the method of closing (the third) is extraordinary; thus it

outdoes the others.

Comments

"In parts of words" means in the choice of root or stem, in the choice of conjugation or

number, and in the choice of prefixes and/or affixes. Anandavardhana writes, "suggested

meaning ... shines forth in phonemes, words, etc., as well as in a sentence, in texture, and

in a complete work."358 Mammata's exposition follows the reverse order. As we saw, he

ignores the suggestiveness of a complete work and that of the texture and structure of

larger passages. Chapter four began with suggestive sentences, continued with suggestive

words, and will now finish with suggestive parts of words. Whereas Anandavardhana

treated suggestive phonemes apart from suggestive prefixes and conjugations, Mammata

groups them all together in this section.

The erotic poem here uses a particular stem to effect its suggestion. The poet

could have chosen many adjectives to describe Siva's third eye; that "triumphs" is used

suggests that the eye is not only victorious in its normal ways, but also luckier than the

other two. This in turn suggests the value of Pārvatī's kisses.

337 Gāthāsaptaśatī 5.55. 338 Dhvanyāloka 3.2, translated by Ingalls, et al. (1990:389). 280

Taxonomically, this is the least obvious of the examples from the section because

it seems to turn on whole words ("triumphs" rather than "shines"), not parts of words.

However, as Sanskrit verbs conjugate, any synonymous verb stem would have roughly

the same ending. In other words, only the first half of the word would change if the verb

"triumph" were replaced by "shine."

yathā vā

preyān so 'yam apākrtah saśapatham pādānātah kāntayā dvitrāņy eva padāni vāsabhavanād yavan na yāty unmanāḥ | tāvat pratyuta pānisamputagalannīvīnibandham dhrto dhāvitv eva krtapranāmakam aho premno vicitrā gatih II

atra padānīti na tu dvārāņīti | (1055)

Or this example (of suggestion by the choice of the stem):

She scorns her lover groveling at her feet, But when he goes from her chambers, Not even two or three steps, She runs, dropping the knots of her loosened skirts, And falls worshipfully at his feet. Such are the wonderful ways of love! (98)339

Here (the suggestion comes from the use of) steps and not "doors".

Comments

In this erotic poem, a self-existent fact suggests a fact. The phrase, "not even two or three

steps," suggests that the lady, no less than the groveling man, is full of intense passion. If

339 Dwivedi writes, "Cited in Vāmana's KASV." 281

she had waited until he had reached the doors, instead of only a few steps, it would show

greater patience, and thus less passion.

As in the last example, the poet could have substituted other stems without

changing the syntax of the sentence. As several different stems could have similar

endings, only part of the word would change. Mammata suggests substituting "doors"

(dvārāni) for steps (padāni). The form of plural accusative ending (-āni) is the same in

both words, so only part of the word would change.

pathi pathi śukacañcūcārurābhānkurānām diśi diśi pavamāno vīrudhām lāsakaś ca l nari nari kirati drāk sāyakān puspadhanvā puri puri vinivrttā māninīmānacarcā II

atra kiratīti kiraņasy sādhyamānatvam nivrtteti nivartanasya siddhatvam tinā supā ca

tatrāpi ktapratyayenātītatvam dyotyate l (1059)

O the paths! the paths that shine with tender shoots as brilliant as parrot beaks! From the east, from the west, the winds make dance the vines to and fro. Man after man receives swift arrows showering from the flowered bow. In town after town honorable women have quit speaking of honor renounced. (99)

In this poem the progressive ending (-ing) in showering shows that the shower (of

arrows) is in mid-process. Likewise the pluperfect tense in have quit shows that the

quitting is already finished. (The having finished is also shown another way in Sanskrit).

282

Comments

I have tried to translate or hint at the line-initial repetitions of the Sanskrit, which might

sound more odd in English than they do in Sanskrit. However, some of the charm of the

poem derives from this unusual construction, which occurs in other traditions as well.

Ezra Pound does a wonderful job translating a Chinese poem of similar construction that

he calls, "The Beautiful Toilet." He begins, "Blue, blue is the grass by the river / And the

willows overfill the closed garden."340

The verb tenses suggest that love-play has begun and will continue. One could

substitute "have stopped" or "have let off" for "have quit" without loosing the suggestion.

And as in the English, only part of the compound verb would change.

yathā vā

likhann āste bhūmim bahir avanatah prānadayitah nirāhārāh sakhyah satataruditocchūnanayanāḥ I parityaktam sarvam hasitapathitam pañjaraśukaih tavāvasthā ceyam visrja kațhine mānam adhunā II

atra likhann iti na tu likhatīti tathā āste iti na tu āsīta iti api tu prasādaparthantam āste

iti bhūmim iti na tu bhūmāv iti na hi buddhipurvakam aparam kiñcil likhatīti

tinsubvibhaktīnām vyangyam | (1064)

The love of your life sits, scratching the dust with head bowed, Your friends won't eat, their eyes swollen from endless weeping, Even the parrots have abandoned their funny calls:

340 Pound (2003:249). 283

This is your situation. Cast off you anger, cruel girl! (100)341

In this poem the conjugations and declensions are suggestive: scratching (the present

participle) is used and not "scratches" (to suggest an ongoing activity), sits is used and

not "is seated" to suggest that he will stay until you show favor. The dust is used instead

of "in the dust" to suggest that he is not writing something deliberate.

Comments

Mammata shows that three of the words could have been conjugated or declined

differently. The poet uses these particular forms to suggest that the lover is penitent and

the woman should forgive him and give her love again. The verb forms suggest that the

lover has been seated for some time and will go on doing so. The poet has the lover idly

scratching the dust itself, not scratching something in the dust. The use of the accusative

instead of the locative with the word "dust" effectively removes the intentionality from

the scratching.

sambandhasya yathā

gāmāruhammi gāme vasāmi ņaaraththiim na jānāmi | ņāariānam paiņo haremi jā homi sā homi ||342

341 Amaruśataka 7. 342 The Sanketah commentary has the following Sanskrit translation: grāmaruhā 'smi grāme vasāmi nagarasthitim na jānāmi | nāgarikāņām patīn harāmi yā bhavāmi sā bhavāmi II

284

atra nāgarikāņām iti șasthyāh I

ramaņīyah kśatriyakumāra āsīt iti kālasya l eșā hi bhagnamaheśvarakārmukam

dāśarathim pratikupitasya bhārgavasyoktih I (1068)

Now an example of (the suggestiveness of) a relation:

I was raised in a village, I live in a village, I don't know the ways of the city. I am what I am I steal the men of city women. (101)

Here of city women is in the genitive case.

In the verse, "The warrior prince was lovely," the past tense (is suggestive). This

was said by the son of Bhrgu to the son of Daśaratha after the latter had broken Siva's

bow.

Comments

Here the genitive case is used to show possession: the men belong to the city women,

either as spouses or lovers. By means of the usage, the poet suggests that despite the

common impression that city women are superior, this country girl outdoes them by

seducing their men. Contra Jha, I doubt it is primarily the girl's cleverness that is being

suggested. I take it that the girl, by acknowledging her inferior cleverness, suggests that

such cleverness is not all important in sexual matters.343

343 Jha (925:109). 285

Paraśurāma, the son of Bhrgu, uses the past tense to refer to Rāma, the son of

Daśaratha, to suggest that he is as good as dead. The story comes from the famous epic,

the Rāmāyana. Rāma has broken Siva's bow. Thus enrages Paraśurāma, who swears to

kill Rāma in revenge. Anandavardhana uses an episode from the same poem to explain

the suggestiveness of declinations and conjugation in the Dhvanyāloka, 3.16.

vacanasya yathā

tānam gunaggahanānam tānukkamthānam tassa pemmassa I tāņām bhanianam sumdara erisiam jāamavasānam ||344

atra gunagrahanādīnām bahutvam premņaś caikatvam dyotyate | (1074)

An example of (the suggestiveness) of grammatical number:

O beautiful beloved! My adorations, My longings, My soft words, And my love All find their home in you. (102)

In this poem the plural of adorations, etc., and the singular of love are suggestive.

Comments

In grammar, vacana means grammatical number. I believe that the plural suggests that

each individual act of admiring, etc., has a beginning and an end, whereas the singular

344 The Sanketah commentary has the following Sanskrit translation: teșām̧ guņagrahaņānāņ tāsāmutkaņțhānām tasya premņaḥ teșām bhaņitānām sundara irdrśam jātamavasānam l 286

suggests the unchanging nature of the love. It could also suggest that the man has but a

single beloved. In either case, it is the grammatical number that grounds the suggestion.

Examples of the suggestiveness of number are easy to find in English. Compare "He

loves his woman." And, "He loves his women."

puruşavyatyayasya yathā

re re cañcalalocanāñcitaruce cetah pramucya sthirapremānam mahimānameņanayanām ālokya kim nrtyasi | kim manye viharisyase bata hatām muñcāntarāśām imām esā kanthatate krtā khalu śilā samsāravārāmnidhau II |(1078) atra prahāsaḥ I (1078)

An example of (the suggestiveness) of change in grammatical person:

O heart, lover of the tremulous eyed maiden! Why renounce the glory of abiding devotion And dance at the sight of this gazelle-eyed one? Are we thinking of wandering for pleasure? Fool! Give up this despicable craving, This drowning-stone in the ocean of the world! (103)

Here ridicule is suggested.

Comments

According to the Bālacittānurañcanī commentary, Pāņini 1:4:106 states that a shift into

the first person when referring to someone else implies ridicule. The Sanskrit has, "do I

287

think that you will wander for pleasure," where both "I" and "you" refer to the heart, not

the speaker. Through suggestion, the poet is mocking his heart's claim to control him.

To attempt this effect in English, I have used the plural "we" to refer to the heart

instead of the singular "you." Compare: "Are we thinking naughty thoughts?" when said

to a child.

pūrvanipātasya yathā

yeşām dorbalam eva dorbalatayā te sammatās tair api prāyah kevalanītirītiśaraņai kāryam kim urvīśvaraih ye kşmāśakra punah parākramanayasvīkārakāntakramās te syur naiva bhavādrśās trijagati dvitrāh pavitrāh param I

atra parākramasya prādhānyam avagamyate | (1082)

Now an example of (suggestion based on) irregular order (of the words in a compound):

Those armed only with their arms are weak, And what is accomplished by diplomacy alone? O Indra incarnate! In the three worlds there are only a few Who in both valor and politics like you excels! (104)

Here the superior value of valor is understood.

Comments

The superior value of valor is suggest by the fact that the word for "valor" precedes the

word for "diplomacy" in the compound even though it is longer, and would thus normally

come second. The poem flatters a king by likening him to Indra, the king of heaven. The

"three worlds" are heaven, the intermediate realm, and earth. 288

vibhaktiviśesasya yathā

pradhanādhvani dhīradhanurdhvanibhrti vidhurair ayodhi tava divasam | divasena tu narapa bhavān ayuddha vidhisiddhasādhuvādapadam II

atra divasenety apavargatṛtīyā phalaprāptim dyotayati II (1085)

The following shows a special use of case-ending (to effect a suggestion):

On battlegrounds twanging with bravely born bows, Your foes did fight all the day long. But by this day, O protector of men, You won praise from the gods and the good! (105)

In this poem, the instrumental case ending in the expression by this day suggests that the

(king's) goal was achieved.345

Comments

The instrumental case communicates that the day was used by the king. That the king was

able to use the day suggests, without explicitly stating it, that the king won the battle. In

the original, the immediate repetition of the "day" at the line end and the line beginning

positions (divasam / divasena tu) highlights the contrast between the struggling enemies

and the victorious king. This type of suggestion is subtler in Sanskrit, where the case

ending have less poetic presence than prepositions do in English.

bhūyo bhūyah savidhanagarīrathyayā paryatantam

345 The Bālavittānurañjanī quotes Pāņini 2.3.6 [apavrge trtīyām] to underpin this example. 289

drstvā drstvā bhavanavalabhītūngavātāyanasthā I sāksātkāmam navam iva ratir mālatī mādhavam yat gādhotkanthālulitalulitair angakais tāmyatīti II

atrānukampāvrtteḥ karūpataddhitasya (1088)

From the high windows of her palace, Malitī watches Mādhava, Like the goddess Passion watching the young god Love. Each time he passes in the street below, Her slender legs tremble more. (106)3 346

Here the use of the nominal suffix ka (added to the word for "legs") signifies pitifulness

(with regards to the legs).

Comments

The use of the particular suffix to signify weakness, together with the situation, suggests

that Mälitī's passion is so great that it weakens her limbs. Compare the English

expression, "become weak at the knees."

The Bālavittānurañjanī quotes Pāņini 5.3.67 [anukampāyām] to explain how the

suffix "ka" can have this meaning.

paricchedātītah sakalavacanānām avisayah punarjanmany asminn anubhavapatham yo na gatavān l vivekapradhvamsād upacitamahāmohagahano vikārah ko 'pyantar jadayati ca tāpam ca kurute II

atra praśābdasyopasargasya l (1091)

346 Mālatīmādhava 1.15 (D). 290

My passion burns and freezes me! My reason crumbles away, Deep stupor sets in steadily.

Never in my life have my feelings gone Beyond the power of words - And even beyond understanding. (107)347

The pra- prefix (in "pradhvamsāď" - "away-crumbles") is suggestive.

Comments

The Sanketah commentary explains that the prefix suggests that the suffering described is

rooted in the mind.348 Jha says it suggests "the erotic in separation."3 These two

opinions can be combined: the suggestion is that the cause of the suffering is mental and

thus an absent lover. The prefix, which very generally connotes "forward-, forth-,"

suggests that the lover is away.

krtam ca garvābhimukham manas tvayā kim anyad evam nihatāś ca no dvisah I tamāmsi tisthanti hi tāvad amśumān na yāvad āyāty udayādrimaulitām II

atra tulyayogitādyotakasya ca iti nipātasya l (1094)

You set your mind upon glory And your enemies were destroyed. What else was there? When sunbeams summit the eastern peaks

347 Mālatīmādhava 1.30 (D). 348 Mohan (1995:1091). 349 Jha (1925:112).

291

Darkness is no more. (108)

The particle and suggests the ornament "equal joining" (tulyayogitā).

Comments

"Equal joining" is an ornament in which two things are equated in order to praise the

lesser. Here the king is equated with the sun to praise his might. Just as the sun

effortlessly destroys the darkness with its mere presence, the king had but to set his mind

on glory and his enemies were destroyed. This ornament suggests the effortlessness of the

king's victory and thus his extreme prowess. The final result is the evocation of the c rasa.

rāmo 'sau bhuvanesu vikramagunaih prāptah prasiddhim parām asmad - bhāgyaviparyayād yadi param devo na jānāti tam bandīvaisa yaśāmsi gāyati marud yasyaikabāņāhatiśreņībhūta- viśālatālavivarodgīrnaih svaraih saptabhih II

atrāsāv iti bhuvanesv iti sarvanāmapratipadikavacanānām na tvad iti na mad iti api

asmad ity asya sarvāksepiņah bhāgyaviparyayād ity anyathāsampattimukhena na tv

abhāvamukhenābhidhānasya l (1097)

The valorous virtues of Rama, the Rāma, Across many lands resound. By a twist of our fate, you do not know him, But his single shaft pierced those seven ancient palms Fashioning a flute for the wind. Listen, my lord,

292

Like a minstrel, it plays the seven tones to glorify him! (109)350

There are several suggestive elements in this poem: the use of the Rāma, the use of

across many lands, the use of virtues, the use our (fate), which implies "all," rather

than "your" or "my," and in the expression twist of fate, which does not imply a lack of

fate, but rather that our fate has gone astray.

Comments

The character in the poem address Rāvana, the demon king. By means of extreme

austerities, Rāvana provoked the supreme deity to grant him invulnerability from the

gods, demons, and other immortals. Through distain, he failed to ask for protection from

men. To curb his increasing arrogance, Nārayana accepts to be born in human form as

prince Räma. In this form he is destine to fight and kill Rāvana. Rāvana first hears of him

after Rāma single-handedly slays a great army of demons.

The poet uses the phrase "the Rama" to suggest that he is already well known.

There is a similar usage in English. Consider the following dialogue: "Bill Clinton was

there. Which Bill Clinton? The Bill Clinton." That his virtues resound not just in a limit

place, but, "across many lands," suggests that Rāma has already achieved great feats of

heroism. The use of the plural in "virtues" suggests that Rāma's capacities are numerous.

That it is not accidental that Rāvana does not know Rāma is suggested by the expression

"twist of fate." That "our fate" is spoken of, not Rävana's fate or the character's fate,

350 Found in the Rāghavānandanātaka, according to Dwivedi. 293

suggests that Rävana's whole army will suffer from fate's working. These suggestions

work together to evoke the heroic rasa.

taruņimani kalayati kalām anumadanadhanurbhruvo pathaty agre l adhivasati sakalalalanāmaulim iyam cakitahariņacalanayanā II

atra imanijavyayībhāvakarmabhūtādhārānām svarūpasya taruņatve iti dhanusah samīpe

iti maulau vasatīti tvādibhistulye esām vācakatve asti kaścit svarūpasya viśeso yaś

camatkārakārī sa eva vyañjakatvam prāpnoti | (1102)

Eyebrows taught by Love's own bow To mimic the skittish eyes of a doe. She is beyond comparison In youth's full flush! (110)

In this poem, suggestion appears from the use of certain grammatical forms: the -imanic

affix, the use of an avyayībhava compound, and the use of the locative as an accusative.

The same meaning could be communicated by other forms, but the forms chosen by the

poet possess a certain beauty.

Comments

The forms add to the charm of the poem, but this is not the main point. In an erotic poem

like this, the forms are aesthetically justified because they are more effective in

suggesting the erotic rasa. The Bālacittānurañjanī commentary explain the details of the

individual suggestions in the Sanskrit, which presuppose too many Sanskrit grammatical

categories to be usefully elucidated here. The forms work together to suggest the intensity 294

of the lover's emotion. A comparable effect in English is found when a lover says to his

beloved, "You are my world," rather than, "You are the most important person in my

world."

evam anyesām api boddhavyam I varņaracanānām vyañjakatvam guņasvarūpanirūpane

udāharisyate l apiśabdāt prabandhesu nātakādisu l evam rasādīnām

pūrvagaņitabhedābhyām saha șadbhedāh I

(The suggestiveness) in other (forms of words) should be understood likewise.

The suggestiveness of certain sound combinations (varna) and styles (racanā)

will be described in the chapter on qualities (chapter eight). The word etc. (in kārikā

43ab.) means that rasa, etc., are found in literary works and in dance. Thus, along with

the two types already described, there are six types of suggestion of rasa, etc.

Comments

The first two types were: suggestion arising from a word and that arising from a sentence.

In the two types we have just seen suggestion arises from context and from parts of

words. Adding the two types that will be examined in chapter eight (suggestion from

sound combinations and regional dialects), we arrive at six, the number mentioned.

295

bhedās tad ekapañcāśat I

vyākhyātāh ll

tesām cānyonyayojane I

samkareņa trirūpeņa samsrstyā caikarūpayā I

na kevalam śuddhā evaikapañcāśadbhedā bhavanti yāvat tesām svaprabhedair

ekapañcāśatā samśayāspadatvenānugrāhyānugrāhakatayaikavyañjakānupraveśena ceti

trividhena samkareņa parasparanirapeksarūpayaikaprakārayā samsrstyā ceti caturbhir

gunane I

vedakhābdhiviyaccandrāh

śuddhabhedaih saha

śaresuyugakhendavaḥ II (1107)

tatra dinmātram udāhriyate I

(4.8. Poems with More than One Suggestion.)

43c. There are fifty-one types of suggestion.

They have been explained.

43d - 44abc. When these fifty-one are squared by means of the three forms of fusion

and that of association they come to number 10,404.

There are not only the pure types. Each type can be combined with (a second suggestive

factor of any of) the fifty-one types. These combinations are possible in four ways: when

it doubtful which (suggestive factor) is dominant, when one factor serves the other, when

296

both work equally together, and when both are independent, thus multiplying the number

(of suggestive types) by four.

When these are combined with the pure types,

44d. (The total number of types of suggestion) are 10,455.

These will be only partially illustrated.

Comments

Each of the fifty-one types Mammata discusses can be combined with any other. This

results in fifty-one squared or 2,601 types. However, each combination can be done in

four ways, resulting in 4 times 2,601 or 10,404 types. Add to that the original 51 simple

types and we arrive at Mammata's 10,455 types. The point seems to be that the

suggestive power of words is practically pervasive since there are so many types.

Both of Mammata's examples of complex suggestion are taken from

Anandavardhana. Indeed, Mammata devotes far less attention to poems of complex

suggestion than does Anandavardhana. For his detailed exposition, please refer to the

Dhvanyāloka 3.43.

khanapāhuņiā deara jāāe suhaa kimapi de bhaņiā l ruai padoharavalahīgharammi anuņijjau varāri ||351

351 The Sanketah commentary has the following Sanskrit translation: kșaņapraghuņikā devarajāyayā subhaga kim api te bhaņitā | roditi śūnyavalabhīgrhe anunīyatām varākī ll 297

atrānunayah kim upabhogalaksaņe 'rthāntare samkramitah kim

anuranananyāyenopabhoge eva vyangaye vyañjakah iti samdehah I1352

Brother-in-law, your dear wife Somehow insulted your festival guest. Now she hides in the empty loft. Go comfort the unfortunate girl. (111)

In this poem the reader cannot decide if the phrase go comfort has its meaning

transformed into another meaning, i.e., go sexually enjoy, or if the suggestor suggests the

dalliance, like an echo.

Comments

Abhinavagupta explains that the woman here speaking loves her brother-in-law, who has

been having an affair with this guest. 353 She is either suggesting that he go make love to

her in order to make him fight with his wife, and thus come to her, or she is showing him

that she knows about the affair and hopes that he will return to her instead of cavorting

with the new woman. Both readings are possible, and the outcome (that the speaker wants

the brother-in-law to be her lover) is the same, thus the reader is left in a pleasurable state

of doubt. Otherwise put, the poem profits from the ambiguity.

On the first reading there is a total transfer of meaning. The speaker does not want

her brother-in-law to comfort the guest, merely to sleep with her. On the second, the

352 I have separated this poem and commentary from the kārikā text above, departing from Mohan in order to make the comments easier to follow. 353 For an alternate translation of the poem and Abhinavagupta's commentary, see Ingalls (1990:645&647).

298

meaning of comfort is retained, but the suggested knowledge is dominant. Thus the doubt

includes which form of suggestion is intended.

snigdhasyāmalakāntiliptaviyato velladbalākā ghanāh vātāh śīkariņah payodasuhrdām ānandakekāh kalāḥ kāmam santu drdham kathorahrdayo rāmo 'smi sarvam sahe vaidehī tu katham bhavisyati ha hā hā devi dhīrā bhava II

atra lipteti payodasuhrdām iti ca atyantatiraskṛtavācyayoh samsrastih | tābhyām saha

rāmo 'smīty arthāntarasamkramitavācyasyānugrāhyānugrāhakabhāvena

rāmapadalaksaņaikavyañjakānupraveśena cārthāntarasamkramitavācyarasadhvanyoh

samkarah I evam anyad apy udāhāryam II

iti kāvyaprakāśo dhvaninirņayo nāma caturtha ullāsah II (1129)

White herons circle against dark clouds That paint the sky with their wet lustre. Winds carry the small rain. The peacocks, friends of the clouds, cry out with joy. Let all this be: my heart is hard; I am Rama and can bear it all. But Vaidehī, how shall she live? Alas, my queen, alas, be brave! (112)354

In the expressions paint and friends of the clouds there is a fusion (of two suggestions)

in which the literal meaning is given up. With this there is also a fusion (in which one

suggestion) serves another, namely in the expression, I am Räma, where the literal

354 * This translation is from Ingalls, et al. (1990:204). Judging it admirable, I have not attempted to improve on it (however, line three could be better rendered, it seems to me, "Already soft showers are blowing in.") Clearly the poet, in painting the arrival of the rainy season, wants to present those delicate first showers that precede the real rains. As Abhinava points out, these light rains are pleasant to the skin. 299

meaning leads to another meaning. (Finally) there is a fusion of the suggestion of rasa

and the (suggestion brought about by) transferring the literal meaning by combining (all

that is) indicated by the word Rāma into a single sentence355.

The other (thousands of) varieties could be similarly illustrated.

Comments

In the previous poem, Mammata illustrated a fusion of two suggestions in which there is

doubt about which form of suggestion is meant. In this poem he illustrates three other

types of fusion. First, there is a fusion of two independent suggestions. Both the word

"paint" and the compound "friends of the clouds" are used metaphorically. Nothing can

either literally paint the sky or be friends with the clouds. Each metaphor suggests a

further meaning: women are "lipta" -"painted" or "decorated," and lovers are also said

to be friends of the clouds, for the rainy season is the season of love. These suggestions

reinforce one another in suggesting love, while remaining semantically independent.

Second, there is a fusion in which one suggestion helps or serves another. In "I

am Rama," the literal meaning is the person named. This is of course the Räma of the

Rāmāyana, who is separated from Sītā, his wife. The name "Rāma" suggests all the

qualities that the hero is known to possess. In this case (in conjunction with the

expression, "my heart is hard," as Abhinava points out)556, the quality that is suggested is

Räma's steadfastness. This, however, serves to suggest that he will be able to endure the

355 Mammata's description of the final form of fusion is both terse and ambiguous. I have been guided in my rendition by the corresponding Abhinava passage as translated by Ingalls, et al. (1990:204 - 206). 356 Ingalls, et al. (1990:204- 206). 300

separation from Sīta during the rainy season, which help clarify the contrast given in the

last two lines.

Finally, there is fusion of the poem's rasa and the many other possible suggestive

meanings of the word "Rama." In this fusion both the rasa and the reverberations of the

word "Rama" play an equal and combined role. Abhinava explains that the many

qualities of Rama are all indistinctly suggested in a single cognition, just as the various

ingredients of a cake are simultaneously, yet indistinctly, present when tasting the cake.

Indeed, the flavor of the cake and the combined flavors of all the ingredients are mutually

dependant (if not identical). Likewise in the poem, there is a rasa that is gained by the

richness of the name itself. This rasa arises exactly because the literal sense bursts in

every direction with reverberations from the rest of the legend.357

This concludes Mammata's exposition of what he considers to be the best type of

poetry: poetry in which the suggested sense predominates. Instead of exploring the many

forms of combined suggestion, he moves on to the second best type of poetry, poetry of

subordinate suggestion. In the next chapter he will first explain the ways in which

suggestion, although present, can become secondary. Once the subtypes of suggestion are

fully explained, he will turn to a philosophic defense of the necessity of suggestion as a

semantic tool.

357 Given that Mammata clearly had Abhinava in mind in this section, I would need a good reason to adopt any reading other than Abhinava's. Jha claims that the suggestion referred to is the suggestion of frustrated love conveyed by the whole sentence (1925:118). Unfortunately, he does not explain how he ties this in with the richness of the word "Räma", so it is impossible to judge the merit of his claim. 301

atha pañcama ullāsaḥ

iti kāvyaprakāśe dhvanigunībhūtavyangyasankīrņabhedanirņayo

nāma pañcama ullāsah |358

Chapter Five

Investigation of the Types of Combinations of

Suggestive Poetry and Poetry of Subordinate Suggestion.

evam dvanau nirņīte guņībhūtavyangyasya prabhedān āha

agūdham aparasyāngam vācyasiddhyangam asphuțam I

sandigdhatulyaprādhānye kākvākșiptam asundaram II

vyangyam evam guņibhūtavyangyasyāstau bhidāh smṛtāḥ | (1143)

(5.1. The Eight Types of Poetry of Subordinate Suggestion.)

358 This line is found at the end of the chapter (Mohan 1409).

302

Having investigated suggestive poetry, we now state the types of poetry of subordinate

suggestion:

45 - 46ab. There are eight types of poetry of subordinate suggestion (classified

according to whether the suggestion is) (1) non-concealed, (2) dependant on

another, (3) dependant on the establishment of the literal meaning, (4) unclear, (5)

of doubtful prominence (compared to the literal meaning), (6) of equal prominence

(to the literal meaning), (7) implied by intonation, and (8) non-striking.

Comments

This chapter is broken into two large sections. The first explains poetry of subordinate

suggestion, as the chapter title announces. The second gives a general defense of

"suggestion" against various philosophic efforts to eliminate the concept by showing it to

be unnecessary in linguistic analysis. The defense is relevant to both chapters four and

five and might have claimed a chapter of its own.

Mammata criticizes all the poems of subordinate suggestion on one point or

another. Thus they might better be called poems of flawed suggestion. A suggestion can

be flawed either because another semantic factor dominates (thus the name

"subordinate") or because of a weakness in the suggestion itself. Mammata divides the

flaws into eight basic types, but examines several auxiliary types in addition to the big

eight. He defines each type of flaw and provides examples. I will comment on them

individually.

303

As explained in the notes to chapter one, Mammata's desire to incorporate

Anandavardhana's system of rasa into the older poetics inspired him to rank the poems

according to the prominence of their suggestive content. This sometimes leads to the

unfortunate result of depreciation of some very fine poems and oversight of important

flaws in others.

kāminīkucakalaśavat gūdham camatkaroti agūdham tu sphutatayā vācyāyamānam iti

guņībhūtam eva l agūdham yathā

yasyāsuhrtkṛtatiraskratir etya tapasūcīvyadhavyatikareņa yunakti karnau I kāñcīgunagrathanabhājanam eśa so 'smi jīvan na samprati bhavāmi kim āvahāmi II

atra 'jīvannity arthāntarasamkramitavācyasya I

unnidrikokanadarenupiśāngitāngā gāyanti mañju madhupā grhadīrghikāsu l etac cakāsti ca raver navabandhujīvapuspacchadābhamudayācalacumbi bimbam II

atra cumbanasyātyantatiraskrtavācyasya II (1148)

(5.1.1. Poetry of Non-Concealed Suggestion.)

Like the jug-shaped breasts of a maiden, the hidden (suggestion) gives pleasure, while the

obvious (suggestion), being totally clear, is like the literal meaning. Thus it is called

subordinate. (Three) examples of this follow.

The insults of my enemies are scalding needles Burning my ears! But what can I do? Though living, I am not alive. I might as well be knitting lacy lingerie. (113) 304

In this poem the transfer of the meaning of living to another meaning (is too obvious).

The rising sun A dandelion359 Kissing the eastern mountains.

Our garden pond Whose dark pink Lotus blossoms attract

The buzzing bees Usually black Now by pollen turned red. (114)

Here the complete setting aside of the literal meaning of the word kissing (is too

obvious).

Comments

The word "living" in the first poem has its literal meaning of being alive transferred into

mean living is such a lamentable state. The word "kissing" in the second poem

completely loses its literal meaning and comes to mean only touching. Both of these are

considered too artless to have any charm. While I agree with Mammata in the first case,

the main fault of the poem appears to me to be the failed connection between the needle

like insults and knitting as a symbol of impotence. This connection fails, of course,

because the poem insists on the scalding nature of the insults, a quality notably lacking in

knitting needles. It would have been better to focus on the pointedness of the insults and

carry over the image into embroidery.

359 The flower in question here is not a dandelion, but a Bandhjīva, which, according to Dwivedi is a Pentapetes Phoenicea (1977:139). A dandelion is an appropriate substitute because it is familiar and resembles the sun. 305

Mammata's analysis of the second poem seems equally weak. The literally

meaning of "kissing" is completely blocked, giving rise to a metaphor: the sun is

touching the ridge like a lover's mouth touches his beloved. This metaphor suggests the

erotic rasa, reinforcing the tone that is everywhere present in the poem. If this metaphor

fails (as I think it does), it is not because the analogy is too obvious, but rather because it

is too weak. The caressing motion of kissing is too different from that steady progress of

the sun. Furthermore, at dawn the sun moves away from the ridge, making it a possible

metaphor for a parting kiss, but not for consummated love. The metaphor evokes no

clear mental image and thus adds no visual detail to the poem. Finally, gratuitously

comparing the sun to a flower utterly ruins the metaphor.360

atrāsīt phaņipāśabandhanavidhih śaktyā bhavad devare gādham vaksasi tādite hanumatā droņādritratrāhrtaḥ I divyair indrajidatra laksmaņaśarair lokāntaram prāpitah kenāpy atra mrgāksih rāksasapateh krttā ca kanthātavī II

atra kenāpyatretyarthaśaktimūlānurananarūpasya l tasyāpyatreti yuktah pāthah II (1155)

My gazelle-eyed one, Here the serpent coil bound your brother-in-law, Deeply wounded in the chest by an arrow. Hanuman brought the Drona Mountains here. Here the divine arrows of Laksmana sent to heaven Indrajit.

360 A similar metaphor is used with great force by Han-shan. Compare: "On a bed of stone I sit, alone in the night / While the round moon climbs up Cold Mountain." Here the personification of the moon adds a clear image. The fact that the rising moon appears to be climbing the ridge Cold Mountain conveys to the reader the shape and steepness of the mountain. The personification also resonates with the lone spectator. Han- shan reverses the traditional idea that the moon is a friend to hermits. Here the moon is off climbing a mountain and only serves to heighten his sense of loneliness. 306

And here someone felled the Demon King's forest of necks. (115)361

In this poem the suggestion is based on the literal meaning of And by someone here. A

better reading would be And here, of that.

Comments

The point is that the suggestion is too obvious. Räma is speaking and refers to himself

with the word "someone." The suggested improvement would change the pronoun to the

impersonal that and change the case from the instrumental to the genitive - here used to

indicated the topic (as in "Of that there is redness" instead of "That is red"). Thus the

final line could read, "The felling of the forest of the Demon King's necks occurred

here." The overall effect of the change in Sanskrit would be to make the suggested

reference more abstract and thus more humble and appealing.

Aparasya rasāder vācyasya vā (vācyārthībhūtasya) angam rasādi anurananarūpam vā |

yathā |

ayam sa raśanotkarī pīnastanavimardanah I nābhyūrujaghanasparśī nīvīvistramsanah karaḥ

atra śrngarah karuņasya | (1159)

(5.1.2. Poetry of Dependant Suggestion.)

361 From the Bālarāmāyana according to Dwivedi.

307

(The second type of poetry of subordinate suggestion is that in which) the rasa is

subservient to either the literal meaning of the sentence or to the suggestion of another

rasa. For example, (a poem in which a wife of Bhūriśravas is gazing at the hand of her

dead husband):

This hand used to loosen my belt, And the folds of my skirts it used to untie. This hand caressed my breasts, Slid down across my navel, Around my thighs and into my treasure. (116)362

In this poem the erotic is subservient to the pathetic.

Comments

It is true that the overtly erotic description here is subordinate to the portrayal of a

pathetic situation. However, it is hard to see this as a failure. This portrayal of a deeply

confused emotional response to deep shock is not only vivid and realistic but also

moving. Mammata's analysis of rasa subservience misses the fact that the erotic

language here does not suggest the erotic rasa (how perverse!). The language works to

reveal the confusion, shock, love, and woundedness of the woman. Properly read, the

poem shows only one rasa: the pathetic rasa. Anandavardhana seems to advance a

similar position when he writes of this very verse, "Even if this tragic rasa is made

362 Mahābhārata 11.24.17. For an alternate translation, see Merwin and Masson (1977:3). 308

predominant, the joining to it of an erotic element by some special turn of speech will

work toward the strengthening of the tragic rasa."363

kailāsālayabhālalocanarucā nirvartitālaktakavyaktih pādanakhadyutir giribhuvah sā vaḥ sadā trāyatām I spardhābandhasamrddhayeva sudrdham rūdhā yayā netrayoḥ kānti kolanadānukārasarasā sadyah samutsāryate II

atra bhāvasya rasaḥ l (1164)

May the brilliant red of Parvatī's toenails protect you forever! They shine ever stronger in the light of Siva's third eye, Defiantly draining away Her eyes' lotus-red luster. (117)

Here rasa is subservient to the emotion (of devotion).

Comments

Mammata's judgment seems to be driven by the classificatory system. This is a

devotional poem, and, as we saw in chapter four, devotion is not considered a rasa.

Whatever the poem's faults, the problem is not the subservience of the rasa to an

emotion. Given that the poem is devotional, the subservient rasa, the erotic, is

intentionally and successfully subservient. The poem succeeds in placing emphasis

exactly where a devoté would want it, i.e., on the surge of faith the poet feels for Pārvatī.

The poem's real "fault" seems to be that the conceit is too recherché. Allow me to

explain.

363 Translated by Ingalls, et al. (1990:499). 309

Although complicated, the conceit is well-constructed and admirably daring in its

total identification of an emotion with one of its symptoms. Pārvatī is angry and her eyes

burn lotus-red from rage. Siva is trying to assuage her. As he bows to her feet, the fire

from his third eye falls on the red paint of her toenail. The poet cleverly hopes that the

growing red of the nails will suck dry the wrathful red of the eyes through poetic color

alchemy. The anger is equated with it symptom, the fiery red of the eyes. Red is red,

whether in the eyes or in the nails. If the fire of the red goes from the eyes to the toenails,

it can no longer be equated with anger. Thus Siva is saved by the nail polish, which, of

course, receives its power from Parvatī's touch. May it protect you too!

atyuccāh paritah sphuranti girayah sphārās tathāmbhodhayas tān etān api bibhratī kim api na klāntā'si tubhyam namah āścaryeņa muhur muhuh stutim iti prastaumi yāvadbhuva - stāvadbibhradimām smrtas tava bhujo vācas tato mudritāḥ II atra bhūvisayo ratyākhyo bhāvo rājavişayasya ratibhāvasya II (1168)

Lofty mountains encircle us. Vast oceans yawn before us. All these you support without fatigue, All honor to you! With awe I sing and sing these praises, Until I remember your arm, Which bears even this earth, And my lips are sealed. (118)364

Here the emotion of love of the earth is subservient to the emotion of love of the king.

364 From the Pañcāksarī, according to Dwivedi.

310

Comments

The first half of the poem praises the strength of earth, which bears all things. However,

in Sanskrit, the king is often called the bearer of the earth. This allows the poet to claim

that the mere arm of the king is mightier than the whole of the earth. The effect is a

hyperbolic praise of the king's power. Like all hyperbole, the effect rests in the audacity

of the comparison, which must yet remain comprehensible. The currency of the

expression "king as earth-bearer" allows this figure to work better in the original than in

translation. Note that the poet does not use the expression to refer to the king, thus

avoiding the fault of non-concealed suggestion.

It is curious to find this poem, and several that follow, in this section on poetry of

subservient rasa, as love (rati) is not considered a rasa unless directed toward a woman.

Mammata himself correctly calls them emotions in his explanation. We saw in chapter

four (kārikās 35bc-36a) that these emotions can ground suggestive poetry. Mammata

seems to be claiming that love of the earth could ground a rasa but that it is

overshadowed by devotion to the king. Despite what he says, I do not think this was his

intention.

I believe that Mammata chose this poem as an example of the heroic rasa being

overshadowed by either the poet's love of the earth or his love of the king. As a eulogy,

the poem suffers from being too personal. The poem should praise the king, not focus on

the poet and his admiration. As we saw in the introduction, Sanskrit poetry does not

generally serve as an expressive medium for the poet's emotions.

311

bandīkrtya nrpa dvisām mrgadrśastāh paśyatām preyasām slişyanti pranamanti lānti paritaś cumbanti te sainikāh l asmākam sukrtair drornipatito'syaucityavārāmnidhe vidhvastā vipado'khilāstaditi taiḥ pratyarthibhiḥ stūyase l

atra bhāvasya rasābhāsabhāvābhāsau prathamārdhadvitīyārdhadyotyau II (1170)

O King, with their husbands looking on, Your troops have seized their doe-eyed women, Embracing and praising them, Carrying them off and kissing them.

Yet these very foes praise you saying, "Sea of virtue, our past goodness has brought you before our eyes. All our troubles are destroyed!" (119)

Here pseudo-rasa in the first stanza and pseudo-emotion in the second stanza are

secondary to the emotion (of the poet for the king).

Comments

The erotic rasa of the first stanza is called pseudo-erotic because of its inappropriateness.

Likewise the admiration of the enemies is an aberration of genuine admiration, given the

situation. The poem is intended as hyperbolic praise of the king. Indeed, a king would

have to be truly great for his enemies to sing his praises while his troops raped their

women. However, if this hyperbole is supposed to convey the heroic rasa, Mammata

seems right that it fails. As in the preceding poem, the heroic rasa does not shine forth,

being secondary to the poet's patriotism.

312

Aviralakaravālakampanair bhrukuțītarjanagarjair muhuh I dadrśe tava vairiņām madah sa gatah kvāpi taveksaņe ksanāt II

atra bhāvasya bhāvapraśamah II (1173)

Waving swords, Arching eyebrows, Defiant cries, All show your enemies' valor. But with one look at you, It flees. (120)

Here the disappearance of emotion is subservient to the (poet's) emotion.

Comments

The description of the disappearance of the enemies' valor is used as a means to assert

the valor of the king. Mammata rightly recognizes that the goal of this poem is to evoke

the king's heroism, and is not a parody of the enemies' valor: strength needs a strong

enemy to prove its strength. Unfortunately, the poet gives us only three external signs of

valor and a description of valor failing. It is hard (even in the Sanskrit) to resist the

feeling that the signs are mere pretense. The poem would be more effective with three

examples of past valor (victorious here, resisted these, survived that, etc.) or three signs

of past valor (scares, blood, dented weapons, etc.). After all, brave words, grimacing, and

sword-waving do not make a man brave. As Mammata points out, missing its true goal

the emphasis of the poem falls on the poet's enthusiasm.

313

sākam kurangakadrśā madhupānalīlām kartum suhrdbhir api vairiņi te pravrtte | anyābhidhāyi tava nāma vibho grhītam kenāpi tatra vișamāmakarod avasthām II

atra trāsodayah II (1176)

With his friends And his antelope-eyed woman Your foe reveled in drink. But your name, Mentioned in passing, Terrified him. (121)

Here the arising of terror is subservient (to the poet's delight in the king).

Comments

The poet seeks to portray the fear on the part of the king's enemy. The king is so mighty

that even mention of his name can ruin a beautiful, safe, and happy moment. However,

the poem winds up eulogizing the king, with the poet's enthusiasm again unfortunately

dominating.

asodhā tatkālollasadasahabhāvasya tapasah kathānām viśrambhesv atha ca rasikah śailaduhituh pramodam vo diśyāt kapatabatuveșāpanayane tvarāsaithilyābhyām yugapad abhiyuktah smaraharah II

atrāvegadhair yoyoh sandhih II (1178)

Unable to abide the Mountain Daughter's weakening austerity And yet uncontrollably delighted with her loving advances, The Slayer of Love was both anxious and reluctant To remove his young-priest disguise. 314

May he command your joy! (122)

Here there is a combination of emotional haste and steadfast calm.

Comments

The Slayer of Love, Siva, is in love with the Mountain Daughter, Pärvatī, who is proving

herself through austerities. Thus Siva wants to keep his disguise and help her succeed in

her task, but he also want to throw it off so that she will come to him. The poem tries to

capture this state of contrary desire by expressing the desires' simultaneity in terms of the

simultaneity of two much simpler actions, removing or retaining the disguise.

The poet has exploited the relative freedom of word order of Sanskrit to mix the

expression of the two desires, thus adding to the effect. The story is familiar enough to

Sanskrit readers for the poem to stand on its own.

We encountered vacillation between love and asceticism in chapter four (poem 52

and notes), where it was praised. The flaw here, once again, is that the Siva's emotional

tension comes across as secondary to the poet's devotion. The poem falls unhappily

between the erotic and the fully devotional.

paśyet kaścic cala capala re kā tvarā 'ham kumārī hastālambam vitara ha ha hā vyutkramah kvāsī yāsi ittham prthvīparivrdha bhavadvidviso'ranyavrtteh kanyā kañcitphalakisalayānyādadānā 'bhidhatte II

atra śankāsūyādhrtismrtiśramadainyavibodhautsukyānām śabalatā II

315

ete ca rasavadādyalankārāḥ l yadyapi bhāvodayabhāvasandhibhāvaśabalatvāni

nālankāratayā uktāni tathāpi kaścid brūyād ityevam uktam I yadyāpi na nāsti kaścid

vişayahyatra dhvanigunībhūtavyangyayoh svaprabedādibhih saha sankarah samsrstir vā

nāsti tathāpi prādhānyena vyapadeśā bhavantīti kvacitkenacidvyavahārah II (1181)

The daughter of your enemy lives in the forest. She speaks with someone while gathering fruits and buds: Someone might see! Get away, you impertinent! What's your hurry? I am a virgin. Lend me your arm. O! What an outrage! But now you leave like this? Where are you Lord of the Earth? (123)

Here (several emotions) are mixed together: doubt, displeasure, respect, remembrance,

fatigue, depression, realization, and desire. These are poetic ornaments such as the

ornament "having rasa." And even if the arising of emotion, the joining of emotions, and

the mixing of emotions should not be spoken of as ornaments, this has been said in case

someone does speak this way.

Even though poetry of subordinate suggestion and suggestive poetry, along with

their subclasses, are never found not mixed, still, poems are classified as one or the other

according to the maxim: names are (given) by predominance (of one quality over the

others).

316

Comments

The poem itself merits little comment. Although one of its compounds (bhavadvidviso) is

delightfully assonant, it has been chosen because of the variety of emotions. The

confusion of the young woman is shown by her successive statements. However, the

girl's statements are not woven together into a convincing whole. As Mammata points

out, the various emotions evoked remain mere decorations.

Mammata has discussed the arising, joining, and mixing of emotion in chapter

four (kārikā 36cd). There he claims that they can serve as the basis of successful

suggestive poetry. In this capacity, they are the suggestive essence of the poem.

However, any emotion, even those that can become rasas, can become a mere ornament

if treated in a superficial manner. Mammata returns to the point in chapter ten where he

takes these ornaments.

The final statement admits the delicate nature of the entire classificatory

enterprise and points to the skill necessary do evaluate poems. Mammata shows the

reader that poetics is more art than science and that quality depends on far more than pure

rasa.

janasthāne bhrāntam kanakamrgatrsnāndhitadhiyā vaco vaidehīti pratipadam udaśru pralapitam kṛtālankābhartur vadanaparipātīsu ghatanā mayāptam rāmatvam kuśalavasutā na tv adhigatā II

atra śabdaśaktimūlānuraņanarūpo rāmeņa sahopamānopameyabhāvo vācyāngatām

nītah II (1192)

317

I have wandered I have wandered The cities of men, In the Janastāna forest My mind deluded My mind deluded By visions of gold. By golden deer. With tearful eyes I cry at every step With tearful eyes I cry at every step "Give me, please!" "Princess of Videhi!" I tried enough to flatter I shot arrows at the many heads of The niggardly rich. Lankā's king I am depressed,3 I am truly Rāma, and I have not but I have not gotten rich. found the mother of Kuśa and Lava (124)366

Here there is a relation of the object compared with the object of comparison - Rāma-

based on the power of the words to echo (several meanings). However, it is subordinate

to the literal meaning.

Comments

The speaker of the poem is probably the poet himself; there is a tradition of poets

complaining about their lack of worldly success. Through the complex punning, the

speaker is compared to Räma during his chase of the magic deer that Rāvana sent to trick

him and his subsequent loss of, and search for, his wife Sīta. The suggestion is that if the

poet had stuck to honest work, he would not be in his present miserable position.

Mammata claims that the literal meaning overshadows the pun and the

suggestion. As he pointed out in the preceding section, deciding whether the literal sense

365 The second meaning of mayāptam rāmatvam = darkness (or pleasure) is obtained by me. 366 Dwivedi reports, "Batta Vācaspati, acc. to Sūktimuktāvalī." 318

or the suggestion dominates in a given poem requires skill and judgment on the part of

the critic. Mammata's judgment may be based on the personal nature of the literal

meaning, which is not generally accepted in Sanskrit poetry. The theme of this poem,

however, is common. Ingalls notes that many of the most personal poems in Sanskrit treat

the disillusionment and poverty of poets.36/

āgatya samprati viyogavisamsthulāngīmam bhojināmkvacid api ksapitatriyamah etām prasādayati paśya śanaih prabhāte tanvangi pādapatanena sahasraraśmih II

atra nāyakavrttānto 'rthaśaktimūlo vasturūpo

nirapeksaravikamalinīvrttāntādhyāropenaiva sthitah II (1198)

The thousand-rayed sun Passed the night elsewhere. Forlorn, the water-born lotus Closed herself in fear. But look my slender-limbed one! Now he comes to reassure her, Gently falling at her stem. (125)

Here the behavior of the lovers, grounded in the literal meaning, is erroneously

transferred to the behavior of the sun and lotus, which is independent.

367 Ingalls (1965:385-386). 319

Comments

The poet compares the lovers to the sun and lotus. Just as the lotus folds itself into a

skinny bud each night in the absence of the sun and opens back up at the sun's first rays,

the lady is said to have grown thin worrying that her lover was elsewhere for the night

but will take heart as he throws himself at her feet and pleads forgiveness. Many of the

words have double meanings, but the poem is not a punning poem, because neither the

word for sun, nor that for lotus, has double meaning.

Mammata rightly claims that the objects of comparison are described with words

only applicable to the subjects of comparison. The poet uses actions appropriate only for

lovers in his description of the sun and lotus (giving joy, bowing at the feet, etc.). Having

done so he uses the (personified) sun and lotus to describe the lovers. The process is as

circular as claiming that bricks are brick red. Mammata agrees that the sun and lotus can

by metaphorically described as lovers, i.e., "The sun, after passing the night elsewhere,

returns to comfort the lotus." However, to go on to say that the lover is similar to the sun

so indicated closes the circle.

We see here that this chapter includes not only poems in which the suggestion is

subordinate to another poetic element, but also some in which it is subordinate to a poetic

flaw. The following four poems all contain flaws that mar their suggestive power.

vācyasiddhyanga yathā

bhramimaratimalasahrdayatām pralayam mūrcchā tamaḥ śarīrasādam I maraņañ ca jalagabhujagajam prasahya kurute visam viyoginīnām I

320

atra hālāhalam vyangyam bhujagarūpasya vācyasya siddhikrt yathā vā

gacchāmy acyuta darśanena bhavatah kim trptir utpadyate kim tv evam vijanasthayor hatajanah sambhāvayaty anyathā | ity āmantranabhangisūcitavrthāvasthānakhedālasām āślisyanpulakotkarāñcitatanur gopīm harih pātu vah II

atrācyutādipadavyagyam āmantraņetyādivācyasya l etac caikatraikavaktrgatatvena

aparatra bhinnavaktrgatatvenety anayor bhedah II (1204)

(5.1.3. Poetry Whose Suggestion) Depends on an Established Literal Meaning.

The poison-water from serpentine clouds Violently strikes separated women with Dizziness, apathy, sloth, and stupor, Prostration, paralysis, coma, and death. (126)368

Here the suggestion of deadly poison makes clear or present the literal meaning of

serpentine.

"I am going, O incorruptible one! What's the fun of just looking at you? If we stay here alone together, Base people will imagine things." Thus the cowgirl, Expressing the deep sorrow Of her sterile visit by an epithet. Then Hari embraces her, His body hair tingling. May He protect you! (127)

368 8 Ingalls, et al., attribute this verse to Sakavrddhi (1990:298). The list of symptoms is, of course, quasi- medical. I have changed the order of the eight to more closely resemble an English list. The repetition of j and g in the third pāda provids striking consonance. 321

The suggestion of incorruptible one makes clear or present the literal meaning of

by an epithet, etc.369

The difference between these two poems (126 and 127) is that in the first there is

a single speaker, while in the second there are distinct speakers.

Comments

Mammata rightly points out that the word "viśam" (poison water) is too strongly

suggestive of snakes to be used as a word for rain in the first verse. It ruins the metaphor

by forcing the secondary meaning to overshadow the primary meaning. The poem would

be more successful either as a simile ("Rain from serpentine clouds is like poison to a

woman alone ... "), or with a word that really evoked both rain and snakes ("Cobra clouds

spit rain / Striking forlorn woman with ... "). Abhinava quotes this verse twice in the

Dhvanyālokalocana. Interestingly, the second time he specifically asserts that the pun

does not flaw the suggestion, because the rasa is firmly established.370

Mammata claims that the second verse lacks charm because the suggestion of the

first half is subordinate to the explanation of the second half. To my ear it is the second

half that is at fault. The cowgirl seeks to dally with Hari, another name for Krsna, and

expresses her desire and frustration in the epithet, "incorruptible one." Hari is charmed by

the wit of the cowgirl and embraces her. Both the epithet and the explanation ("Thus the

369 The words "make clear or present" are elided in this sentence. I have carried them over from the explanation of the last example, which parallels this one. 370 Translated in Ingalls, et al. (1990:490). 322

cowgirl ... ") are necessary for the suggestion to function. However, the suggestion would

be stronger without the heavy-handed phrase, "by an epithet."

In the first example only the poet speaks, while in the second the cowgirl's words

are followed by the poet's. As we saw in chapter four, this difference is relevant to

Mammața's classificatory system.

asphutam yathā

adrste darśanotkanthā drste vicchedabhīrutā l nādrste na drstena bhavatā labhyate sukham II

atrādrsto yathā na bhavasi viyogabhayam ca yathā notpadyate tathā kuryā iti klistam II

(1213)

(5.1.4. Poetry Whose Suggestion) is Unclear.

When I see you not, I long to see you. When I see you, I am afraid of separation. Thus seen or unseen, I have no happiness. (128)

The suggestion here is that your actions should be such that neither do you remain unseen

nor is there fear of separation. The suggestion is tortured and obscure (klistam).

Comments

Mammata hits the nail on the head here. There are other possible meanings suggested by

the concluding phrase (I have no happiness). Thus the desired suggestion becomes

323

obscure and hard to find. Mammata cleverly describes the poem with an adjective

("klistam") that plays on the same sound pattern as the poem itself.

sandigdhaprādhānyam yathā

haras tu kiñcitparivrttadhairyaś candrodayārambha ivāmburāśih I umāmukhe bimbaphalādharosthe vyāpārayām āsa vilocanāni II

atra paricumbitum aicchad iti kim pratīyamānam kim vā vilocanavyāpāranam vācyam

pradhānam iti sandehaḥ II (1216)

(5.1.5. Poetry Whose Suggestion is) of Doubtful Prominence.

Hara, his calm slightly disturbed, like the ocean at moonrise, Focused his eyes upon Uma's face and her sweet, full lower lip. (129)371

Here it is doubtful whether the suggested sense, that he wants to kiss her, or the literal

sense, that the eyes were turned, is prominent.

Comments

Hara, literally "seizer" or "destroyer," is a name for Siva. In this episode of the

Kumārasambhava, Siva is mourning his previous wife Satī. Satī, however, has been

reborn as Umā, and is destine to remarry Siva. She appears before Siva and her beauty

371 Kumārasambhava 3.67. The poem literally speaks of the "bimba fruit lower lip." This is a common trope in Sanskrit. I could not include it in the translation without calling too much attention to it, thus ruining the poem's value as an example. The focus of the translation must remain on the simile.

324

shakes his ascetic resolve and fills him with a desire to kiss her. Kālidāsa compares

Śiva's asceticism to the night ocean: both are vast, unaffected by the world, and

seemingly utterly tranquil. Against the rising moon, however, one can see the gentle

undulation of the ocean, giving rise to the idea that the moon causes the ocean's

undulations. As a beautiful woman's face is often likened to the moon in Sanskrit poetry,

Kālidāsa has no difficulty extending the comparison: Uma's appearance is analogous to

moonrise. So finally, as the moon "disturbs" the ocean, the moonfaced one is said to

disturb Siva's ocean of tranquility.

The simile is striking and Mammata claims it may overpower the suggestion that

Śiva wants to kiss the goddess. If it does, it would become a case in which an ornament

(the simile) overpowers the ornamented (the erotic rasa). This would be a fault because

the essential must dominate the accidental, as we saw with the analogy of a woman and

her jewelry. The uncertainty as to which is dominant leads Mammata to claim that the

suggestion of the rasa is of "doubtful prominence." As this example shows, together with

the next, the suggested sense must be clearly prominent in order for a poem to qualify for

the highest category.

tulyaprādhānyam yathā

brāhmaņātikramatyāgo bhavatām eva bhūtaye I jāmadagnyas ca vo mitram anyathā durmanāyate II

atra jāmadagnyah sarvesām ksatriyānām iva raksasām ksaņāt ksayam karisyatīti

vyańgyasya vācyasya ca samam prādhānyam II (1219)

325

(5.1.6. Poetry Whose Suggestion and Literal Sense are of) Equal Prominence.

You have better Quit offending the priests, so that Jāmadagnya remains your friend! Otherwise, he will get angry. (130)372

The suggested meaning that Jämadagnya will kill all the demons like he killed the

warriors (ksatriya) and the literal meaning are equally prominent.

Comments

Jāmadagnya is another name for Paraśurāma, son of Jamad-agni and sixth avatāra of

Visnu. He fought for the priestly caste (brāhmana) against the warrior caste (ksatriyas),

according to Puränic legend. Here he has written a letter to Mälyavant, a minister of the

demon king Rävana, to warn him that if the demons do not respecting their truce he will

become angry. The suggestion is that he will slaughter them has he previously

slaughtered a group of warriors hostile to the priests.

Mammata points out that the literal meaning that Paraśurama will be angry and

the suggested meaning that he will slaughter the demons are equally charming.

kākvāksiptam yathā

mathnāmi kauravaśatam samare na kopād duhśāsanasya rudhiram na pisāmyur astah l

372 Mahāviracarita 2.10. 326

sañcūrņayāmi gadayā na suyodhanorū sandhim karotu bhavatām nrpatih panena II

atra mathnāmyevetyādivyangyam vācyaniședhasahabhāvena sthitam II (1222)

(5.1.7. Poetry Whose Suggestion is) Implied by Intonation.

I will not wrathfully bash the hundred Kauravas in battle, Nor will I drink the blood of Duhśāsana from his heart. Duryodhana's thighs I will not shatter with my club. Make our king sue for a peaceful settlement! (131)373

In this poem the suggestion that he will bash the hundred Kauravas, etc., remains

connected to the literal denial (of these activities).

Comments

Before the grand battle between the Pāndavas and Kauravas, Krsna attempts to make a

truce between the two sides. One of the Pändava brothers, Bhīmasena, outraged by the

proposal, recites the poem above. Literally he expresses his willingness to follow his duty

and respect his elder brother's desire for peace. But if read with an ironic intonation, it

suggests Bhīmasena's intention to fight the Kauravas regardless of whether the truce is

successfully arranged. The two possible meanings remain in tension because neither is

blocked by textual or contextual information. Perhaps Bhima is still fixed on revenge,

perhaps he has reconsidered. Only the actor's intonation informs the audience that the

second meaning is meant.

373 Venīsamhāra 1.15. 327

This poem can be compared with poem 24 ("What a great help! What can I say? /

Your kindness is widely renowned. / Behaving ever like this, my friend, / May you enjoy

a hundred autumns.") In the earlier poem the literal meaning is blocked by contextual

information and completely set aside. The suggestion arises when one searches for the

reason for the indirect speech. The literal meaning of the current poem, on the other hand,

is not blocked. Intonation alone carries the suggestion.

asundaram yathā

vāņīrakudamguddīnasaunikolāhalam sunamtīe gharakammavāvadāe bahue sīanti angāim ||374

atra dattasanketah kaścillatāgahanam pravista iti vyangyāt sīdanty angānīti vācyam

sacamatkāram II (1227)

(5.1.8. Poetry Whose Suggestion is) Non-Striking.

The confusion of sparrows In the cane thicket Teases the daughter-in-law. Busy with housework, Her limbs droop dejectedly. (132)375

Here the literal meaning of her limbs droop dejectedly is more striking than the

suggestion that someone intent on a rendezvous has entered the thicket of creepers.

374 The Sanketah commentary has the follow Sanskrit translation: vānīrakuñjoddīnaśakunikolāhalam śrnvantyāḥ | grhakarmavyāprtāyā vadhvāḥ sīdanty angānī ll 375 Hāla, Sattasaī 874 (Weber). This verse is also cited in the Dhvanyāloka. See Ingalls, et al. (1990:361). 328

Comments

The confusion of sparrows suggests that the lover has arrived at the romantic rendezvous

in the cane thicket, but housework prevents the young wife from going to meet him. With

her desires frustrated, the work becomes doubly hard and her enthusiasm for it dissipates.

There is no particular charm in the suggestion that the lover has arrived. The description

of the young wife's reaction holds far more interest.

eşām bhedā yathāyogam veditavyās ca pūrvavat Il (1232)

yathāyogam iti

vyañyante vastumātreņa yadā 'lakrtayas tadā l dhruvam dhvanyangatā tāsām kāvyavrttes tadāśrayāt II

iti dhvanikāroktadiśā vastumātreņa yatrālankāro vyajyate na tatra

guņībhūtavyangyatvam II

46cd. These types (of suggestion) should be understood, as far as possible, like the

former case (i.e., as suggestive poetry).

We should understand as far as possible according to the following:

When the ornaments are suggested by the very thing or situation, They invariably form a variety of suggestive poetry, Because the workings of poetry is founded on these (ornaments).376

376 Dhvanyāloka 2:29. 329

As indicated by the words of the author of the Dhvani (āloka, i.e., Ānandavardhana),

there is no subordination of suggestion when the poetic ornament is suggested by the

very thing or situation (described in the poem).

Comments

Ānandavardhana adds, in his self-commentary, that if this sort of ornament is absent, we

are left with a mere statement of fact. Ingalls explains the reasoning as follows, "When

one alankāra (directly expressed) suggests another alankāra, the question can arise

whether the second alankāra is predominant or the first. But when the mere fact or

situation suggests an alankāra, the suggested alankāra must be predominant, for if it

were not, the verse would not be poetry at all."377

Mammata's point is that one should not consider a suggestion subordinate without

clear reason. Usually a poem in which the ornament is predominant will fall into poetry

of subordinate suggestion. As we have seen, if a metaphor, for example, is the most

striking element of a poem, the poem is classified as poetry of subordinate suggestion.

However, Mammata, following Anandavardhana, says this is not the case if the ornament

is suggested by the literal meaning of the poem. Mammata gave many examples of such

poems in chapter four (poems 61, 65, 69, etc.). More generally, Mammata urges that any

poem that might fall into one of the eight types of poetry of subordinate suggestion

377 Ingalls, et al. (1990:356).

330

should first be considered as suggestive poetry. In other words, the aspiring critic should

follow the principle of charity.

This concludes Mammata's presentation of eight basic types of poetry of

subordinate suggestion. Allow me to list them again here: (1) non-concealed or overly

obvious suggestion, (2) suggestion dependant on another rasa or emotion, (3) suggestion

dependant on the establishment of the literal meaning, (4) unclear suggestion, (5)

suggestion of doubtful prominence compared to the literal meaning, (6) suggestion of

equal prominence to the literal meaning, (7) suggestion implied by intonation, and (8)

non-striking suggestion.

sālankārair dhvanes taiś ca yogah samsrstisankaraiḥ I (1239)

sālankārair iti tair evālankārair alankārayuktaiś ca taiḥ I tad uktam dhvanikrtā I

sa gunībhūtavyangyaih sālankāraih saha prabhedaih svaih I sańkarasamsrstibhyām punar apy udyotate bahudhā I

iti l

(5.1.9. Further Varieties of Suggestive Poetry.)

47ab. Suggestive poetry is joined with (these eight forms of subordinate suggestion)

as ornaments, and these can be joined by either fusion or association.

As ornaments means that these (subordinate suggestions) can be ornaments, or that they

can be joined to (other) ornaments. As put by the author of the Dhvani (āloka):

331

Suggestive poetry appears by fusion and association with its own sub-divisions, with poetry of subordinate suggestions, with ornaments, and in many other ways.

Comments

Mammata adopts Abhinava's use of the technical words "fusion" and "association."

Interested readers are referred to the Dhvanyāloka 3:43, which is far too long to

summarize here (30+ pages in translation).8 The overall result is to multiply greatly the

number of possible types of both suggestive poetry and poetry of subordinate suggestion.

anyonyayogād evam syād bhedasankhyātibhūyaso II (1245)

evam anena prakāreņa avāntarabhedaganane 'tiprabhūtatarā gaņanā l tathā hi

śrngārasyaiva bhedaprabhedaganāyām ānantyam kā gaņanā tu sarvesām II

47cd. Thus by combining with each other the number of types (of suggestion)

becomes extremely large.

Thus by counting all the sub-divisions (of suggestion) by the above method, the number

(of types) would become extremely large. For example, there are an infinite number of

types of the erotic rasa alone, what of all the others?

Comments

378 I have followed Ingalls, et al., in the translations of "samsrsti" and "sankara."

332

Mammata calculated 10,455 types of suggestion in chapter four. Here he points out that

these can be further combined with subordinate suggestions. For all practical purposes,

the possible types of suggestion are endless.

sankalanena punar asya dhvanes trayo bhedāh vyagyasya trirūpatvāt l tathā hi kiñcid

vācyatām sahate kiñcit tv anyathā l tatra vācyatāsaham avicitram vicitram ceti I

avicitram vastumātram vicitram tv alankārarūpam I yady api prādhānyena

tadalankāryam tathāpi brāhmanaśramananyāyena tathocyate II

rasādilaksanas tv arthah svapne 'pi na vācyaḥ I sa hi rasādiśabdena srngārādiśabdena

vābhidhīyeta I na cābhidhīyate l tatprayoge 'pi vibhāvādyaprayoge tasyāpratipattes

tadaprayoge 'pi vibhāvādiprayoge tasya pratipatteś cety anvayavyatirekābhyām

vibhāvādyabhidhānadvāreņaiva pratīyate iti niścīyate l tenāsau vyangya eva |

mukhyārthabādhādyabhāvān na punar laksanīyah II (1245)

(5.2. Demonstration that Suggestion is Necessary.)

In brief, there are three types of suggestive poetry, because there are three kinds of things

suggested (simple objects, ornamented objects, and passions/rasas, etc.). Of these (three),

some are denotable (simple objects and ornamented objects), while some are otherwise

(passions/rasas, etc.). The denotable is either non-decorated or decorated. Mere facts

comprise the non-decorated, while the ornaments comprise the decorated. Even though

the object is principally the one embellished, it is called that (an ornament) on analogy

with the brähmana-monk. 333

However, the meaning whose scope is rasa, etc., is not denotable, even in a

dream.39 (If it were) it would be denoted by the words "rasa," etc. or by the words

"erotic," etc. But it is not so denoted. Even when you use such words, if you do not use

the determinants of emotion, etc., there is no cognition of rasa, etc. (Furthermore,) even

when you do not use such words, if you use determinants of emotion, etc., (the rasa, etc.)

is cognized. Hence, by both negative and positive correlation, it is proven with certainty

that the cognition (of rasa, etc.) is only brought about by the denotation of the

determinants of emotion, etc.

It is not metaphorically indicated because the necessary conditions of metaphoric

indication, such as the blocking of the literal meaning, are absent.

From these (considerations), it follows that rasa, etc., are suggested.

Comments

The maxim of the Brahman monk was introduced while discussing poem 57 in chapter

four. Although a man loses caste in becoming a Buddhist monk, mention of his previous

caste can still be used to distinguish him from other monks. Thus calling him a Brahman,

while literally false, is permissible in an extended sense.

In this passage, Mammata distinguishes denotation of non-decorated objects

(mere facts) from denotation of objects qualified by some ornament. The latter objects are

called "ornaments," even though they are not ornaments but rather objects ornamented, in

order to distinguish them from the non-decorated objects.

379 For this use of laksana, see Monier Williams (1899:892). 334

The passage echoes the chapter-four summary of Abhinava's view, in particular,

the claims that rasa is not an object and that rasa experience does not involve cognition

of an object, either simple or complex. However, Mammata makes clear that Abhinava's

description applies only to the most important type of suggestion, suggestion whose

psychological sequence is not cognized (e.g., suggestion of rasa and of certain emotions).

We saw in the second half of chapter four many types of suggestion that suggest a fact or

a figure, either of which could be denoted. In the rest of chapter five, Mammata will try

to prove that the linguistic power of suggestion is necessary for all the original eighteen

types of suggestion. Furthermore, he will try to do so according to both schools of

Mīmāmsā.

Mammata starts with a simplified presentation of his proof, focusing on the

necessity of suggestion for rasa. Are other semantic powers capable of communicating

rasa? The method of positive and negative correlation to disqualifies denotation. Not all

poems that use the names of the rasas evoke cognition of rasa. Furthermore, many

poems that do not explicitly mention rasa do evoke cognition of them. These facts show

that the theory that rasa is denoted is both too narrow and too wide.

Mammata then observes that in many poems that evoke cognition of rasa, the

necessary conditions for metaphor are absent. Mammata will develop this point while

addressing the fourth objection below. For now, he only states briefly a single reason why

metaphoric use of language cannot account for all rasa experience: rasa is sometimes

evoked even when there is no blocking of denoted meaning.

335

Given that there are only three serious candidates for semantic powers

(denotation, metaphoric indication, and suggestion), and given that the first two are not

eligible, the semantic power that communicates rasa must be suggestion. Mammata will

now develop this eliminative argument in greater detail, considering a number of

objections.

arthāntarasamkramitātyantatiraskrtavācyayor vastumātrarūpam vyangyam vinā

laksanaiva na bhavatīti prāk pratipāditam I śabdaśaktimūle tu abhidhāyā

niyantraņenānabhidheyasyārthāntarasya tena sahopamāder alankārasya ca nirvivādam

vyagatvam I arthaśāktimūle 'pi viśeșe sanketah kartum na yujyata iti sāmānyarūpānām

padārthānām ākāmksāsannidhiyogyatāvaśāt parasparasamsargo yatrāpadārtho 'pi

viśesarūpo vākyārthas tatrābhihitānvayavāde kā vārttā vyangyasyābhidheyatāyām II

(1258)

It has been explained above (in kārikā 10) that a metaphor in which the literal meaning is

either transferred to another meaning or completely abandoned is not possible without

suggestion of a fact.

But with respect to suggestions based in the power of an individual word, any

additional meaning that the literal meaning is too limited to denote undoubtedly results

from suggestion. Likewise, ornaments like simile and such, when they are not denoted,

result from suggestion.

336

According to those who hold the "relation-of-the-designated" theory

(abhihitānvayavādin), the sentence meaning is a unit different from the word meanings. It

is the mutual relation of the word meanings, which are universal in character. The

relation results from (the linguistic requirements being met, viz.,) syntactic expectation,

semantic fittingness, and correct diction. As linguistic conventions cannot function with

respect to individual words, how could a suggestion based on the power of the meaning

of an individual word be denoted?

Comments

Mammata takes up the two types of suggestion that top his list of chapter four (see

above). These two are based on certain types of metaphor, which Mammata claims

always contain a suggestion. Mammata does not repeat the argument for this claim,

which filled most of the second half of the second chapter (kārikā 14c-18d).

Mammata has just addressed his third type of suggestion in the proceeding

paragraph.

Mammata presents his argument that suggestions based on an individual word

(type four and five) cannot be communicated by denotation. He will prove this according

to both Mīmāmsa views, and starts here with the Bhatta view. He has already introduced

the "relation-of-the-designated" theory in kārikā 6d (see that section and comments for

more details about this view as well as its rival). Mammata points out that according to

the Bhātta theory, denotation functions through linguistic conventions that cannot

337