1. Movement and Mimesis The Idea of Dance in the Sanskritic Tradition Mandakranta Bose

Movement

and

Mimesis

The

Idea

of

Dance

in

the

Sanskritic

Tradition

by

Mandakranta

Bose

The

University

of

British

Columbia,

Asian

Centre,

Vancouver,

Canada

SPRINGER-SCIENCE+BUSINESS

MEDIA,

B.V.

Library of Congress Cataloging-in-Publication Data

Bose, Mandakranta, 1938-

Movement and mimesis : the idea of dance in the Sanskritic tradition / by Mandakranta Bose.

p. cm. -- (Studies of classical India ; v. 12)

Includes bibliographical references and index.

ISBN 978-94-010-5594-9 ISBN 978-94-011-3594-8 (eBook)

DOI 10.1007/978-94-011-3594-8

1. Dancing--India--History. 2. Sanskrit language--Texts.

I. Title. II. Series.

GV1693.B595 1991 91-21902

793.3'1954--dc20 CIP

ISBN 978-94-010-5594-9

Printed on acid-free paper

All Rights Reserved

© 1991 Springer Science+Business Media Dordrecht

Originally published by Kluwer Academic Publishers in 1991

Softcover reprint of the hardcover 1st edition 1991

No part of the material protected by this copyright notice may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and retrieval system, without written permission from the copyright owner.

Movement and Mimesis

Studies of Classical India

VOLUME 12

Editor:

Bimal K. Matilal

Spalding Professor of Eastern Religions & Ethics, Oxford University, U.K.

Editorial Board:

R. P. Goldman, Daniel H. H. Ingalls and A. K. Ramanujan

The aim of this series is to publish fundamental studies concerning classical Indian civilization. It will conclude editions of texts, translations, specialized studies, and scholarly works of more general interest related to various fields of classical Indian culture such as philosophy, grammar, literature, religion, art, and history.

In this context, the term ‘Classical India’, covers a vast area both historically and geographically, and embraces various religions and philosophical traditions, such as Buddhism, Jainism, and Hinduism, and many languages from Vedic and Epic Sanskrit to Pali, Prakrit, and Apabhramśa. We believe that in a profoundly traditional society like India, the study of classical culture is always relevant and important.

Classical India presents an interesting record of deep human experience, thoughts, beliefs, and myths, which have been a source of inspiration for countless generations. We are persuaded of its lasting value and relevance to modern man.

By using extensive and for the most part unexplored material with scientific rigor and modern methodology, the authors and editors of this series hope to stimulate and promote interest and research in a field that needs to be placed in its proper perspective.

The titles published in this series are listed at the end of this volume.

To

my

mother

Bina

Roy

Choudhury

CONTENTS

Acknowledgements

ix

Abbreviations

xi

Preface

1

Introduction

5

The

Literature

of

Dance

13

The

Nāṭyaśāstra

and

the

Concept

of

Dance

108

Lāsya:

A

Dramatic

Art

131

Nṛtya

and

Uparūpaka

154

Bandha

and

Anibandha

194

The

Deśī

Tradition

216

Conclusion

256

Glossary

261

Bibliography

289

Index

305

ACKNOWLEDGEMENTS

My first guide to the study of literature on dancing was Professor

Thomas Burrow whose kindness and guidance has always been my

inspiration and under whom I began to explore this field at Oxford twenty-

five years ago. The first fruit of this was my B. Litt. dissertation, which

later led to a guide-book to the technical terms used in the Sanskrit

literature on dancing. Following upon that beginning, through the last two

decades I gathered material to write this history of the evolution of

concepts of dancing as recorded in the Sanskritic tradition.

It is a pleasure to record my thanks to my friends and colleagues.

During my research at Oxford the critical advice of Alexis Sanderson and

Jonathan Katz was invaluable. To Professor Bimal Matilal I am indebted

for both moral and scholarly support. Professors Richard Gombrich and

Samjukta Gupta have helped me with many details as has Professor G.C.

Tripathi of Allahabad. I am grateful to Professor Ken Bryant of the

University of British Columbia for the use of computer typography

designed by him. One of my deepest debts is to Professor Arindam

Chakravarti for suggesting the title of this book.

The Spalding Trust has placed me under a debt of gratitude by

awarding me a Trustee's grant. I am also grateful to the Social Sciences

and Humanities Research Council of Canada for a fellowship. To the

Department of Asian Studies, University of British Columbia I owe

thanks for the use of computer facilities.

My greatest debt is to my husband Tirthankar, whose criticism and

help sustained me through the period of writing. My daughter and my son,

Sarika and Pablo, have patiently borne my frequent absences. Without the

support of my family I could not have completed this book.

ix

ABBREVIATIONS

Abh.Dar.

Abhinaya Darpana

A.B.

Abhinavabhāratī

Ad. Lib.

Adyar Library

AP

Agni Purāṇa

A. Soc. MS.

Asiatic Society Manuscript

DR.

Daśarūpaka

G.O.S.

Gaekwad Oriental Series

KĀ

Kāvyādarśa

KAnu.H.

Kāvyānuśāsana of Hemacandra

KAnu.V.

Kāvyānuśāsana of Vāgbhaṭṭa

KAlam.

Kāvyālamkāra

KS

Kāmasūtra

NN.

Nartananimaya

NLRK

Nāṭakalakṣaṇaratnakośa

N.D.

Nāṭyadarpana

NŚ.

Nāṭyaśāstra

NR.

Nṛtyaratnāvalī

NRK.

Nṛtyaratnakośa

Nṛ. Adh.

Nṛtyādhyāya

Bh. Ar.

Bharatārnava

Bh.P.

Bhāvaprakāśana

Mā.

Mālavikāgnimitra

Mānas.

Mānasollāsa

RK.

Rasakaumudī

VDP.

Viṣṇudharmottara Purāṇa

ŚTR.

Śivatattvaratnātakara

Śṛ.P.

Śṛṅgāraprakāśa

xi

xii

ABBREVIATIONS

SKA.

Sarasvatīkanthābharaṇa

SadC.

Sadrāgacandrodaya

SUS.

Sañgitopaniṣatsāroddhāra

SC.

Sañgīta Candra

SN.

Sañgītanārāyaṇa

SDar.

Sañgīta Darpana

SDām.

Sañgīta Dāmodara

SMak.

Sañgīta Makaranda

SMak. V

Sañgīta Makaranda of Vedasūri

SR.

Sañgīta Ratnākara

SSam.

Sañgīta Samayasāra

SSār.

Sañgītasārasaṁgraha

S.D.

Sāhitya Darpana

HM.

Hastamuktāvalī

HR.

Hastaratnāvalī

PREFACE

The most comprehensive view of the evolution of dancing in India is one that is derived from Sanskrit textual sources. These texts are the basic material that students of the dance in India must examine in order to uncover its past. Since the rebirth of informed interest in dancing in early twentieth century, its antiquity has been acknowledged but precisely what the art was in antiquity remains unclear. Discovering the oldest forms of dancing in India requires, as do other historical quests, a reconstruction of the past and, again as in other historical investigations, the primary sources of knowledge are records from the past. In this case the records are treatises and manuals in Sanskrit that discuss and describe dancing. These are the sources that the present work sets out to mine.

These texts taken collectively are more than records of a particular state of the art. They testify to the growth of the theory and practice of the art and thus establish it as an evolving rather than a fixed art form that changed as much in response to its own expanding aesthetic boundaries as to parallel or complementary forms of dance, drama and music that impinged upon it as India's social and political situation changed. When we place the Sanskrit treatises in chronological sequence it becomes clear that the understanding of the art has changed through time, in its infancy as well as in maturer periods. Following these changes through the treatises is essential to an historical investigation because not only do theoretical discussions in the texts reflect dancing as it was but the practice of dancing in turn seems sometimes to have followed the precepts laid down in the texts correlating dancing with other performing arts and setting its aesthetic objectives.

1

2

PREFACE

These texts are part of a tradition of discourse in the arts that was at

once scholarly and practical. The earliest extant example of this tradition

is the Nātyaśāstra of Bharata Muni, in which all the performing arts of the

time are discussed and in which dancing is regarded as a technique for

adding the beauty of bodily motion to dramatic performances. An

ancillary to drama rather than an independent art, dancing carried no

meaning and elicited no emotional response. In the works of later writers,

however, its autonomy was recognized as also its ability to express

thoughts and feelings and it began to be discussed fully in works devoted

particularly to it rather than in works on drama or poetics-a clear sign of

its growing importance in India's cultural life.

Bharata's work, however, remained of seminal importance. His

description of the body movements in dancing and their interrelationship

not only provided the taxonomy for all subsequent authors on dancing but

much of the information on its actual technique. But Bharata described

only what he considered to be artistically the most cultivated of all the

existing dance styles, leaving out regional and popular varieties. These

styles, similar in their basic technique to Bharata's style but comprising

new types of movements and methods of composition, began to be included

in later studies. These later works, most notably the Sangītaratnākara of

Śārṅgadeva and the Nartananirnaya of Puṇḍarīka Viṭṭhala, indicate a shift

of emphasis towards previously neglected styles whereby the art of

dancing underwent substantial structural and aesthetic changes. By the

16th century the newer styles came to occupy the central position in the

accounts of contemporary dancing and coalesced into a distinct tradition

that has remained essentially unchanged to the present time. This is the

tradition, the present study argues, from which the present-day dance

styles of India are directly derived, for modern styles, such as Kathak and

Odissi, show striking parallels to techniques found only in the treatises of

the later tradition, not in works of the earlier tradition of Bharata.

MOVEMENT AND MIMESIS

3

Necessarily, this study rests on an understanding of the technical terms of

dancing, to facilitate which a glossary is appended at the end of the book.1

This study begins by setting out the broad periods into which the texts

fall. Next, chapter 2 offers a chronological review of all the extant and

available Sanskrit works on dancing, including works that discuss dancing

as part of a larger interest in the performing arts in general. In doing so,

the dating and interrelationship of some of the texts are considered This is

followed in chapter 3 by a close examination of Bharata’s views on

dancing, including its legendary origin, its purpose and function, its types,

and its relationship to drama and acting. Chapter 4 deals with lāsya, a type

of performance basic to Bharata’s understanding of the dramatic arts,

which has caused much confusion because of the attempt of later authors

to categorize it as dancing. In chapter 5 we turn to nrtya and uparūpaka,

which became major elements of the conception of dancing after Bharata

but, like lāsya, were variously understood. Chapter 6 discusses a new

method of categorizing dance forms that was devised in the early

seventeenth century: the categories were bandha and anibandha, which

seem to have been adapted from the literature of music to indicate two

distinct systems, the formulaic and the non-formulaic, of structuring

dance performances. These terms manifest an altogether new approach to

dancing that signalled a widening of the dancer’s technical and aesthetic

horizons and one that corresponded with the growing strength of new

forms. These forms, broadly classified under the rubric desi, coalesced

into a distinct tradition that assimilated older forms; this later tradition is

the subject of chapter 7. Chapter 8 concludes the study by summarizing the

historical process of the development of dancing in India as found in the

texts examined here and arguing that this process was one of building upon

rather than imitating the past. The process, therefore, was a flexible one

1For fuller explanations of these terms, see my Classical Indian Dancing: A Glossary

(Calcutta: General Printers, 1970).

4

PREFACE

that permitted the assimilation of new forms and techniques and the

redefinition of its purpose, function and aesthetic scope. This flexibility

explains why the dance in India has not merely survived but has continued

to develop, even though sometimes at an uneven pace, through centuries of

social and political turmoil.

Chapter 1

INTRODUCTION

The literature of dance in India is vast, with a wealth of manuals and theoretical treatises in Sanskrit dating back to the early Christian era. Although the majority of these works have been available for a long time, systematic research on them is a comparatively recent development. As a result, these works have not been exhaustively mined for material relating to the evolution of dancing. A number of authors in the early part of this century briefly noticed dancing in course of discussing various aspects of drama and music, as for example, D. R. Mankad in his The Types of Sanskrit Drama or Ananda Coomaraswamy in The Mirror of Gesture,1 but it did not receive the same degree of scholarly treatment that music and drama enjoyed. Some attempts were indeed made to classify and describe the major types of classical dancing in works such as those by Kay Ambrose, Faubion Bowers and Beryl de Zoete.2 But rigorous research into the art and its history came later with the work of scholars such as V. Raghavan, K.M. Varma and Kapila Vatsyayan. A highly analytical study of the basic concepts of the dance drawn from such texts was that of K. M. Varma.3 In a number of articles on saṃgīta literature and in his editions of the Sṛṅgāraprakāśa and the Nṛttaratnāvalī, V. Raghavan has shed light upon aspects of dance as described in Sanskrit sources.4

1 Mankad,1936; Coomaraswamy,1917.

2 Ambrose,1950; Bowers,1953; Zoete,1953.

3 Varma,1957.

4 Raghavan,1956; 1957; 1960; 1961; 1965; 1978.

5

6

INTRODUCTION

The most extensive study so far of the non-specialist literature of

dancing is that of Kapila Vatsyayan who has examined the Vedas,

Brāhmaṇas, Upaniṣads, epics, purāṇas, grammatical literature, lexicons,

grhyasūtra, dharmasūtra, the Arthasāstra, the Kāmasūtra, Buddhist and

Jaina texts and kāvya literature.5 On the basis of this extensive search

Vatsyayan presents a view of classical Indian dancing that is as analytical

as it is informative. But essential as her work is to the understanding of the

place of dance in Indian cultural history, her book does not offer a

comprehensive survey of the Sanskrit manuals on dancing, concentrating

rather on works from the earliest till the thirteenth century and selectively

highlighting texts from the seventeenth and eighteenth centuries to show

the continuation of the Nātyasāstra's tradition. Surprisingly, she leaves out

Abhinavagupta's commentary on the Nātyasāstra. Nevertheless, her

survey of the vast body of early non-specialist literature forms a necessary

complement to the account of specialist dance manuals attempted in the

present study.

A systematic and exhaustive study of all the available Sanskrit treatises

with a view to tracing the evolution of classical Indian dance concepts has

not been undertaken so far. Yet in their detailed descriptions and careful

commentaries these treatises offer us the fullest range of materials to

reconstruct the body of dance theory and technique from the earliest times.

Further, because these works span at least fifteen hundred years, they also

mark the stages in the growth of the art-form. They thus form the primary

source of the view presented here of classical Indian dancing.

References to dancing and dancers appear in the earliest literature of

India even before the dance became a subject of study in its own right.

First mentioned in the Vedas, dancing is referred to in a wide range of

works, including the purāṇas and works in such varied disciplines as

5 Vatsyayan,1968; see also Bose,1970

MOVEMENT AND MIMESIS

7

arthaśāstra, smṛtiśāstra, alaṅkāraśāstra,6 and in seminal works such as

Pāṇini's writings and the Kāmasūtra.7 Technical terms of dancing occur

in many works that are not specifically on dance or music,8 suggesting

that dancing was a well-known and developed art in early Hindu society.

That there were also manuals of dancing in this early period is evident

from Pāṇini's reference to the naṭasūtras, or guidelines for naṭas, by

Śilālin and Kṛśāśva.9 However, these are merely passing references and

offer no clear view of the nature of the dancing of the time.

The earliest extant textual source for dancing is the Nāṭyaśāstra of

Bharata Muni, which deals with dancing within the general area of the

dramatic arts. While it is possible that works on dancing existed before

Bharata's time, none has survived, possibly because Bharata's

authoritative account made previous ones redundant. The authority that the

Nāṭyaśāstra has exerted over all subsequent works on dancing is due not

only to its chronological priority but also to the range of subjects it covers,

from stage architecture to body movements. Scholarly opinion differs as to

the date of the work, generally placing it between the 2nd century B.C. and

the 2nd century A.D.10 Whether it was the inspiration provided by

6 Arthaśāstra. 2. 27. 25; Manusṃṛti. 4. 64, 12.45; dancing is discussed in a number of

works on poetics, starting with the Daśarūpaka in the tenth century, down to the

Sāhityadarpana in the fourteenth.

7 Kāmasūtra.1.6.

8 Vidhurapaṇḍita Jātaka. (according to Varadapande, 1975, pp. 29-34); Rājapraśnīya

Sūtra. 1. 23.; Kuṭṭanīmatam, pp.220-29; Karpūramañjarī. Act I. 6, Act III. 14, Act IV.

9,10; Mālavikāgnimitram. Act I. prose passage following the prastāvanā; prose

following verse 19; Ratnāvalī. Act I. prose passage following verse 5, 12, 16;

Nāgānanda. Act. I. prose passage following verse 3.

9 Pāṇini, Aṣṭādhyāyī. 4.3.110-11.

10 Kane,1961, p. 47; Krishnamachariar,1974, p.852. It is difficult to determine the date

of the Nāṭyaśāstra with any certainty. But it is likely that it was written before the

chapters on dance in the Viṣṇudharmottarapurāṇa, since the latter-like all texts

subsequent to the NŚ— understands the term lāsya to mean a form of dance, whereas

8

INTRODUCTION

Bharata’s scholarship or the heightened popularity of dancing that inspired

interest in dancing as a subject cannot be known but the Nāṭyaśāstra was

followed by a considerable body of literature on the subject. Sanskrit

literature of the early Christian era has a great many references to works

that evidently described dance techniques. Most of these works are lost but

extensive quotations from many of them found in other works indicate the

wealth of the material.11

In discussing dancing Bharata’s concern was with a style that he

evidently considered as one that formed the central and dominant tradition

of the art in his time.12 In describing it he established a framework of

classification and description that all subsequent writers followed in

principle, although in the details of the actual body movements they

sometimes differed. This tradition of scholarship survived till the

nineteenth century, after which no original work in Sanskrit is known to

have been written, and the works that have come down since the eighteenth

century are mostly in the vernaculars of the regions in which dancing

flourished. Most of the Sanskrit works now extant were written between

the eleventh and the seventeenth centuries, and these, while they follow

the Nāṭyaśāstra in the main in their accounts, often attempt to fit into

Bharata’s framework altogether new types of body movements

representing regional variations. One of the most important contributions

Bharata considered it to be a dramatic presentation. The portion on dance in the

VDP. is believed to have been written around the fifth century A.D.

11 Almost all the manuals on dancing as well as the commentaries on the manuals quote

from earlier authorities. Abhinavagupta, Sārṅgadeva, Kallinātha, Mahārāṇā Kumbhā,

Vedasūri are some of the authors who quote extensively from earlier authors. One of

the most often quoted sources is the lost work by Kohala.

12 NS. 9.163-4; 25.116; 119, 123. Sculptural evidence from different regions of India

dating from the first century A. D. gives us a good idea of how extensively the art had

spread throughout India. Among these the dance figures at the gates of the Sāñchī

Stupas,in the caves of Udayagiri or in Amarāvatī are early representatives of the

tradition found in the Nāṭyaśāstra. See Vatsyayan,1968, pp. 262-332.

MOVEMENT AND MIMESIS

9

to the discussion of regional styles is that of the Sangītaratnākara. The

texts that followed it from the sixteenth to the eighteenth century drew

heavily upon it, although with their own variations. Evidently, the example

of the Sangītaratnākara encouraged interest in regional varieties, for its

successors began more and more to incorporate regional terms.

Interestingly enough, these later works show two shaping forces at

work. First, the framework of movement they describe is essentially that

of the Nāṭyaśāstra; second, breaking away from the principal interest of

that work, they describe regional styles following the Sangītaratnākara.

Recognizing the difference between these types of movements, the later

authors term Bharata's tradition as mārga, that is, the way established as

of old, and the later tradition as deśī, that is, provincial or regional.

Given that the early as well as the later works on dancing move within

the framework of description set up by Bharata, it is essential to examine

that framework closely. Yet, given also that by his own admission Bharata

does not deal with the many styles peripheral to his central tradition (NŚ. 9.

163-64), it is equally essential to examine the later works thoroughly in

order to understand the evolution of the dance concepts in India, especially

since, as will be argued in this study, it is the later, regional tradition that

has directly shaped the styles of dancing we see in contemporary India.

The literature of dancing carries out two main tasks. In the first place,

the works identify and describe the body movements required in dancing.

Secondly, they also employ, though few discuss with clarity, certain

concepts about the nature and categories of dancing. On both levels

developments can be seen at various times and I am persuaded by the

literature to divide them into three periods. The first period is that which

ends with Abhinavagupta in the tenth century of the Christian era. Few

works from this period are extant but this scarcity is more than adequately

met by the most important of all works in the genre, the Nāṭyaśāstra, with

which the period began. Appropriately, the end of this period is marked by

yet another major contribution to the discussion, Abhinavagupta's

10

INTRODUCTION

commentary on the Nāṭyaśāstra. Also, some light on the art is thrown

indirectly by references to it in the literatures of alaṅkāra and kāvya, as

also in the general prose works of the era. At this time dance was mainly

considered to be an auxiliary to drama. The legend of its birth as recorded

by Bharata suggests that it existed as an independent art-form.13

However, while dancing was indeed an independent art-form, it is evident

that it was considered chiefly as an art that lent beauty to a dramatic

performance.

Although we do not have too many works on dancing from this early

period, there is reason to believe that quite a few were actually written.

Bharata himself says that the material he does not cover in the Nāṭyaśāstra

will be dealt with by Kohala in his study :

शेषमुत्तरतन्त्रेण कोहलस्तु करिष्यति। (N.S. 37. 18.)

However, Kohala will explain [lit:do] the rest through

a supplementary treatise.

This work by Kohala, unfortunately lost, was evidently most

authoritative, for virtually every Sanskrit writer on dancing quotes from it

or at least refers to it. Also mentioned, though not to the same extent, are

treatises by Bhaṭṭatota, Bhaṭṭataṇḍu, Saṅkuka and Lollaṭa, none of which

has come down to us. Dattila and Mataṅga are important writers on music

who are believed to have written on dancing as well though their works on

dance again have not survived.

The second period spans the eleventh to the fifteenth century. It was in

this period that most of the extant works on dance and music were written

and new trends in dancing recorded. Dance began to be discussed in its

13 भयावपीडं स्मृतं नृत्तं सन्ध्याकालेपु नृत्यता।

नानाकरनसंयुक्तैरःकृतैरविभूषितम् ।।

पूर्वकृविधावसंसक्त्वासम्प्रकृतं प्रयोजयताम् ।NS. 4. 13-14.

[See p. 109 infra for a translation.]

MOVEMENT AND MIMESIS

own right rather than as an adjunct to drama, just as music, both vocal and

instrumental, was being similarly regarded as an art that had an

independent existence outside dramatic performances. These two art-

forms comprised a particular discipline that came to be known as saṅgīta,

and since they were seen as complementary arts, the manuals on that

discipline dealt with both dance and vocal and instrumental music,

although the emphasis might vary. Indeed, the Nṛttaratnāvalī of Jāya

Senāpati is the only work from this period that is wholly devoted to dance

alone, and even this is very likely part of a larger project covering music as

well, for Jāya himself refers to the companion piece to the work, his

Gītaratnāvalī.14

It was during this second period that a distinction began to be made by

writers on dancing between the mārga and deśī styles. The classical style

was acknowledged to be the one codified by Bharata in the Nātyaśāstra and

taken to form the core of the entire tradition of the art. But it was also

recognized at this time that the many regional and popular styles were

equally part of that tradition and therefore needed to be recorded. Another

major development recorded in the texts from this period was the

appearance of dance dramas. These, discussed in a later chapter, were not

simply dramas to which dancing was added as an ornament as in an earlier

age but an altogether new genre in which dance permeated the whole form.

In the third period, beginning in the sixteenth century and ending in the

nineteenth, the texts record a further development of the tradition, and as

we shall see in the course of this study, it is the shape that dance took at this

time which we see today. In brief, the texts from this period testify to the

growing popularity of the dance, for not only is an increasing variety of

regional forms recorded in the manuals but a number of manuals in

regional languages begin to appear. In Andhra, in particular, dance

flourished throughout this period, as we find from Raghavan's extensive

14 Raghavan,1965. Introduction, pp. 12,36; text, p. 172.

12

INTRODUCTION

account of the dance literature in Telugu.15 Another region from which a

number of texts came was Orissa, as noted in a survey by D. N. Patnaik.16

Dance manuals were produced also in the regional languages of Manipur,

Maharashtra and the region now known as Tamil Nadu, and although no

systematic account of these texts has been made as yet, their existence

attests as much to the codification of the art as to the widespread interest in

it. However, while these works in the vernaculars provide a great deal of

detailed information about dances of the time, they show no development

in the fundamental concepts of dancing. The importance of these regional

texts lies in the information they give about regional dance forms, thus

providing supplemental information unavailable from Sanskrit sources.

However, for the main body of the dance traditions of the time, the extant

Sanskrit texts still remain our most comprehensive source. An account of

these texts, their particular contributions and their relationship to one

another will therefore form the subject of the next chapter.

15 Ibid., Introduction, pp. 18-34.

16 Patnaik,1971.

Chapter 2

THE LITERATURE OF DANCE

The earliest extant Sanskrit work on music, drama and dance, and also

the most influential as a source of both technical description and aesthetic

understanding is the Nāṭyaśāstra, which is believed to have been written

not later than the 2nd century A. D.1 The work deals with dramaturgy and

the arts allied to drama in thirty-seven chapters, of which those of interest

to students of dancing are chapters 4, 5, 8-12, 19, 21, 22, 25 and 31. The

fourth chapter discusses tāndavavidhi, that is, the rules of tāṇḍava or

dancing.2 In this chapter Bharata mentions five concepts that help to

define the art of dancing. These concepts, nṛtta, tāṇḍava, abhinaya,

piṇḍībandha and sukumāraprayoga, are discussed in a separate chapter of

the present study. Bharata's interest in the fourth chapter is centred on

nṛtta, that is, non-representational dancing and valued for the beauty of its

form. He describes this dance style in detail, giving the basic units of

composite movements known as karaṇas, and the basic choreographic

sequences known as aṅgahāras. In addition, he describes group dances and

mentions the use of body movements in acting without giving details.3

Bharata speaks of these movements as necessary parts of the

preliminaries of a dramatic presentation.

The fifth chapter of the Nāṭyaśāstra continues the discussion of those

preliminaries. In the eighth to the twelfth chapter Bharata describes in

detail the movements of every part of the body down to the smallest,

1 Kane,1961, p. 47. The date has not been precisely set as yet. See note 10 , chapter 1.

2 NS. 4.320.

3 NS. 4.279-283.

13

14

THE LITERATURE OF DANCE

dividing them into two classes, the aṅgas and the upāṅgas, or major and

minor parts of the body. He shows how the different movements are

combined into composite movements known as cārīs, maṇḍalas, and

sthānas that are in turn combined into karaṇas, which again are put

together to create aṅgahāras.

The eighth chapter is devoted to the movements of the head, eyelids,

eyebrows, pupils, the nose, cheeks, lips, the chin, the mouth and the neck,

while the ninth is given to the movements of the hands, chest, sides, belly,

waist, thighs, shanks and the feet. Both chapters describe the use of these

movements in conveying meaning, but also included are hand-gestures

meant for abstract dancing rather than for acting. The tenth chapter

describes cārīs, which are movements using one leg and are used both in

dance and drama. Similarly used are sthānas, or postures, which form part

of this chapter; examples of these postures occur widely in temple

sculpture. Bharata goes on to describe maṇḍalas, which are more

complicated movements of the legs involving cārīs. In addition to

describing these movements Bharata speaks of the general principles of

effective exercise, vyāyāma, as well as aesthetic satisfaction, sauṣṭhava.4

The eleventh chapter elaborates on maṇḍalas, while the twelfth describes

gati, or gaits.

Although Bharata's subject is nṛtta, that is, dancing in its abstract

form, in this five chapter long section he explains the meaning attached to

individual body movements and thus recognizes by implication the

possibility of conveying meaning through dancing.5 Such uses of body

movements for enhancing dramatic expression is termed āṅgikābhinaya by

Bharata. That dancing was in fact used for dramatic purposes is evident

4NŚ. 10.92-94, 97-103.

5NŚ.9.219.

MOVEMENT AND MIMESIS

15

from other works on dancing as well as from descriptions of play

production.6

The nineteenth chapter considers the constituents of lāsya, a dramatic

form that is created from delicate body movements and, in Bharata’s

conception, falls between dance and drama. Chapter twenty-one deals

with āhāryābhinaya, that is, the use of costumes, stage properties and

other external aids which are essential both to dance and drama. An

important concept of presentation discussed in this chapter is the

distinction between the lokadharmī and nāṭyadharmī modes of

presentation, corresponding to natural and stylized modes.

In chapter twenty-two Bharata takes up the general technique of

expression in acting, calling it sāṃānyābhinaya, and gives directions for

expressing states of mind and responses to sensory experience, such as

touching or smelling. These movements are considered to be usually self-

explanatory, so that although they are codified into a discipline, they are

not seen as stylized, with special, symbolic meanings attached to them.

The hand-gestures in this category, for instance, are formed with the

purpose of imitating objects. By contrast, chapter twenty-five describes

the special mode of citrābhinaya, in which each movement carries a

particular meaning specific to it. The thirty-first chapter is partly on tāla,

or rhythm, but it also goes back to the subject of lāsya, describing the

movements and the music required for each of its several varieties.

As this survey shows, the framework within which Bharata describes

dancing is largely that of drama and his interest in discussing even nrtta

appears to lie in examining how it can beautify a dramatic presentation.

While the authors who came after him virtually copied his descriptions,

they did not continue this approach and in the later works we no longer find

dancing treated only as a supplement to drama.

6See A.B. on the NŚ. Vol.I. 1956, pp. 181,191.;DR. 1.10; SṭP. vol. II. pp.423-24;NR.

1.29;.and Rāghavabhaṭṭa’s commentary on Abhijñānaśakuntalāᶇ,ed. Kale 1902.

16

THE LITERATURE OF DANCE

The treatise next to the Nāṭyaśāstra in time is the Viṣnudharmottara Purāṇa. It has been variously dated but the editor of the text places it between the fifth and the seventh centuries,7 while Pingree assigns the parts on dancing and painting to the first half of the fifth century; Ludo Rocher has discussed the arguments for the different dates in detail in his volume on the history of Purāṇa literature.8 Close scrutiny of the text, however, leaves little room for doubt that the Nāṭyaśāstra antedates the Viṣnudharmottara Purāṇa. The thirtieth chapter of the Viṣnudharmottara Purāṇa enumerates the rasas as nine; the number given in the Nāṭyaśāstra is eight. It is far likelier that Bharata’s classification was expanded by the author of the Viṣnudharmottara Purāṇa than that his nine classes were shrunk to eight by Bharata.

The Viṣnudharmottara Purāṇa deals with dancing in its third section, chapters twenty to thirty-four. In chapter twenty the author follows the Nāṭyaśāstra in describing the abstract dance form, nrtta, and in defining its function as one of beautifying a dramatic presentation. But the author differs from Bharata in dividing dance into two categories which he calls nāṭya and lāsya. He thus includes both nāṭya and lāsya in the same genre, that of dance, which is a conceptual departure from Bharata. This chapter also specifies the appropriate places for the performance of each category, discussing aspects of the stage and the presentation of the preliminaries. The discussion includes the characteristics of actors , the four different types of abhinayas, namely, āngika, vācika, sāttvika and āhārya, and the names of all the complicated movements necessary for the composition of a dance sequence. In addition, the author briefly touches upon the piṇḍibandhas or group dances mentioned by Bharata and goes on to describe vṛtti, pravṛtti and siddhi, that is style, the means of application and the nature of competence.

7 VDP, 1958, G.O.S. ed. P.Shah, vol. I, Introduction, pp.xxvi.

8 Rocher, 1986, pp. 250-52.

MOVEMENT AND MIMESIS

17

The twenty-first chapter discusses sthānas or postures while lying down, while the twenty-second deals with the sthānas assumed while sitting. The focus of these two chapters seems to be on dramatic presentation. In this the Viṣṇudharmottara Purāṇa seems to agree with Bharata in considering the whole subject as part of dramatic presentation. The twenty-third chapter is devoted to postures meant for both men and women. The twenty-fourth chapter lists the movements of the major limbs, the aṅgas, along with the meaning attached to each of them. The major limbs, according to this text, are the head, the neck, the chest, the sides, the waist, the thighs, the shanks and the feet. In conclusion, the chapter defines the cārī and the karaṇa, the two vital and complicated movements required in dancing. In the twenty-fifth chapter the movements of the upāṅgas or minor limbs are discussed, including the glances that express rasa and sthāyī and vyabhicāribhāvas, the movements of the pupils, eyebrows, nose, tongue and lips as well as the application of these movements.

The twenty-sixth chapter describes three types of hand-gestures, those made with one hand, those made with both-along with the meanings they can convey-and hand-gestures meant for dancing, which convey no meaning. The twenty-seventh chapter is devoted to the explanation of different kinds of abhinaya and the costumes and decorations necessary for a performance. The next chapter, the twenty-eighth, deals with sāṃānyābhinaya, giving general directions for expressing different moods and responses to seeing, touching and smelling objects. Although the author designates this chapter as a discussion of sāṃānyābhinaya, he includes citrābhinaya, that is, special presentations. In fact, this chapter is a conflation of the contents of chapters twenty-two and twenty-five of the Nāṭyaśāstra and contains extensive quotations from it.9 The twenty-ninth

9NS. 22. 81-5; VDP. 3. 28. 2-4; NS. 25. 2-8; VDP. 3. 28. 10-15.

18

THE LITERATURE OF DANCE

chapter describes the gatis, that is, gaits, the thirtieth discusses the nine

rasas, and the thirty-first the bhāvas.

A new feature of the treatment of body movements that is added to the

discussion of body movements appears in the thirty-second chapter, which

deals with what is termed rahasya mudrās, that is, hand-gestures meant for

mystical and ritualistic purposes. Continuing the discussion in the thirty-

third chapter, the author lists more mudrās, all meant for religious

purposes, and calls them mudrā-hastas,10 and associates them with hymns

to the gods and goddesses.11 We must note that the word used in these

chapters is not hasta but mudrā, a term connoting religious practices. The

use of the term mudrā to denote hand-gestures is found in no other text on

dancing. In the medieval period the term mudrā was used in dance

literature to mean a pleasing appearance (NN. 36a; SDar. 7.107).

However, in common usage today, all hand-gestures used in dancing are

called mudrās and the more accurate term hasta has gone out of use.

The thirty-fourth and final chapter on dancing is devoted to the legend

of the origin of dancing. Since the work is devoted to the worship of Viṣṇu,

it is not surprising that its author should view Viṣṇu as the propounder of

the art of dancing, but in every other text it is Śiva who remains as the god

who introduced dancing to mankind.

Our next source of information is a commentary on the Nāṭyaśāstra by

Abhinavagupta, who lived in Kashmir sometime in the late tenth and early

eleventh century A.D.12 Abhinavagupta's commentary is known as the

Abhinavabhāratī. It is not known whether the commentary was on the

entire Nāṭyaśāstra; but what has come down to us follows the Nāṭyaśāstra

chapter by chapter except for the seventh, the eighth, and the thirty-third

to thirty-seventh. It is available now as part of Ramakrishna Kavi's

10 VDP. 3. 33. 123.

11 Ibid. 3. 33. 124.

12 Krishnamachariar, 1974, p. 748.

MOVEMENT AND MIMESIS

19

edition of the Nāṭyaśāstra in the Gaekwad Oriental Series. The text is

unsatisfactory because of the corrupt state of the mss. from which this

edition was prepared.13 Raghavan has suggested a number of emendations

and corrections,14 which are helpful not only to scholars but to dancers as

well, especially in reconstructing the karaṇas of Bharata as explained by

Abhinavagupta. For our present purposes the relevant sections are the

commentaries on those chapters of the Nāṭyaśāstra that deal specifically

with dance and dramaturgy.

The importance of Abhinavagupta's work can hardly be overstated.

Since a number of works on dancing that are known to have been written

after Bharata are now lost, it is difficult to follow the development of the

technical discussion of the art of dancing through the early period of its

evolution. Abhinavagupta's commentary shows us the changes that took

place in the intervening period between his time and Bharata's. In

Abhinavagupta's commentary we find more terms representing concepts

and categories of dancing than there are in the Nāṭyaśāstra, showing that

such terms and what they represented had by this time passed into general

use, as is also attested by their use in other kinds of literature of the time.15

13Raghavan,1980, p.171.

14Ibid. pp. 171-92. Raghavan arrived at his emendations of Abhinavabhāratī by

collating the text of Abhinavabhāratī with that of Nṛttaratnāvalī in course of preparing

his critical edition of the latter text. However, firmer grounds for emendations would

be reached if Abhinavagupta's commentary were collated with Saṅgītaratnākara,

which Raghvan does only in a few instances.

15Abhinavagupta quotes Viśākila on lāsya, NŚ. Vol. I. 1956, pp. 197, Kohala on kāvya

and past authorities on ṇṛttakāvya and rāgakāvya, NŚ. Vol. I. 1956, pp. 171-84;

Daṇḍin in his Kāvyādarśa refers to chalika, lāsya and ṣamyā, Kāvyādarśa., I. 39;

Vararuci, Kālidāsa, Harṣa use lāsya, chalita, ṣamyā, dvipadī, saṅgītaka etc. in their

plays, Ubhayābhisārikā., pp 122-3,134, 141,142 ; Mālavikāgnimitram., Act. I. prose

passage following verse 19; Ratnāvalī., Act. I. prose passage following verse 5,12,16;

Rājaśekhara in his Karpūramañjarī uses carcarī, lāsya and other technical terms related

to this art-form, Act. I. 6; III.14, IV. 9,10; Dāmodaragupta in his Kuṭṭanīmatam uses

20

THE LITERATURE OF DANCE

It is evident that by Abhinavagupta's time the dance had proliferated into

many more forms than known by Bharata but it is also clear that

Abhinavagupta views them as embodiments of the basic concepts stated by

Bharata, for he often cites instances of such new categories of dancing to

elucidate Bharata's concepts.16

For similar elucidations Abhinavagupta also quotes other authorities

extensively. Beside Bharata the most important early authority on dancing

seems to have been Kohala, whose work, occasionally referred to in the

Nātyasāstra,17 is lost. It is through Abhinavagupta's quotations from

Kohala that we can reconstruct some of the conceptual developments that

were either contemporaneous with Bharata or followed shortly after.

Two other authorities quoted by Abhinavagupta, Bhaṭṭatota and

Utpaladeva, were teachers to whom he owed his knowledge of dancing.

Among other authorities cited are Nandī, Rāhula, Dattila, Nārada,

Mataṅga, Viśāk hila, Kīrtidhara, Udbhata, Lollaṭa, Saṅkuka, Bhaṭṭayantra

and Rudraṭa, all of whom wrote on music and dance. He uses these

authorities critically, drawing upon them for information that helps to

elucidate Bharata's sometimes cryptic statements, but often rejecting

their views when his own observations provide evidence to the contrary.

One of the most illuminating features of Abhinavagupta's work is his

practice of citing different views on the same movement. For instance,

while explaining the ardhanikuṭṭaka karaṇa which employes aṅcita of the

hands, Saṅkuka's description, which is different from Bharata's, is

included.18 Abhinavagupta's citation of the two authorities thus shows us

that this karaṇa was performed in two different ways.

dvipadī 338, 858; carcarī 881; Jayantabhaṭṭa in his Āgamaḍambaram describes a dance

danced to a carcarī song, Act. II.

16A.B. on the NS. Vol.II, 1956, pp.171, 181.

17For example,NS. 1.26; NS.37. 18.

18A.B. on the NS. Vol.II, 1956, p. 102.

MOVEMENT AND MIMESIS

21

From the range of Abhinavagupta's discussion it is quite apparent that

his grasp of the subject was not only extraordinarily thorough but based on

direct experience of the art as it was practised in his time. For this reason

he explains the Nātyaśāstra according to the concepts current in his own

time. Whether this represents Bharata's thought accurately enough may

be debatable, for we have no evidence that the same concepts were

prevalent in Bharata's time. In fact, the contrary may be true. For

instance, Abhinavagupta speaks of minor categories of drama, for which

he uses the terms nṛttakāvya and rāgakāvya,19 which obviously mean,

respectively, plays danced and plays sung. The concept of minor dramas is

absent in the Nātyaśāstra, which makes no mention of nṛttakāvya and

rāgakāvya. The reason Abhinavagupta mentions these new developments

is to explain the distinction Bharata draws between sukumāraprayoga and

tāṇḍava. Again, Abhinavagupta takes lāsya to be a form of dance and not

of drama, unlike the Nātyaśāstra, and his discussion of it shows the

evolution of the notion of lāsya after Bharata's time, as noted in the

discussion, in a later chapter, on the development of the concept of lāsya

through later times. Bharata, as elucidated by Abhinavagupta, is thus not

necessarily Bharata himself, for Abhinavagupta not only expands Bharata

but interprets him in the light of what he himself knows of dancing. It is,

therefore, the evolving reality of the art rather than a description frozen in

Bharata's time that is reflected in Abhinavagupta's commentary.

The proliferation of concepts and categories after Bharata's time

revealed by Abhinavagupta's commentary is seen particularly clearly in

his classification and interpretation of nṛtta, which also throws light upon

the development of dance in general. Abhinavagupta classifies nṛtta into

the following seven categories: śuddha, that is, pure or abstract dance,

gītakādyabhinayomukha, a dance that expresses the meaning of a song,

vādyatālānusāri, a dance that follows instrumental music and rhythm,

19Ibid. pp.175, 182.

22

THE LITERATURE OF DANCE

uddhata, a vigorous dance, sukumāra, a delicate dance,

masṛamiśroddhata, a vigorous dance mixed with delicate movements,

and uddhatamiśritamasṛanṛtta, a delicate dance mixed with vigorous

movements.20 Since many of these dances were expressive, they required

abhinaya or interpretative movements. Such dances, then, fall into the

category that later became known as nṛtya.21 Although Abhinavagupta

does not use the term nṛtya, presumably to be consistent with Bharata,22

who speaks only of nṛtta in the Nāṭyaśāstra, Abhinavagupta’s descriptions

testify to the branching out of the art, as pointed out by K. M. Varma in his

detailed discussion of Abhinavagupta’s views on abhinaya and nṛtta in

detail.23

Abhinavagupta provides the details of several dance forms that are

mentioned but not described in the Nāṭyaśāstra. For instance, he describes

bhadrāsana, one of the group dances termed piṇḍibandha by Bharata but

not described by him (NŚ. 4. 290). Although Abhinavagupta does not use

the term bhadrāsana, his description matches that found in a Jainasūtra,

the Rājapraśnīya, which includes bhadrāsana in its description, in the

twenty-third sūtra, of thirty-two varieties of nāṭyavidhis, that is, the

features of nāṭya. These thirty-two varieties have been discussed by the

editor of the Saṅgītopanisatsāroddhāra and by V. Rāghavan in his study of

the Śṛṅgāraprakāśa.24

The commentary on the fifth chapter expands Bharata’s description of

the preliminaries of a dramatic performance and covers such topics as the

use of tāla, vocal and instrumental music, and the arousal of the śṛṅgāra

20Ibid. p.182.

21DR. 1.9.

22Varma,1957, p21.

23Ibid. pp. 15-22.

24Sūttāgame, 1954, pp. 55-6; Saṅgītopanisatsāroddhāra, 1961, ed. P. Shah, Introduction,

pp. xvi.; Raghavan,1963, pp. 572-73.

and raudra rasas in course of depictions of gods and goddesses.25 The

commentaries on the chapters on āṅgikābhinaya, chapters nine to twelve

of the Nāṭyaśāstra (chapter eight does not have any commentary), do not

offer any additional information, although Abhinavagupta's comments

help us to visualize the body movements required. By contrast, the

commentaries on āhāryābhinaya, sāṃānyābhinaya and citrābhinaya

(chapters tweny-one, twenty-two and twenty-five) provide details that

help us to understand the technicalities of stage presentation. Also treated

at some length are Bharata's remarks on lāsya in chapters nineteen and

thirty-one; since Abhinavagupta's interpretative approach to lāsya has

attracted considerable controversy, it will be considered in the discussion

on lāsya later in the present study.

One must remember that like any other commentary Abhinavagupta's

commentary is exegetical and absolute reliance on it may mislead

attempts to reconstruct the tradition as in Bharata's time. This is

particularly necessary to bear in mind because much of the medieval

literature of dancing, including the infuential Saṅgītaratnākara,is entirely

dependent on Abhinavagupta's commentary. As Raghavan says, “. . . the

Saṅgītaratnākara follows Abhinavagupta so closely that in many places it

forms merely a metrical recast of the Abhinavabhāratī.”26 Jāya Senāpati

does the same in the discussion of the mārga dance in his Ṇṛttaratnāvalī.

So what is often taken today as the influence of the Nāṭyaśāstra in these

texts is in reality the influence of Abhinavagupta.

Speaking of Abhinavagupta's commentary Raghavan states that "its

importance to the student of the history of Nāṭya and Saṅgīta can not be

overemphasised." Of particular significance is the fact that

Abhinavagupta was from Kashmir, for at that time the study of nāṭya, in

Raghavan's words,"had a renaissance in Kashmir in the time of Jayāpīḍa."

25A.B. on the NŚ. Vol.I. 1956, pp.207-59.

26Raghavan,1980, p.171.

24

THE LITERATURE OF DANCE

This is borne out by Kalhana, who refers in his Rājataranginī (iv. 420-84) to a temple presentation of nāṭya by courtesans. Jayāpiḍa, the king of Kashmir, fell in love with one of them, Kamalā, married her, brought her to Kashmir and made her his queen. He was a patron of nāṭya and asked his court-poet Udbhata to write a commentary on the Nāṭyaśāstra of Bharata. According to Raghavan, "Udbhaṭa's commentary is the first and it inaugurated an era of vigorous study of the Nāṭyaśāstra. This period produced innumerable works and it closed with the time of Abhinavagupta, after which the śāstra began to flourish in Central and South India."27

Abhinavagupta, who was a Śaivaite ācārya, turned to dramaturgy and poetics and studied under his first guru Bhaṭṭatota. He extensively quotes from Bhaṭṭatota's work the Kāvyakautuka, which is lost, as is the bulk of the literature of this early period that he uses. His work thus brings together the scholarly thought of an entire age and gives us a picture of the evolution of the art of dancing, although the picture is not as clear as we would have liked. Even though his commentary is illuminating in general, there are places where his explanations are not enough to visualize the movements he describes. Since the edited text is often corrupt, the task of understanding is even harder. The movements are sometimes unclear and impossible to reproduce, as in the case of the piṇḍibandhas. However, as the single extant commentary on Bharata's seminal text, Abhinavagupta's work has exerted great influence on subsequent writers on dance and drama. His work came to be accorded the highest authority and to be regarded as the standard work on the subjects, not only in works on music and dancing but on alaṅkāra as well. Hemacandra in his Kāvyānuśāsana, Rāmacandra and Guṇacandra in their Nāṭyadarpana, and Kallinātha in his commentary on the Saṅgītaratnākara continually refer to the views of

27Ibid. pp. 101-3.

MOVEMENT AND MIMESIS

25

Abhinavagupta.28 The Sangitaratnakara’s chapter on dancing is based

entirely on it. Saradatanaya claims that he based his work Bhāvaprakāsana

on Abhinavabhārati.29 In fact he did not, but that he should so claim only

shows how highly Abhinavagupta was regarded in the medieval period.

Two writers on dancing who are traditionally thought to have preceded

Abhinavagupta are Nandikesvara and Nārada but the chronology is

doubtful as are the identities of the authors. The names Nandī and Nārada

are part of the legendary history of dancing and it is possible that in some

long past period there were indeed sages by these names. Presumably, in

an effort to gain acceptance later authors passed off their own works

under the names of these legendary sages. Two works on dancing are

traditionally attributed to Nandikesvara, the Abhinayadarpana and the

Bharatarnava. The Sangitamakaranda, a work on music and dance, is

attributed to "Nārada." Other works have been attributed to "Nārada" but

they may be by different authors.30

M. Krishnamachariar and V. Raghavan have recorded various works

purported to be by "Nandī," 31 but only two of them, the Abhinayadarpana

and the Bharatarnava, both relevant to our present investigation, are

available in edited form. Nandikesvara has been mentioned by almost

every writer on music and dance and Nandī and Bharata have been

mentioned side by side in the Kāvyamālā edition of the Natyasastra.

According to Raghavan, "There is a music work called Nandībharata

noticed by Rice in Mysore and Coorg Catalogue."32 The Madras

Catalogue too mentions one chapter of a work entitled the Nandībharata

28KavyānusāsanaH,1938, chapter 8; Natyadarpana,1959, p. 25; Sangitaratnākara, Vol.

IV, 1953, 7.880,1062. Kalanidhi: Commentary on the SR. Vol. IV.1953, pp. 326.

29Bhāvaprakāsana,1968, pp. 82, 160, 194-5, 313.

30Nijenhuis,1977, pp. 19-20.

31 Krishnamachariar,1974, pp. 825-27; Raghavan,1957, pp. 22-3.

32Raghavan,1957, p. 22.

26

THE LITERATURE OF DANCE

and another work called the Bharatārthacandrikā that has a Telugu commentary which takes the form of a conversation between Pārvatī and Nandikeśvara.33 His date has not been satisfactorily established so far. In his edition of the Abhinayadarpana M. M. Ghosh placed Nandikeśvara between the fifth and the thirteenth centuries A.D.34 Nijenhuis has argued for narrowing the date to the twelfth century A.D.35

Both the Abhinayadarpana and the Bharatārnava are believed to be by Nandikeśvara,36 but whether they were written by the same Nandikeśvara is doubtful because the contents of the two are at odds with each other. Raghavan's view is that the Abhinayadarpana is a summary of the Bharatārnava, literally, the ocean of Bharata's art,37 but this claim is not borne out by comparison of the texts.

If the two works were by the same author, then one would expect them to agree substantially and differ only in the extent of the information given, or in their emphases, for it defies reason to suppose that the same author would give contradictory versions of the same matter in two works without any explanation. Yet the difference between the two works is substantial. While they are on the same subject, they differ significantly in matters of detail. For instance, there are 28 single-hand gestures in the Abhinayadarpana but 27 in the Bharatārnava; against the 13 double-hand gestures in the Abhinayadarpana, the Bharatārnava describes 16; the nrttahastas (gestures used in abstract dancing) in the Abhinayadarpana

33Ibid. p. 22.

34Ghosh, 1957, Introduction, p. 38.

35Nijenhuis, 1977, p.11. My own view would put him in twelfth/thirteenth century A.D. See p. 30 infra.

36Raghavan,1957, p. 22-3. The edited version of ADar. (Ghosh, 1957), however, does not contain any reference either to the name of the text or the author. The text is incomplete and provides no colophon.

37 Raghavan,1957, p. 2 22-3.

MOVEMENT AND MIMESIS

27

are 13 in number while in the Bharatārnava they are 22. Not only are there

differences in numbers but also in the names, definitions and applications

of the movements. Besides these gestures, the Abhinayadarpana describes

hastas denoting devas, avatāras, family, different castes and the nine

planets. The Bharatārnava omits this entire class of hastas and instead

includes an altogether different set of hastas, nānārthdyotakahastas,

which convey a medley of meanings as the name indicates. 38 The

treatment of the dr̥ṣṭis is different as well.

In describing the hand gestures meant to denote the planet Śani, one of

the navagrahahastas, the Abhinayadarpana prescribes the śikhara and

triśūla hand-gestures for the two hands while the Bharatārnava prescribes

sandamśa and alapadma. Budhagraha of the Abhinayadarpana requires

musti and patāk ā while in the Bharatārnava it requires mukula and

sandamśa. Such discrepancies are common. Nandikeśvara, the author of

the Abhinayadarpana, takes eight darśanakarmas from the Nātyaśāstra

and describes them as eight dr̥ṣṭis, whereas Nandikeśvara, the author of

the Bharatārnava, follows the Nātyaśāstra's treatment of the dr̥ṣṭis and

describes thirty-six dr̥ṣṭis that express rasa and bhāva, aesthetic pleasure

and emotions. 39 Gatis or gaits in the Abhinayadarpana are compared to

animal and human gaits and described in detail following Bharata.40 By

contrast, in its description of the gaits the Bharatārnava focuses on how

they are employed in different kinds of tāṇḍavas; the nature of the gaits too

are different in this text. 41 The description of tāṇḍava and lāsya, divided

into categories, that is given in the Bharatārnava is found in no other text.

38 ADar. 89-92, 172-75, 248-49; Bh. Ar. 1-4, 63-66, 93-96; ADar. 204-15, 216-30, 226-

31, 231-44, 250-58; Bh. Ar. 587-640.

39 ADar. 66-67; Bh.Ar. 235-40.

40 ADar. 309-10.

41 These are described as components of deśī nāṭyas, under deśī tāṇḍavas, which,

according to this text, are seven, Bh.Ar. 791-868.

28

THE LITERATURE OF DANCE

The śṛṅganāṭya and aṅgahāra mentioned in this text are two new concepts

found nowhere else, as we shall see in a later chapter. The term

śṛṅgābhinaya is mentioned in the Ṇṛttaratnāvalī and in the Ṇṛtyādhyāya

but none of these authors describe them (NR. 2.112; Nr.Adh. 141). The

cārīs of the Abhinayadarpaṇa and the Bharatāṛṇava differ not only in

their names but in their definitions as well.42 The movements of the feet

are divided into four types in the Abhinayadarpaṇa, namely, maṇḍala,

utplavana, sthāna and bhramarī. The Bharatāṛṇava describes twenty-two

movements of the feet which are a mixture of Bharata’s pāda (feet

movements) and cārī, although the Bharatāṛṇava has a separate section on

cārī as well.43 Such discrepancies occur throughout the texts and a

comparison even as cursory as this calls into question the assumption of the

single authorship of the two works.

Furthermore, we may assign them to different periods. The contents

of the Abhinayadarpaṇa place it in or close to the medieval period, for we

find in it the division of dance into three branches, nāṭya, ṇṛtta and ṇṛtya,

and this separation did not come about till the early medieval period when

the Daśarpaka created the distinction, a distinction maintained in

theSaṅgītaratnākara. Also, the Abhinayadarpaṇa views tāṇḍava and

lāsya as forms of masculine and feminine dancing, which again was an

approach taken not before medieval times. As we shall see later in the

detailed discussion of lāsya, it came to be taken as a dance form after

Bharata wrote his Nāṭyaśāstra and even after the time of the

Viṣṇudharmottarapurāṇa it continued to be so understood. The use of the

term ṇṛtya to denote a specific art-form came later still. Even

Abhinavagupta does not use the term ṇṛtya, although it is likely that he

restricted himself to using only the term ṇṛtta because in his task of

interpreting the Nāṭyaśāstra he was bound by Bharata's usage. In the

42Bh. Ar. chapters 8-9, 11-14.

43Ibid. chapters 4 and 8.

MOVEMENT AND MIMESIS

29

Nāmalingānusāsana of Amarasimha (10th century A.D.)44 we come across the term nṛtya, but there it is used as a synonym of nāṭya rather than a signifier of an independent art form. Since Amarasimha's date has not yet been established beyond question, his use of the term nṛtya does not help us fix a date for the earliest occurrence of the term except that it is likely to have been in use before Abhinavagupta who came after Amarasimha.

No text in the first period of the evolution of dancing seems to have dealt with the nature of nṛtya. Dhananjaya, who was a contemporary of Abhinavagupta, was the first to recognize nṛtya as a distinct category of dance but he said nothing about its nature except that nṛtya belongs to the mārga tradition and is bhāvāśraya, that is, it expresses emotions. The Abhinayadarpaṇa's fuller treatment of nṛtya suggests a later date when the concept had developed beyond Dhananjaya and Abhinavagupta.

Even more persuasive an argument for placing the Abhinayadarpaṇa in the medieval period is that it contains material and expresses views that are characteristic of that period. For example, it talks about the attainment of mokṣa or liberation through the performance of the art. This view is characteristics of musicological works from the medieval period, including the Saṅgītaratnākara and the Abhinayadarpaṇa, as Katz points out.45 Since the Saṅgītaratnākara has been placed in the thirteenth century it is unlikely that the Abhinayadarpaṇa was composed much earlier. Furthermore, the concepts of nāṭya, nṛtta, nṛtya, tāṇḍava and lāsya are introduced in a similar manner and defined in the two works, the Saṅgītaratnākara and the Abhinayadarpaṇa, and there are verbal parallels as well.46 M. M. Ghosh believes that the Saṅgītaratnākara borrowed

44Vogel,1974, p.309.

45Katz,1987, Vol. II, commentary on verses 18-21.

46The editor of the Saṅgītaratnākara gives a list of these parallels. See Adyar Library edition, Vol. IV, 1953, pp. 585-87.

30

THE LITERATURE OF DANCE

from the Abhinayadarpana but offers no supporting evidence for his claim

and it is impossible to determine which text is the borrower and which the

source. The movements of the pratyangas, that is, the minor limbs, are not

discussed in the early treatises, the Nātyasāstra and the Viṣnudharmottara

Purāṇa being silent about them, but are included in medieval works. Both

the Abhinayadarpana and the Sangīaratnākara describe pratyangas. They

also describe bhramarī as a separate category of movements, which again

is a medieval development. This close correspondence between the two

texts strongly argues for placing the Abhinayadarpana close in time to the

Sangītaratnākara (13th century A.D.), that is, some time in the early

medieval period.

We must, however, note that the similarity to the Sangītaratnākara is

not the only distinctive characteristic of the Abhinayadarpana. In fact, it

occupies an unique position in the literature of classical dancing of India.

The author, unlike any other, is concerned exclusively with the

performing arts as they apply to dancing, specifically the forms of

abhinaya identified by Bharata, with the exception of vācika. While at

least twenty-nine texts over a period of eighteen centuries describe the art

of dancing, only one text, the Abhinayadarpana concentrates entirely on

āṅgikābhinaya or the use of body movements in acting and describes how

such acting techniques is used in dancing. Judging by the author's concern

for performance it seems likely that this text was meant as a textbook for a

dancer to learn the technique of mimetic dance. Interestingly enough, that

is how it is now used in the training regime of at least one classical style of

Indian dancing, Bharatanatyam, although the practitioners claim that they

are thereby following Bharata, after whom the style is named. In a

number of contemporary Bharatanatyam schools, each student has to learn

by rote the entire text of the Abhinayadarpana and perform the movements

described in it.

MOVEMENT AND MIMESIS

31

The text of the Abhinayadarpana, as it is available now, is quite short,

with only 324 verses. 47 The text begins with a benediction and short

description of the origins of drama, sings its praise, and goes on to describe

the different kinds of nātana (dance), namely, nāṭya, nrtta, nṛtya. It then

speaks of the qualities of the audience, the stage, the performer, the

kiṅkiṇīs (ankle-bells), the prayer to be danced. Then the author instructs

the performer to begin the performance with various stylized body

movements. Next, he establishes the importance of abhinaya and briefly

discusses the characteristics of its four kinds. This whole opening section

takes up only forty verses and the rest are devoted to describing the

movements of the individual parts of the body, which, according to the

author, are of vital importance for a performance. That this discussion is

meant for dancers and actors is quite apparent for it includes descriptions

of the kiṅkiṇīs or ankle bells. Very few texts describe kiṅkiṇīs. Another

unique feature of this text is that although in describing the basic hand-

gestures and the eyes the author follows the Nāṭyaśāstra, his treatment of

the movements of the feet comes from a tradition recorded in no other

text. He also includes some new gestures found in no other text. 48

The edition in which the Bharatārnava is at present available has

fifteen chapters with 996 verses. To this the editor has added a pariśiṣṭa,

an appendix, consisting of 251 verses from a manuscript fragment in the

Bhandarkar Oriental Research Institute. 49 The reason for its inclusion is

not clear, for while the fragment bears the titleBharatārnava, there is no

internal evidence supporting this identification. Against the identification

we must note that the material comes from a different school of dancing

and does not belong to the school which is represented in the

47Ghosh,1957.

48ADar. 1957, pp. 204-44. These include devahasta, daśāvatārahasta, jātihasta and

bāndhavahasta.

49Gairola,1978.

32

THE LITERATURE OF DANCE

Abhinayadarpana. To give only one example, the appendix describes a dance sequence which it calls ālāpacārī (Bh.Ar. Pariśiṣṭa 28-37); this sequence is the same as the mukhacāli described in the Nartanairṇaya (NN. 40a-43a), the 16th century work by Puṇḍarīka Viṭṭhala, and has no parallel in any earlier work.

The beginning of the main text of the Bharatārnava is missing and the text starts from the descriptions of single hand-gestures. The second chapter describes double hand-gestures and the third the hand-gestures used specially in dancing. The fourth chapter gives other varieties of single hand-gestures as taught by Bṛhaspati. It also describes glances and movements of the head and the feet, citing the views of other authorities as well. The fifth chapter describes different postures. The sixth deals with the application of the postures and the applications of combinations of hand gestures. Tāla or rhythm is described in the seventh chapter. The eighth chapter deals with cārī. The ninth describes a new kind of aṅgahāra, of seven types, which is described in no other text. At the end of this chapter the author gives an unusual definition of aṅgahāra as the opinion of the experts: प्रातर्नृत्तप्रकटनैरऱ्हारो विधीयते | (Bh.Ar. 9. 584) – aṅgahāra is prescribed as nṛtta performed [lit: displayed] in the morning.

The tenth chapter again deals with more hand-gestures that express a variety of meanings. The eleventh and twelfth chapters deal with yet another new form, śṛṅganāṭya, of which nine types are mentioned. This form, again, is described in no other text. The thirteenth chapter describes seven lāsyas and seven tāṇḍavas, this being the only text to describe these divisions. The names of the seven lāsyas given here are found among the deśī dance pieces described in the Saṅgītaratnākara and the Nṛttaratnāvalī. The description of the deśī dance prerāṇi parallels the descriptions found in the Saṅgītaratnākara and the Nṛttaratnāvalī.50 The fourteenth chapter describes the use of tāla, gati, karaṇa and cārī in

50 SR. 7.1303-25; NR. 7.34-58.

MOVEMENT AND MIMESIS

33

delineating śuddha and deśī tāṇḍava, a type of tāṇḍava found only in this

text. The treatment of tāla etc. is also entirely new since this text, instead

of merely naming the tālas required in dancing as other texts do,

prescribes how tālas are to be used in actual dance sequences. The

fifteenth chapter is on puṣpāñjali, the right manner of flower offerings,

and other such matters relating to presentation. The descriptions of all the

movements includes their meaning and application, expect for the

nṛttahastas, which are not meant for representational performance.

The appendix seems to be the beginning of a treatise since it opens by

prescribing the preliminaries to a performance, including the appropriate

arrangements for holding a performance, the manner in which singers

should make their entrances and the opening music should be played to

tāla, and the kind of competence and training required in the musicians.

General instructions concerning movements come next. Instructions

follow regarding how the actual performance should begin, with citations

from Kohala (Bh.Ar. Pariśiṣṭa 37). The rest of this section deals with

more hand-gestures, many of them new and found neither in the

Abhinayadarpana nor in any other text. Two verses in the appendix match

verses in the Abhinayadarpana,51 and one of them, on the division of

dance into tāṇḍava and lāsya, occurs also in the Saṅgītaratnākara. 52

On the basis of its treatment of several topics,the Bharatārṇava seems

to be of a later date than the Abhinayadarpana, and the appendix later still.

Besides the differences between the Abhinayadarpana and the

Bharatārṇava pointed out earlier, we must also note that the Bharatārṇava

follows the practice of describing dance pieces along with the gaits,

rhythm, cārī and karaṇa and tāla recommended for them, which is typical

of texts that are later even than the Saṅgītaratnākara and the

Nṛttaratnāvalī. This inclusiveness of the description became common in

51Bh.Ar. Pariśiṣṭa. 1-2; ADar.36-7, 247(reiteration of 36)

52Bh.Ar. Pariśiṣṭa. 38, 39; SR. 7. 28, 30.

34

THE LITERATURE OF DANCE

the literature produced from the sixteenth century onwards. Moreover,

some of the technical terms used in the Bharatārnava, as well as in the so-

called appendix, such as udupa, dhuvāda, kuvāda and sūlu, are terms that

began to appear in the Sanskrit works on dancing around the sixteenth

century and not before.

Although the Abhinayadarpana and the Bharatārnava seem

unquestionably by two different authors and from different periods, the

Bharatārnava is as important as the the Abhinayadarpana as an

instructional manual although it is not so used now, nor do we know if it

was ever so used. Nonetheless, its importance as an excellent practical

guide can not be denied. If the Abhinayadarpana trains a dancer in the

basic movements, the Bharatārnava teaches a dancer to compose a dance

piece. Both pursue the same purpose of instructing practising artists and

not merely of recording the art form of the time. Since its own evidence

places the Bharatārnava after the Abhinayadarpana, it may be regarded as

being complementary to the Abhinayadarpana, and this may explain why

the unfounded identification of the two as by the same author has persisted

for so long.

Another controversial text on music and dance is the

Sangītamakaranda by Nārada. Nārada, as we have noted above, is a

mythical name which appears in almost all dance literature in Sanskrit.

Nārada appears in the literature of the medieval period as well as in the

early period, in the works of Bharata, Dattila, Matanga and

Abhinavagupta.53 We have at least two Nāradas, according to Raghavan,

one who wrote the Sangītamakaranda and one who wrote the Nāradiya

Siksā.54 Raghavan places the Nārada of the Sangītamakaranda after the

seventh century since the text refers to Mātrgupta, who is believed to have

53Matañga, Brhaddesí, 2;Lath,1978 p.48, 52, 557; NS. 5. 32; A. B. on the NS. Vol. I.

1956 p. 221.

54 Raghavan,1957, p. 23.

MOVEMENT AND MIMESIS

35

lived around that time, but it does not specify any date for Mātṛgupta.55

Nijenhuis suggests that this text was written after the Sangītaratnākara.

She believes that this author is not the same as the one to whom the

Sangītaratnākara refers.56 Another work by a certain "Nārada,"

Catvāriṃśacchatarāganirūpaṇa, has been placed in the 16th-17th

century.57 Svāmī Prajñānānanda states that there are four works by four

different Nāradas: the Nāradiya Śikṣā (1st century A.D.), the

Pañcamasārasaṃhitā (1440 A.D.), the Sangītamakaranda (14th/15th

century A.D) and the Rāganirūpaṇa (16th/17th century A.D).58 The

editor of the Sangītamakaranda(1920), M. R. Telang, places Nārada, the

author of the Sangītamakaranda, between the 7th and 11th centuries,59

but this dating remains doubtful. The problem with fixing a date for

Nārada is that this is a mythical personage who appears in all the

literature of dance and music as an authority and there is more than one

Nārada. The whole problem has been compounded by the fact that the

edited version is a corrupt text and one that describes a debased style

recorded in no other text.

The Sangītamakaranda has two parts, one on music and the other on

dance, each divided into four sections. The whole text is obscure but the

dance portion is the most difficult part of this generally difficult work.

Many of the difficulties arise from the text as edited. It is hard to get a

picture of the tradition but the little we may glimpse shows a decadent

style.60 This seems to be a deśī tradition that was obviously not very highly

55Ibid. p.23.

56 Nijenhuis,1977, pp. 13-14.

57Ibid. p.19.

58Ibid. p.20; Prajñānānanda,1963. p. 114.

59Sangītamakaranda, 1920, Introduction, p. x.

समगौकति विशेषज्ञाता हृस्तौ वामस्तने धृतौ ।

कैवालं दक्षिणकुचे लग्नौ चुम्बकदर्शनम् ॥ SMak.4.4.13.

36

THE LITERATURE OF DANCE

regarded-possibly because of its vulgar eroticism-since it appears in no other text. The first part of the text is entirely on music. In the first section of the part on dance the author discusses the dance-hall, the audience, the poet, the singer, types of learned spectators, the presiding person, the dance-teacher, the percussionist, the performer, the flower-offering and the origin of the tālas. The second section names the makers of tāla and describes the characteristics of 101 tālas. The third section gives a variety of information on tāla, including the derivation of the word, the essence of tāla, the time, mārga, deśī and other such technicalities of music. The fourth section is devoted to drums but also contains a short final subsection of 33 verses called naṭibhāvanirūpanam which is devoted to dance. It describes five double hand-gestures, five single hand-gestures, eight bramarīs, nine head movements and four feet movements. The text concludes by referring to 33 naṭibhāvas, that is, emotional expressions appropriate for women, but does not elaborate on them. None of the movements described in this text is found in any other text. The technical terms do not match any term found in the existing literature, except for two, cakrabhramarikā and sama (pāda), and even these two are described differently.

As for the date of the Sangītamakaranda,there are a number of reasons why it must be placed in the medieval period. The text describes dance as nṛtya which is a term found in use not before the time of Dhananjaya (10th century A.D.),61 whose Daśarūpaka was the first surviving work to define nṛtya. The concept of mokṣa, which we have already noted as a medieval concept, is present in this text.62 But the strongest reason to place it in the medieval period is that it describes an entire style that falls outside the

[When] the hands hold the left breast it is known as samagaika [and when the hands] touch the right breast to show the nipple it is kaivāla.

61Dasgupta,1962, Vol. I. p. 550.

62See p. 29 supra; SMak. 1.1.24.

MOVEMENT AND MIMESIS

37

tradition described by Bharata and his followers. This new style was part

of the deśī tradition and it was only in the medieval period that the deśī type

of dancing came to be accepted as one that was worth recording.

However, no precise dating of the Sangītamakaranda has been possible

because it contains no material paralleling any other text by means of

which terminal dates might be set. However, the type of material it does

contain allows the conjecture, supported by at least one musicologist,63

that it was not written before the Sangītaratnākara.

The Daśarūpaka of Dhananjaya, a work on dramaturgy, marks the

beginning of the second period. Although Abhinavagupta and Dhananjaya

(924-995 A.D)64 were contemporaries, yet Dhananjaya must be placed in

the second period because in his work we find for the first time how

certain concepts had evolved beyond Bharata. The first period

is dominated by the concepts enunciated by Bharata, and these form the

subject of Abhinavagupta. It is true that Abhinavagupta interprets them in

commenting upon the Nātyasāstra and thereby develops them, thus

indicating that the process of evolution was under way, but the conceptual

framework remains entirely that of Bharata. It is in the Daśarūpaka that

we find the recognition of these new categories and concepts defined for

the first time. The most important of these are represented by the terms

nṛtya, lāsya, mārga and deśī.

The Daśarūpaka has four prakāśas or sections. As the name suggests,

the work is on the ten types into which Bharata had divided dramatic

literature, and it analyzes their technical features, concentrating on plot

construction in particular. A systematic and compact work, it replaced the

Nātyasāstra for the later authors as a guide to the technicalities of plays.

The first chapter, which consists of 68 verses, starts with a benediction

and proceeds to offer short definitions of the terms nātya, nṛtya, nṛtta,

63Nijenhuis,1977, p.14.

64Gerow,1977, p. 263.

38

THE LITERATURE OF DANCE

mārga, tāṇḍava and lāsya. Here Dhananjaya takes nṛtta and nṛtya as

auxiliaries to drama and hence necessary to his study. After defining the

terms the author moves on to his main purpose, that of defining and

describing nāṭya or drama. According to Dhananjaya, vastu, netā and

rasa, that is, the subject matter, the hero and the aesthetic affect, are the

three matters of primary importance and the rest of the first chapter is

devoted to the discussion of vastu in detail and of the structural

components of a play. The second chapter, in 72 verses, describes the

characters in a play, especially the hero, the heroine, their qualities, the

method of expressing various emotions etc. Vṛttis or acting styles and the

use of language by different characters are also discussed in this chapter.

The third chapter, in 70 verses, deals with the varieties of rūpakas or plays

and identifies lāsyāṅgas as components of bhāṇa, a major type of play.

The fourth chapter, in 87 verses, discusses the nine rasas or aesthetic

affects, the eight sthāyībhāvas or mental states, and the thirty-three

vyabhicārībhāvas or transitory moods.

The definitions of the terms nṛtta, nṛtya, tāṇḍava and lāsya in

Dhananjaya mark a distinct stage in the evolution of the understanding of

dance and drama. Although Dhananjaya does not elaborate on the

concepts he defines, one of his most important contributions is the

distinction he draws between nṛtta and nṛtya: . . . भावाश्रयं नृत्यम् (DR. 1.

9.) -- nṛtya is dependant on emotion. He distinguishes nṛtta from nṛtya by

saying: नत्तं ताललयाश्रयम् (DR. 1. 9)-- nṛtta depends on rhythm and tempo.

He is thus the first writer known to have used the term nṛtya to denote

mimetic dance and also dance-dramas. He further states that nṛtya

belongs to the mārga tradition and nṛtta to the deśī.65 While this is a

significant departure from Bharata, Dhananjaya remains anchored to

Bharata's basic view that both nṛtta and nṛtya are auxiliaries to drama.

Thus, in principle, Dhananjaya continues the tradition established by

65 DR. 1969 1. 9-10.

MOVEMENT AND MIMESIS

39

Bharata. Dhananjaya further identifies tāṇḍava and lāsya as subcategories

of both nṛtta and nṛtya, tāṇḍava being a vigorous mode of dancing and

lāsya a delicate one.66

To interpret nṛtta as deśī is a surprising view because nṛtta is accorded

the status of mārga in almost all works on dancing on the ground that

Bharata describes only nṛtta and distinguishes it from the deśī styles

which, he expressly says, do not form his subject.67 It is not till the

thirteenth century that we find another work, the Saṅgītaratnākara,

calling nṛtya a mārga form of dancing. Still later, in the seventeenth

century came the Saṅgītadarpana,68 which was the third and only other

text to categorize nṛtta and nṛtya respectively as deśī and mārga, repeating

verses from the Saṅgītaratnākara. One of the most important pieces of

information we get from the Dasarūpaka concerns bhāṇa, which is one of

the ten major types into which drama is divided by Bharata, who calls it a

one-act play. But the Dasarūpaka extends our understanding of the type by

attributing to it delicate movements and the arousal of śṛṅgārarasa.

According to the Dasarūpaka, such movements are employed in lāsyāṅgas

which thereby generate śṛṅgārarasa. Dhananjaya further states that

lāsyāṅgas are essential parts of bhāṇa.69 He does not specify the

particular aṅgas or features of lāsya, and the ten lāsyāṅgas he names

match those found in the Nāṭyaśāstra, except one, dvigūḍhaka, which is

called dvimūḍhaka in the Nāṭyaśāstra, but this may be the result of a

scribal error in the Nāṭyaśāstra.70 While Dhananjaya follows the

Nāṭyaśāstra in naming the lāsyāṅgas, he ventures in a new direction by

66 DR. 1.10; 1.11.

67 अन्ये तु लोकिकया ये ते लोकाद् ग्राह्या: सदा बुधै:। NS. 25.119.

68 DR. 1.10; SR. 7.27; SDar. 7.94.

69 DR. 4.51-53.

70 DR. 4.52.

40

THE LITERATURE OF DANCE

relating them to the bhāṇa and thus placing this type of expressive

movement in the domain of drama.

The evidence of the Daśarūpaka makes it clear that by the tenth

century śṛṅgārarasa (aesthetic pleasure of the erotic variety) and

kaiśikīvṛtti (graceful style) came to characterize lāsyāṅgas which had

become standard parts of bhāṇa. This is confirmed by Abhinavagupta,

who cites various authors regarding the development of a certain bhāṇa

play.71 This feature was absent from Bharata’s concept of bhāṇa, and he

specifically forbids the use of kaiśikīvṛtti.72 Dhananjaya is silent on

minor dramas, but Dhanika, the commentator on the Daśarūpaka, cites

seven minor dramas, calling them ṇṛtyabhedas, without defining them.73

Apart from the information mentioned above, this text gives one other

that is quite vital to our understanding of the use of gesture language in

drama. Gestures obviously formed a very important technique for

expressing meaning in the performance of a play. Dhananjaya mentions

tripatākākara, the hands in tripatākā, to indicate janāntika, an aside.74

That he does not take the trouble to define the term indicates that the use of

these gestures was common in his time and therefore required no

explanation.

The Daśarūpaka reflects considerable changes in the discourse on

dancing since Bharata’s Nāṭyaśāstra. Dhananjaya’s strength lies

particularly in the fact that he composed a methodical account of the

categories of dance and provided clear, if brief, explanations. Prior to his

work, much of the information available, including what we find in

71 AB. on the NS, Vol. I.1956, p. 181. Abhinavagupta’s use of both bhāṇa and bhāṇaka

for this minor type caused some confusion which continued through the entire

literature on dance and drama.

72 NS. 18.8-9.

73 DR. 1969, p. 8 (Avaloka of Dhanika).

74 DR.1.65.

MOVEMENT AND MIMESIS

41

Abhinavagupta, is fragmentary, existing as quotations from lost works or

from the general body of literature. Sometimes the information comes in

as passing remarks or views not clearly expressed. In Dhananjaya the

concepts and the categories are set down and defined unambiguously

enough to suggest that their meanings had come to be generally accepted.

Our next source is the Śṛṅgāraprakāśa of Bhoja which is believed to be

from the eleventh century.75 As the name suggests, it is an exposition of

śṛṅgārarasa. The term śṛṅgāra here must be understood in the context of

Bhoja's "new found theory of the one Rasa of Ahaṅkāra, Abhimāna or

Śṛṅgāra as well as the lower śṛṅgāra of Rati."76 This text is a work in

thirty-six chapters and deals with both poetics and dramaturgy. As

Raghavan points out, "Bhoja seems to be the first writer, whose work is

available to us now, who embraced both branches and wrote in the same

work on both."77 It seems from the treatment of the subject that Bhoja

wanted to write a work similar to the Nāṭyaśāstra. Like Bharata, Bhoja

discusses the characteristics of the Sanskrit language, but does so even

more elaborately than Bharata, although-surprisingly-Bhoja omits

prosody, which is closely related to poetics.78 However, he discusses

dramaturgy in detail and his analysis of śṛṅgāra is unique, covering as it

does the types of śṛṅgāra and its expression.in dramatic presentation. The

range and depth of the work has been noted by Raghavan who discusses its

contents and compares it with other works on the subject.79

The part of the work most relevant to our present study is Bhoja's

discussion of the minor types of plays. In the eleventh chapter he describes

twenty-four types of drama and their structure which he terms as

75 Raghavan,1956; p. 29.

76 Raghavan,1978, p. 8.

77 Ibid. 1963, p. 9.

78 Ibid. p.9.

79 Ibid. The comparative study continues through the whole work of Raghavan.

42

THE LITERATURE OF DANCE

preksyaprabandhas, that is, poetical compositions to be seen. These he divides into two subcategories according to the kind of acting necessary,

one requiring vākyārthābhinaya and the other padārthābhinaya.80

Vākyārthābhinaya literally means expressing the meaning of utterances, that is to say, the acting technique required for this type of drama brings

out the entire meaning of the play. This is the kind that represents major drama. Padārthābhinaya on the other hand is less demanding and calls for

expressing the meaning of the words (in a song) and is employed in minor dramas. Bhoja does not name the class of drama that requires

padārthābhinaya, but he does name and describe the twelve varieties within the class. These descriptions show that these varieties are full of

delicate and expressive movements. Going back to Dhananjaya's definition of nṛtya we find that he defines nṛtya as an art requiring

padārthābhinaya (DR. 1.9). If padārthābhinaya can thus be correlated with nṛtya, which seems plausible, then it is likely that Bhoja viewed as

varieties of nṛtya the dramatic types that require padārthābhinaya. Dhanika in his Avaloka on the Daśarūpaka and Śāradātanaya in his

Bhāvaprakāśana view these minor dramas as nṛtyabhedas, that is, as various types of nṛtya.81 Later, Śubhaṅkara refers to these varieties as

"nṛtye ratnanakṣatramālā."82 As these references show, these minor dramatic types were known under many names. Later, however, they

were categorized as uparūpakas by Viśvanātha in the Sāhityadarpana (fourteenth century A.D.)83 and we shall examine their relevance to

dance in a later chapter. It is in that context that Bhoja's views on the use of padārthābhinaya in preksyaprabandhas are of particular significance.

80 The text of Śṛṅgāraprakāśa, Vol. II.1963, pp.461, 466.

81 DR. Avaloka, 1969, p. 8; Bh.P. 1968, pp. 255-68.

82 Saṅgītadāmodara, 1960, .pp. 79-80.

83 Sāhityadarpana, 1923, 6. 4-6.

MOVEMENT AND MIMESIS

43

The Nāṭyadarpana of Rāmacandra and Guṇacandra is a treatise on dramaturgy written in the twelfth century.84 As the name suggests, the text reflects the characteristics of drama. After writing the main text the two authors comment upon their own work to elucidate it. The work is in four chapters and in the third chapter, while discussing āṅgikābhinaya, the authors list by name the movements of the different parts of the body as well as extended sequences and compositions. Their method of listing the terms for movements is to select one or two terms to indicate an entire class and then to state the number of movements in that class, as we find, for instance, in phrases such as: हस्तयोः पताकत्रिपताकादयश्चतुःपष्टिः or स्थिरहस्तपर्यस्तकादयोऽष्टत्रिंशत् and so on.85 This list matches the list given in the Nāṭyaśāstra and we find that the division of limbs in this work into aṅga and upāṅga is also a continuation of the tradition of the Nāṭyaśāstra, although the movements of the limbs are merely named, not described. While it follows the Nāṭyaśāstra, in many respects, it also provides additional information. It virtually copies the Nāṭyaśāstra in describing āṅgikābhinaya, but names two additional types of major dramas that are not found in the Nāṭyaśāstra, raising the number of major dramatic types to twelve. Again, the description of bhāṇa is different from that in the Nāṭyaśāstra, and following the tradition that developed after Bharata it includes śṛṅgārasa and ten lāsyāṅgas as features of bhāṇa. One other major topic that this text introduces is a discussion of the minor types of drama for which it does not have a class name although it describes thirteen varieties of minor drama in detail. 86 The Nāṭyaśāstra, as we know, does not describe minor dramas. But that they existed is evidenced by Abhinavagupta’s citations in the Abhinavabhāratī and in Dhanika’s

84 Krishnamachariar,1974, p. 643.

85 Nāṭyadarpana, 1959, pp.168-69.

86 Ibid. pp. 190-92.

44

THE LITERATURE OF DANCE

Avaloka on the Daśarūpaka,87 though Dhanika does not describe them.

Bhoja was the first to include and describe twelve varieties of minor

dramas, giving details, and the Nāṭyadarpana was the next work to include

them in the category of drama. However, these dramatic types still lacked

an accepted class-name. Different names appear in different works, such

as ṇtyakāvya and rāgakāvya in the Abhinavabhārati, ṇtyabheda in the

Avaloka of Dhanika, geyarūpaka in the Kāvyānuśāsana of Hemacandra.

Rāmacandra and Guṇacandra simply categorize them as anyāni rūpakāṇi,

that is, other plays.88 Even Śāradātanaya, who, writing in the thirteenth

century, gives the largest list, consisting of twenty minor dramas, refers to

them as ṇtyabhedas.89 Oddly enough, much later Subhaṅkara includes

both rūpakas and uparūpakas under ṇtya (SDām. pp. 79, 80, 99).

However, the description in the Nāṭyadarpana has lines almost identical

with Bhoja's description except that it adds one other minor dramatic type,

saṭṭaka.90 In a later chapter we shall discuss the development of these

minor types of plays as well as their nature.

From the inclusion of these minor types in the Avaloka of the

Daśarūpaka and in the Śṛṅgāraprakāśa, a practice that is followed by all

treatises on music, drama and dance for several centuries, we are justified

in surmising that these minor types of dramas were popular and had

become well-developed and well established during this second period of

our study. This was no doubt the reason why they were recognized by

scholars as important enough to be included in works on dramaturgy.

87Abhinavagupta mentions ten such minor types (AB. on the NŚ. 1956, pp. 168-84.);

Dhanika mentions seven,1969 (Avaloka on the DR. p. 8).

88AB. on the NS., Vol. I, pp. 171, 175; Avaloka on the DR. 1969, p. 8; KAnu.H. 1938, 8.1;

ND. 1959, p. 190.

89Bh.P. 1968, p. 256.

90ND. 1959, p. 190-92.

MOVEMENT AND MIMESIS

45

However, the precise genre to which they belonged remained a matter of

varied opinion.

Another text of great interest from the twelfth century is Mānasollāsa

by Someśvara, also known as Abhilaṣitārthacintāmaṇi.91 In its one

hundred chapters it deals with a variety of subjects that interest the

royalty. A guide to royal pastimes, this book has five parts, each

containing descriptions of twenty vinodas or pastimes. The fourth part

deals with dance and music in chapters sixteen to eighteen. Three others

are on vocal and instrumental music and dance and the fifth is on

miscellaneous matters. The part called nṛtyavinoda has 457 verses, from

16.4. 949 to 16.4.1406. He introduces us to the subject first by saying when

dances should be performed: at every festive occasion, to celebrate

conquests, success in competitions and examinations as well as occasions

of joy, passion, pleasure and renunciation. Six varieties of dancing are

described next and then six types of nartakas. Nartaka here stands for

performers in general and includes nartakī (danseuse), naṭa (actor),

nartaka (dancer), vaitālika (bard), cāraṇa (wandering performer) and

kollāṭika ( acrobat). He then describes body movements, dividing them

into aṅga, upāṅga and pratyanga. The division of the minor limbs into two

categories, upāṅga and pratyanga, is a new approach. The Nāṭyaśāstra

divides the body into major and minor limbs and the term Bharata uses for

the minor limbs is upāṅga. This division is followed by the

Viṣṇudharmottara Purāṇa. But Someśvara introduces this further

subdivision of pratyangas. This division appears in almost all works on the

subject except for the Nāṭyadarpaṇa which, as we have noted above,

follows the Nāṭyaśāstra.92 The movements of six aṅgas, eight upāṅgas

and six pratyangas are discussed in detail, with the meaning and

application explained. The one type of movement to which meaning is not

91 Mānasollāsa, G.O.S. edition, Vol. I. 1958. p. vi.

92 NŚ. 8.8; VDP. 3.24.65; ND. p. 168.

46

THE LITERATURE OF DANCE

attributed is the nṛttahasta, because these movements are meant for adding

beauty, not meaning. These are discussed in verses 978-1306. Verses

1307-78 are devoted to describing 21 sthānas and 26 cārīs, of which 16 are

described. Verses 1379-99 describe 18 karaṇas of the deśī variety, none of

which is found in earlier works. He finishes this section by saying that

these should be performed either by a dancer or by the king himself to

please his beloved. But performance of the viṣama (acrobatic) and vikaṭa

(comic) varieties of dancing should be avoided by the king since they

cannot generate śṛṅgārarasa. This suggests that generating śṛṅgārarasa

was taken to be a prerequisite for dance. The preferred audience would

be connoisseurs of dancing, and dances should be performed inside the

palace or a house, or in a pleasant courtyard or a garden. Thus ends the

section on dancing.

In the Mānasollāsa we find the recognition of four new categories that

developed after Bharata, namely, nṛtya, lāsya, mārga and deśī. Nṛtya was

first recognized as a distinct category of performance in the Daśarūpaka.

The Mānasollāsa takes the term to represent the whole art of dancing. It is

also the first text with a complete and sustained discussion on dancing

which treats lāsya as a division of dancing. Someśvara also divides

dancing into the mārga and deśī varieties, thus reflecting the growth of the

two traditions. The other important contribution of Someśvara is the

introduction of the deśī karaṇas, eighteen in number and found in no other

text.

However, Someśvara defines neither nṛtya nor nṛtta. The term nṛtta

occurs only a few times in the text some of which might have been due to

scribal errors since the author does not seem to be concerned with nṛtta

except for using the word to form the term nṛttahasta.93 The term he uses

specifically for dancing is nartana, which he divides into six types: nāṭya

(mimetic), lāsya (delicate), tāṇḍava (vigorous), viṣama (acrobatic),

93 Mānas. 16.4. 957 ; 1138 ; 1150; 1238; 1262;1269-70; 1277; 1296.

MOVEMENT AND MIMESIS

47

vikaṭa (ludicrous) and laghu (light and graceful). The inclusion of nāṭya as

a form of dancing is a continuation from the Viṣṇudharmottara Purāṇa's

division of dancing,94 and the use of the term shows that it obviously had a

wider meaning at this time than in later times. We shall reserve its

discussion for a later chapter.

A work from this period, but not dated with certainty, which deals with

drama is the Nāṭakalakṣaṇaratnakośa of Sāgaranandin. From internal

evidence Raghavan suggests in his introduction to the translation of the

text that Sāgaranandin's date could not have been earlier than that of

Abhinavagupta, Dhanañjaya or Bhoja, that is, not earlier than the

thirteenth century, nor later than that of Śāradātanaya.95 The text, as the

name suggests, deals with the definition and nature of nāṭaka as well as

other varieties of drama in the fullest detail. The author devotes a

substantial part of the text to minor dramatic types but does not put them

into a separate category. He follows the tradition of the Nāṭyaśāstra in his

approach to the major dramatic types and also in his discussion of rasa; for

example, following Bharata, he does not include śāntarasa, which was a

category introduced by Abhinavagupta.

One of the most important contributions of the Nāṭakalakṣaṇaratnakośa

is that it provides a systematic overview of the literature that preceded it.

The same task was undertaken, but not performed methodically, by

Śāradātanaya, whom we shall discuss next. The survey in the

Nāṭakalakṣaṇaratnakośa helps us to understand the views of many earlier

writers on many topics of Indian dramaturgy. But Sāgaranandin also has

some noteworthy views of his own.96 We may remember that

94 VDP.3.20.2.

95 Raghavan's Introduction to the translation of the Nāṭakalakṣaṇaratnakośa, Vol. II.

1960, p.7; S. Chattopadhyaya, 1960, puts the lower limit of the date as the eleventh

century (Introduction, p.xxxii).

96 Nāṭakalakṣaṇaratnakośa,1960. Vol. II. p.7.

48

THE LITERATURE OF DANCE

Abhinavagupta did not deal with minor dramas in much detail.

Dhanañjaya paid no attention to them. Dhanika simply acknowledged the

existence of such types by naming some of them. Bhoja was the first

author to discuss the minor varieties of drama and to describe their

characteristics, although he did not treat major and minor dramas

separately. Sāgaranandin gives us a clear picture of the technical aspects

of the plays of his time, noting in particular the technical language of the

theatre: "The nāṭakaparibhāṣā or stage jargon, the conventions, the

technical terms and the mode of address peculiar to drama, is another

subject dealt with in NLRK (section 16) for which we are indebted to it;

the only other works devoting a separate section to this subject are the

Sāhityasāra of Sarveśvara and the Rasārṇavasudhākara of

Singabhūpāla."97 That his interest lay in technical issues of a practical

nature is evident everywhere: "Towards another subject of similar

historical nature, the music and dance items of pūrvarañga, Sāgaranandin

adopts a more practical attitude (1157-1158) which is repeated by

Viśvanātha in a more outspoken form in his Sāhityadarpana."98

The list as well as the descriptions of minor types of drama given in

this text, seventeen in all, are similar to what we find in Śāradātanaya's

discussion of nrtyabhedas in his Bhāvaprakāśana. In a later chapter of the

present study, in which the development of concepts after Bharata is

discussed, we shall see how the minor types of plays developed.

The Nāṭakalakṣaṇaratnakośa follows the Nāṭyaśāstra in all respects

but two. In the first place, as we have seen, it discusses minor dramatic

types and not only major types as Bharata does. Secondly, the

Nāṭakalakṣaṇaratnakośa deviates from the Nāṭyaśāstra in prescribing

lāsyāñgas for the performance of types of bhāṇa. Bhāṇa, a major dramatic

type discussed in the Nāṭyaśāstra, has caused a great deal of confusion in

97 Ibid. p. 8.

98 Ibid. p. 8.

MOVEMENT AND MIMESIS

49

the literature, from which the Nāṭakalakṣaṇaratnakośa is by no means

free. After Bharata, there developed two minor dramatic types, bhāṇaka

and bhāṇikā (often referred to as bhāṇa and bhāṇī). These had developed

from bhāṇa, as the names indicate, and were similar to it inasmuch as they

were one-act plays. What mainly distinguished the two was the

application of kaiśikī vṛtti and the arousal of śṛṅgārarasa, both expressly

forbidden by Bharata. The Nāṭakalakṣaṇaratnakośa, however, reflects the

post-Bharata development by including kaiśikī vṛtti and śṛṅgārarasa as

features of bhāṇaka and bhāṇikā. Of greater interest to the present work is

the fact that the Nāṭakalakṣaṇaratnakośa prescribes the use of ten

lāsyāṅgas in bhāṇa and bhāṇī. The names and descriptions of the lāsyāṅgas

remain virtually unchanged from the Nāṭyaśāstra, though the context is

entirely different. The Nāṭakalakṣaṇaratnakośa thus shows the dramatic

use of the lāsya type of dancing that had evidently become part of the

tradition.99

The Bhāvaprakāśana of Śāradātanaya (1175-1250 A.D.)100 is a

compendium of poetics and dramaturgy based on all critical works written

before the author's time, that is, the period starting with the Nāṭyaśāstra.

The author discusses different schools of literary criticism, sometimes

accepting other's views and sometimes refuting them or giving his own

views. The editor of this text has traced Śāradātanaya's indebtedness to

his predecessors in detail in the introduction to the edition.101 The text

contains ten adhikāras or chapters, of which the first three are on bhāva

and rasa. The fourth and the fifth are on the characters of the hero and

heroine. Of relevance to dancing is a discussion at the end of the fifth

chapter on glances that express rasa and bhāva. The sixth chapter

discusses the relation between word and meaning. Chapters seven to ten

99 NLRK. 2853-98; See also Bose, 1970, p. 15.

100 Raghavan, 1957, p. 28; Bh.P. 1968. G.O.S. ed. Introduction, pp. 72-77.

101 Ibid. pp. 63-71.

50

THE LITERATURE OF DANCE

all contain material of immediate interest to us. The seventh chapter is on

nāṭya which includes dance and music. The eighth chapter discusses

major and minor types of dramas, while the ninth chapter discusses in

detail the minor types, putting them into a separate category which the

author calls nṛtyabhedas. The ninth chapter also includes a discussion on

different types of kāvyas or poetic compositions. The tenth and final

chapter is of particular interest for it explains the distinction between nṛtta

and nṛtya and between mārga and deśī. The information on dance given in

the previous chapters is summarized in the tenth, which further describes

the characteristics of various kinds of performers and audiences. The

chapter also covers a number of topics of general interest, adds a few

more words on music, rasa and bhāva, alaṅkāra, and discusses the land

known as Bhāratavarṣa, her languages and their use in a dramatic

performance.

Although Sāradātanaya is very informative, his treatment of his

subjects (particularly, drama and dance) is unsystematic. He often moves

abruptly from one topic to another and gives inconsistent information on the

same topic at different times, drawing-particularly-upon different

sources or theories, which, however, he does not adequately explain. As a

result one has to scan the entire text and correlate the scattered discussion

to get a cogent overview of his opinions.

An analysis of some topics in the chapters of our interest will clarify

my comment on his work. In the seventh chapter he introduces major

dramatic forms and puts other forms into a category that he calls

nṛtyabheda: नृत्यभेदा भवेयुस्ते डोम्बीश्रीगदितादयः (Bh.P. p.181), that is,

[spectacles such as] ḍombī, śṛṅgadita and so on are divisions of nṛtya. He

does not comment on their nature in this context since his main interest at

this point is the use of rasa and bhāva in these spectacles, in discussing

which he follows Bhoja. The plays, called rūpakas, which are dependant

on vākyārthābhinaya, literally, expressing the meaning of the sentences,

produce rasa and are व्याक्यार्थाभिनयात्मकाः: that is, they aim at bringing

MOVEMENT AND MIMESIS

51

out the total meaning. The पदार्थाभिनयात्मक spectacles, that is, the

spectacles expressing the meaning of particular words, are dependant on

भाव. He further says that नृत्य is dependent on emotion: भावाश्रय, a

view shared by Dhananjaya.102 About नृत्त, however, Sāradātanaya has

his own opinions. He believes that नृत्त is dependant on rasa: नृत्तं

रसाश्रयम् (Bh.P. p. 181.). Not only is this view original to him, it

contradicts Bharata's view that, न खल्वर्थकथनचिन्तनृत्तमपेक्षते (NŚ. 4.263),

that is, नृत्ता does not depend on any meaning. Bharata further says that न

हि रसादृते कश्चिदर्थः प्रवर्त्ते (NŚ. Vol. I, 1956, p.271), that is, no meaning

is expressed without [producing] any rasa. Therefore, Sāradātanaya's

definition of नृत्ता seems to assume that it expresses meaning and does not

merely beautify a dramatic presentation. Accordingly, नृत्त can generate

rasa.

Sāradātanaya differs not only from Bharata but from all his

predecessors and contemporaries. In the Dasarūpaka Dhananjaya views

नृत्त as dependant on tāla and laya, rhythm and tempo: नृत्तं ताललयाश्रयम्

(DR.1. 9) and recognizes only नाट्य as an art-form expressing rasa: नाट्यं

रसाश्रयम् (DR.1. 9). In the Sangītasamayasāra, Pārśvadeva, another

author who flourished around this time, says that नृत्त depends on tāla,

भाव and laya: तालभावलयायत्तः . . . . (SSam. 7.2). That नृत्त is rasāśraya

is Sāradātanaya's own interpretation.

After introducing the dramatic forms in the seventh chapter

Sāradātanaya elaborates on the preliminaries of dramatic performances

which include music and dance. In this part of the discussion he mainly

follows Bharata. Sāradātanaya states that the definition of नृत्त and नृत्य

are many and given by various authors, which he will discuss later. He

views these two art-forms as auxiliary to drama that are used in the

preliminaries of a play. In this context he follows the tradition of Bharata.

102 Ibid. p. 181; DR. 1.9.

52

THE LITERATURE OF DANCE

In the eighth chapter he discusses ten major types of drama that depend

on rasa and twenty others that depend on bhāva. He says that all the thirty

varieties are often referred to as rūpakas.103 He discusses all the major

types of drama, including bhāṇa, in detail. He states that ten lāsyāṅgas are

used in bhāṇa. After describing the ten lāsyāṅgas, following Bharata, he

describes four nrtyabhedas, which he says other authorities discuss. These

he includes in the same section on bhāṇa. Not minor types of drama, these

nrtyabhedas are the four piṇḍibandhas or group dances described by

Bharata in his fourth chapter and viewed by him as parts of the

preliminaries to a play. The descriptions of these four nrtyabhedas in the

Bhāvaprakāśana, however, are much more precise and easy to

comprehend compared to Bharata’s cryptic definition ( NŚ. 4. 290-91).

The discussion ends with instructions for training a dancer in these dances:

भद्रासनेन यन्त्रेण तत्तच्छिक्षा विधीयत` (Bh.P. p. 246) -- the training is

prescribed through bhadrāsana and yantra.104 The actions implied in

these terms are hard to visualize through the bare mention of the terms in

this text as also in the Nāṭyaśāstra and the Viṣṇudharmottara Purāṇa

(VDP.3.20.55). Abhinavagupta and the Rājapraśnīya, one of the

Jainasūtras, clarify the terms to some extent, as will be seen in a later

chapter of the present study in which the conceptual development of the

art of dance will be discussed in detail.

The ninth chapter of the Bhāvaprakāśana is devoted to the nrtyabhedas,

that is, the twenty minor varieties of drama mentioned in the seventh

chapter. One of these nrtyabhedas is what he calls bhāṇa, which is

different from the bhāṇa that is known as a major dramatic type. The use

103 Bh.P. p. 221.

104 The actions denoted by the terms भद्रासन and यन्त्र are not very clear from the

definition of Bharata, and Śāradātanaya has not explained them either. These seem

to be formations of figures on the stage by a group of female dancers. Bharata also

says that students need proper guidance to be able to perform them (NŚ. 4. 290-91).

See pp.113-16infra for a discussion of the piṇḍibandhas.

MOVEMENT AND MIMESIS

53

of the same term to name both major and minor types, each with a

different definition, has caused a great deal of confusion. This chapter

gives a detailed description of each kind of these nṛtyabhedas with

illustrations from literary works, some of which have survived, like the

Karpūramañjarī of Rājaśekhara, Vikramorvaśīyam of Kālidāsa,

Ratnāvalī of Harṣa and others.105 These are major sources for our

understanding of the development of dancing.

The last chapter, the tenth, is used to sum up what the author has dealt

with so far. He begins by saying: उत्तो नाटचस्य नृत्यस्य भेदाः सर्वे यथार्थतः:

(Bh.P. p.284)--the varieties of nāṭya and nṛtta have all been explained [by

me] in an appropriate manner. In the middle section of the chapter (pp.

295- 302) he summarizes what had been said about dance and drama in the

preceding chapters, but adds new interpretations, and these have often

confused the issues. However, we shall try to clarify them as far as

possible when we examine them more closely in a later chapter.

The Saṅgītasamayasāra of Pārśvadeva is the second work devoted to

musicology during the second period of our study. This text discusses

saṅgīta, that is, gīta (vocal music), vādya (instrumental music) and nṛtta

and nṛtya (dance). Besides the Mānasollāsa, other works from the second

period discussed so far are works on dramaturgy and literary criticism.

The Saṅgītasamayasāra is the second work of the period to discuss music

and dance as separate art-forms and not as art-forms auxiliary to drama.

The date of the author is uncertain. According to Raghavan,

Pārśvadeva, was a Jaina writer who must have lived between 1165-1330 A.

D. because he refers to Bhoja (1010-1050 A.D), Someśvara (about 1131

A.D.), Paramardi (about 1165 A.D.) and Siṁhabhūpāla(about 1330

A.D.)106 In the introduction to his edition of Nṛttaratnāvalī, Raghavan

105 Bh.P, pp. 238, 244, 269, etc.

106 Raghavan,1956, p, 29.

54

THE LITERATURE OF DANCE

places Pārśvadeva between Paramardi and Jāya (1253 A.D.)107 The

Sangītasamayasāra was first edited in 1925 from one manuscript by T.

Ganapati Shastri who does not suggest any date for the author. This

edition in the Trivandrum Sanskrit Series is often corrupt. A second

edition came out in 1977, prepared by Acarya Brihaspati and based on the

collation of more manuscripts and of T. Ganapati Shastri's edition.

Comparing the text with other relevant works, the editor places

Pārśvadeva after Śārṅgadeva and in the later part of the thirteenth century.

Pārśvadeva is similarly placed in the thirteenth century by M.

Krishnamachariar.108 From what the Sangītasamayasāra has to say on the

deśī dances it seems likely that Pārśvadeva lived before Jāya in the later

part of the twelfth or the earlier part of the thirteenth century.109

The two editions of the text differ in the organization of the text, but

both versions have essentially the same contents. Nijenhuis has described

the content of the work following the 1925 text,110 one that has too many

errors to be reliable. I have followed the later edition in describing the

text and its content.

The work is in nine chapters and for the most part it is devoted to vocal

and instrumental music. The seventh chapter and the last part of the eighth

are of interest to the study of dance. The author quotes frequently from

older authorities, such as Kohala (SSam. 1.39) and Matanga,

naming the author (SSam. 2.1). Among later authorities he refers to Bhoja

and Someśvara (SSam.1.3 ).

107 Raghavan,1965, NR. Introduction, pp. 77-8.

108 Acarya Brihaspati, 1977, Introduction, p. 21; Krishnamachariar, 1974, p. 855.

109 Pārśvadeva says that he will discuss all the deśyaṅgas, the features of deśī, which

have not been discussed so far by anyone (SSam. 7.194). As against the eighteen

described by Pārśvadeva, Jāya describes forty-six. (SSam. 7. 194-97; NR. 6.118-

22.)

110Nijenhuis,1977, p.13.

MOVEMENT AND MIMESIS

55

Beginning with the usual benediction, the first chapter discusses the intricacies of vocal music, such as svara, mūrchanā, jāti, rāga etc. of the mārga or classical tradition. We have already noted in the Dasarūpaka that the term mārga is used to refer to the classical tradition. Sangītasamayasāra's use of the term in the same sense shows that the term had by this time passed into common usage. The second chapter is devoted to the deśī or regional tradition. In the third chapter the author expands on the theories of various aspects of music, such as svara, ālāpa etc. The fourth chapter deals with rāgas. The fifth discusses the composition of vocal music and the sixth, instrumental music. The seventh chapter is devoted entirely to dancing, which he refers to as nṛtta, its definition and the body movements required, in dealing with all of which it follows the tradition established by Bharata. Like Bharata, Pārśvadeva divides body parts into two:añga and upāṅga. He counts all the movements of the different parts of the body and the karaṇas and aṅgahāras following Bharata but while describing them he does not discuss the cārī, sthāna, karaṇa or aṅgahāra from the Bharata tradition but follows the deśī tradition. He seems to take Bharata's tradition as nṛtta and the deśī tradition as nṛtya. He describes the deśīnṛtyas which, according to him, consist of peraṇa, pekkhaṇa, guṇḍalī and daṇḍarāsa.111 He then discusses the sthānas and cārīs needed for these deśī dances. He uses the term pāla for cārī, a term found in no other text. However, although the discussion starts with the term pāla, it ends with the term cārī.112 Next, the author describes the utplutikaraṇas, also needed for the deśī dances, other karaṇas with different deśisthānas, bhramarīs,113 and moves on to

111 SSam. 7. 130-32.

112 SSam. 7. 157; 7. 180.

113 Ibid. 7. 193-4. While it is mentioned merely as a form of cārī in the NS. bhramarī became an important movement of dance technique in the medieval period. This is

56

THE LITERATURE OF DANCE

describing the aṅgas or features of deśī dances, calling them deśyaṅgāni. Jāya, as we shall see later, combines these deśī aṅgas and the aṅgas of lāsya, and putting them into one category calls them deśī lāsyas. After describing the deśī aṅgas, Pārśvadeva describes the aṅgas or parts of peraṇa. Finally, he discusses the instrumental music, drumming in particular, needed for four kinds of deśī dances, namely, peraṇa, pekkhaṇa, guṇḍalī and daṇḍarāsa. The requirements of a good dancer, her physical appearance and the way she should be dressed are also described towards the end of the seventh chapter. The eighth chapter discusses the tāla or rhythm and its varieties, both mārga and deśī. The ninth chapter discusses the sitting arrangements, the qualifications of the audience, the poets, the singers, the dancers, the qualities and faults of a singer, drummers and their qualities and faults and those of the dancers needed for each type of deśī dance. The author warns against making dance and music subjects of gambling matches and ends the text by saying that music brings liberation or mokṣa, a concept characteristic of the medieval period.114

One of the most influential works on music and dance from this period is the Saṅgītaratnākara of Śārṅgadeva who lived in the thirteenth century under the patronage of the Yādava king Singhana.115 Followed by most of the writers after Śārṅgadeva's time, it is one of the most important sources or our knowledge of music and dance because there is hardly any other work of the time comparable to the Saṅgītaratnakāra in its methodical approach. The special strength of the work lies in the author's clear exposition of both music and dance. On music it remains a standard authority. On dancing Śārṅgadeva has less information to offer than his sub-classes.

the first text to consider bhramarī as a distinct class of dance movement, with several

114 Katz, 1987, commentary on the verses. 18-21.

115 Nijenhuis,1977, p.12; SR. Vol. IV, tr. Raja and Burnier, 1976, Preface. p. vii.

MOVEMENT AND MIMESIS

57

contemporary Jāya Senāpati, but what he offers is thoroughly systematic,

presenting a clear picture of two traditions, mārga and desī.

The text is in seven chapters, of which the first six are on vocal and

instrumental music. The first chapter deals with nāda (the sound), the

second with rāga (musical mode), the third with prakīrṇa (miscellaneous

topics relating to music), the fourth with prabandha (composition), the

fifth with mārga (classical) and desī (regional) tāla (rhythm), and the

sixth with vādya (musical instruments). The seventh chapter, on dance, is

the one that is of interest to us. It has 1678 verses and is an excellent

source of our knowledge of both mārga and desī traditions of dancing in the

author's time. In describing the mārga tradition he follows Bharata but he

extends our knowledge significantly by giving for the first time a

systematic account of the desī tradition of dancing. Though the

Mānasollāsa and the Sangītasamayasāra had previously introduced the desī style into the traditional account of dancing, it is Sārngadeva who

systematizes that account. In doing so he combines the material from the

Nāṭyaśāstra with that from later works and presents a coherent view not

found in previous works. Although he follows Bharata in describing the

movements of the body, he differs from Bharata in dividing the limbs into

three categories, anga, upanga and pratyaṅga, and not only into anga and

upanga. In this he follows the Mānasollāsa. Although the limbs are

divided into three categories, no new movements are added to those in the

Nāṭyaśāstra. He follows Someśvara in using the term nartana for dance,

dividing nartana into three categories, nāṭya, nṛtya and nṛtta (SR. 7. 3). In

narrating the legend about the origin of dance he again follows the

Nāṭyaśāstra, but differs from it in identifying tāṇḍava as Siva's dance and

lāsya as Pārvatī's. According to Sārṅgadeva, nṛtta and nṛtya can both be

of two kinds, tāṇḍava and lāsya (SR. 7. 28). Tāṇḍava requires uddhata

(forceful) and lāsya requires lalita (delicate) movements (SR. 7. 29-30).

Nṛtta has three varieties, viṣama, vikaṭa and laghu, identified respectively

as rope-dancing, a comic dance, and a dance based on easy karaṇas. Here

58

THE LITERATURE OF DANCE

Śārṅgadeva is repeating the classification found in the Mānasollāsa,

except that he uses the term ṇṛtta, not nartana, as Someśvara does and that

he defines them somewhat differently. Śārṅgadeva's description of cārī,

sthāna, karaṇa and aṅgahāras of the mārga type are as in the Nāṭyaśāstra.

But the deśī cārīs, sthānas and utplutikaraṇas are, of course, new, for

Bharata does not record the deśī tradition. Some of the thirty-six

utplutikaraṇas in the Saṅgītaratnākara are the same as those in the

Mānasollāsa which lists eighteen karaṇas of the deśī variety (Mānas. 16. 4.

1384-99). The Saṅgītasamayasāra describes eleven deśī karaṇas and

mentions five bhramarīs. Bhramarīs are included among utplutikaraṇas by

Śārṅgadeva. In the Nāṭyaśāstra bhramarī was the name of a cārī, a not

particularly complicated revolving movement. Later, more than one

bhramarī came into being, all of them being variations of whirling

movements. Gradually, these developed into so much more elaborate

movements that they came to be recognized as constituting a distinct class,

the earliest work to so consider them being Saṅgītasamayasāra (7.193);

the Abhinayadarpana (289-98) too regards them as a distinct group. Their

inclusion among the utplutikaraṇas by Śārṅgadeva shows that by his time

bhramarīs were so developed and important that they had come to be

regarded as a form of karaṇas.

After the description of the sthānas which include sitting and lying

postures that are appropriate to drama, the author discusses vṛttis or styles

of presentation and conventions. This is followed by a description of

maṇḍala (combination of cārīs) and then of ten lāsyāṅgas of the deśī

variety. Jāya, who was a contemporary of Śārṅgadeva, gives us a list of

forty-six lāsyāṅgas, and Pārśvadeva, who preceded both, gives us twenty

deśī aṅgas requiring similar movements,116 but it is Śārṅgadeva's list of

ten lāsyāṅgas that continued to be reproduced by a number of later authors.

116NR. 6. 117-73; SSam. 7. 194-216.

MOVEMENT AND MIMESIS

59

After describing the lāsyāṅgas Śārṅgadeva explains the importance of

aesthetic beauty, lays down the rules of exercise, and describes the

qualities and faults of a performer (including a description of her make-up

and costume), and those of the teacher and the group of supporting

performers. Then he describes the sequential process of a performance,

including the musical accompaniment, in the pure mode or

śuddhapaddhati(SR. 7. 1260-73).

Next described are gaunḍalī and peraṇi. These two dances seem to be

the commonest pieces of dancing in the deśī tradition, for they are found in

almost all the works from the Saṅgītasamayasāra in the twelfth/thirteenth

century down to the Śivatattvaratnākara in the early eighteenth century.

Peraṇi in particular has had a long life, for literature from the Andhra

Province records it till the end of the nineteenth century and a number of

present day dancers have attempted to revive it.117 Its popularity in that

region is also suggested by the widespread use of one of the fast

movements essential to it, gharghara, which requires special movements

of ankle bells. In the sixteenth century work, the Nartananiṛnaya,

Puṇḍarīka Viṭṭhala remarks that gharghara was a distinctive feature of the

deśī dances of the Drāviḍa region (NN. 50a-b). The same movement is an

essential feature of today's Kathak dancing118 in which it is known as

tatkār. Gaunḍalī dance certainly survived till the eighteenth century119

but no current practice of it is known.

After describing these two dance pieces, Śārṅgadeva deals with the

qualifications of the ācārya (the teacher), the naṭa (the actor), the nartaka

(the dancer), the vaitālika (a general entertainer), the cāraṇa (an expert in

understanding gharghara) and the kohlāṭika (a performer who specializes

in bhramarī, rope-walking and dancing with a dagger). Next, he describes

117 NR.,1965, Introduction, p.137-41.

118 Classical and Folk Dances of India,1953, Part III, pp. 25-27.

119 STR. 6.3. 15-16.

60

THE LITERATURE OF DANCE

the audience and the sitting arrangements. Finally, he discusses rasas (nine in number), sthāyībhāvas (thirty-three in number) and the definition of sattva (the essence) and sāttvikabhāvas (eight in number). The chapter concludes with final prayers.

The Sangītaratnākara has a highly important supplement in the form of a commentary on it called Kalānidhi and written in the fourteenth century by Kallinātha. The commentator quotes from Kohala to expand Sārṅgadeva's description of arm movements and adds an entire section on vartanā, of which thirty-one (SR. pp.105-10) are given, and another section on cālaka, which are fifty in number (SR. pp.111-24). Kallinātha quotes Kohala again with respect to cārī, adding a new cārī called madhupacārī, of which twenty-five are described (SR. pp.313-17). Except for the Kalānidhi, the only texts that include these cārīs are the Nṛtyādhyāya of Aśokamalla and the Nṛtyaratnakośa of Mahārāṇā Kumbhā, which, however, use the term muḍupacārī for the same movements. 120

Close in time to the Sangītaratnākara was the Nṛttaratnāvalī of Jāya Senāpati which was written in the thirteenth century A.D. 121 The authors were contemporaries and as Raghavan points out, Jāya must have known of Sārṅgadeva's work. But Jāya makes no mention of Sārṅgadeva's work, 122 very likely because of the rivalry between the two neighbouring states where they lived. Whereas the Sangītaratnākara is a text that deals with saṅgīta, that is, all three musical arts, namely, vocal and instrumental music, and dance, the Nṛttaratnāvalī concentrates on dancing. The

120 Nr.Adh. 1080-1115, NRK. 2. 3. 1-37. NRK. quotes Kohala after Kalānidhi, acknowledging the source at the beginning of the chapter. The lines match those in the edition of the SR. 1953, pp. 313–317; NRK . calls them muḍupa.

121 NR. 1965, Introduction, pp. 8–12; Krishnamachariar,1974, p. 855.

122 NR. 1965, Introduction, pp. 73-4. Sārṅgadeva's patron was the Yādava king while Jāya's patron was the Kākatīya king; these were the rulers of two neighbouring states in South India.

MOVEMENT AND MIMESIS

61

Sañgītaratnākara's single chapter on dancing is a comprehensive study of

dance but a concise one. By contrast, the Nṛttaratnāvalī devotes all eight

chapters to dance and discusses vocal or instrumental music only in the

context of dance.

The Nṛttaratnāvalī has eight chapters devoted entirely to dancing and

covers two traditions. The first four chapters of the text discuss the mārga

tradition following the Nāṭyaśāstra and the other four discuss the deśī. The

tradition of Bharata is regarded as mārga while the tradition comprising

regional and popular styles is called deśī, the latter being a subject on

which Jāya is particularly informative. This is the only work that deals

exclusively with dance in such detail. The text has been critically edited

by V. Raghavan who gives a great deal of related information on the

subject in his introduction. Of particular help is Raghavan's comparison

of Jāya's work with the work of some of his contemporaries in the field.

The first chapter begins with the customary benediction, praising

dance as an art-form and defining nāṭya. The four modes of abhinaya, i.e.,

āṅgika, vācika, āhārya and sāttvika, the six forms of dancing-nṛtta, nṛtya,

mārga, deśī, tāṇḍava and lāsya-are then discussed in detail. Chapter two

deals with abhinaya. Since āṅgika abhinaya or body movements are the

most relevant to dance, Jāya describes in detail the movements of the

major and minor limbs, six aṅgas, six pratyaṅgas and six upāṅgas. He also

describes the hand-gestures in this chapter, in course of which he quotes

other authors, such as Abhinavagupta, Kīrtidhara and others.123 The

third chapter is on cārīs (movements of one leg), sthānas (postures), nyāya

(rules of performance), vyāyāma (exercise), sauṣṭhava (grace), more

sthānas and maṇḍalas (combinations of cārīs). The fourth chapter

describes karaṇas (dance-units) and aṅgahāras (sequences of dance-units)

and ends with recallas (extending movements of the neck, the hands, the

123 NR. 2. 182; 2. 183; 2. 214; 2. 265; 2. 275. Jāya refers frequently to a " muni," which

seems to indicate Bharata, NR. 2. 4, 202, 213, 215, 273.

62

THE LITERATURE OF DANCE

waist and the feet) mainly in the lines of the Nātyaśāstra. In general,

Jāya's treatment of the mārga tradition is faithful to Bharata and is precise

and systematic.

The second half of the text is devoted to the deśī tradition, in dealing

with which the author is just as organized and informative as he is with

mārga. The fifth chapter, which is the first on the deśī tradition, starts by

defining the term deśī. Deśī sthānas, utplutikaraṇas (deśī karaṇas) and

bhramarīs (pirouettes) are described. We have already noted how

bhramarīs became popular and came to be regarded as a separate category

comprising several varieties. We may note that one such variety,

crakabhramarī, a fast pirouette, has survived as a distinctive feature of

the Kathak style. A variations of the movement appears also in the

Manipurī style which employs slow revolving movements, but bhramarīs

are no longer important in other classical styles of dancing in India

today.124 In this chapter, the fifth, the author refers to his other work, the

Gītaratnāvalī, which has not come down to us.

The sixth chapter deals with movements of the feet. These deśīpādas,

as they are called in this work, are often described as deśī cārīs in the later

texts. Cārīs are the movements of one leg which necessarily include foot

movements as well but cārīs and pādas are always treated as separate

movements in the Bharata tradition. Jāya has a separate section on deśī

cārīs, describing cārīs not found in works preceding his. His descriptions

of deśī pāṭamaṇis (foot movements) are his own as well.125 From the

descriptions of the deśīpādas, pāṭamaṇis and cārīs one can say that pādas

involve simple floor contacts of the feet; pāṭamaṇis involve striking the

ground with the feet and cārīs involve movements of one leg extended. He

then describes forty-six varieties of deśī lāsyaṅgas which include the deśī

124 The movement in Kathak comparable to the cakrabhramarī is called cakkar. See p.

125 Raghavan,1965, p.174. fn.

MOVEMENT AND MIMESIS

63

lāsyāṅgas described in the Saṅgītaratnākara and the deśī aṅgas found in

the Saṅgītasamayasāra. Gatis or gaits are described next.

The seventh chapter discusses auspicious dates for beginning dance

lessons (the term he uses here is nṛtya), the characteristics of the stage and

some general discussion on presentation, the time and location of dance

performances, the worship of Gaṇeśa, methods of practice, the

qualifications desirable in a dancer, dance costume, hand-gestures for

practice and the accompanying vocal and instrumental music. The chapter

then concentrates on describing individual dance pieces, calling them

deśīnrtta. These include peraṇi, pekkhaṇa, sūḍa, rāsaka, carcarī,

nāṭyarāsaka, śivapriya, ciṇṭu, kanduka, bhāṇḍikā, ghaṭisani, cāraṇa,

bahurūpa, kollāṭa and gauṇḍalī. We come across only a few of these in the

Saṅgītaratnākara and the Saṅgītasamayasāra. Most of them occur later in

the Nṛttaratnakośa.126 The chapter concludes with the description of the

stage, its types and shapes, the audience and the presiding dignitary. The

eighth and final chapter provides more information regarding

presentation in general, the recital, the appropriate time for its

presentation, the arrival of the chief guest and the welcome accorded the

king, other members of the audience, the qualities required in a dancer,

her costume, the orchestra, the sitting arrangements, the entrance of a

dancer, the use of three curtains on the stage and their removal.127 The

chapter ends with advice on honouring the dancer, the musicians and the

poet.

Although it may seem that the author has not made any significant

contribution while dealing with the mārga tradition since he follows

Bharata, this is not true. The description of each karaṇa of the mārga

126 See p.81 infra.

127 This feature of using and removing the curtains held by performers other than the

dancer is still very much in use in various dance styles of South India. Kathakali, the

classical dance style of Kerala, is the commonest example.

64

THE LITERATURE OF DANCE

tradition is unique. This happens to be the first of the very few texts which

describe karanas along with cārīs, hand-gestures and foot movements in

such a way that by following the directions a karana can be reproduced.

Not only does Jāya elaborate on the actual performance of the karanas, he

quotes the views of earlier writers, which is helpful in tracing the evolution

of dance. In a later chapter we shall discuss his treatment of the karanas

and deśī dances, in which context his contribution is particularly important.

The Sāhityadarpana of Viśvanātha, written in the fourteenth century,

is the last treatise from the second period, one principally on poetics, to be

considered here. A work that deals also with dramaturgy,128 the

Sāhityadarpana is one of the most illuminating works on poetics.

According to Gerow, "With Viśvanātha, the period of poetics beginning

with Mammata, marked by a concern to regularize and codify

alaṅkāravidyā as a śāstra, may be said to have reached its conclusion."

Gerow views this text as the "second best known Indian poetic text."129

As the name suggests, the work is meant to be a mirror of literature, that

is, a work on literary theory. The text is divided into ten chapters: i. the

nature of kāvya (a poetic composition), ii. vākya (a sentence), iii. rasa and

bhāva (aesthetic affect and emotional expression), iv. dhvani (sound), v.

vyañjanā (rhetorical expression), vi. dṛśyakāvya (a poetic composition to

be seen), vii. doṣa (the defects of a kāvya), viii. guṇa (the qualities of a

kāvya), ix. rīiti (style) and x. alaṅkāra (poetics). Of these ten, the sixth

chapter, which is concerned with dṛśyakāvya, is of interest to us.

Viśvanātha divides kāvya into two groups, dṛśya and śravya, that is, poetic

compositions to be seen and poetic compositions to be heard. The

dṛśyakāvyas are divided into two categories again, rūpaka and uparūpaka.

Rūpakas are said to be of ten types, which is a division inherited from

128 Dasgupta, 1947, Vol. I, p. 566.

129 Gerow, 1977, pp. 281, 283.

MOVEMENT AND MIMESIS

65

Bharata. However, Viśvanātha considers lāsyāṅgas as parts of bhāṇa like other writers who associate lāsyāṅgas with bhāṇa.

The major contribution of Viśvanātha is that he gives a class name to minor types of rūpakas, which he calls uparūpakas, a term that became standard in the works of the authors after him and continues to be used by Sanskrit scholars. These uparūpakas are eighteen in number, they employ dancing, and they are described in detail.130 Since they comprise a major genre relevant to the study of dancing, we shall reserve them for detailed study in a later chapter.

The Saṅgītopanisatsāroddhāra is a treatise on music and dance written in the fourteenth century (1350 A.D.) by the Jaina writer Sudhākalasa. The writer says that it is an abridged version of his previous and larger work, Saṅgītopaniṣat, written in 1324 A.D.131 No copy of it is known to exist. The abridged version does not record anything new, or anything that had not already been said by Śārṅgadeva or Jāya who preceded Sudhākalaśa by about a hundred years. Nor does the author give any information on the development of the art that must have taken place during these hundred years. The Saṅgītopanisatsāroddhāra is not concerned with two traditions, mārga and deśī. Surprisingly, the author does not even seem to recognize the development of the separate traditions of mārga and deśī dance. He mostly repeats the information gathered till his time by his predecessors, with a few minor variations. Nevertheless, the text is useful as an organized and succinct report.

The work is in six chapters, the first four of which are on gīta or vocal music and its technicalities, on tāla or rhythm, and on the origin of vādya or musical instruments. The fifth and the sixth are devoted to dancing and these are the two chapters which are of interest to us. In the fifth chapter, which contains 141 verses, the author begins with the praise of Jinapati.

130 SD. 6.3-312.

131 Saṅgītopanisatsāroddhāra, 1961, Introduction, p. v; see also the text, SUS. 6. 152.

66

THE LITERATURE OF DANCE

The term he uses for dance is nṛtya (SUS. 5. 2). After a few verses, giving

the legend of the origin of dance, the author defines three varieties of an

art-form consisting of nṛtta, nṛtya and nāṭya. The author derives the term

nṛtya from the root nṛt and defines the art as gātrasya vikṣepaḥ, i.e.,

movements of the body.132 According to him, nṛtta is danced by men,

nṛtya by women, while nāṭya is nāṭaka, performed by both men and

women. He considers both lāsya and tāṇḍava to be forms of nṛtya and says

that lāsya originated from Gaurī (=Pārvatī) and tāṇḍava from Rudra

(=Śiva). Lāsya is sukumāra or delicate while tāṇḍava, he says, has the

qualities of viṣama.133 Evidently, viṣama had a different connotation for

Sudhākalasa, for he views viṣama as "difficult or vigorous" movements as

opposed to Śārṅgadeva's view of viṣama as an "acrobatics" style that

included such performances as rope-walking. Jāya, however, connects

viṣama with tāṇḍava while referring to viṣamatāṇḍava as a feature of a

deśī dance called peraṇi, although he does not explain its nature.134

The rest of the fifth chapter is devoted to describing the movements of

the limbs, which are divided into aṅga, upāṅga and pratyaṅga, as is

customary in the texts of the medieval period. Sudhākalasa does not say

anything about their application and ends the chapter with a description of

six sthānas. However, his treatment of two classes of movements,

movements of the feet and the postures, differs from that of other texts.

Pādas, or the positions of the feet, and pādakarmas, or the actions of the

feet, are described separately in this text in contrast both with the

Nāṭyaśāstra and the Saṅgītaratnākara, which describe positions as well as

movements of the feet under one category called pāda.135 Further,

Sudhākalasa treats sthānas and sthānakas as separate categories of

132SUS.5. 6.

133SUS.5.10.

134SR.7. 31; NR. 7. 63.

135SUS.5.126-34.

MOVEMENT AND MIMESIS

67

postures which no other writer does. According to him, sthānas are

postures meant for women while sthānakas are postures meant for men.

He adds no new postures, although his descriptions of the movements vary

in details from other accounts.136

The sixth chapter mainly defines karaṇas and aṅgahāras. Karaṇas,

according to Sudhākalāśa, are components of lāsyāṅgas and ṇṛtya. He lists

one hundred and eight karaṇas but actually describes one hundred and five.

In some details the karaṇas occasionally differ from those found in the

Nāṭyaśāstra and the works that followed it, including the

Saṅgītaratnākara. Some have different names and movements.137 The

number of the aṅgahāras is the same as in the Nāṭyaśāstra but some names

are different,138 although what these new aṅgahāras were cannot be

ascertained, for they are described neither in this work nor in any other.

That Sudhākalāśa was recording contemporary traditions is apparent from

his treatment of the bhramarīs, which he puts in a group of six. Most of

them have the same names as in the Saṅgītaratnākara, with a few

variations.139 We have already noticed in the earlier treatises that

bhramarī had developed from a simple cārī movement to a separate

category which included a variety of whirling movements. The

Saṅgītasamayasāra was the first text to collect them into one group. It

places them between the descriptions of the utplutikaraṇas (karaṇas

involving leaps and recognized as deśīkaraṇas) and the deśī aṅgas

136 SUS. 5.135-40.

137 SUS. 6. 2; 6. 6-111; Karaṇas such as ardhasūci and alāṭa, for instance, use different

cārīs in the tradition described in the Nāṭyaśāstra and the Saṅgītaratnākara. See

Bose, 1970, pp. 108-35; karaṇas such as uruśliṣṭa and kaṭirecita etc. are new. These

may have come from a different tradition.

138 SUS. 6.115-19. Haripriya, ālipallava, bhujanga-trasta, jayadarpā, vilasanmada and

mahodvṛtta are new names but since they are not described we do not know what

they were like. They are described in no other text.

139 SUS. 6.120-22.

68

THE LITERATURE OF DANCE

(features of deśī).140 The Sangītaratnākara describes seven bhramarīs

which it places with the utplutikaraṇas, while the Nrttaratnāvalī treats

them separately as a group of thirteen variations.141 The

Sangītopaniṣatsāroddhāra also treats them separately from the utpluti or

deśikaraṇas. After the bhramarīs cārīs are named but not described.

Concluding the descriptions of the dance movements and their

sequences at this point, the author makes an interesting remark. He states

that theory and practice are as important to a dancer as they are to a

theorist. He then passes on to nartakabhāṣā and nrtyabhāvas, that is,

special languages and expressions, and provides examples from the

western part of India where he lived.142 The paddhati or presentation

technique is described next. At the end, the text follows the contemporary

practice of other works by providing the usual description of the presiding

dignitary, the audience, the vāggeyakāra143 (a musician who also

composes music) and the gandharvarañjaka (knowledgeable in a variety

of music and thus an entertaining performer). The text comes to an end

with a final prayer and the verse stating that it is an abridged version of the

author’s larger work, the Sangītopanisat.

The Sangītacandra is a work containing 2168 verses by Śuklapaṇḍita,

also known as Vipradāsa. He has been placed in the fourteenth century by

the editor of the currently available edition who follows Ramakrishna

Kavi. Kavi believes that Kumbhakarṇa based his work on the

140 There are five such bhramarīs. SSam. 7. 193.

141 SR. 7.755; 7.775-82; NR. 5. 83-105.

142 SUS. 6. 129-32. See Bose, 1970, p. 4.

143 SUS. 6. 143. The editor has chosen vāñmayakāraka to refer to a composer musician.

The word for this type of musician found in other texts is vāggeyakāraka. (SR. 7. 347;

NR.8. 29). One of the MSS. of the SUS. reads gāggeyakāraka, very likely a scribal error.

MOVEMENT AND MIMESIS

69

Sangītacandra.144 But it is hard to say for certain who quoted from whom

since the traditions recorded by the two authors seem to have come more

or less from the same period. No evidence, internal or external, is

available regarding the author's date or land of origin. The single

manuscript in which the text exists does name the author but not at the

beginning or the end, which are the sections where the author's particulars

usually appear. The lack of such information may be due to the evidently

unsatisfactory state of the manuscript which seems to have a number of

scribal errors unemended in the current edition.145

The text begins with a benediction and recounts the origin of the

nāṭyaveda. It then lists the proposed subjects of discussion, and gives the

rules for building the dance-hall, performing the pūjā, and describes the

qualifications of the presiding person and the audience. The author then

explains the conduct of the pūrvarañga or the preliminaries following the

Nātyaśāstra and mentions the four abhinayas and the four vṛttis. This

introductory section ends at verse 388. From the next verse, verse 389, the

author begins to describe nṛtya which he takes to represent the entire art of

dance. He first divides nṛtya into two, mārganṛtya which expresses rasa,

and nāṭyanṛtya, which expresses bhāva (389-94). He next describes

natural and stylized movements and then adds a third variety of nṛtya,

calling it deśīya nṛtya. All three varieties of nṛtya can again be divided

into tāṇḍava and lāsya (395-410). Viṣama, vikaṭa and laghu are then

referred to as varieties of nṛtta, a term not used so far (411-13). The

author then moves on to describe movements of the major and minor limbs,

which cover verses 414-1046. In describing the movements of the body he

follows the Nātyaśāstra, but in his list of hand-gestures he adds a few new

gestures that fall outside the Nātyaśāstra. These he takes from other

sources, which he does not, however, consider reliable. He recommends

144 Sangītacandra, Introduction, p. viii.

145 For instance, kolkāṭika is not emended to the correct form, SC. 2. 107.

70

THE LITERATURE OF DANCE

the appropriate dress and make-up for the dancer but gives no details (853-4). To this list of body movements he adds a new one, a movement of the

soles of the feet (1024-25).

After describing the movements of the limbs, he describes the sthānas or postures(1047-1149), which he calls essential to dance since a dance

begins with a posture and proceeds to a cārī. The sthānas include postures meant for men and women, postures for lying and sitting, all of which may

be of both mārga and deśī varieties. Cārīs of the mārga and deśī varieties are described next in verses 1150-1272. Karanas of the mārga and deśī

varieties are described in verses 1273-1517, followed by mandalas in verses 1518-80. Movements in the mārga tradition are described after

Bharata.

Four types of nyāya or rules for depicting battle-scenes are discussed next in 1581- 98, a topic that seems to be the author's own contribution.

Angahāra comes next (1596-1709), followed by descriptions of the four recakas (1710-16). Then the author discusses how the performer may

represent the nine rasas (1717-1936) and the forty-one bhāvas (1937-2027). The description of the dance teacher and the preceptor comes next

(2028-32), followed by the rules of practice(2033-40). In verses 2041-57 the author identifies the qualities required in a performer as well as her

possible strengths and shortcomings. This is followed by a description of the dancer's ornaments (2058-69).

In the next section the author discusses the deśī forms, beginning with ten lāsyāngas (2070-96). While he mainly follows the Sangītaratnākara

here, he adds three new types of rekhā (2097-99). This is followed by a description of the peraṇi dance (2100-2102), after which the author again

discusses (2103-2114) the characteristics of various performers as well as others connected with performances, such as, a naṭa (actor), a nartaka

(dancer), a vaitālika (critic and entertainer), a cāraṇa (musician), a kolkāṭika (correctly, a kohlāṭika, an acrobatic dancer) and the troupe. He

then discusses the techniques, respectively, of a śuddha or pure

MOVEMENT AND MIMESIS

71

performance (2116-2128), and the peraṇī and gaunḍalī, which are dances

in the deśī style (2129-2165). The text ends by saying in verse 2166 that

dance is an erotic art, an observation made by no other writer.

The work differs from its contemporaries in regarding dance merely

as an art auxiliary to drama and in ignoring music. This is surprising

because by the time this work was written, dance had already developed

into a separate art form that made use of vocal and instrumental music and

was no longer merely one of the components of a dramatic performance.

Vipradāsa's treatment of dance as a part of the pūrvaraṅga of a drama is

closer to the position taken in the Nātyaśāstra. However, the division of

the limbs into three categories and the description of deśī dancing are signs

of the medieval approach to dancing in which Vipradāsa is a follower of

Sārṅgadeva.

A major work of the medieval period, which originated in eastern

India and influenced a number of works in that region, is the

Sangītadāmodara by Subhaṅkara. The editors of the text believe that the

author lived sometime in the fifteenth century in Bengal. Nijenhuis

suggests that the author may have lived sometime in the early part of the

sixteenth century.146 That he influenced an extensive body of musical

literature through several centuries, particularly in Bengal and Orissa, has

been established.147 He was quoted by Bhavānanda Thākur, author of

Sadarpakandarpa, by Ghanaśyāmadāsa, author of Bhaktiratnākara, by

Rāmgopāl Dās, author of Rasakalpavallī and by Puruṣottamamiśra, author

of Sangītanārāyaṇa, all from the eastern region.148 Such was its

popularity that not only was it studied by these Vaiṣṇava authors from

Bengal but, as Nijenhuis tells us, it was even translated into Persian. She

also points out the importance of the anthropomorphic images of the rāgas

146 Sangītadāmodara, 1960, Introduction, p. 15; Nijenhuis, 1977, p.18.

147 Nijenhuis, 1977, p.19.

148 SDām.1960, Introduction, p. 15. Katz, 1987, Introduction, p. i.

72

THE LITERATURE OF DANCE

that appear in the dhyānaślokas on rāgas in this text. It is on these images

that the rāgamālā miniature paintings are based.149

Although the Sangītadāmodara is principally a work on music and

dance, it includes substantial discussions on drama as well. It has five

chapters, the first of which is on bhāva or emotional expression. The

second describes the nāyaka or the hero, the nāyikā or the heroine, the

sakhi or the friend of the heroine, and then explains the meaning of nāda or

sound and gīta or vocal music. The third chapter is an exposition of the

intricacies of the grammar of vocal music, which includes svara, rāga,

and tāla among other subjects. The fourth chapter describes vādya, nṛtya

and nāṭya. The fifth chapter is devoted partly to more technicalities of

music and partly to rasa or aesthetic affect.

For our purpose the fourth chapter is the important one, where the

author describes dancing under two headings, aṅgahāra and nṛtya.

Aṅgahāra, in this context, is defined as aṅgavikṣepa, movements of the

body. Under this heading the author includes āṅgikābhinaya because it

means acting by using the movements of the limbs. The limbs are divided

into major and minor limbs, aṅga, upaṅga and pratyaṅga. The author then

names the sthānas or postures, the cārīs or movements executed with one

leg, the karaṇas or the coordinated movements of the hands and the feet

(dance units), maṇḍalas or the combination of cārīs, and aṅgahāras or

sequences of dance units. These are described after Bharata150 but

Subhaṅkara adds one more class of movements to his list. This category,

viṣama alaga, consists of difficult movements and does not belong to the

styles described by Bharata and his followers. Subhaṅkara mentions

thirteen such movements.151 He states that nṛtya is deśī or regional and it

has two subdivisions called tāṇḍava and lāsya. Tāṇḍava is puṃnṛtya, that

149 Nijenhuis, 1977, p. 19.

150 SDām. p. 61-68; cf. NS. chapters 8-12.

151 SDām. p. 69-70.

MOVEMENT AND MIMESIS

73

is, danced by men, and lāsya is strīnrtya, that is, danced by women.

Following the usual practice of most medieval texts, the Sangītadāmodara

subdivides tāṇḍava into peraṇi and bahurūpa and describes them briefly.

But the corresponding subdivision of lāsya into churita and yauvata, again

discussed briefly, is a new feature introduced by the Sangītadāmodara.152

The author then moves on to describe nāṭya, which includes the major

and minor types of drama. While describing the preliminary parts of

nāṭya, the author again speaks of nrtya as it is performed for the ritual of

flower offerings, specifying for it the performance of māṇḍalī, which

involves lāsyāṅgas as well.153 The term māṇḍalī is very likely a scribal

error for guṇḍalī or gauṇḍalī, for the particular dance described in the

Sangītadāmodara as māṇḍalī is similar to the guṇḍalī or gauṇḍalī

described in most medieval works on dancing.154 The author mentions

twenty-seven rūpakas and uparūpakas, major and minor types of dramas.

He classifies them under the heading nrtye nakṣatramālā, the gemlike

starry garland of nrtya, and describes them in detail. The lāsyāṅgas are

described in detail, as is bhāṇa, one of the major types of drama.

Subhaṅkara's treatment of bhāṇa agrees with that of the authors who came

after Bharata rather than with that of Bharata himself and includes

lāsyāṅgas as one of the features of bhāṇa.155 However, in describing the

ten rūpakas he follows Bharata. Seven of the seventeen minor types of

dramas, that is, uparūpakas, described in this text employ dancing. All

seventeen are described in detail. The concluding verse of the fourth

152Ibid. p. 69.

153Ibid. p. 73.

154See Bose, 1970, p. 185, fn. 57. The dancer was called a guṇḍalī and her dance was

gauṇḍalī, but often the dance is also called guṇḍalī.

155SDām. pp. 79; 85- 7.

74

THE LITERATURE OF DANCE

chapter before the closing prayer again refers to all these dramatic types

as ṇṛtye ratnanakṣatramālā.156

Besides Saṅgītadāmodara Śubhaṅkara wrote two other treatises, the

Saṅgānasāra, an unedited manuscript mentioned in the catalogue of the

Asiatic Society of Bengal and the Hastamuktāvalī,157 which adds to the

discussion in the Saṅgītadāmodara. In the section of the Saṅgītadāmodara

in which the author lists hand-gestures he directs the reader to the

Hastamuktāvalī for a fuller discussion (SDām. p.63).

As its title indicates, the Hastamuktāvalī deals only with hand-

gestures. Nijenhuis is not altogether accurate in thinking that this text

contains dancing,158 for it describes hand-gestures that are applicable

both to dance and drama. Beginning with a few verses of benediction the

author states the subject of his discussion, which is hasta, that is, gestures

of the hands. The number and the names of the gestures mentioned in his

Saṅgītadāmodara match the material here, while the actual descriptions

and the application of the gestures in general match the discussion in the

Nāṭyaśāstra. According to the author, there are thirty single and fourteen

double hand gestures that signify words as well as emotions. The author

then describes twenty-seven ṇṛttahastas, hand gestures meant for dancing,

which express no meaning. This text was obviously quite popular since a

certain Rāghava Rāya acknowledges that his Hastaratnāvalī is based on

the Hastamuktāvalī. The Bodleian Library has a copy of the

Hastaratnāvalī which seems to be a virtual copy of the Hastamuktāvalī

insofar as the portions on the single and double hand gestures are

concerned.159 However, Rāghava Rāya's contribution, as Maheśvar

156 SDām. pp. 79, 91-6, 99.

157 Neog, 1952-59.

158 Nijenhuis, 1977, p. 19.

159 See Bose, 1964, pp. 57-87 for the discussion of Hastaratnāvalī, 1970, p. 6.; 47-67.

MOVEMENT AND MIMESIS

75

Neog notes, lies in his explanation of the applications of the gestures.160

The applications of the hand-gestures described in the Hastamuktāvalī are

very elaborate, which shows that the hand-gestures had become so

important that it was necessary to write a separate work on these alone.

Another important work from this period is what is known as the

Nṛtyādhyāya of Aśokamalla, which is part of a larger work. The edition

currently available was prepared from a single manuscript which lacks

the beginning and the colophon of the text.161 Because of the missing

colophon the actual title of the work is not known, and the title by which it

is known, Nṛtyādhyāya, was given by the editor. Fortunately, the author's

name is known from his own references to himself in the text.162 The

editor has placed the author in the fourteenth century A.D.163 and

comparisons with other works, known to be from the fourteenth/fifteenth

centuries supports this view. For instance, most of the deśī lāsyāṅgas,

thirty-seven in all, described in the Nṛtyādhyāya are also found in the

section on deśī aṅgas in the Saṅgītasamayasāra, and among the deśīlāsyas

of the Nṛttaratnāvalī. The Saṅgītaratnākara describes fewer lāsyāṅgas

but they too are found in the list of deśī lāsyāṅgas in the Nṛtyādhyāya.

Furthermore, a number of lines in the Nṛtyādhyāya. closely match lines in

the Saṅgītaratnākara.164 The Nṛtyādhyāya has a reference to the

commentary of Siṁhabhūpāla on the Saṅgītaratnākara (Nr.Adh. verse

695. which, as the editor points out, places the author after Siṁhabhūpāla,

who is believed to have lived in the fourteenth century. The

Nṛtyaratnakośa and the Nṛtyādhyāya both quote the same prākṛt verse

160Neog, 1952-59.

161Nṛtyādhyāya,1963, Introduction, pp. 9-10.

162Ibid. verses 695, 707, 1272.

163Nṛ.AdH., 1963, Introduction, p. 10.

164Ibid. verses1513-65; SSam. 7. 194-216; SR. 7. 1206-15; NR. 6. 117-73; Nr. Adh. verse 1301; SR. 7. 745.

76

THE LITERATURE OF DANCE

while describing the mārga lāsyāṅgas, although neither author mentions

the other. They may both have quoted from the same source. We cannot

determine whether Maharāṇā Kumbhā, who wrote the Nṛtyaratnakośa

and lived in the fifteenth century, was before or after Aśokamalla.

However, the internal evidence strongly suggests that Aśokamalla came

after Pārśvadeva, Sāṅgadeva, Jāya and Siṁhabhūpāla.

The Nṛtyādhyāya consists of 1611 verses, with the beginning and the

end of the text missing. Although the beginning is missing, the contents

reflect the deśī tradition of dance recorded in the Saṅgītaratnākara and the

Nṛttaratnāvalī. The text starts with the hand gestures followed by the

movements of the major and minor limbs, that is, aṅga, upāṅga and

pratyaṅga. These take up 614 verses and resemble the descriptions in the

Saṅgītaratnākara and the Nṛttaratnāvalī. Verses 615 to 707 describe

viciṛābhinaya (various ways of acting), dividing the contents into

bhāvābhinaya (emotional expression) and indriyābhinaya (expressing

through the sense organs). These seem to be summarized versions of the

chapters in the Nāṭyaśāstra (NŚ. chs. 22 and 25) on sāṃānyābhinaya and

ciṛābhinaya. Verses 708 to 753 describe thirty-six vartanās or arm

movements, of which twenty-five are claimed to be derived from

Hanuman and the other eleven from other authors. The first twenty-five

are similar to the arm movements described in Kallinātha’s commentary

on the Saṅgītaratnākara (SR. Vol. IV pp. 105-10) where Kallinātha cites

Kohala as his source. These are also found in the Nṛtyaratnakośa. Cālanas

or cālakas, another type of arm movement, are fifty-five in number and

described from verses 754 to 862. The first fifty are similar to what we

find in Kallinātha (SR.Vol. IV. pp. 111-23). Sthānas or postures are

described next, from verses 863 to 948, which include postures meant for

men (six) and women (eight), reclining postures (six) and deśī sthānas

(twentythree). The deśī and mārga varieties of cārīs, one hundred and

eleven in all, including twenty-five muḍupacārīs, are described in verses

949 to 1115. The same descriptions of the muḍupacārīs are found in

MOVEMENT AND MIMESIS

77

Kallinātha’s commentary on the Sangītaratnākara where the term given is

madhupacāri (SR.Vol. IV, pp. 313-17). In verses 1116 to 1305 the author

describes one hundred and eight karanas after Bharata. Verses 1306 to

1342, describe twenty-seven deśī karaṇas. The text has a lacuna at this

point but the editor has supplied us with descriptions of the other eleven

deśī karaṇas (Nr. Adh. pp. 138-39), taking them from the Bharatakośa. The

descriptions of the first few aṅgahāras are also missing, which the editor

has supplied from the Sangītaratnākara (SR. 7. 789-808; Nr. Adh. pp. 140-

41.). Verses 1343 to 1415 describe the rest of the aṅgahāras, thirty-two in

all. In the next five verses, 1416-1420, the author talks about the musical

accompaniment for the aṅgahāras. Four recakas are described next in

verses 1421 to 1426. Twenty maṇḍalas are described next in 1427 to 1486.

Verses 1487 to 1512 describe mārga lāsyāṅgas while verses 1513 to 1565

describe 37 deśīlāsyāṅgas. The text ends with descriptions of the closing

movements known as kalāsas which are covered in verses 1566 to 1611.

The last variety of kalāsa is missing from the text but the editor has

supplied the missing part the description of the last variety of kalāsa from

the Nrtyaratnakośa (Nr. Adh. pp. 172-73; NRK. 4.1.83-5).165

This is one of the few texts to refer to śṛṅgābhinaya (Nr.Adh. verse.

141). The Bhartāṛṇava has one whole chapter on śṛṅgābhinaya (chapter

12). The Nrttaratnāvalī also refers to śṛṅga in passing while describing the

application of the kartarīmukha hand (NR. 2.112) but offers no detail.

Kalāsa is another movement which is described in the Nrtyādhyāya in

detail, the only other text that describes them in similar detail being the

Nrtyaratnakośa. Kalāsas are not defined but this term is in use today in

Kathākali as a concluding part; the same function is indicated by

Kallinātha’s explanation of the term.166

165An exhaustive discussion of the last three movements have already been noted by the

present writer. See Bose, 1970, pp. 167-83.

166SR. Vol. IV, p. 381, commentary on verse 1292: कलासयेत् नर्तनं समापयेदित्यर्थः।

78

THE LITERATURE OF DANCE

The Nrtyaratnakośa of Mahārāṇā Kumbhā can be taken as the last

work of the second period. It is part of a larger work, the Saṅgītarāja,

which closely follows the Saṅgītaratnākara. According to Nijenhuis, who

gives an account of the work in her monograph, "This large work

comprising 16,000 verses is divided into five chapters on: 1. pāṭhya

(recitation), 2. gīta (vocal music), 3. vādya, 4. nṛtya and 5. rasa."167

Mahārāṇā Kumbhā, who was a scholar king of the fifteenth century, is

generally acknowledged as the author and the case for this attribution has

been made at length by the editor of the text.168 The same attribution

appear in both Raghavan and Nijenhuis but Kunhan Raja believes the

author to be Kālaseṇa.169

The fourth chapter, which is on nṛtya, is divided into four ullāsas or

parts, each consisting of four parikṣaṇas or sections. Each section starts

with a benedictory verse. The first section of the first part, in 811 verses,

discusses a variety of subjects and concludes with a description of the

movements of aṅgas or major limbs. Beginning with the benediction, the

first 123 verses describe the origin of the śāstra of nāṭya, the rules of

building the performance-hall, the qualifications of the person presiding

and the audience. Also discussed are the preliminaries to the

performance, which are described in detail following the Nāṭyaśāstra.

The author then defines nāṭya, nṛtya and nṛtta (verses 124-170). Nāṭya,

according to him, is to be understood as abhinaya, नाट्यशब्देन च अभिनयम्

(NRK.1.1.128), nṛtya as rasa, नृत्यशब्देन च रसं पनः (NRK.1.1.128) and

nṛtta consists of the combination of karaṇas and aṅgahāras,

करणाङ्गहारनिचयैर्‌नृत्तम् (NRK.1.1.129). Verses 171 to 212 describe

piṇḍibandhas or group dances, which are performed by sixteen female

167Nijenhuis, 1977, p.16-18.

168Nrtyaratnakośa, Vol. II, 1968, pp. 2-5.

169Raghavan, 1960, pp. 11-12; Nijenhuis, 1977, p. 16; NRK. Vol. II, 1968, Introduction,

pp. 2-5.

dancers as part of the preliminaries. More details about the preliminaries are given in verses 213 to 276. The terms nṛtya, and nṛtta are explained in verses 277 to 286, nṛtya being identified as the mārga style and nṛtta as deśī. In so classifying them the author follows the Daśarūpaka and the Saṅgītaratnākara.170 Lāsya and tāṇḍava, and viṣama, vikaṭa and laghu as varieties of tāṇḍava, are described in verses 287 to 296. Sāmānyābhinaya or the general style of acting comes next (verses 297-315), followed by citrābhinaya which means a special acting style (verses 316-18), āhāryābhinaya or the use of costume, make-up and stage-properties (verses 319-25), vṛttis or styles (326-31) and sāttvikabhāva or the representation of emotions(verses 332-433). The rest of the verses are devoted to āṅgikābhinaya or the movements of the body (434-811). The author repeats the same verses in many places in the same chapter (285-86 repeated in 448, 287-92 repeated in 459-64). Although the author has followed the Saṅgītaratnākara in listing the aṅgas or the major limbs, he adds a new one, which is dhammilla or braid (502-4). The dramatic applications of the movements of the aṅgas or major limbs are also described. Some additional hand-gestures are also described, the source of most of which is identified as the Bṛhaddeśī. They are also found in the Saṅgītaratnākara, which, however, acknowledges them as borrowed from an unidentified source(SR.7.284-5; NRK.1.1.513-15; 755). The Nṛtyaratnakośa uses the term nṛtyahasta and not nṛttahasta for hand-gestures employed in dancing, which include three new hand-gestures, aṅjana, jayanta and candraka (NRK. 1.1.774-8), found in no other text.

The second section of the first part discusses in 89 verses nine pratyāṅgas or minor limbs. It includes vartanās or the movements of arms, which the author acknowledges as taken from the Kalānidhi, Kallinātha's commentary on the Saṅgītaratnākara (NRK. 1.2.27.). The third section of the first part, consisting of 159 verses, is on upāṅgas, another division of

170 DR. 1.9; SR. 7.26-28.

80

THE LITERATURE OF DANCE

minor limbs, which are said to be twelve. The fourth and final section of

the first part, consisting of 43 verses, is on āhāryābhinaya or costume,

make-up and stage properties. The first twenty eight verses discuss

āhāryābhinaya but the rest deal with various actions and positions of the

hands, namely, pracāra, karaṇa, karma, kṣetra (NRK. 1.4. 29-42).

The second part begins its first section, of 92 verses, with a

benediction, which is followed by a description of the sthānakas or

postures, which include mārga and deśī types of postures meant for men

and women, for sitting and reclining. The second section of the second

part, in 63 verses, is on mārga cārī. The third section of this part, in 61

verses, is on deśī. An appendix of 37 verses to this section, which,

according to the author, quotes from the Kalānidhi, gives descriptions of

the movements of recakas, that is, oscillating movements, and the muḍupa

or deśī cārīs.171 The fourth section, consisting of 62 verses, is on

maṇḍala or the combination of cārīs. All of these descriptions so far,

follow the Saṅgītaratnākara in the main.

The third part begins its first section, which consists of 188 verses,

with a set of three benedictory verses, after which karaṇas or dance-units

are described, according to the author, after Bharata (NRK.3.1.2). The

author refers to this set of karaṇas as śuddha (literally, pure) karaṇas, by

which he means the mārga variety. The second section of the third part

describes the deśī variety of karaṇas in 51 verses that include descriptions

of nine bhramarīs. This section has an appendix of 23 verses attached to it,

describing thirteen more bhramarīs which the author says he has taken

from Ānandasañjīvana (NRK. Vol. II, p. 169-72). The third section of the

third part discusses aṅgahāras in 106 verses and at the end of the section

the author states that the aṅgahāras are meant specifically for the

prelimināries to a play (NRK. 3.3. 103-5). In the last verse of this section

171 Muḍupacārī is the same as madhupacārī. NRK. pp. 134-38; SR. pp. 313-17; Nr. Adh. pp.108-12.

MOVEMENT AND MIMESIS

81

(NRK. 3.3. 106) Kumbhā gives directions for the application of four aṅgahāras, directions that had never appeared before in any previous work. The fourth section of this part describes in 8 verses the recakas or oscillating movements.

The fourth part starts its first section, of 85 verses, with a long benediction, which is followed by detailed discussions of four vrttis or styles and six kālāsas (dance movements with which a performance concludes) with twenty-two sub-varieties. There are 7 more verses appended to this section giving the definitions of an upādhyāya (instructor), an ācārya (preceptor), a naṭa (actor), a nartaka (dancer) and a vaitālika (a critic who is knowledgeable in music and can amuse the audience). The second section is in 52 verses. In this the author discusses different presentation techniques and rules, then describes a dance called peraṇī, and ends with the definition of a kohlāṭikā, that is, a performer who is able to perform acrobatic movements. The third section of this part has 193 verses, the first 85 of which describe the lāsyāṅgas, of the mārga variety (twelve in number) as well as of the deśī (thirty-six in number). In verses 86-105 the author discusses different gatis or gaits, in describing which he says he is following Bharata. This is followed by a description of the deśī nrtyabhedas in verses 106-123. The term nrtyabheda is used by Dhanika and Śāradātanaya to denote minor dramas that require dancing.172 But Kumbhā uses it to mean dancing. Of the six nrtyabhedas described in this text, śivapriya comes from the Nṛttaratnāvalī, which describes it as a deśī dance. It occurs in no other text. Dohaka is also a new name which is not mentioned in any other text. The rest are described in all other texts that discuss minor dramas.173 Verses 124-128 discuss the technical language used by the dancers. Four features of deśīnṛtyas are

172 See pp. 167-9 infra.

173 The nrtyabhedas are śivapriya, rāsaka, nāṭyarāsaka, daṇḍarāsaka, carcarī and dohaka; NRK.106-123.

82

THE LITERATURE OF DANCE

described next, followed by deśigītanṛtyavidhi which describes six

varieties of regional dances following specific tālas and the music

appropriate to them, along with general directions for such presentations

in verses 129-144.

From the above discussion it seems that this author divides the minor

dramas that employed dance and expressive dance of the deśī variety into

two different categories, respectively, deśī ṇṛtyabheda and deśinṛtya, the

latter being a category used by what he calls "other" authorities. The rest

of the verses are on the nine rasas and their expression through body

movements. The fourth and the final section of the fourth and last part of

the treatise has 88 verses. In this section the author attempts to define the

ideal performer, explains the aesthetic concept known as rekhā or the

lines of body movements that enhance beauty, enumerates the qualities

and faults of a performer, discusses make-up, different schools of

performing artists, their qualities and faults, the śuddapaddhati or the

pure way of presentation, the gaunḍalīvidhi or the regional variation

called gaunḍalī, explains how instructions are given to performers and

ends with a final prayer.

The beginning of the third period in the evolution of the concept of

dance is marked by the Nartananiṛnaya of Punḍarīka Viṭṭhala who lived in

the late sixteenth to the early seventeenth century.174Till Mahārāṇā

Kumbha's Ṇṛtyaratnakośa the major divisions of the art as conceived by

174Krishnamachariar, 1974, p. 865. This text remained unedited and unpublished until

recently. Dr. S. Satyanārāyaṇa brought out an edition in early 1987 in Kannada

script prepared from one full and four fragmentary MSS.. A detailed account of the

fourth chapter of the text appears in the present writer's dissertation (Oxford, 1964)

and in her book on classical Indian dancing (1970). A critical edition based upon 15

MSS. is under preparation by the present writer. References to the 4th chapter of the

text in the present study are to the India Office Library, London manuscript (MS.

5197). References to other chapters are to the Asiatic Society, Calcutta MS.(MS. III.

D.5) and to the Jaipur Khasmahal Library MS.(MS. 6885).

MOVEMENT AND MIMESIS

83

scholars were mārga and deśī, that is, the canonical style and the regional

or popular styles. The Nartananiṛnaya marks a major conceptual

departure by dividing dance primarily by structural principle into two

divisions, namely, bandha, or styles that rigidly adhered to set rules of

composition, and anibandha, styles that did not do so and allowed

innovations by the dancer. Many of the works on dancing written after the

Nartananiṛnaya still followed the approach of the Nāṭyaśāstra as found in

the Saṅgītaratnākara but the Nartananiṛnaya's approach to categories of

dancing became part of their conceptual framework.

Puṇḍarīka Viṭṭhala is believed to have come from southern India and

later he became a court poet in the courts of Man Singh of Jaipur and his

brother Madho Singh. He wrote several texts on music under the

patronage of these rulers but still later moved on to Akbar's court. His

three works on music, the Sadrāgacandrodaya, the Rāgamālā and the

Rāgamañjarī are well-known. He is believed to have written two more

texts on music, the Rāganārāyaṇa and the Saṅgītavṛttaratnākara. He also

wrote a lexicon titled the Sīgrabodhinīnāmamālā and a treatise called the

Dūtīkarmaprakāśa on the characters of the heroine and the hero.175

The Nartananiṛnaya was written in the sixteenth century to please the

emperor Akbar, according to Puṇḍrīka's statement at the end of the

text.176 Like most works before it, the Nartananiṛnaya discusses the

various types of acting, namely, āhāryābhinaya, sāttvikābhinaya and

āṅgikābhinaya, and in doing so it follows the Nāṭyaśāstra as interpreted in

the Saṅgītaratnākara. However, instead of following the usual practice of

reproducing the Nāṭyaśāstra's (and in the case of later texts, the

Saṅgītaratnākara's rendering of the Nāṭyaśāstra's material) descriptions

of the 108 karaṇas and the 32 aṅgahāras created by combining them, the

175 Krishnamachariar, 1974, p. 865; Raghavan, 1960., pp. 15-16.

176 अकबरनृपसच्‍चर्यं भूष्लोके सरलसंगीतं ।

कृतमिदं बहुत्रभेदं सुहृदां हृदये सुबं भूयात् ॥ (Nartananiṛnaya. 53b).

84

THE LITERATURE OF DANCE

Nartananirnaya selects only 16 of the karanas as those needed in bandhanṛtya, of which it describes several varieties. The text then proceeds to describe the distinctive features of the various kinds of anibandhanṛtya. From these descriptions of dance compositions there emerge striking similarities with the classical dance styles of the present time, as demonstrated in a later chapter of the present study. This text thus provides the link so far missing between the older and the present day traditions of classical Indian dancing.

At the beginning of the first chapter, the author declares his plan to write on five topics, namely, tāla or rhythm, vādya or instrumental music, gīta or vocal music, nartana or dance and nāṭya or drama.177 However, he never wrote the fifth chapter, that on nāṭya, nor gave any reason for the omission, leaving the treatise with four chapters, the last of which is the one on dance.178 The text begins with a set of 34 verses, written in a variety of metres, in praise of Akbar and his ancestors.179 The first chapter, consisting of 259 verses, is on rhythm, the second, in 116 verses, is on drums and the third, in 579 verses, is on vocal music. The fourth chapter, the largest one, has 916 verses that deal with dancing. This chapter starts by defining nartana, a term used by the author to mean dance. Nartana is divided into nāṭya, ṇṛtya and ṇṛtta, of which the last is again divided into three, viṣama, vikaṭa and laghu. All the types are defined and the author reproduces in the first ten verses the Sangītaratnākara's view that ṇṛtya and ṇṛtta may both have varieties of tāṇḍava and lāsya. In verses 11 to 206 abhinaya is discussed, with the sāttvika, āhārya and āṅgika types of abhinaya described in detail but not vācika as it is not employed in nartana. Citrābhinaya is then described in

177 Nartananirnaya. 1a (Calcutta.MS.).

178 Nartananirnaya. 53b (London.MS.).

179 This part is contained in the Jaipur.MS. but not in the Kannada edition, which is based on only one full and four fragmentary manuscripts.

MOVEMENT AND MIMESIS

85

verses 207 to 238. The author does not divide the movements of the limbs

into the movements of añga, upāñga or pratyañga but in verses 239 to 244,

where he lists his topics of discussion, he mentions the movements of the

parts of the body which, in his view, are of importance. These include the

movements of the head, the eyes, the eyebrows, the arms, the hand-

gestures and other actions of the hands, the waist and the feet. It also

discusses the function of the colour of the face. The list further includes

more complicated movements generated from the combination of the

movements of the parts of the body, such as the sthānas or postures, cārīs

or the movements of one leg, karaṇas or dance-units and recakas or

oscillating movements. Also in the list are the dance-hall, the

characteristics of a good dancer, rekhā or the lines created by the

movements of the body, the lāsyāñgas or features of lāsya, sauṣṭhava or

standing without any movement, citrakalāsa or concluding movement,

mudrā or natural grace, pramāṇa or harmony, the audience, the person

presiding, sitting arrangements, the troupe of musicians, the flute, the

entrance of a dancer and various dance-sequences. The actual discussions

of these topics is in verses 245 to 656. Most of the material comes either

from the Nāṭyaśāstra or the Saṅgītaratnākara. The original matter that the

author contributes to our knowledge comes after this when he begins to

describe the opening dance item, the dancer's entrance into the stage.

This is where he divides nṛtta into two categories, bandha and anibandha.

Under bandhanṛtta he puts mukhacāli, urupa, dhuvāda, viḍulāgava,

śabdacāli (later discussed as śabdanṛtta), śabdaprabandha, svaramaṇṭha,

gītaprabandha, cindu, dharu and dhruvapada. Their descriptions in verses

668 to 874 show them to be highly structured dance pieces.180 A group of

five bhramarīs is also discussed (794 to 98) between the discussions on

viḍulāgava and śabdanṛtta.

180See Bose, 1970, pp. 149-161 for details.

86

THE LITERATURE OF DANCE

Next, anibandha dance is discussed in verses 875 to 898 with its forms

given, namely, nāmāvalī, yati, different neris, kaivartana, murū,

tālarūpaka, guṇḍāla, kamala, natajānuka, maṇḍī, muḍupa, muraṇdarī,

kuḍupa, tiryakaraṇa, lāvanī and batu. These have fewer details compared

to the discussion of the bandhanṛttas. At the end of these descriptions the

author refers to these sequences as anibandha urūpas, evidently using the

term urūpa to denote a broad category of dance. Urūpa is described only in

two works, in this text and in a later work, the Saṅgītamakaranda of

Vedasūri.181 Finally, Puṇḍarīka Viṭṭhala ends the work with two more

dance sequences, jakkaḍī and rāsa,which he includes under anibandha

dance (875 to 912). Throughout these descriptions the terms nṛtta and

nṛtya are used interchangeably. The last four verses of the text are in

praise of the patron and contain a final prayer.

The Rasakaumudī of Srīkaṇṭha is another text from the third period.

The author was a contemporary and student of Puṇḍarīka182and was a

court poet of Śatrusālya (Jam Sattarsal of Navanagar near Dvārakā) who

lived in the late sixteenth and the early seventeenth centuries.183 He

refers to Puṇḍarīka as his teacher whom he follows in defining music.184

The editor of the text thinks that Srīkaṇṭha, like his teacher, was originally

from the South and moved to western India.185

The Rasakaumudī is a work in ten chapters divided into two parts,

pūrva and uttarakhaṇḍa, each consisting of five chapters. The first part

deals with vocal and instrumental music and dance. The second part deals

181 In some of the copies of the manuscripts of this text the term urūpa is spelt as urupa or

udupa.

182श्रीमद् विट्ठलदेवस्य प्रसादाच्च गुरोर्मया।

श्रीकण्ठेनोच्यते कामं तालकादिनिर्णय: ॥RK. 4.96.

183Rasakaumudī,1963, G.O.S. ed. 143. Introduction, pp. 8-9.

184Ibid. p. 8. RK. 4.80.

185 RK. Introduction, p. 8.

MOVEMENT AND MIMESIS

87

with drama, rasa, the seasons and the proper conduct of a king. The first

chapter, in 91 verses, deals with general matters pertaining to music. The

second, in 219 verses, explains the nature of rāga and the technicalities of

making and playing the vīṇā. The third chapter consists of 150 verses

discussing further intricacies of vocal music. The fourth chapter, in 190

verses, discusses vādya or musical instruments, with details borrowed

extensively from the Nartananiṛnaya.

Chapter five is the part that holds our interest most since it deals with

dancing. Consisting of 356 verses, this is the longest chapter. Beginning

with verses in praise of Kṛṣṇa, the author relates the origin of nāṭya

following the Sangītaratnākara. This is followed by a description of the

stage, and of the initial presentation by the principal dancer and others.

The author recommends the ārabhaṭī style of presentation for the initial

part of the performance. Dancers from different regions are described

next, followed by āṅgikābhinaya or the use of body movements in acting,

which include the movements of the head, the hands and the feet. The

author lists the number of movements prescribed for the minor limbs

without giving any details. He mentions ten divisions of nāṭya but calls

only the first variety nāṭya. The ten divisions of nāṭya that he mentions are

nāṭya, nṛtya, nṛtta, tāṇḍava, lāsya, viṣama, vikaṭa, laghu, peraṇi and

gaundalī. He describes four kinds of abhinayas and then reproduces

verbatim verses from the Nartananiṛnaya that give general instructions on

acting, for which Puṇḍarīka had drawn upon the discussion on

sāmānyābhinaya in the Nāṭyaśāstra (chapter 22).186 Before describing

any dance-sequence, Śrīkaṇṭha describes the essence or prāṇa of a nṛtya

(nṛtya here stands for the art of dance). Next, puṣpāñjali or the flower-

offering on the stage, the rules of exercise, make-up, and presentation

techniques are described, followed by the definitions of 108 karaṇas. The

author then presents a list of cārīs, maṇḍalas and sthānas, adding that there

186RK. 5.145-161; NV. 3b-12a.

88

THE LITERATURE OF DANCE

can be many more varieties of karanas, aṅgahāra, cārī and maṇḍala. Then

he proceeds to describe utplutikaraṇas, aṅgahāras, cārīs, recakas, sthānas,

maṇḍalas and lāsyaṅgas. In describing lāsyaṅgas he follows the

Nartananirnaya. He divides gatis or gaits into three different types,

uttama, madhyama and adhama and refers to the use of three tempos.

Thus ends the chapter on dancing and the first part of the text, in which he

keeps referring to Bharata as the authority followed.

The sixth chapter is very short, consisting of 5 verses and, as the editor

of the text points out, it serves as a connecting link between the first and

the second parts of the text with a benedictory verse and an introduction to

the subject of the next chapter, which is rasa. The seventh chapter, in 26

verses, is devoted to the discussion of the nine rasas. The author discusses

śṛṅgārarasa, giving its varieties in detail and pays less attention to the other

rasas. The eighth chapter, in 31 verses, begins with a discussion of how to

enhance the beauty of a woman in sixteen different ways. This he refers

to as śṛṅgāravarṇana. It includes taking a bath, putting on make-up,

wearing jewellery, dressing and giving the finishing touch by chewing on a

betel leaf (in order to redden the lips). The author adds shrewdness to the

list of qualities necessary in a good performer and instructs the performer

to use appropriate emotional expressions. The ninth chapter describes the

six seasons in 59 verses after justifying the inclusion of this subject.

According to the author, the seasons should be enjoyed and the joy of the

seasons can be expressed through different costumes worn to represent

different seasons. The tenth and last chapter has 48 verses and deals with

royal conduct and the daily programme of a king. The reason given for

including this subject is that a king should know and enjoy the art of

dancing. The chapter and the text end with a final prayer.

The contribution of Śrīkaṇṭha to dance scholarship is not substantial in

terms of the definitions and categories of dance that he presents, nor of the

descriptions of body movements. His chief contribution lies in his

discussion of the concept of prāṇa, or the essence of a performance, which

MOVEMENT AND MIMESIS

89

sums up what a dancer should aim at while performing. The ten prāṇas are

listed in the following couplet:

रेखा च स्थिरता वेगो भ्रमरी दृष्टि+रश्मय: +।

प्रीतिर्मेधा वचो गीतं नृत्ये प्राणा दशोदिता: ॥

(RK. 5. 162)

The line, the steadiness, the swiftness, the pirouettes,

the glances, the desirous [smile], the pleasing

[appearance], the intellect, the speech and the song

constitute the ten vital spirits of a nṛtya.

It is interesting to note that the fourth essential prāṇa is bhramarī, which

by this time must have become a movement of such vital importance in

dance that it was exalted to the status of a prāṇa. It is unfortunate that

Śrīkaṇṭha leaves it undiscussed, especially as he is the first author to speak

of prāṇa.

Śrīkaṇṭha's other contribution is his detailed instructions for make-up

and stage presentation, his approach to both subjects being original.

Regarding the second, for instance, he says:

माल्यकारो यथा माल्यं वध्नाति कुसुमोत्करै: ।

सुधी कुर्यात् तथा नाट्यमज्ञोपाधै रसादिभि: ॥

(RK. 5. 183)

Just as [the way] a garland maker makes a garland

with the help of flowers a performer [literally, the

skilled one] should act in the same way with [the

movements of] the major and minor limbs and

[expressions of] rasa etc.

The work also has the virtue of being a systematic and organized approach

to the subject. However, as we have noted in summarizing the work, it is a

90

THE LITERATURE OF DANCE

derivative piece of writing, the author relying heavily on Puṇḍarīka for the

most part.

The Saṅgītadarpana of Dāmodara is a work on music and dance. He is

believed to have been a descendant of Kallinātha, the commentator on the

Saṅgītaratnākara. Dāmodara was a poet at Jahangir 's court, which places

him in the seventeenth century.187 A complete manuscript is known to

exist in the Bibliotheque Nationale of Paris; a fragment containing three

chapters on instrumental music, rhythm and dancing is in the Bodleian

Library at Oxford.188 The readings of the two manuscripts are virtually

the same. One edition of the full text, an unsatisfactory one, as Rāghavan

points out, was brought out in 1952 by the Saraswati Mahal Library of

Tanjore. Nijenhuis ascribes the work to Haribhaṭṭa, who was indeed, as

reported by Raghavan, the author of a work called the Saṅgītadarpana.189

The two works are almost identical, which permits the conjecture that

either Haribhaṭṭa and Dāmodara were one and the same person, or

Haribhaṭṭa "made his own version of Dāmodara's work."190 According

to Raghavan, the Tanjore edition is from the manuscript that bears the

name of Haribhaṭṭa. Generally speaking, the Saṅgītadarpana is ascribed

to Dāmodara by most scholars, including Raghavan. Comparison between

the chapter on dancing of the edited version with that of the Bodleian MS.

shows that the two are identical.

The Paris manuscript has seven chapters while the manuscript

examined by Raghavan is said to have six. The edited text of the

187 Krishnamachariar, 1974,.p.866.

188 See Bose, 1964, pp.16-18; 1970, p. 5. Saṅgtadarpana of Dāmodara, Bibliotheque

Nationale, Paris, MS.No. 771; Bodleian Library , Oxford, MS. No. Mill. 47.d.

189 Raghavan, 1960, pp. 20-21.; Nijenhuis, 1977, p. 27.

190Raghavan, 1960, p. 21.

MOVEMENT AND MIMESIS

91

Sañgītadarpana has seven chapters,191 the first on svara or the science and philosophy of sound, the second on rāga or musical modes, the third on

prakīrṇa or miscellaneous technicalities of music, the fourth on prabandha or composition, the fifth on vādya or instrumental music, the sixth on tāla

or rhythm and the seventh on nṛtya or dancing. The first chapter has 150 verses and starts with a benediction and proceeds to define the term

saṅgīta to explain its division into mārga and deśī. The origins of rāga or musical mode, nāda or sound and other technical intricacies of vocal

music are discussed, and the chapter ends with a verse that lists seven birds and animals that represent the seven musical notes. The same verse

is also found in the Nartananiṛnaya.192 The second chapter (verses 151-289) begins with a verse defining rāga, which is found in the

Sañgītaratnākara and the Nartananiṛnaya. This chapter describes the anthropomorphic images of the rāgas and rāgiṇīs as well as their

structures. Kriṣṇamāchariar and Rāghavan state that Dāmodara borrowed his anthropomorphic description of musical modes from Somanātha's

Rāgavibodha.193 Chapter three (verses 290-352) discusses a variety of subjects, including ālāpa, the use of the voice, the qualities of a singer and

the characteristics of a musical composition. The fourth chapter (verses 353-531) discusses the details of musical composition. The fifth chapter

(verses 532-615) is on vādya or instrumental music and describes the vīṇā and other stringed instruments, different kinds of flutes, their

characteristic sounds, the compositions appropriate to them, drums and so on. Here the author again quotes extensively from the Nartananiṛnaya,

though he does not name it as his source. The sixth chapter is on tāla. In

191Raghavan reports that six chapters and the chapter on tāla is missing from his list, 1960, p 21. The Paris MS. has seven chapters.

192SDar.1. 150; cf. NN. 20a. There are a number of verses which seem to be quoted from the Nartananiṛnaya.

193Raghavan, 1960, p. 21; Krishnamachariar, 1974, p.866.

92

THE LITERATURE OF DANCE

verses 616-798 it defines the term tāla, explains its derivation, and

describes ten prāṇas, mārga tālas and, finally, deśī tālas. At the end of this

section appears a description of 120 tālas, of which the source is the

Sañgītaratnākara. This is followed by another set of 103 tālas described

according to other authorities (pp. 149-70). In the section that follows,

other technicalities of tālas are described.

The final and seventh chapter, in 282 verses, is the one on dancing.

The chapter begins with a benediction followed by a detailed description

of the opening dance called mukhacāli which is very similar to the

mukhacāli described in Nartananiṛnaya in greater detail. That the

descriptions in the two works come from the same tradition is shown by

the fact that the song to be illustrated by hand-gestures is the same in both

the texts. Next, verse 53 lists ten types of dancing but does so confusingly.

The verse identifies nātya, nṛtya, tāṇḍava, nṛtta and lāsya as types of

nartana. It then divides nṛtya into five types, namely, viṣama, vikaṭa,

laghu, peraṇi and goṇḍalī (a variant form of the word gaunḍalī), which are

deśī forms. This division might thus seem to indicate that nartana is the

general class name and that nṛtya is a subdivision. But the verse concludes

by stating that all ten types mentioned are types of nṛtya. Evidently, then,

the text is using the terms nartana and nṛtya interchangeably in this verse.

Verses 54-93 describe bhāva, rasa, mukharāga, the four types of

abhinaya, and explains the meaning of lokadharmī or natural and

nāṭyadharmī or stylized presentations. Verses 94-96 define nṛtya and

nṛtta. Here, nṛtya is considered to be a mārga style which portrays

emotions through body movements. Nṛtta, on the other hand, is based on

tāla and laya and expresses no meaning. Verses 97-103 list the movements

of aṅga, upāṅga and pratyaṅga, followed in verses 104-11 by definitions of

rekhā (the imaginary lines described by the body in motion), pramāṇa

(harmony), and the ten prāṇas (the vitalizing power of a performance) to

be instilled into a dance by the dancer. Prāṇa is discussed, as we have

seen, also in the Rasakaumudī. Gatis are described from verses 112-23 in

MOVEMENT AND MIMESIS

93

the manner of the Nartananirnaya. Tāṇḍava and lāsya are defined with

quotations from the Saṅgītaratnākara in verses 124-25. The author views

them as dances for men and women, respectively. In verses 126-30

mukharāgas or the colour of the face, nāndī or the opening verse and the

characteristics of an actor are described. The rest of verses 131-282

describe dance pieces. Most of these descriptions are taken from the

Nartananirnaya. The author then describes yatinṛtya, śabdacāli, uḍupas,

dhruvāḍas, śabdanṛtya, camatkāranṛtta, śabdanṛtta, gītanṛtya,

svarābhinaya, svaramanṭhanṛtya, sūdanṛtya, dhruvagītanṛtya, one after

the other, classifying them under the śuddha or pure dance style (verses

131-240). Deśīnṛtyas are described next, which include cindu, kaṭṭarī

(dharu is a variety of this type) and vaipota (verses 241-61). Finally,

bandhanṛtya, kalpanṛtya, jakkarī, peraṇī and goṇḍalī are described (verses

261-82).

While it is evident that there is no original material on dancing in the

Saṅgītadarpana, either in terms of information or interpretation, the work

is valuable as an illustration of an approach to dancing that began with the

Nartananirnaya. As shown above, the author follows the Nartananirnaya

as well as the Rasakaumudī, which itself borrowed the views of the

Nartananirnaya, in showing a growing interest in such aesthetic elements

as rekhā, pramāṇa and prāṇa. Generally speaking, then, we find from the

Nartananirnaya onwards a great stress laid on the responsibility of the

dancer for creating the artistic effect.

Saṅgītanārāyaṇa is a seventeenth century text by Purṣottama Miśra, a

poet at the court of Gajapati Nārāyaṇadeva of Orissa.194 The work is in

four chapters and deals with music and dance. Chapter one (546 verses),

called gītanirnaya, is devoted to the technicalities of vocal music, such as

nāda, śruti, svara, grāma, rāga, gīta and tāla. The second chapter,

vādyanirnaya (131 verses), discusses vādya or instrumental music. The

194 Katz, 1987, Introduction, p. iii -vi.

94

THE LITERATURE OF DANCE

third chapter, nātyamirṇaya, in 828 verses, deals with nātya or mimetic art,

which includes dancing. The fourth chapter (verses unnumbered, pp.176-

200), śuddhaprabandhodāharaṇa, is on prabandha or composition. The

first edition of the text was brought out by the Saṅgīt Nāṭak Academy of

Orissa in 1966. A scholarly edition of only the musicological parts of the

text has been recently prepared by Jonathan Katz but not yet published.

There is no satisfactory edition of the part on dancing. However, the

edition prepared in Orissa gives us a fair idea of the contents of that part.

The third chapter, on dance and mimetic art, begins by relating the

origin of the nāṭyaveda. Similar verses are also found in the

Abhinayadarpaṇa (SN.3.2,3; ADar.2, 3). Next, the author divides nartana

into nāṭya, nṛtya and nṛtta and defines them, citing various authorities. He

then defines mārga and deśī along the lines of the Saṅgītaratnākara and

also refers to the views of Kohalā (3.3-11). As examples the author names

twenty mārga-nāṭyas, which include ten rūpakas and ten other varieties of

dramatic presentation, for which no class-name is given.195 Sixteen

varieties of deśīnāṭyas are given as those identified by Dattila.196

Twelve names of deśīnṛttas are given next, some of which are the same as

types of mārganāṭya and deśīnāṭya.197 The names of ten mārganāṭyas

(excluding rūpakas), sixteen deśīnāṭyas and twelve deśīnṛttas are cited as

found in various texts as examples of minor dramas (3.12-20). The names

of the deśīnṛttas are those of the nṛtyabhedas in the Avaloka of Dhanika, in

195 Ten varieties of mārganāṭyas, not included in the rūpakas, are Nāṭikā, Prakaraṇikā,

Bhāṇikā, Hāsikā, Viyoginī, Ḍimikā, Utsāhavatī, Citrā, Jugupsitā and Vicitrā. SN. 3.

14-15.

196 The sixteen varieties of deśīnāṭyas are: Saṭṭaka, Troṭaka, Goṣṭhī, Vṛndaka, Śīlpaka,

Prekṣaṇa, Saṁlāpaka (sallāpaka?), Hallīśa, Rāsaka, Ullāpyaka, Śrīgadita,

Nāṭyarāsaka, Durmallī, Prasthāna, Kāvya and Lāsikā. SN.3. 15-18.

197 Domikā, Bhāṇikā, Prasthāṇaka, Bhāṇaka, Lāsikā, Rāsikā, Durmallikā, Vidagdhā,

Śīlpinī, Hasti(Daṇḍi)nī, Ulmukī (Bhillukī), Tumbikā are cited as the twelve

deśīnṛttas. SN.3. 18-20.

MOVEMENT AND MIMESIS

95

the Bhāvaprakāśa and in the Sangītadāmodara.198 The use of the term

nṛtta in this context in the Sangītanārāyaṇa seems to be a scribal error.

Not only do the varieties mentioned above appear in the category of nṛtya

in other texts, but the Sangītanārāyaṇa itself recognizes nṛtta as a

different branch, as we see only a few verses later where the types of

nṛtta are discussed along the lines of the Sangītaratnākara. Although

Puruṣottama borrows frequently from others, the classification of the

dramatic arts into three categories is entirely his own, as is his mention of

so many dramatic types, which we find in no other work.

Tāṇḍava and lāsya are next defined and described as two types of nṛtta.

Two kinds of tāṇḍava, preranī and bahurūpa, and two kinds of lāsya,

sphurita and yauvata, are discussed in detail (3. 21-28). Nṛtta is then

described, with its three varieties, viṣama, vikāṭa and laghu (3.29-30).

Tāṇḍava and lāsya are designated as dances for men and women following

the Sangītadāmodara (3.31; SDām. p.69). A new set of nātyas, known as

prakāranāṭyas are mentioned next and their names are given as kāṣṭhā,

jākaḍī (same as the jakkaḍī mentioned in the Nartanirnaya ), śāvara,

kurañjī and mattāvalī. These are in fact dance pieces, of which only

jākaḍī was recorded by earlier authors. After describing these dances the

author refers to them as deśīnṛtya (3. 31-36), of which a definition is then

given (3.37). We may note that this author records two kinds of deśī

dances, namely, deśīnṛtta and deśīnṛtya.

The author then describes the nāṭyaśālā or the dance-hall and the

sitting arrangements for the audience (3. 38-52). The qualifications

desirable in spectators are enumerated (3. 53-61). The author then

describes the talents and defects of performers, particularly the principal

dancers, reviewing dancers from different regions. Then he describes

how the training of a dancer begins(3. 62-83). He speaks of the work of

the teacher and mentions the use of the curtain (3. 84-90). Four types of

198 DR. 1969, p.8; Bh.P. 1968. pp. 255-69 ; SDām. 1960, pp. 91-96.

96

THE LITERATURE OF DANCE

abhinayas are also mentioned. The type of abhinaya known as āhārya or

the use of costume, make-up and stage accessories is described at great

length (3. 90-175). The languages used for different characters in a play

and the forms of addressing one another are then described (3.176-201).

The author also suggests names for characters which will express their

distinctive traits (3. 202-13). Vṛttis or styles of presentation come next (3.

214-19). The author says that since sandhi, bindu, patākã etc. (terms that

mark the structural divisions of a play) and the pūrvarañga (the

preliminaries to a play) are not relevant to nartana or dancing, he will not

even mention them (3. 220). This suggests that the use of nartana or dance

in the pūrvarañga of a play was no longer of vital importance when this

text was written. Instead, the art of dance must have been very much a

part of the total presentation of saṅgīta, that is, a comprehensive recital of

vocal and instrumental music and dance. The interdependence of dance

and music can be seen from the early medieval period and as we analyse

the texts it becomes much more apparent that in the later medieval period

there was no need for any author to describe saṅgīta primarily as a part of

a dramatic presentation.

The author proceeds with the characteristics of the best, mediocre and

worst performers. The method for entering the stage follows, which this

text borrows line for line from the Saṅgītaratnākara and the

Nartananiṛnaya.199 The actions that are to be avoided on stage are

specified next. The ideal time for the presentation of a nāṭya is then

mentioned.200 The author adds that the spectator has to remember to

encourage the dancer by showing his appreciation; otherwise many

misfortunes may befall him (3. 221-255). Āṅgikābhinaya is then described

in detail (3. 255-728) with the divisions of limbs into añga, upāñga and

199 SN. 3. 235-8; SR. 7. 1260-63; NN. 40a.

200 Nāṭya in this context refers to dance, a common practice of the writers on dancing through its history, which is followed in naming the present-day style Bharatnatyam.

pratyańga.

Next, the author identifies five features of nṛtya, namely,

sthānaka, cārī, karaṇa, maṇḍala and ańgahāra, and he does so, as he

himself says, by following the path shown by Bharata. Sthāna is then

defined and its different types are described with the names of 52 mārga

and deśī sthānas given but no details (3. 731-43). Cārī is then defined and

two lines are quoted from the Sańgītadāmodara and the

Sańgītaratnākara.201 The names of 32 mārga and 54 deśī cārīs are given

without any description (3. 744-64) but with a passing reference to

Bharata. Karaṇas are explained next, Śārńgadeva being cited as the

source and the names of 108 karaṇas are given (3. 765-82). The author

prescribes the talapuṣpapuṭa karaṇa for the worship at the beginning of the

performance and the gańgāvataraṇa karaṇa as part of the benediction with

which a performance should conclude (3. 784-5). The names of thirty-six

utpluti or deśī karaṇas are given next (3. 786- 92). Seven bhramarīs are

included in the list of utplutikaraṇas as in the Sańgītaratnākara.202 Next,

in verses 793-96, thirty-two kalāsakaraṇas are mentioned. Kalāsas occur

in other works but are never thought of as karaṇas, which is a very

different way of looking at these movements and unique to this text.203

The author then defines maṇḍala and lists twenty varieties, quoting lines

from the Sańgītadāmodara (3. 797-801; SDām.p. 67). The author

describes ańgahāras next (3. 802-17), quoting lines from Śārńgadeva and

Subhańkara.204 Recakas are then described in detail. The chapter ends

with verses, again from Dāmodara, prescribing methods of practice,

eating light food etc.205 The author finally talks about nyāya or the use of

201 SN. 3. 745; SDām. p. 64; SN. 3. 783-4; SR. 7. 748-9.

202 SN. 3. 791-2; SR. 7. 755.

203 Kalāsas are described in detail in the Nṛtyādhyāya and in the Nṛtyaratnakośa. Nr.Adh.

1570-1613; NRK. 4. 1. 37-85.

204 SN. 3. 797-8; SDām. p. 67; SN. 3. 804; SR.7. 791; SN. 3. 808; SDām. p. 68.

205 SN. 3. 823; SDām. p. 66.

98

THE LITERATURE OF DANCE

appropriate movements and ends his discussion out of concern for the length of the text.

The new information given by this author relates to the different varieties of mārga and deśin̄ātya, deśinṛtta (3.12-20), five varieties of prakāraṇātya, also known as deśinṛtya (3.31-36), and kalāsakarana (3. 793-96). Although he devotes a large number of verses to descriptions, the information does not add much to the knowledge we have already gathered from other texts insofar as the movements of the limbs are concerned.

The next work on our list is the Sangītamakaranda of Vedasūri which was written in the early seventeenth century. From Raghavan’s account, written in 1932-33 (repr. 1956,1957, 1960, 1961), it appears that the only extant part is the chapter on dancing, which has several prakaranas or sections. Raghavan reports that on examining the manuscripts he found that the chapter on dance discussed rasadṛṣṭis, gatīs, cārīs, hastas, various dances and lastly rasas. He also mentions that S̄āhajī, father of the Maratha leader S̱ivājī, was Vedasūri’s patron.206 Krishnamachariar repeats this information but adds that Vedasūri was the son of Ananta and a grandson of Dāmodara, author of the Sangītadarpana.207 The text was edited by K.V. Vasudeva Shastri and published in several parts in the Journal of the Saraswati Mahal Library, Tanjore in volumes VI.3; X. 1; X. 2 and X. 3 and XII from 1955 to 1957.

Of these, only some parts have been available to me. The first section available (X.2) starts from verse 65 and describes the sixth gati, mrgī. This description is different from any discussion of gati that we have come across so far. The author treats each gati like a dance sequence and describes the gati with all its components of movements. For instance, while describing the mrgī gati the author gives all the movements necessary for its presentation, such as the appropriate karana, sthāna,

206 Raghavan, 1961, pp. 21-23.

207 Krishnamachariar, 1974, p.867.

MOVEMENT AND MIMESIS

99

cārī, the hand-gestures, the head movements and glances. The rest of the gatis are described in the same way. In all he describes the eleven gatis that, he says, are recognized by schools of dancing. This is followed by a description of two cālakas of the arms, laharīcakra and nīrājitapadma (the names are new) that are necessary for maṇḍalāsthanaka. At the end of this section the author describes the abhinaya or miming of a verse in the same way that Puṇḍarīka Viṭṭhala in Nartananiṛnaya has described the abhinaya of a śloka,208each word of which is to be expressed by an appropriate hand-gesture. This first prakaraṇa or section ends with a description of the mukhacālī or the opening dance-sequence. The first description of mukhacālī in dance literature occurs in the Nartananiṛnaya. The second prakaraṇa of the Saṅgītamakaranda describes the aṅgahāras which are to be performed in the preliminaries of a play. According to the author, this dance requires twelve karaṇas, which are described in detail.209The rest of this section is devoted to the description of the aṅgahāras.

In the next prakaraṇa the author says that after performing the aṅgahāranartana, uḍupa in lakṣmītāla is to be performed and describes in detail the dance along with the required tāla. Next to be described is the dance called nāmāvalī which, according to the author, is taken from Kohala's description. A closer source, however, is Puṇḍarīka's Nartananiṛnaya. Nāmāvalī can be performed in four, five or six khaṇḍas or units of tālas, each of which is described in the Saṅgītamakaranda with its appropriate rhythmic syllables, for which the Saṅgītadarpana is cited

208See Nartananiṛnaya. 42b.

209The twelve karaṇas are vaisākharecita,vṛścika, vṛścikakuṭṭita, ākṣiptaka, cakramaṇḍala, uromañḍala, āvarta, kuñcita, dolāpāda, vivṛta, vinivṛtta and kaṭicchinna. Interestingly, although these karaṇas are from the list found in the Bharata tradition, they are different from the group of sixteen described in the Nartananiṛnaya. Obviously, the style that is described in the Saṅgītamakaranda is different from the style described in the Nartananiṛnaya. NN. 32b-33b.

100

THE LITERATURE OF DANCE

as the source. Next comes śabdacāli nṛtta, followed by the dances termed neris. A new dance sequence called hastaneri, not found in any other text, is described. This dance requires twenty-four kinds of hand-gestures, ten varieties of sthānas, six types of cārīs, various movements of the different parts of the body and different gatis done in a slow tempo and to āditāla.

The author begins to describe the twenty-four hand gestures required for this dance. Unfortunately, the edited text to which I had access breaks off after the description of the seventh hand-gesture.

Although the whole text has not been available to me, the portions examined here are sufficient to show that the author's approach to his subject was unusually full in the detailed instructions he gives for the actual performance of dance movements. He seems to have been interested mainly in the structure of dance compositions as combinations of smaller movements. He describes these movements step by step, and includes with each movement the appropriate rhythm and tempo.

Yet another seventeenth century work that deals with, among other subjects, music and dance is the Śivatattvaratnākara. The author was Bāsavarāja, a king who reigned between 1684-1710 A.D. The text has been edited by Narayanaswamy Sastry in three volumes and in the introduction to the first volume H. Deverappa tells us that Bāsavarāja's kingdom was known as Keladi which stretched across "the whole coast of Kanara from Goa in the North to Cannanore in the South as also parts of the Shimoga district of Mysore.210

The text consists of nine kallolas or sections with several tarangas or subsections each. Altogether a hundred and one such subsections cover a variety of subjects. Dance is described in the sixth chapter in several tarangas or subsections numbering 3-6 while music is described in 7-9. The third taranga of the sixth chapter has 56 verses, the fourth has 114, the

210 Śivatattvaratnākara, 1964, Vol. I, Introduction, p. ii.

MOVEMENT AND MIMESIS

101

fifth has 101 and the 6th and final taranga, the one that is most relevant to

our study, has 100 verses.

The third taranga of the sixth chapter begins with a definition of dance

for which the author uses the term natana (6.3.1) and which he equates

with both nrtta and natya (6.3.5; 6.3.7). Here he explains the nature of

dance, writes in praise of the art, relates its origin and discusses verbal

acting. Nātya and nrtta are continually used interchangeably (6.3.5-9).

Next, marganrtta is described, divided into tāndava and lāsya, followed by

a description of desinrtta. Then the author describes the seating

arrangements for the king, the queen, and the rajavilasinis or women of

the royal court. Next he provides a traditional description of the

characteristics of the nartaka and nartaki or the male and female dancers

(6.3. 9-33), but adds a personal note to it by stating that the nartakis who

come from Lata, Gurjara, Saurashtra and Maharastra are the best

(uttama), presumably in terms of competence, while the dancers from the

central and northern regions are mediocre (madhyama). The worst

(adhama) dancers, according to him, come from Karnata, Dravida and

Andhra (6.3. 33).211 The author then discusses at length the appearance

and the qualities of a nartaki (6.3. 34-39). Next he describes how to begin a

performance by bowing to Siva (6.3. 40-45), and gives the number of the

drummers needed, naming the different drums as hudukkā, kāhala,

karadā, cancu ( 6.3.45-48). For initiating the performance the author

recommends a set of six ragas, beginning with sri or dhruva and following

up with sri, madhyama, nata, karnatgolaka and chayānāta (6.3.48-50). In

the same subsection the terms for the movements of anga, upanga and

pratyanga are mentioned, as are the sthanas, caris, mandalas, padas,

211 This seems surprising in view of the widespread cultivation of the art, which can be

inferred from the fact that at least from the medieval period a number of manuals

were produced from the south and were still being produced when this text was

written. Further, we find references to highly developed dance styles from those

regions. See Raghavan, 1965, Introduction, pp. 18-34.

102

THE LITERATURE OF DANCE

angahāras and recakas. Departing from the usual practice of dance manuals, the author also includes in this section a brief discussion of the method of teaching dancing. All of these the author promises to discuss in the following sections (6.3.50-56).

The fourth taranga deals with the movements of the head and its parts in 114 verses. The fifth taranga describes, in 101 verses, hand-gestures and the movements of the arms and prescribes their applications. Unlike other works, the Śivatattvaratnākara first describes the nrttahastas or the hand-gestures specially meant for dancing, and then describes the others, that is, the single and double hand-gestures. The sixth taranga discusses the rest of the major and minor limbs ( 6.6.1-43), finishing with the movements of the feet and moves on to describing sthānas ( 6.6.43-85). In discussing the movements of the limbs the author mostly follows the Nātyaśāstra and Sangītaratnākara.

The author mentions deśī sthānas as they appear in the Sangītaratnākara but does not describe them. Cārīs and mandalas are discussed next but no deśī varieties (6.6.86-99) are given. The author at this point expresses concern about the length of his work and says that there are 108 karanas but does not name them. However, he states that 16 of these karanas are better known than the others; we may recall that in the Nartananirnaya, Punḍarīka describes a set of 16 karanas that constitute bandhanrtta (6.6.100; NN. 32b). Whether the 16 karanas considered by Bāśavarāja as the most prominent were the same as the karanas required for bandhanrtta cannot be known but the coincidence is worth noting. As we shall see in a later chapter, the bandhanrtta described in the Nartananirnaya is very likely closely related to the style now known as Odissi. If indeed the 16 karanas selected by Bāsavarāja were elements of bandhanrtta, then there is reason to believe that this style was practiced in Western as well as Eastern India at the time this author wrote.

The Sangītasārasangraha is a late work by a "Bengali Vaiṣṇava lyric poet and composer Narahari Cakravartī, alias Ghanaśyāmadāsa" who

MOVEMENT AND MIMESIS

103

lived in the early nineteenth century.212It is in six chapters which deal

mostly with music, and as Nijenhuis points out, it is a compilation of

materials drawn from older works on music from the eastern part of India,

mainly from Orissa and Bengal. The text offers very little to add to our

knowledge of the concept or descriptions of movements or sequences. But

it does show that informed interest in certain dance styles continued from

the medieval times till the early nineteenth century in Eastern India.

The chapters deal with the following subjects: vocal music,

instrumental music, dance and drama, body movements, language and

metre. Of these, chapters 3 and 4 are of interest to us. The third chapter

(pp. 63-69) starts with a benediction, relates the origin of nāṭyaveda,

which the author refers to as the fifth veda, and mentions the division of

nartana (dance) into nāṭya, nṛtya and nṛtta. These terms as well as mārga

and deśī are explained with quotations from Kohala. Ten mārganāṭyas,

sixteen deśīnāṭyas and twelve nṛtyas are mentioned, the information being

taken from the Saṅgītanārāyaṇa213Nṛtta is divided into the usual three

categories, namely, viṣama, vikaṭa and laghu. Tāṇḍava and lāsya, which

the author says can be both nṛtta and nṛtya, are discussed in detail along

with their varieties. Tāṇḍava and lāsya are referred to as dances for men

and women respectively. Tāṇḍava has two varieties, preraṇi and

bahurūpa, and lāsya has sphurita214 and yauvata; these are taken from the

Saṅgītadāmodara. The author quotes from the Saṅgītadāmodara,

Saṅgītakaumudī and the Saṅgītasāra to describe these dances. He gives a

list of dance pieces from the late medieval period, which includes kāṣṭhā,

jākaḍī, śabda, karaṇījī and mattāvalīnṛtya. These again are from the

Saṅgītanārāyaṇa.215 Jākaḍī and śabdanṛtya are found in other texts as

212 Nijenhuis, 1977, p.35; Katz, 1987, Introduction, pp. xxii-xiii.

213SN. 3. 12-20; SSār.Sam. pp.63-69.

214 SDām. has churita, p. 69.

215SN.3. 31-36; SSārSam. pp.63-69.

104

THE LITERATURE OF DANCE

well.216 In describing dance movements the author mainly follows the

Sangītadāmodara and the Sangītaratnākara. He quotes from Kaumudī

(i.e., Sangītakaumudī) to say that different regions have different varieties

of nṛtya according to the taste of the particular region.217

The fourth chapter (pp. 70–90) discusses aṅgahāra, by which the author

means aṅgas, upaṅgas and pratyangaas, or the major and minor limbs. In

this he follows the Sangītadāmodara.218 The applications of some of

these movements are also described. The author finishes the chapter by

mentioning the five components of a dance which are sthānaka, cārī,

karaṇa, maṇḍala and aṅgahāra; the term aṅgahāra, in this context, stands

for the basic dance sequence. He does not discuss them in detail since he is

concerned about the length of the work.

No survey of Sanskrit treatises on dancing can hope to be entirely

comprehensive since many of the earliest works are lost. As the present

survey has indicated, the lost work by Kohala was clearly a source of

abiding authority on which contemporaries as well as later writers

continued to draw. The medieval period has fared better, for most of the

treatises written in that period are extant and most are available in modern

editions. In the later period a number of manuals were produced, some in

Sanskrit and some in the vernacular languages, many of which are still

unedited and available only as manuscripts scattered through libraries all

over India and a few in libraries outside India. Raghavan has described

some of these manuals on dancing in his survey of saṅgīta literature in

Sanskrit, and some vernacular works in the introduction to his edition of

the Nṛttaratnāvalī.219

216 NN. 47b–48a (śabdanṛtta), 53a (Jakkaḍī); SDar.7. 268–71 (Jakkaḍī); SMakV. ch.2.

217 SN.3.37.

218 SSārSam. .pp.70–90.; SDām. .pp. 61–9.

219 Raghavan, 1956–61; NR.1965, Introduction, pp. 18–34.

MOVEMENT AND MIMESIS

105

The present survey covers the majority of Sanskrit works that directly

deal with dancing, leaving out only those from the late period (16th -17th

century) that fail to provide new information or to show any originality of

approach. Also omitted, for fear of needless duplication, are certain

eighteenth century texts that have been exhaustively described by Kapila

Vatsyayan. Her survey, as earlier mentioned, may be used to great

advantage to complement the present survey.220 The present survey has

also left out works in regional languages since they do not form part of the

Sanskrit tradition that is the subject of this study. But the considerable

number of these vernacular dance manuals produced in the late medieval

times and through the nineteenth century in Andhra, Tamilnadu, Kerala,

Orissa, Bengal and Manipur attest to the importance and popularity of the

art.221

The relevance of dance treatises is not limited to dance historians and

critics but extends to the actual dancers. At the present time, practitioners

of each of the modern styles claim to have their own manuals which they

follow, some written in Sanskrit some in the languages of the regions

where they originated. The Bharatanatyam dancers follow the text of the

Abhinayadarpana of Nandikeśvara for learning the movements of the

body, Odissi dancers seem to follow the Abhinayacandrikā of Maheśvara

220 Vatsyayan, 1968, pp.32-37, has used the Nātyaśāstra of Bharata, the

Abhinayadarpana of Nandikeśvara, the Sangītaratnākara of Śārṅgadeva, the

Sangītamakaranda of Vedaśūri, the Nātyaśāstrasaṃgraha of Govindācārya, the

Hastalaḳṣaṇadīpikā (followed by the Kathākali dancers), the Hastamuktāvalī of

Śubhaṅkara(followed by the Manipurī dancers) and the Balarāmabharata of

Balarāma Varmā (followed by the Mohinīāṭṭam dancers) to compare the theories

offered. The present researcher has given an account of the Nāṭyaśāstrasaṃgraha in

her earlier studies (1964, 1970). This text is left out f rom the present study. The

Hastalaḳṣaṇadīpikā and the Balarāmabharata are left out since Vatsāyayan's book

provides the details of their content.

221Patnaik, 1971; Classical and Folk Dances of India, 1963, part i, 'Bharatanatyam,' p.

29; part ii, 'Kathakali,' p. 10; part iv, 'Odissi,' p.20; part v, 'Manipuri,' p.39.

106

THE LITERATURE OF DANCE

Mahāpātra, Kathak dancers refer to Nartanasarvasva as their guidebook (though no copy of it has been found so far), and Kathakali and Mohinīattam dancers follow - among other manuals - the Hastalaksanadīpikā and Balarā̄mabharatam as their guidebook.222 This reliance of dancers on the Śāstra of dancing is a product of the tradition of discourse established in the Sanskrit treatises, which has resulted in simultaneously providing theoretical exposition and practical guidance.

The literature of dancing surveyed here is important as the major repository of information but more important still as the only record and the only measure of the changes through which dancing has evolved in India. In terms both of techniques and concepts we see the phenomenon of steady change in the art and it is only by studying the treatises in their chronological sequence that we may trace the details of that evolution. The present survey thus forms the necessary background to the study undertaken here of the particular features and underlying principles of classical Indian dancing. To understand these features and principles, this study will closely examine the terms, relating to both concepts and techniques, found in the texts of the subject, which are charted below.

222The sources for the various styles are as follows. Kathak: 'Kathak,' Vatsyayan, 1956, pp. 74-88, and 'Bhārater Śāstrīya Nṛtyakalā,' Kothari, 1987, p. 60; Mohinīattam: 'Mohiniaattam,' Rele, 1981, pp. 69-70; Kathakali: 'The Tradition: A Brief Historical Survey,' Khokar, 1963, pt. II, p. 10; Odissi: 'A Glimpse Into Odissi Dance,' Miśra, 1981, p. 90; Bharatanātyam: 'Śāstra and Prayoga: The Use of Abhinayadarpaṇa,' Bose, 1988.

MOVEMENT AND MIMESIS

107

Work

Date(approximate)

Author

Nāṭyaśāstra

2nd/ 3rd c. A. D.

Bharata

Viṣnudharmottara Purāṇa 5th c. A. D.(section on dance)

Abhinavabharatī

10th-11th c. A.D.

Abhinavagupta

Daśarūpaka

10th c. A.D.

Dhanañjaya

Śṛṅgāraprakāśa

11th c. A.D.

Bhoja

Nāṭyadarpana

12th c. A.D.

Rāmacandra and Gunacandra

Mānasollāsa

12th c. A.D.

Someśvara

Nāṭakalakṣaṇaratnakośa

12th/13th c. A.D.

Sāgaranandin

Bhāvaprakāśana

around13th c. A.D.

Śāradātanaya

Sañgītasamayasāra

12th-13th c. A.D.

Pārśvadeva

Sañgītaratnākara

13th c. A.D

Śārṅgadeva

Nṛttaratnāvalī

13th c. A.D

Jāya Senāpati

Abhinayadarpana

around13th c. A.D.

Nandikeśvara

Sañgītamakaranda

13th/14th c. A.D.

Nārada

Sāhityadarpana

14th c. A.D.

Viśvanātha Kavirāja

Sañgītopaniṣatsāroddhāra

14th c. A.D.

Sudhākalaśa

Sañgītacandra

14/15th c. A.D.

Vipradāsa

Sañgītadāmodara

15th c. A.D.

Subhaṅkara

Hastamuktāvalī

15th c. A.D.

Subhaṅkara

Nṛtyādhyāya

14th/15th c. A.D.

Aśokamalla

Nṛtyaratnakośa

15th c. A.D.

Mahārāṇā Kumbhā

Bharatarṇava

16th c. A.D.

Nandikeśvara

Nartananiṛṇaya

16th-17th c. A.D.

Puṇḍarīka Viṭṭhala

Rasakaumudī

16th-17th c. A.D.

Śrīkaṇṭha

Sañgītadarpana

17th c. A.D.

Dāmodara

Sañgītanārāyaṇa

17th c. A.D.

Puruṣottama Miśra

Sañgītamakaranda

17th c. A.D.

Vedasūri

Śivatattvaratnākara

17th/18th c. A.D.

Bāsavarāja

Sañgītasārasaṁgraha

19th c. A.D.

Ghanaśyāmadāsa

Chapter 3

THE NĀṬYAŚĀSTRA AND THE CONCEPT OF DANCE

The Nāṭyaśāstra of Bharata is not only the earliest but also the most extensive work on all aspects of drama, dance and music. Its influence has been so great on all subsequent works on the performing arts that not one of the many treatises on dancing has attempted to describe dancing without adhering to the framework of discussion laid down by Bharata and often borrowing substantially from his work. That the impact of the Nāṭyaśāstra is so deep and permanent is due to the fact that it combines theory and practice, formulating concepts of dance as it describes its technique. These concepts must be closely examined since they have remained central to all discussions on dancing.

These concepts may be set in the context of Bharata’s comprehensive view of drama, dance and music by noting the place that Bharata accords these arts in the social structure. According to him, the drama is the fifth Veda, which was created by Brahmā for those castes that did not have access to the Vedas proper:

न वेदव्यवहारार्हायैः संश्राव्यः शूद्रजातिषु ।

तस्मात् सृजापरं वेदं पञ्चमं सार्ववर्णिकम् ॥

एवमस्त्वति . . . .

(NS. 1. 12-13)

The Vedas [lit: the use of this Veda ] are not to be heard by [lit:in] the śūdra castes. Therefore [please] create another Veda, the fifth, meant for all the varṇas. "Let it be thus," [said]. . . .

108

109

MOVEMENT AND MIMESIS

Bharata calls this fifth Veda the nāṭyaveda: nāṭyākhyam pañcamam

vedam . . . (NŚ. 1. 15).

Dance is seen by Bharata as an introduction to dramatic performance.

It was Śiva, he says, who prescribed the addition of dance to the

preliminaries of a dramatic performance in order to add beauty to it.

मयापीडं स्मृतं नृत्तं1 सन्ध्याकालेषु नृत्यता ।

नानाकरणसंयुक्तैरर्हारैरविभूषितम् ॥

पूर्वरङ्गविधावस्मिन्स्त्वया सम्यक् प्रयोज्यताम् ।

(NŚ. 4.13 -14)

And I have also propagated this [art of] dancing

embellished with aṅgahāras consisting of a number of

karaṇas whilst performing dances in the evening.

Apply this art (i.e., dance) in an appropritatie manner

in the rites of these preliminaries of [a drama].

Although references to dancers and dance are scattered through early

literature,2 it was Bharata who first treated dance systematically,

defining it as an art-form which consisted of beautiful movements of the

body performed to rhythm and to vocal or instrumental music or to both,

and which may be mimetic or purely decorative. This definition rests upon

five major terms: nṛtta, tāṇḍava, piṇḍībandha, abhinaya and

sukumāraprayoga. Although most of these are familiar terms, the

concepts behind them have been obscured by time and often by confusing

1 Nṛtta is the term, not nṛtya, that Bharata uses consistently throughout his work to denote

dancing. The term nṛtya appears only twice in the G. O. S. edition of the NŚ (4.13, and

30), and seems to have been a scribal error. See Bose, 1970, p.8; Varma, 1957, p.21.

The Kāśī edition has nṛtta, which supports this view. Nṛtta and nṛtya are different

concepts and the latter did not become current till the tenth century A.D.

2 See Bose, 1970, pp. 1-4.

110

THE NĀṬYASĀSTRA AND THE CONCEPT OF DANCE

commentary. To understand the tradition of art described by Bharata, one

must clarify the concepts.

Bharata's most significant statements on dance occur in the fourth

chapter of the Nāṭyāstra, which is devoted entirely to this subject.

Dance, we are told, was created for beauty's sake: शोभां प्रजनयेदिति नृत्तं

प्रवर्त्तितम् ।3 The terms he uses for this dance are two-nṛtta and tāṇḍava:

रेचकाङ्गहाराश्च पिण्डीबन्धास्तथैव च ॥

सृष्ट्वा भगवता दत्तास्तण्‌डे मुनये तदा ।

तेनापि हि ततः सम्प्रगानभाण्डैः समन्वितः ॥

नृत्तप्रयोगः सृष्टो यः स ताण्‌डव इति स्मृतः ।

(N.S. 4. 259-61)

Recakas, aṅgahāras and the piṇḍībandhas were

created by god [Śiva] and then given to the sage

Taṇḍu. That method of dancing which was then

created by him [=Taṇḍu] accompanied by appropriate

songs and drums, is known as Tāṇḍava.

Taken as a karmadhāraya samāsa, the compound nṛttaprayoga in the

last line of this passage equates tāṇḍava with nṛtta.4 Nṛtta is described as

an art-form which is beautified with aṅgahāras made of various karaṇas.

He then describes these karaṇas and aṅgahāras in detail: हस्तपादसमायोगो

नृत्तस्य करणं भवेत् । - the coordinated movement of the hands and feet

makes the karaṇa of a dance.5 A karaṇa is the basic unit of dancing and

the movements that constitute karaṇas are clearly prescribed. A

combination of karaṇas, numbering six to nine, makes an aṅgahāra, which

3 N.S. 4. 264.

4 ताण्‌डवमिति सर्व नृत्तमुच्यते । (AB. on the N.S. G.O.S ed., 1956, Vol. I, p. 180).

5 N.S. 4.30

MOVEMENT AND MIMESIS

111

can be described as a basic dance-sequence.6 Bharata states that such

sequences beautify a dramatic performance when they are included in the

pūrvarañga or preliminaries, which can be presented with or without

dance. When there is no dance but songs are sung, a pūrvarañga is known

as śuddha or pure. The kind that has dancing in it is called citra or mixed.7

In this fourth chapter Bharata deals with the citra type of pūrvarañga. In

the same chapter Bharata says that sthānas, cārīs and nṛttahastas

prescribed in the discussion of vyāyāma or exercise (in the chapters on

āṅgikābhinaya, chapters eight to twelve) should be applied to a karaṇa.8

Sthānas are the basic postures or stances in a dance.9 Cārīs are executed

with one leg: एकपादप्रचारो यः सा चारीत्यभिसंज्ञिता10 The different

gestures of hands meant solely for the use in nṛtta are known as

nṛttahastas: करणेऽ तु प्रयोक्तव्या नृत्तहस्ता विशेषतः11 -- in a karaṇa

nṛttahastas are to be applied in particular.11 Bharata describes these

karaṇas and aṅgahāras in detail and then describes another movement

called recaka. Recakas, defined as the basic movements of the feet, hips,

hands and neck, seem to be quite fundamental to dancing, for Bharata has

included them in his chapter on dance.12 He states his view clearly about

the nature of nṛtta by saying that aṅgahāras, recakas and piṇḍībandhas

done together in an appropriate manner form nṛtta (NŚ. 4. 259-61).

After listing and discussing the constituent movements of dancing,

Bharata returns to the structure of the pūrvarañga as a whole. Here he

6 NŚ. 4.33.

7 NŚ. 4.15-16.

8 NŚ. 4. 169 -71.

9 NŚ does not define sthāna as such but from the descriptions of the sthānas in chapter ten

of the NŚ the movement can be understood (NŚ. 10. 50-71).

10 NŚ.10.3.

11 NŚ.. 9.210.

12 NŚ. 4.248- 49.

112

THE NĀTYAŚĀSTRA AND THE CONCEPT OF DANCE

introduces another term, piṇḍībandha. Piṇḍībandhas are group dances that constitute a distinct phase of the preliminaries following the introductory dancing and they are of four types: piṇḍī, latā, śṛnkhalā, and bhedyaka (NŚ. 4. 287-89). The purpose of these dances is to invoke the blessings of the gods. Bharata relates how, after disrupting Dakṣas's sacrifice ceremony, Maheśvara danced in the evening with aṅgahāras set to rhythm and tempo. Evidently, Bharata is taking it for granted that Śiva was accompanied in his dancing by his followers, and that this performance created a number of group dances or piṇḍībandhas, for Bharata goes on to say that Śiva's gaṇas or disciples, such as Naṇḍī and Bhadramukha, saw the piṇḍībandhas and named the variations of one type of piṇḍībandha known as piṇḍī, that is, a cluster. They named each variation of the cluster after a god or a goddess. These piṇḍīs, named in this way by Śiva's disciples, are then listed by Bharata who adds that other piṇḍīs should be similarly named after other gods or goddesses. Emphasizing the mythological status of Taṇḍu as the archetypal dancer among mortals, Bharata tells us that recaka, aṅgahāra and piṇḍībandhas were created and given to Taṇḍu and it was thus that with songs and drums the art of dance was created. While describing the performance of tāṇḍava he again elaborates on piṇḍīs.13

The presentation of the preliminaries seems to be an elaborate performance which calls for, beside the singers and drummers, one principal female dancer and a group of other female dancers. The group of dancers is required to enter and dance in a variety of formations which are known as piṇḍībandhas. Such formation dancing is prescribed only for female dancers and considered appropriate only for this particular part of the entire stage presentation. Piṇḍībandhas are dedicated to different gods who are denoted by their emblems, which are represented by the formations created by the dancers. In addition, the dancers form into

13 NŚ.4.279-91.

MOVEMENT AND MIMESIS

113

ritualistic diagrams. Bharata states that in order to be able to create such

formations in an appropriate manner, these dancers require careful and

thorough training.14 Piṇḍībandhas thus have a auspicious purpose in the

main, to invoke the blessings of the gods, and that Bharata should include

them as necessary parts of the stage-presentation adds to our

understanding of the function of dancing as an art in Bharata's time.

Elaborating on the piṇḍībandhas, Bharata says that different yonis,

bhadrāsana and yantra are to be applied in composing piṇḍībandhas. The

use of the first two refers to the formation of auspicious diagrams.

Yantra is a term for mystical diagrams in tāntric and Buddhist usage. If

we take the word in that sense then it fits in with yoni and bhadrāsana as

the name of a figure formed by a group of dancers. Raghavan, however,

takes yantra to mean mechanical aids in his discussion of the uparūpakas

of the Jaina tradition. The editor of the Saṅgītopaṇiṣatsāroddhāra gives a

list of thirty-two varieties of dance dramas which includes bhadrāsana.

These are from the Jainasūtra named Rājapraśnīya, as noted also by

Raghavan. Following this text, Raghavan has described this bhadrāsana

as the "execution of rows or garland-like formations made up of groups

resembling moon, sun, swans, stars, pearls, gems and so on."15 In

Abhinavagupta's commentary we find references to a formation like a

stick grasped in the beak of a swan.16

Bharata divides piṇḍībandha into four classes:

पिण्डीनां विधयश्चैव चत्वारः सम्प्रकीर्तिताः ॥

पिण्डी शृङ्खलिका चैव लताबन्धोऽथ भेद्यकः ।

पिण्डीबन्धस्तु पिण्डत्वाद् गुल्मः शृङ्खलिका भवेत् ॥

जालोपनद्धा च लता सन्नृतो भेद्यकः स्मृतः। (NŚ. 4. 287-89)

14 NŚ. 4. 252- 59; 4. 287-91.

15 Raghavan 1963, pp. 572- 74. See also SUS.,1961, Introduction,xvi-xvii.

16 AB. on NŚ. 1956,Vol.I.,p.191.

114

THE NĀṬYASĀSTRA AND THE CONCEPT OF DANCE

The practice of piṇḍīs can be of four types:

piṇḍī,

ṣṛṅkhalikā,

latābandha and bhedyaka.

Piṇḍībandha is

[lit.: from] a roundish mass,

ṣṛṅkhalikā is [like a]

cluster,

latā is entwined [lit.: bound] in a net and

bhedyaka is known to be [performed with] ṇṛtta.

The precise nature of the formations is not easy to understand from this

passage.

Abhinavagupta has a fairly long commentary on it which is not

much clearer.

17

Attempting a better explanation,

V.

Rādhavan in his

introduction to

Nṛttaratnāvalī

has suggested an improved reading of the

edited version of the commentary on piṇḍībandhas as given in

Ramakrishna Kavi's edition of the

Nāṭyaśāstra.

The passage in Kavi's

edition is:

नर्तकీयोज्य:

परस्परसम्बन्ध

एव

पिण्डीबन्धद्रयप्रकार:

सजातीयो

वा

एकता(ना)लावबद्धकमलयुगलवत्

विजातीयो

वा

हंसवदनपरिगृहीतानालनलिनवत्

गुल्म:

शृङ्खलिकाशब्दवाच्य:

।

नर्तकीतयप्रयोज्यस्तु

ततोडपि

वैचित्र्यसहिष्णुत्वाज्जालवद्विचित्रतां

गच्छत्पूर्ववत्सजातीयविजातीयात्मा

लताबन्ध:

चतुष्टयप्रयोज्यस्तु

.

.

.

Raghavan emends it thus:

नर्तकీयोज्य:

परस्परसम्बन्ध

एव

पिण्डीबन्ध:

सजातीयो

वा

एकनालबद्धकमलयुगलवत्,

विजातीयो

वा

हंसवदनपरिगृहीतानालनलिनवत्

।

गुल्म:

शृङ्खलिकाशब्दवाच्य:

नर्तकीतयप्रयोज्यस्तु।

ततोडपि

17

Ibid.

MOVEMENT AND MIMESIS

115

वैचित्र्यसहिष्णुत्वात् विचित्रतां गच्छत्

सजातीयविजातीयात्मा लताबन्धश्रृङ्खतुष्ठयप्रयोज्यस्तु . . . 18

The employment of female dancers joining with one another [in a dance figure] is Piṇḍībandha, the symmetrical [form of the figure] like a pair of lotuses joined in one stalk, the assymertic like a lotus stalk held in the beak of a swan. Gulma is expressed by the word chain to be performed by three female dancers.

Through its capacity for diversity, latābandha of symmetrical and asymmetrical nature, [and] demonstrated by four [female dancers] becomes more entertaining . . . .

In his study of Śṛṅgāraprakāśa, Raghavan further attempts to explain the characteristics of piṇḍībandhas by citing Bhoja. Bhoja, he says, defines the piṇḍībandhas as group dances used in presenting the rāsaka, nāṭyarāsaka and lāsya types of minor drama (that is, padārthābhinayātmaka preksyaprabandha).19

We may recall that in Bharata's account piṇḍībandhas are parts of the preliminaries to a dramatic performance (as is all dancing), not of the presentation of the dramas themselves. Neither Bhoja's mention of piṇḍībandhas nor Raghavan's commentary throws any more light on the nature of piṇḍībandhas and all we know to any degree of certainty is that these were group dances that were used in Bharata's time in the citra type of pūrvaraṅga but had come to be part of the technique of minor dramas by Bhoja's time.

As we shall find later in the course of discussing the minor types of drama such as rāgakāvyas, ṇṛttakāvyas, ṇṛtyabhedas and uparūpakas, these group dances formed a major part of dramatic

18 Ibid. 191; NR.1965, Introduction, pp.143.

19 Sr.P. 1963, pp. 563; 588-89.

116

THE NĀṬYASĀSTRA AND THE CONCEPT OF DANCE

presentation for centuries. Even today some of them survive, as for

instance in the rāsa dance of Gujrat, Rajasthan, Vrindavan and Mathura

and in Kathak and Manipuri.

In describing nṛtta, tāṇḍava and piṇḍibandha Bharata is concerned

with the process of dancing as described in terms of the physical structure

of movements. Our understanding of the process is augmented by two

other concepts that deal with the affective quality of the art, namely,

sukumāraprayoga and abhinaya.

The concept of sukumāraprayoga is invoked in the context of Śiva's

dance. According to the legend cited by Bharata, Śiva's dance, which

comprises aṅgahāras and recakas, inspires Pārvatī who demonstrates a

different way of dancing by employing graceful and delicate movements:

रचकैरङ्गहारैश्च नृत्यन्तं वीर्यशोभनम् ।

सुकुमारप्रयोगेण नृत्यन्ती चैव पार्वतीम् ।

(NŚ. 4. 249-50)

On seeing Śaṅkara dancing with recakas and

aṅgahāras and on seeing Pārvatī dancing with

delicacy . . .

Beyond the use of the adjective sukhumāra Bharata does not provide any

clue to the nature of Pārvatī's dance and goes on to describe the whole

sequence of the dance that is included in the pūrvaraṅga along with

drumming, singing etc. Abhinavagupta is not much help either. All he says

is that since Saṅkara's dance lacked grace, Pārvatī demonstrated that

quality by giving a performance that was characterized by delicate

movements.20 This implies that Pārvatī's dance falls into a separate

category. However, Pārvatī's dance is different not because she uses

movements different from those of Śiva, but because she uses the same

20 AB. on NŚ. 1956, Vol. I., p. 164.

MOVEMENT AND MIMESIS

117

movements more delicately. The difference is thus stylistic not

substantive.

The significance of sukumāraprayoga is not entirely clear in this

passage. Later on in his discourse Bharata does offer some explanation of

sukumāra when he says:

स्त्रीपुंसयostu संलापो यस्तु कामसमुद्भव: ।

तज्ज्ञेयं सुुकुमारं हि श्रृङ्गाररससम्भवम् ॥

(NŚ. 4. 303)

The conversation, which is generated from the

passion between a man and a woman is known as

sukumāra that arises from the erotic sentiment.

Such a sentiment, he says, can be represented in a performance by

applying the aṅgahāras shown by Pārvatī: देवीकृतैर्झर्झहारैरलितैस्तत्

प्रयोजयेत् । -- which is to be applied with delicate aṅgahāras created by the

goddess (=Pārvatī).21 The emphasis here is on the affective function of

sukumāra. Necessarily, then, by relating sukumāra to dancing Bharata is

shifting conceptual areas, from the structure to the function-at least one

function-of dancing. By itself Bharata's statement does not clarify the

nature of the application. Abhinavagupta tries to clarify the nature of

sukumāra by giving some examples from the rāgakāvyas which later, in

the medieval period, came to be known as uparūpakas or minor types of

drama. From the examples of the rāgakāvyas given by Abhinavagupta one

may say that these required graceful and delicate movements. He further

cites some aṅgahāras which employ less vigorous movements as examples

of delicate aṅgahāras22 but no other detail can be found. If one tries out

these so called aṅgahāras, the delicate nature of the movements can be

21 NŚ. 4. 312.

22 AB. on NŚ. 1956, Vol. I. p. 201.

118

THE NĀṬYASĀSTRA AND THE CONCEPT OF DANCE

perceived, although it is difficult to fully understand through verbal

description an art-form which relies entirely on visual experience for its

effect. That is why sukumāraprayoga and its nature gave rise to a great

deal of confusion in the later texts concerning what constituted the

delicacy of movements. In the next chapter of the present study this

confusion has been discussed in detail.

The other concept that relates to the quality rather than the structure of

dance is the vital one of abhinaya. The term abhinaya denotes the method

of expressing ideas through gestures and is explained in the Nāṭyaśāstra as:

. . . प्राप्त्यर्थमथानां तज्ज्ञैरभिनय: कृत: ।

(NŚ. 4.261.)

. . . abhinaya is created by the experts to understand and the

meaning [lit.when the meaning is aimed at] [of the

kāvya].

Bharata attempts to address questions asked by the sages who seem

puzzled when they find that nṛtta has no connection either with the total

meaning of the accompanying songs, nor even with the meaning of

individual words or sentences. They ask again, why then was nṛtta

created?

न गीतकार्यसम्बन्धं न चाप्यर्थस्य भावकम् ।

कस्मात् नृतं कृतं . . .

(NŚ. 4.262-63)

[Nṛtta] does not have any connection with the

meaning of the song nor does it express the meaning of

the [words]. Why is nṛtta created?

To this Bharata's answer is:

. . . . . न खल्वयं कांचिन्नृततमपेक्षते ॥

MOVEMENT AND MIMESIS

119

किं तु शोभां प्रजनयेदिति नृत्तं प्रवर्तिततम् ।

(NŚ. 4.262-63).

..... nṛtta requires no meaning at all but [it was] to

produce beauty [that] nṛtta was created.

Abhinaya is explained by Bharata not only as a particular mode of

performance but also as a distinct phase of stage-presentation. We have

seen how in the preliminaries the aṅgahāras are performed at first and

then the piṇḍibandhas. To perform the aṅgahāras the main nartakī enters

and dances with pure karaṇas. She then performs abhinaya to a song

without the accompaniment of drums:

यत्राभिनेयं गीतं स्यात् तत्र वाद्यं न योजयेत् ।

(NŚ. 4.276)

Where a song is to be acted drumming should not be

applied.

She first performs nṛtta and then abhinaya, after which she makes an

exit. Then female dancers enter and make different group formations.

Bharata says that until the piṇḍīs are formed, paryastaka (i.e., body

movements that do not express ideas) is performed by these dancers.23

Bharata describes in the fullest detail the entire sequence of this

performance of the preliminaries with songs, instrumental music, tāla etc.

He then clarifies his concept of nṛtta, tāṇḍava and abhinaya.

Bharata's statement that nṛtta has no connection either with the total

meaning of the accompanying songs, nor even with the meaning of

individual words or sentences leaves unanswered the question whether

nṛtta has any function of communication. Here we are confronting a basic

problem of art, that of its semantic purpose. David Smith has tried to give

23 Ibid. p.188: अभिनयगूनमा(आ)कृतिबिधि: पर्यस्तक: ।

120

THE NĀṬYASĀSTRA AND THE CONCEPT OF DANCE

an answer to this question posed by the sages while discussing tāṇḍava in

his study of Haravijayam. He views nṛtta as a superior art-form that does

not look for any purpose: "The dance does not imitate anything in real life,

but is a self-subsistent creation free from any practical aim. It is the

natural expression, through the movements of the limbs, of a given state of

mind. . . . Even if Śiva dances abhinaya, mimetic dance, there are still no

grounds for enquiry into the meaning of this dance, since dancing Śiva is

all there really is."24 In support of this point of view Smith quotes the

following verse:

करणाङ्गहारविधिभिः सविस्तरैः

सकलासु शृङ्गार ! निशासु नृत्यता ।

क्रियते त्वयानुकृतिरात्मनो विभो !

सच्चराचरं जगदव्ये तस्थुषः ॥

(Haravijayam, 6. 180)

O beneficent Śiva !

every night you dance

with the whole range of karaṇas and aṅgahāras,

yet, O Lord!

there is nothing for you to imitate

but yourself

for you pervade, you are, the entire universe.25

Given the self-subsistent nature of Śiva that Smith points out, the

question of meaning (or the lack of it) seems beside the point. However,

the question of purpose still remains. In fact, Bharata attributes purpose to

nṛtta. This is clear in the distinction he draws between the dance as

demonstrated, respectively, by Śiva and Pārvatī. The dance taught by Śiva

24 Smith, 1985, p. 250.

25 Ibid. p. 251.

MOVEMENT AND MIMESIS

121

to Tanḍu, that is, tāṇḍava, aims at devasvuti, that is, praise-offerings to the

gods. It does so because dancing of this type creates beauty of form and is

like any other beautiful object, such as flowers, that are offered to the

gods. Thus, tāṇḍava has purpose but no meaning because it is not

conceived by Bharata as a vehicle of feelings or emotions.

Emotional signification seems to be the business of the dance

demonstrated by Pārvatī. The respective functions of the two styles are

quite explicitly stated :

देवस्तुत्याश्रयकृतं यदृक्षं तु भवेदथ ।

माहेश्वरैरङ्गहारैरैस्तत्तत् प्रयोजयेत् ।

यत् तु शृङ्गारसम्बन्धं गानं स्त्रीपुरुषाश्रयम् ।

देवीकृतैरङ्गहारैरैलिङ्गितैस्तत् प्रयोजयेत् ॥

(NŚ. 4. 312)

Whatever part [of a play; here, the pūrvarañga] is for

praising gods, that [one] should perform with the

vigorous aṅgahāras created by Maheśvara, while a

song related to erotic sentiment involving [lit.:

depending on] man and woman should be performed

with the aṅgahāras created by Devī.

While tāṇḍava is exclusively intended for worship as an offering

valued for the beauty of its form, Pārvatī’s dance is used for human

communication and is valued for its ability to express emotions. As we go

from Śiva’s dance to Pārvatī’s, we find the emphasis shifting from the

metaphysical to the material, from form to content. Bharata 's concept of

sukumāraprayoga sets Pārvatī’s kind of dancing in a context of human

relations where सुकुमारप्रयोगश्च शृङ्गाररससम्भव:-

the application of delicate (movements) generates the erotic sentiment (NŚ. 4.269). This

does open up the possibility of the expressive use of nrtta which is seen in

the references to the application of delicate movements in the dramatic

122

THE NĀṬYASĀSTRA AND THE CONCEPT OF DANCE

literature of all periods. The dramatic potentiality of sukumāraprayoga

brings dance movements within the orbit of acting or abhinaya.

In speaking of dancing while discussing pūrvarāṅga in the fourth

chapter, Bharata prescribes two styles of presentation: uddhata and lalita.

Uddhata aṅgahāras are used for devastuti in a pūrvarāṅga. In a

pūrvaraṅga, lalita or delicate movements are employed to depict the

feelings between a man and a woman. This distinction between the styles

that are appropriate, respectively, to the portrayal of gods and the

depiction of human beings in love is drawn again in the discussion of the

pūrvarāṅga in the fifth chapter. There Bharata says:

शृङ्गारस्य प्रचारणाच्चारी सम्प्रकीर्तिता ।

रौद्रप्रचारणाच्चापि महाचारीति कीर्तिता ॥

(NŚ. 5.27)

Cārī is known to manifest śṛṅgāra [rasa] and mahācārī

is known to manifest raudra [rasa] as well.

It is important to bear in mind that Bharata places the entire discussion

on vigorous dancing in the context of the pūrvarāṅga, not in that of drama

proper. As the following discussion shows, the nature of abhinaya

pertaining to the dance in the pūrvarāṅga may seem confusing if

the. uddhata and lalita styles are extended to drama proper. As

Abhinavagupta explains, uddhata, which is the style of Śiva, obviously

indicates vigorous dancing with forceful aṅgahāras while lalita is the style

of Pārvatī and is characterized by delicate aṅgahāras.26 Uddhata is used

in devastuti and lalita is used in presenting love scenes.

The classification of dance into uddhata and lalita also suggests a

distinction between masculine and feminine modes of expression, first

because of their physical characteristics and second, because of their

26 AB. on NŚ. 1956, Vol. I. p. 201.

MOVEMENT AND MIMESIS

123

creation, respectively, by a male and a female deity. Though Bharata does

not explicitly correlate uddhata with manly actions or lalita with feminine

moods, the correlation seems implied in the use of the terms themselves

and has been recognized by later authors, following the lead of

Abhinavagupta. As an example of tāṇḍava Abhinavagupta cites a minor

dramatic form in which the action is performed by a masculine figure,

whereas for the lalita style he cites a dramatic form in which the

corresponding figure is feminine. The association of uddhata and lalita

with, respectively, masculine and feminine qualities is evident, though it is

stated somewhat obliquely, specially in the case of the uddhata style.

Explaining that tāṇḍava is usually directed at devas and tutelary deities

and expressed through uddhata aṅgahāras, Abhinavagupta refers to their use in bhāṇakas

(or bhāṇas, as he later calls them), which, according to him, are

rāgakāvyas. In support of his view he cites the cirantanā s, that is,

traditional authorities, according to whom avatāras such as Nṛsiṁha (the

man-lion incarnation of Viṣṇu) and the Boar are signified by uddhata

aṅgahāras. To illustrate sukumāra Abhinavagupta refers to ḍombikā,

a minor dramatic form which shows how a woman performs the actions of

flattering a king.27

Bhāṇaka and ḍombikā later came to be known as types of ṇṛtyabhedas

or uparūpakas, minor dramatic types28 in which the emphasis is on body

movements. Whether Bharata himself had actual dramas such as these in

mind is doubtful. In the first place, he is speaking of dancing as part not of

the drama proper but of the citra type of pūrvaraṅga. In the second place,

it is not known whether there were any such types as bhāṇaka or ḍombikā

in Bharata's time, for in his discussion of dramatic types he does not

include them, nor indeed the class of uparūpakas to which they belonged.

27 Ibid. pp. 180-81.

28 DR. 1969, p. 8; BhP. 1930, p. 255.

124

THE NĀṬYASĀSTRA AND THE CONCEPT OF DANCE

The genre that Bharata discusses is rūpaka, that is, major dramatic types. One such type is bhāṇa. Because of the similarity of the terms bhāṇa and bhāṇaka, one might be led, as Abhinavagupta was, to assume the use of uddhata aṅgahāras in bhāṇa. But this seems highly unlikely. Bhāṇa is a play acted by a single character. The actor can either present his own feelings or those of others by conversing with imaginary persons with suitable movements. The character considered by Bharata as appropriate for this kind of play is that of a viṭa, a knave, or a dhūrta, a rogue.29 It is hard to believe that Bharata is thinking of bhāṇa as the proper place for the use of uddhata aṅgahāras, which he considers suitable for devastuti. If bhāṇakas employed the uddhata style as mentioned by Abhinavagupta, such a use was evidently a later development for which the Nāṭyaśāstra provides no basis. Ḍombikā is nowhere mentioned in the Nāṭyaśāstra. It would seem that by Abhinavagupta's time the uddhata and lalita styles referred to genres not included in Bharata's own framework of discussion. However, because of a confusing statement in a later chapter of the Nāṭyaśāstra, it is possible to assume mistakenly that Bharata authorizes the use of uddhata and lalita in bhāṇa.30 In that chapter Bharata describes the types of lāsya which involve delicate and expressive movements and quite often use nṛtta as well. Here Bharata compares lāsya with bhāṇa. The similarity lies in the fact that both forms use one character. But the types of lāsya as described by Bharata are not such as to accommodate a character like the viṭa,31 and there is no justification for equating lāsya with bhāṇa or for attributing to bhāṇa the same qualities of abhinaya. A fuller discussion of the term lāsya is reserved for a later chapter but at this point we may note that this mainly involved female performers who often expressed their feelings with delicate and

29 NS. 18. 108-10.

30 NS. 19. 117-18.

31 NS. 19. 121-35.

MOVEMENT AND MIMESIS

125

meaningful body movements and sometimes performed pure nṛtta as well.

What is important here is that lāsya possessed qualities of both dance and

drama. Abhinaya was, therefore, specially appropriate to it.

What, then, was abhinaya thought to be? The use of the term abhinaya

is of vital importance in the Nāṭyaśāstra, which covers more than one idea

under the term. Bharata discusses abhinaya first in the fourth chapter and

then in chapters eight to thirteen. In the fourth chapter he uses the term

32

to denote expressive movements of the body in the preliminaries of a play.

Abhinaya is performed to a song where drumming is specifically

prohibited. Tāṇḍava, that is, nṛtta is performed to the beat of drums while

abhinaya is done to express the meaning of the songs. Both are used in the

preliminaries.

33

Beyond this Bharata does not deal with the nature of

abhinaya in the fourth chapter.

After the discussion of nṛtta Bharata discusses the performance of

pūrvarañga in full detail in the rest of the fourth and in the fifth chapter. In

the sixth and the seventh chapters the subject is rasa and bhāva-aesthetic

affect and emotion. The next five chapters are devoted to abhinaya. In his

eighth chapter he defines abhinaya and describes its nature. Abhinaya, he

says, can be of four kinds:

आङ्गिको वाचिकश्चैव आहार्यः सात्विकस्तथा ।

श्रेयस्त्वभिनयो विप्राश्चतुर्धा परिकीर्तितः ॥

(NS. 8. 10)

O brahmins! Āṅgika, vācika, āhārya and sāttvika-

these are known to be the four abhinayas.

32 NS. 4. 261; 276; 283; 294; 297; 300; 316.

33 NS. 4. 276-77; 300.

126

THE NĀṬYASĀSTRA AND THE CONCEPT OF DANCE

Reminding the reader that sāttvika has already been discussed in the

chapter on bhāva, here he takes up āṅgika.34 He devotes five chapters to

the body movements (āṅgikābhinaya). Every movement of each part of

the body that can be moved to beautify or express or do both is codified

here. These show the emphasis given to the body movements at the time

when this text was written.

From Bharata’s arrangement of the topics it appears that he considers

the creation of beauty in a stage-presentation as of primary importance.

That is why he discusses nṛtta first because it creates śobhā or beauty and

thus captures the mind of the audience. The next important issue in his

mind is the arousal of feelings, emotions, aesthetic pleasure and affect.

Accordingly, rasa and bhāva are the subjects of the succeeding chapters.

How are emotions and aesthetic pleasure recreated and communicated ?

Bharata answers, through the movements of the body, through glances,

hand-gestures etc. He describes in minute detail the movements of the

smallest parts of the body, such as the upper lip, lower lip, eyelid, etc., all

of which can express feelings. It is hard to believe that these movements

can express any feelings by themselves but once they are put together they

may indeed form a harmonious whole to enhance communication.

However, the most expressive of all are the movements of the hands and

the glances, according to Bharata, and he devotes most of his discussion of

āṅgika to them.35 The entire literature of dance has followed Bharata in

this. Although most of these movements are used to express meaning he

does not forget about nṛtta here. He devotes a section of hand-gestures

reserved specially for abstract nṛtta. He describes sthānas and cārīs

which are components of the karaṇas and therefore of the aṅgahāras as

well.

34 NŚ. 8.11.

35 NŚ. chapters 8 & 9.

127

Āṅgikābhinaya, according to Bharata, can be of three kinds: śārīra,

mukhaja and ceṣṭākṛta-that is, by means of the limbs, the face and the

movements of the entire body.36 He also refers to śākhā, ṇṛtta and aṅkura

as components of āṅgikābhinaya.37 By śākhā he refers to āṅgika, which he

does not elaborate upon. Abhinavagupta does not have any commentary on

the topic. It seems that the term śākhā here refers to gesticulation that

expresses meaning. Sārṅgadeva, the most influential author on dance and

music of the medieval period, supports this view.38 Ṇṛtta is again defined

as the creation of beauty by means of the movements of the body, these

movements being structured into karaṇa and aṅgahāra. Finally, the term

aṅkura, that is, pantomiming through gestures, is mentioned as the other

component.39 Concerning that abhinaya which is created through facial

movements Bharata discusses in detail the movements of the different

parts of the face as well as the application of these movements in

expressing feelings. Next he discusses the movements of the parts of the

rest of the body. Among these, hand-movements are the most important.

By moving the fingers in specific ways the meaning of words as well as

emotions can be expressed and the movements and their application are

discussed in detail. At the end of the discussion of ṇṛttahastas, that is, the

hand-movements meant for ṇṛtta, he adds a verse:

नृत्तेऽभिनययोगे च पाणिभिरवर्तनाश्रयैः |

मुखभ्रूनेत्रयुक्तानि करणानि प्रयोजयेत ||

(NŚ. 9.219)

In ṇṛtta and in the discipline of abhinaya (=nāṭya),

karaṇas should be applied with hands [moving] in

36 NŚ. 8.12.

37 NŚ. 8.15.

38 SR. 7.35-8.

39 Ghosh, 1951, p. 152.

128

THE NĀṬYASĀSTRA AND THE CONCEPT OF DANCE

vartanā along with [the movements] of the face, the

eyebrows and the eyes.

To nṛtta, then, Bharata is adding the features of abhinaya, not in order

to express meaning but to beautify the production. He directs the

performer to use the same kinds of movements in both nṛtta and abhinaya.

In abhinaya, facial action will express emotion while in nṛtta it will

enhance beauty. That the technique of abhinaya is used for both purposes

is asserted also in the discussion on hastapracāra in the ninth chapter,

which deals with āṅgikābhinaya. Bharata calls the hastas

nāṭyanṛttasamāśrayāḥ -dependent on nāṭya and nṛtta.40 The association of

techniques of abhinaya with nāṭya indicates that they express emotion,

while their association with nṛtta indicates that they serve the needs of that

art-form, that is, the creation of beauty by means of physical action.

Bharata’s view here is that abhinaya is differently used in different art-

forms. In the body of a nāṭya abhinaya is of all four kinds, namely, āṅgika,

vācika, sāttvika and āhārya, while in the pūrvaraṅga vācika is left out of

the abhinaya performed by the dancer, who employs āṅgika, sāttvika and

āhārya to express the meaning of the song as it is sung by the singers. In the

pūrvaraṅga it is the singer who uses vācika-appropriately, since singing is

verbal expression. Thus, abhinaya is an art that employs body movements

to express meaning in nāṭya while it lends beauty to nṛtta. That is why

Bharata defines abhinaya as an art which is so called because it “brings

forward” the presentation:

अभिपूर्वस्तु नीतौधातुराभिमुख्यार्थनिणये ।

यस्मात् प्रयोगान् नयति तस्मादभिनय: स्मृत: ॥

(NŚ. 8.7.)

40 NŚ. 9. 181.

MOVEMENT AND MIMESIS

129

The root nī is preceded by abhi in the sense of towards. Abhinaya is so called because it brings [forward] the presentation.

It seems that in the evolution of dancing abhinaya played a central role as the bridge between nātya and nṛtta. As we have seen above, Bharata recommends the use of abhinaya not only in nātya but also in nṛtta. He treats nātya and nṛtta as different but complementary art-forms, nātya being used as a term encompassing a larger area of presentation. In the early period, it seems, nātya included abhinaya, which was of four kinds, and āṅgika with its stylized movements was a necessary part of nātya. Nṛtta was a separate art-form employed to beautify the pūrvarāṅga of a drama. But it is doubtful whether nṛtta was ever used in a play towards the development of a theme. However, this is an entirely different issue which is beyond the scope of the present investigation. All we can say is that the concept of abhinaya provided a wide scope for interpretation and speculation in expressing new ideas.

It seems likely that it was due to the versatility of abhinaya that an entirely new art came into being, although it did so after Bharata’s time. This was nṛtya, which evolved out of the synthesis of nṛtta and nātya. This mimetic dance form is not mentioned by Bharata, nor by his immediate follower, the author of Viṣṇudharmottara Purāṇa.41 That the concept of dance was undergoing change is evident from the fact that although the Viṣṇudharmottara Purāṇa does not use the term nṛtya it uses another term, lāsya, as one category of nṛtta. The absence of the term nṛtya in that work shows that mimetic dance developed even later. The term lāsya, as we shall see in our discussion in the next chapter, was used by Bharata to indicate a delicate style of dramatic presentation. But by the time the

41 VDP.3.20.1.

130

THE NĀṬYASĀSTRA AND THE CONCEPT OF DANCE

Viṣnudharmottara Purāṇawas written lāsya had come to be recognized as

a form of nṛtta or dancing.

In earlier times, before nṛtya was established as a separate art form

two terms, nṛtya and nāṭya may have been used interchangeably or

synonymously, as it is evident in the Amarakoṣa.42 In modern times we

find the term Bharatanāṭyam, the classical dance of Tamilnadu, which

includes both abstract and mimetic dance. The term nāṭya applied to

dance is a perfect instance of equating nṛtya with nāṭya. Although the

name Bharatanāṭyam was given only in the thirties of the present century,

the usage is traditional. The word for nāṭya in Tamil is āṭṭa, which has

always been used to mean dance and Bharatanāṭyam developed from

Dāsiāṭṭam of Tamilnāḍu. Mohiniāṭṭam of Kerala is another dance style

that uses āṭṭam to mean dance.

The category known as nṛtya or mimetic dance thus evolved from the

concepts both of nṛtta and nāṭya in the Nāṭyaśāstra. Before we consider

the process of that evolution, it will be useful to examine one of the most

basic of these concepts, that of lāsya, because the various meanings

attributed to the term in the post-Bharata śāstric tradition substantially

influenced the understanding of the nature and technique of dancing.

42 Amarakoṣa,1808, p. 42.

Chapter 4

LĀSYA: A DRAMATIC ART

In the critical literature of the performing arts of classical India, there

is some confusion about what is meant by the term lāsya. It is usually

regarded today as a feminine dance described by Bharata in the

Nāṭyaśāstra.1 There is, however, no authority for this view in the

Nāṭyaśāstra which, on the contrary, compares lāsya with bhāṇa, that is, a

form of drama:

अन्यान्यपि लास्यविधावकाशानि तु नाटकोपयोगीनि ।

अस्माद्विनि:सृतानि तु भाण एवैकप्रयोज्यानि ॥

भाणाकृतिवल्लस्यं विश्रेयं त्वेकपात्रहार्यं च ।

+प्रकरणवद्द्वा कार्यासंस्तवयुक्तं विविधभावम्+॥ 2

(NŚ.19.117-18)

In the practice of lāsya,there are other elements [lit: limbs] that are suitable for nāṭaka. But when they

have emerged from it, they, like the bhāṇa, are to be

performed by one person. Lāsya is to be so defined as

to have a form similar to that of bhāṇa and it is to be

1 Chattopadhyaya, 1978, p. 56; Lath,1978, p. 32.

2"प्रकरणवदूhya कार्यासंस्तवयुक्तं " This reading in the G.O.S. edition seems very obscure.

The Kāśī edition gives a better reading: "प्रकरणवदूhyकary संसतयुक्तं"..which was also independently suggested by S.J. Sanderson of Wolfson College, Oxford. The last two lines are found in the G.O.S. and Kāśī editions of the Nāṭyaśāstra; the Kāvyamālā edition does not have them, nor does Abhinavagupta have any commentary on them. I have followed the reading of the Kāśī edition in translating this line. However, the readings still remain unsatisfactory.

131

132

LĀSYA: A DRAMATIC ART

acted by one person. As in prakarana, its plot [lit:the

function] is a conjectured one which relates to

praising and a variety of emotions.

The emphasis here is on the dramatic quality of lāsya. It can be divided

into parts that lend themselves to dramatic representation and constitute

dramatic action. More than just a performing style, lāsya is compared in

this passage to such dramatic genres as nāṭaka, bhāṇa and prakaraṇa. Its

association with dancing is not evident from this passage, nor its identity -

later taken for granted-as a feminine art.

The dramatic quality attributed by Bharata to lāsya excludes it from

his definition of tāṇḍava, which is Bharata's term for dance because it

was Tanḍu who received the art of dancing from Śiva. As we have noticed

in the previous chapter, the method of dancing that was created, Bharata

says, is known as tāṇḍava: नृत्तप्रयोग: सृष्टो य: स ताण्डव इति स्मृत: (NŚ. 4.

261); here nrttaprayoga is taken as a karmadhāraya samāsa. By this

definition nothing that falls outside the category of tāṇḍava, that is, nrtta,

could be called a dance.

If lāsya is comparable to nāṭaka, bhāṇa and prakaraṇa, then it cannot

be a component of nrtta which is defined by Bharata as a non-referential

art:

अर्थोच्चयते न खल्वर्थ कश्चिन्नृत्तमपेक्षते ।

(NŚ. 4. 263)

Here, it is said that dance not require any

[representational] meaning.

MOVEMENT AND MIMESIS

133

That Bharata regarded lāsya as a dramatic art becomes even clearer

from his description of the ten forms of lāsya that he mentions:3

गेयपदं स्थितपाठचमासीनं पुष्पगणिडका ।

प्रच्छेदकं त्रिमूढं च सैन्धवाख्यं द्विमूढकम् ॥

उत्तमोत्तमकं चैवमुक्तप्रत्युक्तमेव च ।

लास्ये दशविधे ह्येतदज्ञनिर्देशलक्षणम् ॥

आसनेषूपविष्टैर्यात्तन्रीभाण्डोपबृंहितम् ।

गायनैर्गीयते शुक्रं तदृ गेयपदमुख्यते ॥

प्राकृतं यद्वियुक्ता तु पठेदात्तरसं स्थिता ।

मदनालत्पात्करी स्थितपाठचं तदुच्यते ॥

आसीनमास्यते यत्र सर्वातोद्यविवर्जितम् ।

अपसारितगात्रं च चिन्ताशोकसमन्वितम् ॥

वृत्तानि विविधानि स्युःगेयं गाने च संश्रितम् ।

चेष्टाभिश्चाश्रयः पुंसां यत्न सा पुष्पगणिडका॥

प्रच्छेदकः स विज्ञेयो यत्र चन्द्रातपाहता: ।

स्त्रियः प्रियेषु सज्जन्ते ह्युपि विप्रियकारिषु ॥

अनिष्टुरश्लक्षणपदं समवृत्तैरलड्कृतम् ।

नाटचं पुरषभावाढचं त्रिमूढकमिति स्मृतम् ॥

पात्रं विभ्रष्टसङ्केतं सुव्यक्तकरणान्वितम् ।

प्राकृतैर्वचनैर्युक्तं विदुः सैन्धवकं बुधाः ॥

मुखप्रतिमुखोपेतं चतुरश्रपदक्रमम् ।

शिलष्टभा: रसोपेतं वैचित्र्यार्थ दिमूढके ॥

उत्तमोत्तमकं विद्यादनेकरससंश्रयम् ।

विचित्रैः श्लोकबन्धैश्च हेलाहावविचित्रितम् ॥

कोपप्रसादजनितं साधिकेषपदाश्रयम् ।

उक्तप्रत्युक्तमेवं स्याच्चित्रगीतार्थयोजितम् ॥(NS. 19. 121 -35)

3In his edition of the Nāṭyaśāstra, Rāmakṛṣṇa Kavi includes two extra types of lāsya on the

strength of Abhinavagupta's commentary. In view of Bharata's categorical statement

that the number is ten, the addition seems to be an interpolation.

134

LĀSYA: A DRAMATIC ART

Geyapada, sthitapāṭhya, āsīna,

puṣpagaṇḍikā, pracchedaka, trimūḍhaka, saindhava,

dvimūḍhaka, uttamottamaka and uktapratyukta-these

ten features characterise lāsya.

Where the male singers, being seated in the seats

sing dryly [i.e., uttering meaningless syllables] being

supported by stringed instruments and drums, it is

known as geyapada.

Where a separated woman, her body burning with

the fire of love, remains withdrawn and reads Prākṛt

[verses], it is sthitapāṭhya.

When [a woman] sits, without any musical

instrument around her, without stretching out her body

and is overcome with anxiety and sorrow, it is Āsīna.

Where songs are sung in different metres, where

geya [the composition to be sung] is dependent on gāna

[the regional style, as opposed to traditional

gāndharva style 4] abounding in manly movements, it

is puṣpagaṇḍikā.

That is known as Pracchedaka in which women,

although wronged by lovers, are attached to their

lovers, being struck by the moonlight [in love].

Trimūḍhaka is a nāṭya [performed by women]

composed of gentle and tender words, adorned with

even metres and abounding in manly emotions.

4 Lath, 1978, pp. 183-4; SR. 4.3.

MOVEMENT AND MIMESIS

135

Where a person has failed to keep a tryst, recites

Prākṛt to the accompaniment of karaṇas, it is

Saindhavaka.

Where a song is presented with mukha and

pratimukha [=exposition and progression] [sung] in a

series of caturaśrapada adhering to bhāva and rasa and

arriving at a different meaning, it is known as

dvimūḍhaka.

Uttamottamaka is known to be dependant on many

rasas [expressed through ] a variety of ślokas and is

adorned with [feminine movements], such as helā and

hāva.

Uktapratyukta is derived from anger or pleasure

and [often] contains words of censure. It [should also]

contain a variety of expressive songs.

From these descriptions it seems that lāsya was the representation of

brief segments of dramatic action which called for the expression of the

softer emotions. It is to be noted that lāsya is discussed in connection with

one kind of drama, not with dance. The dramatic nature of lāsya is further

attested by Bharata’s remark in a later chapter:

लसनाल्लास्यमित्युक्तं स्त्रीपुंभावसमाश्रयम् ।

(NŚ. 31. 333)

Since it is playful [ from the verbal root las] it is

called lāsya. It is based on the emotional relations of

men and women.

Yet, against Bharata’s explicit statements, later authors categorized

lāsya as a dance and a feminine one at that. The first work to include lāsya

under dancing was the Viṣṇudharmottara Purāṇa, which, of the works that

136

LĀSYA: A DRAMATIC ART

deal with dancing, seems to be chronologically the closest to the

Nāṭyaśāstra. 5 This text states:

नृत्तं तु द्विविधं प्रोक्तं नाट्यलास्याश्रयिं सदा ।

(VDP.3.20.2.)

Nṛtta, always based on nāṭya and lāsya, is known to be

of two types.

This classification seems debatable in a work that follows the

Nāṭyaśāstra so faithfully. Since nṛtta is agreed to be a non-referential

dance form,6 it is a contradiction in principle to state that it is always

based on nāṭya and lāsya, which are specifically characterized by Bharata

as containing plot elements. Evidently, by the time the Viṣnudharmottara

Purāṇa was written, lāsya had acquired a strong association with dancing.

In part, the reason for this may be found in the Nāṭyaśāstra. Bharata

states,

रेचकैरड्गहारैश्च नृत्यन्तं वीक्ष्य शङ्करम् ।

सुकुमारप्रयोगेन नृत्यन्तीं चैव पार्वतीम् ॥

(NŚ. 4.249-50.)

Seeing Śaṅkara dancing with recakas and aṅgahāras

and seeing Pārvatī dancing with delicate movements.

5 This text has been variously dated but its editor places it between the fifth and seventh

centuries (P.Shah, ed. Viṣnudharmottara Purāṇa,vol. I. Introduction, p. xxvi.). Pingree

assigns this part of the text to the first half of the fifth century; Ludo Rocher has

discussed the arguments for the different dates in detail in his volume on his history of

Purāṇa literature. However, since the author of this text adds a ninth rasa, śānta, to the

list of eight rasas in the Nāṭyaśāstra, there can be little doubt that at least this part of the

text postdates the Nāṭyaśāstra.

6 NŚ.4.263.

MOVEMENT AND MIMESIS

137

Sukumāraprayoga means the use of less strenuous angahāras as

opposed to the vigorous movements of Śiva. By associating Pārvatī with

delicate movements, this passage provides authority for correlating

feminine grace with a delicate style of dancing, although it does not call

that style lāsya. The association between femininity and delicate

movements is further reinforced by a later verse:

देवस्तुत्याश्रयकृतं यदृङ् तु भवेदथ ।

माहेश्वरेऋद्धानिर्द्धतेस्ततं प्रयोजयेत् ॥

यत् तु शृङ्गारसम्बन्धं गानं स्त्रीपुरुषाश्रयम् ।

देवीकृतैरऋद्धैरलितेस्ततं प्रयोजयेत् ॥

(NŚ. 4.311-12.)

Whatever part [of a play, i.e., pūrvarañga] is for

praising gods, that [one] should perform with the

vigorous angahāras created by Maheśvara, while a

song related to erotic sentiment involving

[lit:depending on] man and woman should be

performed with the angahāras created by Devī.

Abhinavagupta explains this passage as follows:

. . . उद्धतैरिति । विद्युद्भान्तगरुडप्लुतकादि प्रधानैः । अत्र

हेतुर्यतस्ते माहेश्वरप्रयुक्ता:। . . . ललितैरिति ।

तल्पुष्पपुटलोलीनितम्बाधारबन्धः। मत्र हेतुः यतस्ते

देवीकृता:।

(Abhinavagupta on the Nāṭyaśāstra, G.O.S. ed.,Vol.I,

p. 180)

. . . By uddhata [=energetic] is meant mainly by[the

use of karanas such as] vidyudbhrānta or garuḍapluta.

The reasoning here is that these are employed by

Maheśvara. . . . By lalita [=delicate] is meant done by

Devi.

138

LĀSYA: A DRAMATIC ART

[karanas such as] talapuṣpapuṭa, līna and nitamba.

The reasoning here is that these are created by Devī

[Pārvatī].

While commenting on tāṇḍava and its connection with devastuti,

Abhinavagupta says that whenever tāṇḍava is mentioned, lāsya has to be

understood implicitly as its complement in the art of dance. To explain this

complementarity and as a comparable instance he cites the maxim of

गोबलोर्दनन्याय in which the correlation between the cow and the bull is

derived from the logic of associative inference:

ताण्डवमिति सर्व नृत्तमुच्यते। लास्यशब्देन सन्निधौ

गोबलोर्दनन्यायेने प्रवर्त्तते।

(N.S. G.O.S. ed.,1956,Vol. I, p. 180)

Tāṇḍava is said to refer to [lit:to be] all nṛtta

(=dancing). By proximity the word lāsya is taken as in

the maxim of gobalīvarda.

Commenting on गोबलोर्दनन्याय the Mīmāṃsākośa says,

गामानय बलीवर्दञ चानय

इत्यत्र गोपदेनैव

बलीवर्दबोधसिद्धौ

बलीवर्दपदं

दुर्दम्यत्वज्ञापनपरत्वेन सफलमिति।

(Mimamsakosa, p. 1617)

Bring a cow and bring a bull as well-here, although by

using the word 'go' alone, the implication of

'balīvarda' is achieved [yet] the effectiveness of the

word 'balīvarda' [consists] in indicating

uncontrollability.

Even though the entire phrase gobalīvarda is needed for emphasis, the

word go suffices on its own to express the totality of meaning. By analogy

MOVEMENT AND MIMESIS

139

with this formula of reasoning, any reference to tāṇḍava may be taken to

include lāsya as well. In the complementary relationship thus formed,

tāṇḍava represents virility and energy while lāsya represents delicacy and

grace. Given such characteristics, the categorization of tāṇḍava and lāsya

as masculine and feminine, respectively, is not surprising. It is on this

basis that Abhinavagupta takes lāsya as a style of dancing. As the

evidence of the Viṣṇudharmottara Purāṇa suggests, after Bharata's time

the term lāsya had generally come to mean a feminine style of dancing and

Abhinavagupta was interpreting the Nāṭyaśāstra to provide support for

current usage. The association of Pārvatī with lāsya is thus an outcome of

her use of delicate angahāras in the legend given in theNāṭyaśāstra in

which she is said to have used delicate angahāras.7

The same view is repeated by other early writers. Nandikeśvara

reports in his Abhinayadarpana that lāsya was taught by Pārvatī.8

Whether he was the first author to ascribe lāsya specifically to Pārvatī

cannot be ascertained since the date of the Abhinayadarpana has not been

put prior to Abhinavagutpta's commentary on the Nāṭyaśāstra.

Nandikeśvara may well have borrowed the idea from the existing

literature on the subject. He has little to offer in the way of description or

analysis, for he stops at including lāsya in a brief review of nāṭya, nṛtta

and nṛtya, and stating merely that it was an art demonstrated by Pārvatī.9

The Saṅgītamakaranda of Nārada talks about lāsyāṅgas, elements of

lāsya, in connection with the qualities required in a dancer but does not

describe the nature of these lāsyāṅgas.10

7 Ibid. 4. 249-50.

8 ADar.4.

9 Ibid.5.

10 SMak. 2. 1. 54.

140

LĀSYA: A DRAMATIC ART

Among the early writers on alaṁkārasāstra, Daṇḍin specifically

mentions lāsya as a form of dance.11 The dramatists, Kālidāsa,

Bhavabhūti and Harṣadeva, have used lāsyāṅgas,that is, elements of lāsyas

in their dramas. Abhinavagupta has cited them in in his commentary on

lāsyāṅgas.12 Another early text, the Dasarupaka of Dhananjaya, calls

lāsya one of the two major classes of dance styles, the other being

tāṇḍava.13These works show that soon after the writing of the Nātyasāstra

tāṇḍava and lāsya came to be associated with masculine and feminine

qualities, respectively, and with male and female dancers, a segregation

that never appears in the Nātyasāstra.

That lāsya was commonly recognized in the early period as an art

lending feminine grace to stage actions is attested by the general literature

of the period. Dāmodaragupta in his Kuṭṭanīmatam identifies lāsya as a

form of movement related to a female performer, calling it strīlāsya.14

Rājaśekhara in his Karpūramañjarī refers to lāsya without describing its

nature.15 Although it was so widely known, lāsya was never described in

detail until much later. As we have seen, even Abhinavagupta does not

provide descriptions. He refers to lāsya while commenting on lāsyāṅgas in

the nineteenth chapter and also while discussing tāla in the thirty-first

chapter, the tālādhyāya, refers to the style of lāsya; but does not tell us

what this lāsya was like except that it used delicate movements.

In the medieval period lāsya continued to be discussed, and discussed

at some length, but whether it should be regarded as dance or drama

11 K.A.1.39.

12 Abhinavagupta on the NŚ. 1954, Vol. III, pp. 75-77.

13 DR. 1.9-10.

14 Kuṭṭanīmatam, 853. Dāmodaragupta lived in the eighth century. See Kṛṣṇamāchāriar,

1974, p.353.

15 Karpūra­mañjanī, 4.10. Rājaśekhara is believed to have lived in the last quarter of the

ninth and the first quarter of the tenth century. See Poddar,1974, p. 174.

MOVEMENT AND MIMESIS

141

remained uncertain. Bhoja categorizes lāsya as a variety of nartanaka, an

uparūpaka, which uses graceful and delicate movements to express bhāva

(emotions).16 The other three in this category are identified by Bhoja as

śamyā, chalika and dvipadī:

यत्र पदार्थाभिनयं ललितलयं सदसि नर्तकी कुर्वते ।

तन्नर्तनकं शम्यालास्यछलिकद्विपद्यादि ।

(Śr.P. Vol. II, p. 468)

Where in an assembly, a female dancer performs [in]

a graceful tempo to act out the meaning of the word, it

is nartanaka [which comprises] śamyā, lāsya, chalika

and dvipadī etc.

Here the association between lāsya and drama seems strongly suggested.

At the same time, Bhoja continues like others before him to emphasize the

graceful quality of lāsya inherent in the erotic sentiment, stating, तदिदं

शृङ्गारप्रधानत्वाल्लास्यमित्त (It is [called] lāsya because it deals mainly with

śṛṅgārarasa [erotic sentiment]). Since he also categorizes lāsya as a form

of dṛśyakāvya (i.e., a poetic composition to be seen), another form being

tāṇḍava, he is apparently attributing to it the qualities both of dance and

drama. Its suitability as a vehicle of dramatic communication is further

indicated in the śabdālaṅkāra section of Bhoja 's Sarasvatīkaṇṭhābharaṇa,

where he divides compositions into preksya (to be viewed) and śravya (to

be heard). Preksya is subdivided into lāsya, tāṇḍava, chalika, śamyā,

hallīsaka and rāsaka, while lāsya is defined as one containing mainly

śṛṅgārarasa.17 The dramatic nature of lāsya is recognized by Bhoja also

in his comments on piṇḍibandhas or group dances. In the Śṛṅgāraprakāśa

he calls them necessary parts of the rāsaka and nāṭyarāsaka types of

16 Śṛṅgāraprakāśa, 1963,Vol. II, p. 468.

17 SKA. 1969, pp. 124-25.

142

LĀSYA: A DRAMATIC ART

uparūpakas rather than of lāsya.18 Here Bhoja seems to be following

traditional thinking; Abhinavagupta quotes Kohala and the cirantanas, that

is, ancient authorities,19 to explain that these two nṛtya types of

uparūpakas contained both delicate and vigorous nṛtta movements but he

mentions lāsya nowhere in the context.

The tradition regarding lāsya as a dance form is carried on in a

major work on music and dancing from the next century, the Mānasollāsa

of Someśvara, which views lāsya as a dance consisting of aṅgahāras,

specifically lalita aṅgahāras.20 But this work provides no further

information. A much fuller discussion appears in the Bhāvaprakāśana of

Śāradātanaya, which discusses dramaturgy and allied arts. This

compendium of critical works, extending from that of Bharata down to the

writings of Kṣemendra in the eleventh century, was written before the

Saṅgītaratnākara and the Nṛttaratnāvalī. Śāradātanaya's treatment of

lāsya is rather confusing. He defines lāsya as a dance composed of

delicate aṅgahāras:

ललितैरङ्गहारैश्च निर्वर्त्यललितैरलैयै: ॥

वृत्तिः स्यात्कैशिकी गीतिर्यत्र तल्लास्यमुच्यते ।

(Bh.P. p. 296)

That is known as lāsya in which the style is kaiśikī,

[where] there is a song and which is composed of [lit:

being done by] graceful aṅgahāras done to delicate

layas.

According to him, lāsya is of four types, namely, śṛṅkhalā, latā, piṇḍī,

and bhedyaka:

18 Sr.P. 1963, Vol. II, p.468.

19 NS.1956, Vol. I, p. 181-2.

20 Mānas. 16. 4. 962 - 63.

MOVEMENT AND MIMESIS

143

सुकुमारप्रयोगो यो नियतो लास्यमुख्यते ।

तच्छृङ्खलालतापिण्डीभेदकः स्याच्चतुर्विधम् ॥

(Bh.P. p.297)

What is delicately applied is called lāsya, which can

be of four varieties-śṛṅkhalā, latā, piṇḍī and

bhedyaka.

Śāradātanaya goes on to enumerate further sub-divisions and describe

the actual movements of these four types. These types, though not the

identical terms, are taken from the Nāṭyaśāstra:

पिण्डीनां विधयश्चैव चत्वारः सम्प्रकीर्तिताः : ॥

पिण्डी शृङ्खलिका चैव लताबन्धोऽष्ठ भेदकः।

(NŚ. 4.287-88)

The ways [of forming] piṇḍīs are known to be four:

piṇḍī, śṛṅkhalikā, latābandha and bhedyaka.

Śaradātanaya's attempt, unsupported by the Nātyaśāstra, to include

these movements under lāsya is inconsistent with his own statement

earlier in the treatise:

अपरैनृत्यभेदास्तु गुल्मशृङ्खलितालताः।

भेद्यकश्चेति चत्वारः कथ्यन्तेऽत्र मनीषिभिः ।

पिण्डीबन्धश्च गुल्मश्च पर्यायाविति केचन ॥

(Bh.P. p. 246)

Here, other experts have described four nṛtyabhedas

[such as] gulma, śṛṅkhalitā, latā and bhedyaka. . . .

Some say that piṇḍībandha and gulma are synonymous.

In this passage Śāradātanaya clearly identifies piṇḍībandhas as

dramatic types, for nṛtyabhedas are varieties of dramatic action. It is

144

LĀSYA: A DRAMATIC ART

possible that he took lāsya primarily as a dramatic art, for he states on the authority of Kohala that:

कोहलादोभराचैरुक्तं भाणस्य लक्षणम् ।

लास्याङ्गदशकोपेतं . . . ॥

(Bh.P. p. 245)

The characteristics of bhāṇa are described by experts [such as] Kohala and others as being arrived at by ten lāsyāṅgas . . .

By telling us that lāsya forms part of the presentation of bhāṇa, Sāradātanaya is again placing lāsya in the context of dramatic art. As subdivisions of that art, piṇḍibandhas would necessarily be dramatic pieces.

One of the first manuals on dancing to describe lāsya in detail is the Nṛttaratnāvalī of Jaya Senāpati. In the first chapter of the first section of the book he defines nṛtta and says that

लास्यताण्डवभेदेन द्वयमेतद् द्विधा पुनः ।

सुकुमारं तयोराच्यं भवेदपरमुद्दतम् ॥

(NR.1. 1.56)

By the distinction between tāṇḍava and lāsya this [dance] is again of two types. The first of the two is delicate and the other is energetic.

He then goes on to describe the parts of lāsya as they appear in the Nāṭyaśāstra. These lāsyāṅgas, according to him, are elements of traditional forms of dance (mārga). In expressing this opinion he is obviously not following Bharata, whose emphasis is on the dramatic nature of lāsya. The first section of the Nṛttaratnāvalī deals with mārga and the second with deśī. In the second section the author deals with deśīlāsya, that is, the regional variations of lāsya. He lists forty-six such

MOVEMENT AND MIMESIS

145

deśīlāsyas.21 Since these movements employ only āṅgika, not vācika

abhinaya, they may be categorized as dance movements.

The Saṅgītaratnākara of Śārṅgadeva defines lāsya as a delicate dance

that stimulates erotic sentiments:

लास्यं तु सुकोमलाङ्गं मकरध्वजवर्धनम् ।

(SR. 7.30)

Lāsya has delicate features [lit: limbs] and arouses

erotic [sentiments].

But Śārṅgadeva does not treat the subject in full, discussing only deśī

lāsyāṅgas, that is, only the local or regional varieties of lāsya. He follows

the Nāṭyaśāstra and Abhinavagupta's commentary in discussing and

defining āṅgika abhinaya, that is, movements of the aṅga and upāṅga, cārī,

sthāna, maṇḍala, karaṇa and aṅgahāra, but omitting from his discussion

the forms of lāsya described by Bharata. This seems perfectly justifiable

since those forms were specified by Bharata as dramatic actions, not

dance movements. Śārṅgadeva seems to be the first author to realize that

the elements of lāsyas described by Bharata are meant for dramatic

actions. The deśī lāsyāṅgas he describes in detail are ten and they are

contemporaries, but Saṅgītaratnākara was written before the

Nṛttaratnāvalī, and as V. K. Raghavan points out, there is internal

evidence that Jāya knew about the existence of theSaṅgītaratnākara when

he wrote his text.22 However, that does not minimize the importance of

Jāya's work, which deals with dance in even greater detail than

Śārṅgadeva's, giving us a fuller picture of the deśī tradition, although in

21 NR. 6. 117- 73.

22 NR. Introduction, pp. 73-77.

146

LĀSYA: A DRAMATIC ART

the later literature after the fifteenth century on dance we find the

influence of Sārṅgadeva, not Jāya.

The meaning of lāsya has been further clouded by the use of the term

lāsyāṅga in several of the texts that came after the Nāṭyaśāstra. The term

was never used by Bharata in the Nāṭyaśāstra Abhinavagupta did so in his

commentary on the Nāṭyaśāstra but he used it not to mean a class of

movements distinct from lāsya but simply in the sense of the different

components of lāsya. That the word lāsyāṅga was no more than a

derivative of [the word] lāsya is clear from Abhinavagupta's

commentary.23 It is worth noting that the Nāṭakalakṣaṇaratnakośa of

Sāgaranandin, the only work that faithfully follows the Nāṭyaśāstra

directly rather than Abhinavagupta's commentary in describing the type

of lāsya, does not use the term lāsyāṅga.24 We must also remember that

Bharata's discussion of lāsya and its aṅgas forms part of his description of

the elements of a dramatic composition rather than a dance. This is a

categorization also followed, among the later writers, by Viśvanātha who,

in his Sāhityadarpana, includes the aṅgas of lāsya among the elements of

dramatic representations, such as sandhi etc., just as Bharata did.25

However, the term lāsyāṅga found general acceptance among the later

authors, most of whom viewed lāsya and lāsyāṅga as different classes of

movements, drawing a distinction that has caused some confusion. The

confusion is worst in the Bhāvaprakāśana, which reproduces in chapter

eight the same ten movements described in the Nāṭyaśāstra as parts of

lāsya. Later, however, in chapter ten, the Bhāvaprakāśana divides

dancing into tāṇḍava and lāsya, going on to enumerate four types of lāsya,

two of which–ṣṛṅkhalā and bhedyaka–are divided each into ten lāsyāṅgas,

23 Abhinavagupta on the NS. 1954,Vol. III,19. 117-135.

24 NLRK. pp. 270-76.

25 SD. 6.270.

MOVEMENT AND MIMESIS

147

which are the same ten movements previously specified as lāsya in

chapter eight.26

We have noted before that Saṅgītaratnākara defines lāsya as a dance

that arouses erotic sentiments. It also describes ten types of delicate

movements, calling them lāsyāṅgas but not equating them with lāsya:

लास्याङ्गानि दशैतानि देश्यां देशीविदो विदुः ।

कोमलं सविलासं च मधुरं ताललास्ययुक्तकं ।

नातिद्रुतं नातिमन्दं नृत्यश्रुताप्रचुरं तथा ।

पादोः कटिबाहूनां योगपद्येन चालनं ।

चालिः सा शैव्यसांमुख्यप्रायाः चालिवडो भवेत् ।

सुकुमारं तिरश्चीनं विलासरसिकं च यत् ॥

यगपकटिबाहूनां चालनं सा लडिर्मतिः ।

कर्णयोर्हावबहुलं लसल्लीलावतंसयोः ॥

विलम्बेनाविलम्बेन सूक्ष्मं तल्तलचालनं ।

विलम्बेनाविलम्बेन कुचयोर्हुज्जशीर्षयोः ॥

ललितं चालनं तिर्यकं तज्ज्ञाः प्राहुरोरुक्षणं ।

धसकः स्यात्सुललितं स्तनाधोनमनं लयादृत् ॥

सतालललितोपेता क्रमात्कायार्धयोर्नतिः ।

धनुर्वदृदृहारः स्यादिति निःशङ्कभाषितं ॥

किश्चित्तिर्यंगधो मूर्ध्नो गतिरोयारको मतः ।

स्मितं स्याद्विद्रिसी यस्तु शृङ्गाररसनिर्भरः ॥

अभ्यस्तादन्य एवातिस्वरूपप्रत्ययप्रभुः ।

गीतादेरागतः

स्थायस्तल्लयात्तन्मनो मतम् ॥

(SR. 7. 1207- 15)

These are ten lāsyāṅgas in the deśī [style]

regarded by those who are knowledgeable in deśī.

[They are] delicate, playful, attractive and

[adorned]with tāla and lāsya [rhythm and grace].

26 Bh.P. p. 297.

148

LĀSYA: A DRAMATIC ART

Cāli is the simultaneous movement of the feet,

hips, thighs and arms. It should be neither done too

fast nor too slow, and [should be] full of oblique

movements done in tāla.

The same action performed in a quick tempo,

mainly facing the front, is known as cālibada.

Ladhi is known to be the simultaneous horizontal

movements of the hips and arms which is soft and

graceful.

Sūka is a rhythmic movement full of hāva

[feminine gestures], of the ears shining with dangling

earings and performed either slowly or quickly.

Urongana is known as a graceful and oblique

movement, slow or fast, of the shoulders and the

breasts.

Dhasaka is the graceful and rhythmic downward

movement of the breast.

Angahāra is the successive bending of the two

halves of the body, slowly and gracefully to tāla, and

in the likeness of a bow, so says the confident one [i.e.,

Sārngadeva].

Oyāraka is a slightly oblique and downward

movement of the head.

Vihasi is a smile full of the sentiment of love

different from a trained one. It has an extremely

subtle and novel charm.

Mana is known [to be performed] to songs etc. set

to rāga in the sthāya tempo.

These movements do not constitute lāsya as Bharata understood the

term, for the Sangītaratnākara classifies them as deśī forms, that is,

MOVEMENT AND MIMESIS

149

popular or regional as opposed to mārga or refined classical forms, while

the Nāṭyaśāstra bypasses the whole subject of deśī forms. It is evident that

scholarly opinion, presumably following performing practice, had

changed considerably since the time of Bharata. An important point

established by the description of lāsya movements in the Sangītaratnākara

is that by the time that work was written, lāsya had definitely come to

mean dancing and had a defined form. It is also evident that lāsyāṅgas had

come to be regarded as deśī forms.

Pārśvadeva in his Sangītasamayasāra gives us a list of nineteen

deśyaṅgas which, from their description, seem to be the same kind of

movements as the lāsyāṅgas given in the Sangītaratnākara. This is

supported by the fact that Jāya lists forty-six deśī lāsyāṅgas, which are similar

in character to those described in the Sangītaratnākara, although only ten

of them match Śārṅgadeva's list. The nineteen deśyaṅgas mentioned by

Pārśvadeva and the ten lāsyāṅgas of the Sangītaratnākara are all included

in Jāya's list of forty-six deśī lāsyāṅgas. Pārśvadeva is earlier than Jāya,27

and it is possible that the latter has taken similar movements from the

Sangītasamayasāra and the Sangītaratnākara and combined them into one

category, calling them lāsyāṅgas of the deśī variety and adding others. In a

similar way, the Nṛtyādhyāya of Aśokamalla and the Nṛttaratnakośa of

Mahārāṇā Kumbha give lists of thirty-six deśī lāsyāṅgas, again a

combination of the deśyaṅgas and the lāsyāṅgas described respectively by

Parśvadeva and Śārṅgadeva. Subhaṅkara's Sangītadāmodara describes

the ten categories of movements included under lāsya in the Nāṭyaśāstra,

but calls them lāsyāṅgas which, according to the author, are necessary to

bhāṇa.28 Subhaṅkara later defines lāsya as: ...स्त्रीनृत्यं लास्यमुच्यते।29—

feminine dancing is known as lāsya—which he calls a deśī form of dancing.

27 See NR. 1965, Introduction, pp. 79-80.

28 SDām.. pp. 73-4.

29 Ibid. pp.69.

150

LĀSYA: A DRAMATIC ART

When we examine the dance-sequences described by the medieval authors, those who came after Kumbha, we again find the use of these deśī lāsyāṅgas.

On reviewing these texts it seems that from the time of the Bhāvaprakāśana lāsya and lāsyāṅga begin to be considered as two different types. These later texts also state that lāsyāṅgas form a part of bhāṇa. But this position is not entirely satisfactory. Bharata, as we have seen above, says that lāsya is suitable for dramatic performances. In that connection he refers to bhāṇa, a type of drama which is performed by only one person. Lāsya is similar to it because this too is done by one person. But mere similarity does not mean that bhāṇa may include any form of lāsya. Bharata never claims anything of the sort when he describes bhāṇa in chapter nineteen of the Nāṭyaśāstra30 On the other hand, he says specifically that kaiśikī vṛtti should not be applied to bhāṇa:

उत्सृष्टिकाङ्को व्यायोगो भाण: प्रहसनं डिम: ।

कैशिकीवृत्तिहीनानि रूपाणयेतानि कारयेत् ॥

(NŚ.18. 8-9)

Utsṛṣṭikāṅka, vyāyoga, bhāṇa, prahasana and ḍima-kaiśikīvṛtti [the delicate and graceful style].

This excludes delicate movements from bhāṇa. Since both lāsya and bhāṇa involve one performer, and since in the Nāṭyaśāstra the ślokas about lāsya occur just after the passage on bhāṇa, some relation between lāsya and bhāṇa may be conjectured. But no relationship is actually established by Bharata. The proximity of the ślokas in the Nāṭyaśāstra was probably the reason why later authors were led to think that lāsya or its aṅgas were auxiliary to bhāṇa.

30 NS. 18. 108-10.

MOVEMENT AND MIMESIS

151

In his commentary on bhāṇa, Abhinavagupta states that bhāṇas are

chronicles of prostitutes and men who live by their wits.31 But the

Nātyasāstra does not mention any such thing. Later treatises, however,

seem to accept Abhinavagupta's view. They may have also tried to

connect bhāṇa with lāsyāṅga because a lāsyāṅga, being a delicate

movement, might pertain to the arts as they were practised by prostitutes.

The Dasarūpaka prescribes vīra and śṛṅgārarasa for bhāṇa.32 The

Bhāvaprakāśana mentions Kohala's view and says that bhāṇa should only

have śṛṅgārarasa.33 The Nātyadarpana of Rāmacandra and Guṇacandra

states that śṛṅgāra should be predominant in bhāṇa.34 The

Nāṭakalakṣaṇaratnakośa does not say that the various forms of lāsya are

necessary to bhāṇa, though according to it bhāṇa can also be performed by

a woman; when it is so performed, aṅgahāras must be used. In this text,

the description of ten lāsyāṅgas follows its account of bhāṇa, because a

lāsyāṅga, like a bhāṇa, is performed by a single actor.35

The history of lāsya makes the evolution of the performing arts of

India particularly complex. As we have seen in the foregoing survey, the

term lāsya is never used in the Nātyasāstra to denote dancing of any kind.

Instead, that text accords lāsya the same status as a performing art as it

gives to bhāṇa, although it does not equate them. For later authors, by

contrast, lāsya is a style of dancing distinguished by its feminine qualities,

while tāṇḍava appears in their works as a virile style embodying

masculine qualities. This recognition of lāsya as an independent style of

dancing and its separation from tāṇḍava is not found in the Nātyasāstra but

in all likelihood came about in the following way: Bharata's description of

31 Abhinavagupta on the NS., 1954. Vol. III, 18. 109-10.

32 DR. 3.50.

33 Bh.P. p. 245.

34 ND. p.127.

35 NLRK. pp. 270-71.

152

LĀSYA: A DRAMATIC ART

lāsya movements shows them to be delicate and graceful; elsewhere in the

text he mentions that Pārvatī adorned her dance with "sukumāraprayoga,"

that is, with delicate grace.36 When the Sangītaratnākara or the

Abhinayadarpana attributed lāsya to Pārvatī, they did so, presumably,

because of the delicacy and grace of her movements, and gave it the status

of proper dance, a status not conferred by Bharata himself. The

association of erotic sentiments with lāsya may thus be seen as a result of

its association with Pārvatī. It was evidently because of that association

and also because of its essentially gentle grace, that lāsya was considered

to be suitable for female performers. Tāṇḍava, on the other hand, appears

from all descriptions to be a strenuous, physically demanding style of

dancing. That is why it seemed to call for male performers.

But, again, such a division of labour is not authorized by the

Nāṭyaśāstra. Whenever Bharata talks about a dancer, he uses the feminine

noun nartakī.37 He could, then, hardly exclude female dancers from

performing tāṇḍava, the form of dancing on which he focuses his attention.

Iconographic support for this may be found in the Cidambaram temple

sculptures depicting the compositions described in the Nāṭyaśāstra.38

Bharata states that uddhata or vigorous aṅgahāras are suitable for

devastuti, or praise-offerings to gods, while lalita, or graceful aṅgahāras

are best suited to the depiction of erotic sentiments.39 This distinction

may have given later authors yet another reason for correlating tāṇḍava

and lāsya respectively with men and women. Lāsya, however, does not

necessarily involve erotic sentiments and can also be used in devastuti.40

Abhinavagupta supports this view when he says that lāsya is meant to

36 NŚ. 4.250.

37 NŚ. 4.118-20.

38 Naidu,V.,Naidu, S. and Pantulu,V.R.,1971.

39 NŚ. 4. 268-69 and 4.311-12.

40 NŚ. 4.302.

MOVEMENT AND MIMESIS

153

please the gods or kings.41 In the context, therefore, of the earliest source

of the Indian tradition of the performing arts, it seems likely that any

allocation of tāṇdava and lāsya separately to men and women would be a

matter of convenience rather than of principle.

The evidence of the Nāṭyaśāstra clarifies the meaning of lāsya in two

important ways. First, lāsya-as originally practised-was neither nṛtta, the

pure form of dancing, nor was it bhāṇa. Instead, it was an art form

possessing qualities of both and was a wholly independent art bridging

dance and drama. Secondly, lāsya was not necessarily an art reserved for

female performers, although it could be the representation of feminine

moods. Despite the various-and often confusing-opinions of Bharata's

successors regarding lāsya, the continued acknowledgement of its

suitability for representing emotional status reflects a growing trend

towards recognizing the dramatic potentiality of dancing. In the following

chapter we shall examine the connection between forms of dancing and of

drama perceived in musicological literature.

41 A.B. on the NS. 1956,Vol. I, 4. 268, .p. 180-81.

Chapter 5

NṚTYA AND UPARŪPAKA

In the last two chapters we have seen how dance was conceived in the

literature of the earliest period, especially in the Nāṭyaśāstra. It is

important to bear in mind that Bharata describes dance mainly in the

context of the preliminaries of a play. He also recommends the use of the

karaṇas (components of nṛtta) in the body of the drama, for scenes of

battle or for gaits which he mentions only once (NŚ. 4. 56). Finally he

describes the karaṇas, but does not specify which karaṇas are to be used

for which purpose. However, in his commentary on the Nāṭyaśāstra

Abhinavagupta correlates some of the karaṇas with particular dramatic

characters (A.B.Vol.I, p.206). Bharata’s own interests seem to have been

in the broader principles of stage presentation, which included elements of

dance.

Bharata classifies dancing by using comparatively few terms. These

were, as we have seen in chapter 3, nṛtta, tāṇḍava, abhinaya, piṇḍībandha

and sukumāraprayoga. Nṛtta and tāṇḍava are used synonymously to mean

abstract dance, that is, compositions of movements that please the senses

but carry no meaning (NŚ. 4. 261). Abhinaya, in the context of the

preliminaries of a play, is the art of miming the meaning of a song through

gestures and body movements (NŚ. 4. 283). Piṇḍībandhas are group dances

forming different emblems to represent different gods for the purpose of

showing reverence to them (NŚ.4. 252-59; 4. 287-90). Sukumāraprayoga is

the graceful and delicate way of presenting dance movements (NŚ.4. 250).

To these terms he adds lāsya, which he applies to a form of dramatic art.

However, as we have seen in chapter 4, the meaning of lāsya became

much more complex after Bharata’s time.

154

MOVEMENT AND MIMESIS

155

That this growing complexity was not restricted to the understanding

only of lāsya but extended to the whole art of dancing is shown by the

persistent attempts to expand its classification by adding new terms to its

conceptual vocabulary. With the Nāṭyaśāstra as its starting point, the

discourse on dancing began to widen its scope and soon after the

Nāṭyaśāstra, new terms that represented the proliferation of conceptual

classification began to appear in dramatic as well as other literature.1

The literature of the early period used these terms extensively, but later,

in the medieval period, most of these terms were absorbed in more general

terms that represented broader concepts. To set the later, broader terms in

their historical context, it is necessary first to consider the earlier terms,

although the focus of the present chapter will be on the two terms into

which most of them were assimilated, namely, nṛtya and uparūpaka.

Nṛtya and uparūpaka do not appear before the tenth and the fourteenth

centuries, respectively.2 That does not, of course, mean that the art forms

represented by these terms were not known before the medieval period,

but rather that they were not important or popular enough to be the subject

of serious discussion. They might have been simply left undescribed, just

as regional varieties of dancing were ignored by Bharata even while he

acknowledged their existence (NŚ. 9. 164). Their evolution becomes

clearer when we examine other terms that appear in the early literature

where they are used for various art forms involving dancing. They are

never defined in any work but what they stood for may usually be

conjectured from the content of the discussion if we consider the contexts

in which they appear. The descriptions are often cryptic and brief but in

the absence of any detailed discussion, we may form a picture from the

1 Mālavikāgnimitra, Vikramorvaśīya of Kālidāsa, Ratnāvalī of Harṣa, Kuṭṭanīmata of

Dāmodaragupta and Arthaśāstra of Kauṭilya are some of the relevant texts.

2 The term nṛtya first appears in the Daśarūpaka and uparūpaka appears first in the

Sāhityadarpana.

156

NṚTYA AND UPARŪPAKA

little information we have before we examine the meaning of nṛtya and

uparūpaka.

The terms we examine here are eight: dvipadī, śamyā, rāsaka,

skandhaka, chalika, carcarī, lāsya and saṅgītaka. Most of these terms

disappeared from dance literature after the early period. One that

remains well-known is lāsya, which, as we have seen already, was treated

as a separate, special category denoting both movements that conveyed

meaning and those that did not. Rāsaka is another that has survived and is

now applied to a group dance mostly connected with religious festivals in

different regions of India and depicts episodes from legends of Kṛṣṇa,

Rādhā and the gopīs.

When they first appeared, each of the eight terms had different

connotations in different works but they all signified dance presentations.

They have been examined by D. R. Mankad in his study of drama.3 The

occurrence of these terms in various textual sources, excepting that of

saṅgītaka, has been noted by V. Raghavan. He has also described the

characteristics of the dances represented by these terms while discussing

minor plays in his study of the Śṛṅgāraprakāśa of Bhoja.4 A. K. Warder

has discussed them at length, giving instances from different texts in

which they occur.5 To avoid duplicating these studies, the present work

will deal very briefly with the use of these terms in the early texts. These

terms stand for short dance pieces in which the emphasis falls on

portraying an emotion (bhāva) or creating an aesthetic affect (rasa).6 Not

surprisingly, these terms were mostly used in the kāvya and alaṅkāra

literature, although references are found in other literature as well, but

they are never clearly defined.

3 Mankad,1936, pp. 91-144.

4 Raghavan, 1963, pp. 545-74.

5 Warder,1972,Vol. I, pp.138-68.

6 Raghavan,1978, p. 539.

MOVEMENT AND MIMESIS

157

Dvipadī literally means a composition of two metres or two verses. In

Bhāmaha's Kāvyālaṃkāra this term represents a mimetic presentation:

abhineyārtham (Kalarin.1.24). However, in other works the term appears

to have different meanings. For instance, in the Vikromorvasīya of

Kālidāsa, it is taken as a graceful body movement. In Act IV, Citralekhā

moves about in dvipadikā while looking at different directions and getting

up from a reclining position (Act IV, prose passages following verses 1, 5

and 7). The directions are: प्रवेशान्तरे द्विपदिकया दिशोऽवलोक्य (after entering looks in all directions in dvīpadī);

दिशोऽवलोक्य (afterwards looks in all directions in dvīpadī) and

द्विपदिकयोत्थाय नि:श्वस्य (after getting up in dvīpadī and sighing).7 Each of these stage directions is followed by a verse. It is perhaps the presence

of these verse passages that has prompted Raghavan to take dvipadikā as a laya.8

But again on the same page we find the stage direction: द्विपदिकया

दिशोऽवलोक्य। नि:श्वस्य । साश्रु। (afterwards looking in all directions in

dvīpadī and sighing with tears ). A prose line follows which also refers to

body movements. In his commentary on the Mālatīmādhava Jagaddhara

refers to Mādhava's entrance in dvīpadikā as described in the text: तत:

प्रविशति यथानिर्दिष्टरुपो माधव: (then Mādhava enters in the manner

indicated) and Jagaddhara comments: माधवप्रवेश इह द्विपदिकया

(Mādhava here enters in dvīpadīkā),9 which indicates a slow movement

(गमनमलसं. . . ). Rāghavabhaṭṭa takes dvipadikā as laya or tempo: यं

. . . द्विपदीयामा लयभेद: (here . . . dvīpadī is a division of tempo).10

Raghavan says that "from the laya to the song and from the song to the

dance, the name Dvīpadī has had its semantic extension." Raghavan also

7 Vikromorvasīya, Velankar ed. 1961, pp. 61, 65, 66.

8 Raghavan, 1963, p. 560.

9 लोकविभ्रमयुक्ते तु व्याधिचिन्तासमाश्रिते ।

श्रुतवार्तादिवेलस्ये योज्या द्विपदिका बुद्धे: ॥ Bhandarkar edition, 1876, p.36.

10 Rāghavabhaṭṭa's commentary on Abhijñānaśakuntalā, Kale edition, 1902, p. 8.

158

NṚTYA AND UPARŪPAKA

refers to a dvipadikā metre and a dvipadikā song still current in Telugu and

Tamil.11 Warder refers to dvipadikhaṇḍa as a sub-variety of dvipadī and

cites Harṣa's Ratnāvalī where two actresses are directed to act and sing a

dvipadikhaṇḍa.12 In his Śṛṅgāraprakāśa, Bhoja refers to dvipadī as one of

the features of nartanaka, a minor drama ( padārthābhinayātmaka

preksyaprabandha, that is, a composition which is to be seen and which

expresses the meaning of the words).13 Śārṅgadeva defines dvipadī as a

musical composition sung to karuṇa tāla in the Saṅgītaratnākara, and in the

Nāṭyadarpana Rāmacandra and Guṇacandra take it as a metre. Dvipadī is

described as a laya by Dāmodaragupta in his Kuṭṭanīmata.14 In the

Nartananiṛnaya Puṇḍarīka merely follows Śārṅgadeva in defining dvipadī

(NN. 39b). According to K. S. Upadhyaya, dvipadī is still used in

yakṣagāna, one of the folk dance-dramas of Andhra.15 We find no

instance of dvipadī as a drama or a dance in late medieval literature.

Therefore, it seems that although Bhāmaha defines it as a dramatic

presentation, in later times it was identified with musical composition,

metre and tempo, to which a delicate dance was sometimes performed.

Another term from early times is samyā, which literally means a stick.

In the Nāṭyaśāstra samyā is described as a time-beat (NŚ.31.36-7). The

Hrdayamigamā on the Kāvyādarśa of Daṇḍin describes a dance so named,

which involves the use of a coloured stick (Hrdayamigamā on the

Kāvyādarśa.1. 39). It developed as a group dance which can be identified

with daṇḍarāsaka (rāsa dance with coloured sticks). Daṇḍarāsaka is still

practised in Gujarat. Bhāmaha refers to samyā as a form to be acted

11 Raghavan, 1963, p. 561.

12 Ratnāvalī, Shastri edition,1978, p.6; Warder, Vol.I, 1972, p.153.

13 Raghavan, 1963, p. 555.

14 SR. 4. 213-19; ND. 1929, p. 214; Kuṭṭanīmata, Kāvyamālā ed.,Vol. III, 338, 858.

15 Upadhyaya, Sangeet Natak, No.11, p. 39.

MOVEMENT AND MIMESIS

159

(abhineyārtha) and Daṇḍin as one to be seen (preksyārtha).16 In the Svarasvatīkanṭhābharana, Bhoja describes samyā as one of six kinds of

dramatic presentations, others being tāṇḍava, lāsya, chalika, hallīsaka and rāsa (SKA. 1969, p.124) and he says that chalika becomes samyā when it

relates to kinnara (तदिदं छलिकमेव किन्नरविषयं शम्या - SKA. 1969, p.125). In the Śṛṅgāraprakāśa, it is described as a feature of the nartanaka

type of minor drama (ŚrP. 1963,Vol. II, p. 468). In the latter instance lāsya and samyā both appear as features of nartanaka. The authors of the

Nāṭyadarpana refer to samyā as a lāsyaṇṛtta pertaining to the kinnaras (किन्नरविषयं लास्यं नृतं शम्या - ND. 1959, p.191). Raghavan has cited

many instances from the Nāṭyaśāstra, the Amarakoṣa and the Rāmāyaṇa as evidence that the term means a time-beat, an opinion in which Warder

concurs.17 In the Saṅgītaratnākara samyā is used as a time-beat (SR.5.6; 9). M. Lath quotes Dattila and says that samyā was a concept central to the

gāndharva tāla.18 Except for Bhāmaha, then, samyā is not described by any other writer primarily as a dramatic presentation and Daṇḍin merely

mentions it as a visual presentation. However, in later times the writers followed the interpretation of the Nāṭyaśāstra and in medieval works on

saṅgīta samyā is described as a tāla.

The definition of rāsaka first appears in Bhāmaha who lists it as a dramatic presentation. It is generally described as an elaborate dance in a

number of early texts. Daṇḍin does not mention the term but Kohala (quoted in the commentary on Haravijaya)19 describes it as a play.

Abhinavagupta quotes authorities who treat it primarly as a dance which is performed by a number of women and is full of a variety of rhythm and

tempo. Since he states this while describing ṇṛttāmakaprabandhas,

16Kāvyālaṅkāra, 1.24; Kāvyādarśa, 1.39.

17 Raghavan, 1963, pp.561, Warder, Vol. I, 1974, p.152.

18 Lath, 1978, p.101;Dattilam, 1930, 110.

19 Commentary on Haravijayam, XVII, p.108.; Warder, Vol. I, 1974, p.145.

160

NṚTYA AND UPARŪPAKA

rāgakāvyas and gīyamānarūpakas,20 it would seem that he himself did not

think of rāsaka exclusively as a dance but as a dramatic composition

presented through music and dance movements. A similar term used by

Abhinavagupta is rāsakāṅka (an example being the Rādhāvipralambha by

Bhejjala),21 which means a play in a regional language depicting Kṛṣṇa

and Rādhā. Rāsaka, according to Bhoja, is a padārthābhinayātmaka

preksyaprabandha or minor drama (ŚrP. Vol.II, pp. 468-69) and Bhoja says

that in rāsaka, piṇḍībandhas or group dances forming different patterns

were performed. In his Sarasvatīkaṇṭhābharaṇa and Śṛṅgāraprakāśa

Bhoja says that hallīsaka becomes rāsaka when danced to definite tālas.22

This comment suggests that rāsaka was primarily a dance. The

Nāṭakalakṣaṇaratnakośa (NLRK. 3205-9) describes rāsaka as a one-act

play using a variety of languages and five characters, etc. This text

prescribes delicate and forceful emotions (maśṛṇodāttabhāvabhūṣitam)

for rāsaka but does not mention any tāla and laya. It seems that the author

of this work, Sāgaranandin, takes it primarily as a drama. In his

Nartananirnaya, Puṇḍarīka describes rāsaka as a composition employing

rāsatāla that had four different varieties (NN. A Soc. Ms. 46b).

Rāsaka is an important term which appears to have come down from an

early period and it represents a dance-drama or a dance with dramatic

overtones. In two of the present-day classical styles of India, Manipuri

and Kathak, it is still a major part of the repertoire. In folk-dances in

several regions of India rāsa continues to be performed in some form or

other. Gujarat, Tamilnad and Andhra, for instance, have their individual

regional versions of rāsa.

Skandhaka appears as a dramatic presentation in Bhāmaha who

considers its use to be abhineyārtha: to be acted (KAlauṁ.1.24). However,

20NS. Vol. I, 1956, p.181.

21NS. Vol. III, 1954, pp. 63, 72.

22SKA. p. 125; Śr.P. Vol. II, p. 468; Raghavan, 1963, p.562.

MOVEMENT AND MIMESIS

161

Dandin refers to it as a metre (KĀ.1. 37). Bhoja says that it was composed

in Prākrt (ṢrP. 1963, Vol. II, p. 480). Very little information can be

gathered about it "beyond the fact that it is the Prākrit name for the metre

āryāgīti, a form of gaṇachandas (musical 'bar' metre)."23 It seems to

have been used for a performance that included songs and dances. Warder

describes it as a single musical strophe used in Mahārāṣṭrī. 24

Chalika or chalita appears in Kālidāsa's Mālavikāgnimitra as a.dance

composition that forms part of a drama (चलितं नाम नाट्यम् -Act I, prose

passage following the prastāvanā). Mālavikā dances to a song in prākṛt,

expressing the emotion of love.25 Dandin refers to this form as a

prekṣyārtha kāvya (KĀ.1. 39). Somadeva in his Kathāsaritsāgara

describes chalita as a dance performed by Rambhā in heaven (KSS.

Vol.III, iii. 20). In the Svarasvatīkanṭhābharana, Bhoja describes chalika

as one of six dramatic presentations and states that it creates both vīra and

śṛṅgārarasa (SKA. p. 124-25). In the Ṣṛṅgāraprakāśa, he describes it as a

part of the minor dramatic type called nartanaka (Ṣr.P. Vol.II, p. 468). The

Hrdayamiṅgamā on the Kāvyādarśa follows Kāṭayavema (Raghavan, 1963,

p.556) in interpreting chalika but adds that instrumental music is a feature

of chalika (Hrdayamiṅgamā on KĀ.1.39). In the Bhāvaprakāśana it is

described as an element of prekṣaṇaka, a minor dramatic type (Bh.P. p.

263). Raghavan takes chalika, not chalita, to be the correct older form of

the term (Raghavan, 1963, p. 557). A similar variant of the term occurs in

the Harivaṁśa (II, chapters 88, 89, 93) which mentions chālikya; chapter

89 of the work is called chālikyakṛḍā and relates how at the request of

Kṛṣṇa and Balarāma apsarases dance to vocal and instrumental music

(Harivaṁśa, II, 89.5). The dancers use gestures and dance rāsa

23 Warder, Vol. I, 1954, p.153.

24 Ibid. p.176.

25 Kāṭayavema, the commentator on Mālavikāgnimitra explains it as a piece based on

chala or deceit; Raghavan, 1963, p.556.

162

NṚTYA AND UPARŪPAKA

(Harivaṁśa, II, 89.24, 30) singing songs in different local dialects and

keeping rhythm by the beats of the palms and wearing regional costumes.

The songs tell legends of Kṛṣṇa. Chālikya is also referred to as a song

(Harivaṁśa, II, 89.67). There are more instances of such uses of the term

chālikya.26 The authors of the Nātyadarpana refer to chalita, which they

say delineates vīra, śṛṅgāra and raudrarasa (ND.1959, p.191). In his study

of minor dramas, Warder describes chalika as a form containing four

strophe, called vastu.27 However, the descriptions in the relevant

literature do not warrant the view that chalika was taken as a minor

drama; rather, it was considered to be a performance featuring songs and

delicate body movements that created rasa.

Carcari is yet another term that appears in a variety of meanings in

different texts. In a number of dramas it appears in stage directions that

seem to indicate body movements as well as singing. In the

Vikramorvasīya of Kālidāsa, for instance, a stage direction indicates:

चर्चरीकया विचिन्त्य (thinking with [movements in] carcari), and a prose

passage follows which represents what the king is supposed to say while

seated. From this position he gets up and again the stage direction states:

अनन्तरे चर्चरी (afterwards carcari), and a verse follows which is either to

be recited or sung, accompanied by some body movements.28 In the

Ratnāvalī (Act I, pp. 4, 5-6) carcari appears in a spring festival as a piece

of music with a dance performed to suit it.29 The Karpūramañjarī of

Rājaśekhara describes carcari as a dance (Karpūramañjarī 4. 9). The

Kuvalayamālā of Udyotana does the same.30 Bhoja equates carcari with

26 Ibid. pp.556-58.

27 Warder, Vol. I,1972, p.151. चतुष्पादोदृंवं छलितमू.... (Mālavikāgnimitra, Karmakar ed.

1950, p. 22).

28 Vikramorvasīya, Velankar ed.1961, Act. IV, p. 67.

29 Prabhā on Ratnāvalī, 1978, describes it as a song (Act. I, pp. 22, 30.).

30 Warder, Vol.I, 1972, p.147.

MOVEMENT AND MIMESIS

163

nāṭyarāsaka and describes it as a dance performed in springtime in which

dancers perform in pairs and groups (ŚrP. Vol.II, pp. 468-69). The

Bhāvaprakāśana follows Bhoja (Bh.P. pp. 264-66). In Kuṭṭanīmata,

Dāmodaragupta speaks of carcarī performed in a spring festival (886) and

later refers to it as a song (904). Raghavan describes carcarī as a laya and

later, following Bhoja, he equates carcarī with nāṭyarāsaka.31 Ajay

Mitra Śāstrī has given an extensive account of carcarī, its meaning and

incidence in various places. He equates it with staff-dance and calli dance

(merry making in general).32 Calli or cilli appears in Abhinavagupta's

commentary and in Hemacandra's work but its nature is not clear from the

references. The confusion regarding carcarī is further compounded later

by its use in saṅgīta literature as a tāla.33 However, from the various

references and discussions mentioned here one may conclude that carcarī

was taken as a song performed during festive occasions, particularly in

spring, when not only music but dance was often performed as well.

Lāsya is a term that appears in the earlier literature in very many

different connotations, as we have seen in the extensive discussion in

chapter 4. But it may be briefly noted again in the present context that

initially the term lāsya signified a general rather than a specific class of

movements, movements that were associated with emotion and delicacy.

Among the eight terms discussed here, this is the only one that developed

into a distinct category and was not absorbed in either nṛtya or uparūpaka.

Lāsya, as we have noted before, appeared in the Nāṭyaśāstra as a delicate

movement used in dramatic presentations, and later became identified as a

dance presentation, delicate, graceful and at times erotic as well, which

has survived as a major feature of dancing till the present time.

31 Raghavan, 1963, pp.562-65.

32 Śāstrī, 1975, p. 160.

33 SR. 5.241; 5. 266.

164

NṚTYA AND UPARŪPAKA

Sañgītaka, the last term in this group, seems to refer to a mimetic dance. However, when Vararuci refers to saṅgītaka he means a dramatic presentation involving dance, music and acting performed during the spring festival: भगवानो नारायणस्य भवने मदनसेनया मदनाराधने संगीतके यथारसमभिनीयमाने (in acting out the saṅgītaka named Madanārādhanā with appropriate rāsa in the house of Lord Nārāyaṇa -Ubhayābhisārikā, 1959, 3. 9; pp. 122-23) and कुसुमपुरपुरन्दरस्य भवने पुरन्दरविजयं नाम संगीतकमभिनेतव्यम् (the saṅgītaka named Purandaravijaya is to be enacted in the house of Purandara of Kusumapura -Ubhayābhisārikā 28.7; p.134).

In both instances the author seems to have taken saṅgītaka as a dance-drama. By contrast, in two other instances the term saṅgītaka may have been used by the same author to mean a song or a mimetic dance, but the passages are not clear enough to indicate what the author specifically meant by the term (Ubhayābhisārikā, 28.7, pp.134, 141). In the Mālavikāgnimitra of Kālidāsa saṅgītaka is used as a dance-piece: प्रवृत्तं संगीतकम् ( the saṅgītaka has begun -Mālavikāgnimitra, 1950, p. 24) and a verse follows, to the accompaniment of which Mālavikā performs. It seems that Mālavikā sang and mimed to that verse. In Harṣa's Ratnāvalī (Act I, after verse 5) and in Śūdraka's Mṛcchaṭika (Act I, after verse 8) we find the use of saṅgītaka, which in these instances seems to mean mimetic dancing. Bhoja seems to refer to a dramatic presentation when he says that saṅgītaka should be performed by women (Śṛ.P. Vol.II, p. 466). Śāradātanaya refers to saṅgītaka as a metre (Bh.P. p.295). According to Katz, "The post Nātyaśāstra term itself suggests a concerted form. The word saṅgītaka mentioned in the Ubhayābhisārikā of Vararuci and once in Śrīharṣa's Ratnāvalī was a form of music in which all three elements were present. It is possible that the term saṅgīta came into use alongside or even after this theatrical form."34 From the early medieval period on, saṅgīta is defined in saṅgīta literature as an art composed of three arts: vocal and

34 Katz, 1987, Vol. II, note on verse 22, p. 9.

MOVEMENT AND MIMESIS

165

instrumental music (gīta and vādya) and dance (nṛtta and nṛtya).35 It is

possible that the saṅgītaka mentioned by Vararuci was not a full-fledged

drama but a long piece with song and dance as part of it which required the

performer to mime a story. Warder and Raghavan offer no explanation

but Mankad takes it as a dramatic presentation.36 Since we do not have

any instance of saṅgītaka as a dramatic presentation other than Vararuci’s

unclear references, Katz’s view seems justified.

From the early medieval period onwards, six of these eight terms

disappeared from the literature of dance, with the exception of lāsya and

rāsaka. It is worth considering briefly what happened to the others. We

have seen that all of these terms represented performances that used body

movements set to rhythm and tempo, but with this additional feature in

common, that all of them mimed emotional states or presented stories. It

was this mimetic character that differentiated them from Bharata’s nṛtta.

They were presumably different also from abhinaya performed in the

pūrvaraṅga as Bharata describes it, because they followed rhythmic beats

as opposed to the abhinaya portion of the pūrvaraṅga, which could not have

been set to rhythm since Bharata prohibited musical accompaniment for it

(NŚ.4.276).

It is likely that the performances denoted by the terms discussed above

were developments after Bharata. It is important to remember that for

Bharata nṛtta was the only art form that depended upon both body

movements and tāla and laya, and that nṛtta was expressly defined by

Bharata as a non-representational art. Yet later we find an art called

nṛtya which, like nṛtta, is an art of body movements to the accompaniment

of tāla and laya but, unlike nṛtta, has referential meaning. No doubt it was

to fit this art into Bharata’s conception of the performing arts that Kohala

35 SR. 1.21.

36 Mankad, 1936, p.35.

166

NṚTYA AND UPARŪPAKA

(if we place him after Bharata) added in his account the following

explanation, reported by Abhinavagupta, of the origin of nṛtya:

सन्ध्यायां नृत्यतः शम्भोर्भक्त्याद्रौ नारदः पुरा ।

गीतवान्स्त्रिपुरोन्माथं तच्चित्ततत्त्वस्थ गीतके ।

चकाराभिनयं प्रीतस्तततस्तण्डं च सोऽब्रवीत् ।

नाट्योक्त्याभिनेयेदं वत्स योजय ताण्डवम् ।

(A.B.on NS.Vol.I, p.180)

In time past, while Sambhu was dancing in the

evening, Nārada sang, immersed in devotion, a song

about [Śiva's] victory over Tripura, and setting his

mind in the song. He [=Śiva] mimed it [i.e., the

content of the song]. Being pleased [with this, He]

said, Tandu, my boy! Please add this tāṇḍava to acting

as described in the Nātya [śāstra].

Evidently, an expansion of Bharata's conception of dancing was in

progress in the thought of his followers, and the eight terms examined here

marked that process. They denoted minor art forms used in a drama, not as

a part of the pūrvarāṅga, but as modes of graceful miming in the main body

of the drama. That the stage directions call for rhythm, tempo and

presumably songs to accompany body movements in various situations,

suggests that abhinaya and ṇtta, variously combined into a new type of

presentation element, became part of a dramatic performance. The terms

chalika, samyā, dvipadī, skandhaka and carcarī seem to have been

associated more with tempo and singing, while rāsaka, as found in the

early works, began as a dance-drama and later on.became identified in

medieval texts as a dance and is still performed as a dance in various parts

of India. Lāsya, the graceful dramatic presentation mentioned by

Bharata, later became a graceful and delicate style of dance and the term

came to be used generally to denote all forms of performance which were

MOVEMENT AND MIMESIS

167

particularly known for their delicacy and grace. Such forms included

many that had been earlier denoted by some of the eight terms discussed

here. The term saṅgītaka seems gradually to have evolved into the term

saṅgīta which came to denote the three allied arts of vocal and

instrumental music and dance taken together. This change took place in

medieval times and saṅgītaka ceased to be used in the narrower sense of a

dance-drama or dramatic presentation.

Raghavan’s view of the history of these terms is worth noting here.

According to him, the terms that were associated entirely with laya and

metre gradually became associated with songs and then with dances,

which were set to those songs.37 Later, only the term lāsya came to used

generally for all dance movements that were particularly graceful and

delicate. From the tenth century onwards nrtya appears in the literature

first as a term for minor drama38 and then, from the thirteenth century, as

mimetic dance.39 Since nrtya could have two variations, lāsya and

tāṇḍava, if a dance had qualities of lāsya and/or nrtya, that is, delicate

and/or mimetic features, any other term became unnecessary to denote

them, and the terms carcarī, chalika, samyā, dvipadī and skandhaka

became obsolete in dance literature even though they continued to be used

in the literature of music. Rāsaka remained a variety of group dance. It is

interesting to note that from the time the terms lāsya and nrtya begin to

appear in dance literature, carcarī, chalika, samyā, dvipadī and skandhaka

are no longer found in association with dancing.

The growing importance of the broad genre known as nrtya is first seen

in the Dasarūpaka of Dhananjaya in the tenth century. Dhananjaya

describes nrtya as: भावाश्रयं नृत्यम् (nrtya is dependent on emotion -DR.1.

9). But then he adds that it is padārthābhinaya, which means that nrtya

37 Raghavan, 1963, p.561.

38 DR. 1.9.

39 SR. 7.26-7.

168

NṚTYA AND UPARŪPAKA

calls for mimimg the meaning of words (DR.1. 9). What, then, did the

term nṛtya exactly mean in the Daśarūpaka? If we are to judge by

Dhanika's commentary on the Daśarūpaka, nṛtya was an art that had some

features of drama without being drama proper. Dhanika mentions seven

divisions of nṛtya: ḍombī, śṛṅgadita, bhāṇa, bhāṇī, prasthāna, rāsaka and

kāvya (DR p. 8), all of which are known from other texts to be minor types

of drama. Since Dhanika takes them to be varieties of nṛtya, he evidently

understands Dhanañjaya to take nṛtya as an art with dramatic overtones.

We may note in passing that Abhinavagupta never uses the term nṛtya.40

Bhoja does not use the term nṛtya but he does provide support for defining

nṛtya as a dramatic art. He places all dramatic representations under one

term, preksyaprabandhas (compositions to be seen) which he divides into

two classes, padārthābhinaya (i. e., one that expresses the meaning of a

word) and vākyārthābhinaya (i. e., one that expresses the meaning of an

entire sentence - Śr.P. Vol.II, pp. 461, 466). If, then, he takes all

padārthābhinaya to be dramatic, he must take Dhanañjaya's nṛtya as

dramatic as well because it is padārthābhinaya (DR. 1.9). That here he

does have in mind the same types of composition as those in Dhanañjaya's

conception of nṛtya is proved by the fact that his own list of types of

padārthābhinayātmalaka preksyaprabandhas -twelve in all- includes the

seven mentioned by Dhanañjaya's commentator Dhanika. While nṛtya

was considered to be a dramatic art, it was clearly not drama proper. This

becomes evident from Dhanañjaya again, for he distinguishes nṛtya from

rūpaka—which was synonymous with drama-by calling nṛtya bhāvāśraya

and rūpaka rasāśraya. Dhanika strengthens the distinction by commenting

that rūpaka is vākyārthābhinaya. Since nṛtya is padārthābhinaya it has a

40

One may well ask why, if nṛtya was at all a valid category for Dhanañjaya, it is not

mentioned by his contemporary Abhinavagupta. The answer is that Abhinavagupta

was commenting on Bharata's views and since Bharata does not speak of nṛtya, there

was no occasion for Abhinavagupta to include it in his commentary.

MOVEMENT AND MIMESIS

169

narrower scope of dramatic expression than rūpakas and hence could be

no more than minor dramatic presentation. That rasa was a broader

category than bhāva was of course a principle established by Bharata

himself: नानाभावोपगमाद् रसनिष्पत्तिः (NS. Vol.I, p. 287). Vākya has a

broader scope inasmuch as it is a compound of many bhāvas, while

padārthābhinaya is a representation of only one bhāva.

That Dhananjaya was using nṛtya to denote minor dramatic types is

supported also by Sāradataṇaya’s use of the derivative word, nṛtyabheda,

to signify types of minor dramas that he classifies under

padārthābhinayātmaka rūpakas (Bh.P. 255). He also refers to them as the

"other" rūpakas,that is, other types of dramatic presentation. The same

use also occurs in the Nāṭyadarpana (अन्यान्यपि च रूपकाणि – 1959, pp.

190-92) of Rāmacandra and Gunacandra. The list of rūpakas in the

Nāṭyadarpana matches the list given by Bhoja. Sāgaranandin, another

author who discusses minor dramas, has no name for the class but gives the

same names of the individual types. Nṛtya and nāṭya are used

synonymously in the Amarakoṣa, which shows that at some point in the

development of the performing arts in India nṛtya was equated with

nāṭya.41 The association continues today in the terms Bharatanāṭyam and

Mohiniāṭṭam ('āṭṭam' being the Tamil for the Sanskrit nāṭya).

It was not till the 13th century that the term nṛtya came to be

dissociated from nāṭya and to be used exclusively for the art of dancing.

The first work to define nṛtya categorically as 'dance' was the

Sangītaratnākara (SR. 7.26-7). From that time on saṅgīta literature also

began to divide dancing into two types, nṛtta and nṛtya, the former being

abstract dancing and the latter mimetic dancing (SR. 7.26-8). By the 14th

century the sister arts of dance and drama had become clearly separated

as different genres and dance was no longer considered to be a kind of

minor drama. The term used from this time on for minor dramatic types

41 Colebrooke, 1808, p. 42.

170

NṚTYA AND UPARŪPAKA

was uparūpaka, a term first used by Viśvanātha (SD. 6.6). The later

medieval texts all followed Śārṅgadeva in using nṛtya for mimetic dance

while they followed Viśvanātha in using uparūpaka for drama.

To sum up, then, in alaṅkāra literature, such influential authors as

Dhanañjaya, Dhanika and Śāradātanaya use the term nṛtya to mean minor

types of drama. Bhoja does not use the term nṛtya at all, but it is clear that

he does recognize a dramatic type of performing art as a definite category

which expresses meaning. He draws a distinction between two types of

preksyaprabandhas (compositions to be seen) by designating one as

vākyārthābhinaya (acting that brings out the sense of entire sentences) and

padārthābhinaya (acting that brings out the sense of words). Bhoja’s

concept of this second type of composition parallels Dhananjaya’s

definition of nṛtya as an art that employs padārthābhinaya. Further, Bhoja

lists the same nṛtyabhedas those we find in Dhanika and Śāradātanaya,

thus tacitly endorsing the categorization of nṛtya as a dramatic type.

Turning to saṅgīta literature, that is,works on dance and music, we find

that Mānasollāsa is the first work of this kind to mention nṛtya, which

forms the title of an entire section called Nṛtyavinoda. The author,

Someśvara, uses the terms nṛtya and nartana synonymously (16.4.950) and

includes nāṭya among the six varieties of nartana (16.4.959-60). Stating:

साम्प्रतं वर्णयिष्यामि नृत्यं लक्षणसंयुतमः (I shall now describe nṛtya with its

characteristics - 16.4.950), Someśvara describes nartana as:

नाट्यं लास्यं ताण्डवं च लाघवं विषमं तथा ।

विकटं चेति निर्दिष्टं नर्तक(नं) षट्प्रकारम् ॥

(Mānas. 16.4.959-60)

Nartaka(na) is said to be of six kinds: nāṭya, lāsya,

tāṇḍava, lāghava, viṣama and vikaṭa.

Throughout this work he uses nṛtya as a generic term for dancing.

MOVEMENT AND MIMESIS

171

The sense of nṛtya is more restricted in the Saṅgītasamayasāra of

Pārśvadeva, which seems to take nṛtya as a deśī or regional form of

dancing (S.Sam. 7.130-32), although it gives no precise definition. The

first definition of nṛtya in saṅgīta literature came from the

Saṅgītaratnākara of Śārṅgadeva, according to whom nṛtya is one of three

divisions of nartana, a term he uses in the general sense of stage

presentation. Here he differs from Someśvara in taking nartana as the

broader term, not nṛtya, which he considers to be one of the divisions of

nartana, the other two being nāṭya and nṛtta (SR. 7.2-3). He states that

nāṭya is both vākyārthābhinayātmaka and padārthābhinayātmaka,

although it employs all four types of abhinaya, and that it creates rasa and

bhāva (SR. 7.33-4). Nṛtya is defined as body movements that express

emotion: आङ्गिकाभिनयैरेक भावानेव नयनकृत यत् । तन्नृत्यं (that is nṛtya

which expresses emotion by miming through body movements - SR. 7.26-

7). It is interesting too that about the same time Jāya, the author of

Nṛttaratnāvalī, defines nṛtya as bhāvāśraya and padārthābhinayātmaka

(NR.1.50) and cites examples of nṛtya as: नृत्यं श्रीगदितादि स्यात् (nṛtya is

[represented by]śṛṅgadita etc. - NR.1.52). As in the Saṅgītaratnākara,

here the emphasis on bhāva and padārthābhinaya suggests that at this stage

nṛtya was viewed as a mode of dramatic expression distinguished by

singing and dancing, although it was no longer thought of as a category of

drama. From this time on, almost all the works that are concerned with

music consider nṛtya primarily as a form of dancing. The only exception is

the Saṅgītadāmodara which at one point classifies as nṛtya all twenty-

seven types of drama recognized by it (SDām. p. 79). However, when the

author defines the minor dramatic types, he calls them uparūpakas rather

than nṛtya. Moreover, when he comes to describe dance movements, he

classifies them under aṅgahāra and nṛtya, taking nṛtya as regional dance

forms. Saṅgītopaniṣatsāroddhāra gives an unusual definition of nṛtya by

identifying it as a dance danced only by women (SUS. 5.8), although the

author uses the term as a generic label for the art of dancing (SUS. 5, 6).

172

NṚTYA AND UPARŪPAKA

In the Saṅgitacandra nṛtya is again taken to represent the art of dancing in general (SC. 389). Similarly, Mahārāṇā Kumbhā in his Ṇṛtyaratnakośa discusses nṛtya as a form of dancing and refers to it as a mārga form although he does not define it (NRK.1.1.291). Moreover, he uses the term nṛtyabheda to indicate different varieties of dance (NRK.1.1.466; 4.3.106). We may recall that the term nṛtyabheda was earlier used by Dhanika and Śāradātanaya to denote minor types of drama. The change in the meaning of nṛtyabheda thus shows that by Mahārāṇā Kumbhā’s time nṛtya had come to be identified as a form of dancing. It would seem that to begin with, nṛtya was regarded as part of a dramatic presentation, but later it moved away from the domain of drama, and though it retained its mimetic character, it no longer employed speech and evolved into a dance form.

Closely associated with nrtya is the term uparūpaka, which is a category of major significance in dance literature. We have already noted that the term came into existence fairly late, at the very end of the śāstric tradition of alaṅkāra. In these later works it denotes certain types of drama but it is not clear precisely which types. We may recall that Bharata himself deals only with the major types of drama. But that there were others is clear from the Nāṭyaśāstra, which contains a reference to a treatise that Kohala was supposed to be writing, presumably covering grounds untrodden by Bharata.42 That these included the minor types of drama left out of his description by Bharata is evident from the commentary of Abhinavagupta who mentions these types under various class names when speaking of Kohala. Abhinavagupta further cites other authorities who evidently dealt with minor types of drama.43 Since we have none of these texts, we have no knowledge of the dramatic types they discussed. We do know that they existed but it is often hard to distinguish them from the nṛtya forms on the basis of their treatment in the texts.

42 NŚ 36.65 (Kaśī), NŚ 37.18 (G.O.S.), NŚ 37.18 (Kāvyamālā).

43 NŚ Vol. I, pp. 171, 181, 182

MOVEMENT AND MIMESIS

173

Bhāmaha is the earliest extant writer of alaṁkāra literature to name some of the types of stage presentations of a dramatic character. As already pointed out, although these types (dvipadī, samyā, rāsaka and skandhaka) are stated to be abhineyārtha, that is, to be acted and in that sense similar to major drama, they are in fact not even minor dramatic forms but mimetic dance pieces. Daṇḍin, the next writer on alaṁkāra, defines the terms lāsya, chalika and samyā as prekṣyārtha (i.e., meant to be seen), which again seems to indicate that these were mimetic dances, although neither Bhāmaha nor Daṇḍin describes the dances. The minor dramatic types are named by Abhinavagupta, who quotes various authorities to list the names as well as their characteristics. He lists ten types in all: ḍombikā, bhāṇa, prasthāna, ṣidgaka, bhāṇikā, preraṇa, rāmākrīḍa, hallīsaka, rāsaka (all these being taken from older authorities) and kāvya (from Kohala). Even from Abhinavagupta's brief discussion of these types it is evident that they were short dramatic pieces that presented emotional situations by miming short stories and required at least one, sometimes several female dancers, their dances being set to particular tāla and laya. In some of these presentations male performers also took part.

Abhinavagupta was not the only early writer to refer to these types of stage presentations. Vātsyāyana (believed to be from the 3rd century A.D.)44 in his Kāmasūtra mentions hallīsaka, nāṭyārāsaka, prekṣaṇaka and goṣṭhī.45 Kumārila (believed to have lived between 590-650 A.D.) 46 in his Tantravārtika mentions dvipadī and rāsaka.47 Since both Vātsyāyana and Kumārila pre-date Abhinavagupta, the latter's references show a continuing tradition of recognizing minor dramatic

44 Warder, Vol. I, 1972, p. 11.

45 KS. ; I.6, II.10.25.

46 Krishnamachariar, 1974, p. 618

47 Tantravārtika, Chowkhamba edition, p. 279.

174

NṚTYA AND UPARŪPAKA

types, although we have no detailed characteristics since neither

Vātsyāyana nor Kumārila elaborate on them. The Agnipurāṇa48 lists 27

such varieties of dramatic presentation but does not give them a class

name. From the names given we can identify the first ten as belonging to

the class of rūpakas or major types of drama. The other seventeen are

minor dramas (AP. 338.2-4). All of these are presumably the abhineya

kāvyas mentioned in the preceding chapter of the text (AP. 337.1, 39) but

they are not defined anywhere. We have noted that Dhanika’s avaloka on

the Daśarūpaka names seven classes of nṛtya (DR. p. 8), but he does not

explain what they are. Abhinavagupta’s brief discussion of ten minor

dramatic types remains the only descriptive account of these little known

forms till the time of Bhoja, from whom we get a fuller picture. He lists

twelve types. He further mentions dombalikā, which may be the same as

the dombikā mentioned in the Abhinavabhāratī.49 The Nāṭyadarpaṇa

describes fourteen such types, calling them 'other' kinds of rūpakas and

following Bhoja in describing them.

A different approach is taken by Hemacandra, who places minor

dramas under the class geya (KAnu.H. pp.327-29). In this he is closely

followed by Vāgbhaṭa (KAnu.V. p.18). In his commentary on his own

work Vāgbhaṭa uses the term geyarūpaka as the class name. Both

Hemacandra and Vāgbhaṭa quote the Abhinavabhāratī for the views of the

cirantanas; they also quote Bhoja’s lines on goṣṭhī. Their lists include

śṛṅgadita, to define which they add the following lines:

यस्मिन् कुलाङ्गना पत्या: सख्यग्नै वर्णयेद् गुणम् ।

उपालम्भं च कुर्वे गेये श्रीगदितं तु तत् ॥

(KAnu.H. p.446;K.Anu. V. p.18)

That geya in which a virtuous woman

48De, 1960, p. 99

49 Raghavan, 1963, p. 567.

MOVEMENT AND MIMESIS

175

describes the qualities of her husband and

censures him as well in front of her friend, is

known as śṛṅgadita.

The Nāṭakalakṣaṇaratnakoṣa lists fifteen minor types of drama and

describes them briefly (pp. 126-34). The author includes two types called

nāṭikā and toṭaka which, however, he does not place in any particular

category. Śāradātanaya provides the longest list by far of minor dramatic

types, all twenty of which he describes (Bh.P. p. 255). The Sāhitya

Darpana lists eighteen, calling them uparūpakas and describing them in

detail (SD. 6.6).

The nature of uparūpakas, their names and their numbers have been

discussed by several modern scholars. Mankad made the first detailed

study of minor types of drama that involved dancing, comparing their

characteristics in tabular form.50 However, he had insufficient data and

occasionally arrived at observations that have since proven to be incorrect.

That lacunae, it is hoped, will be remedied by the comparative

descriptions provided in the present chapter, based on additional material.

It will also help to distinguish the two genres that we are considering here:

one that employed only body movements and carried dramatic overtones;

and a second which used both body movements and speech as in drama

proper, though forms of this genre were narrower in scope than major

dramatic types. The first kind consisted of mimetic dances and, from the

time of Śārṅgadeva, were designated as nṛtya. The second kind were

minor types of drama which were distinguished from the nṛtya forms by

being termed uparūpakas.

Besides Mankad two other modern scholars have dealt with

uparūpakas extensively. V. Raghavan has discussed them at several

50 Mankad, 1936, pp. 129-31.

176

NṚTYA AND UPARŪPAKA

places,51 most exhaustively in his study of Bhoja's Śṛṅgāraprakāśa. Bhoja

lists twelve types of major dramas, of which two, nāṭikā and saṭṭaka, are

regarded as minor types by other early writers. Raghavan's discussion of

uparūpakas includes the twelve minor types termed

padārthābhinayātmaka preksyaprabandha by Bhoja. In course of his

discussion Raghavan also cites the views of other writers on these types.

The other important work in which minor types of drama are studied is

by Warder (1972, Vol. I, pp. 137-68). He takes each type of minor drama

and dance mentioned in alaṅkāra literature and lists what the different

authorities have said about it, thus offering an invaluable compendium of

the material available. It is unnecessary to duplicate here the information

given by Warder but since he restricts himself to alaṅkāra literature for

descriptions of the types, it is necessary to draw material from dance and

music manuals. This has been attempted here in two parts. First, the

references to minor dramatic types in all relevant sources-from both

alaṅkāra and saṅgīta literature-have been tabulated, with the sources

arranged chronologically. Second, the descriptions of each type as they

appear in the sources have been summarized, noting in particular the

major differences if any. This survey will thus help in discovering how the

understanding of the minor dramatic types evolved through time.

Before going on to the descriptions of nṛtya and uparūpaka, which

included musical plays, dance dramas, dramatic dances and mimetic

dances, we may briefly note the many terms used by the early authors for

this broad spectrum of dramatic presentation, which often confuse by their

variety. Abhinavagupta uses the following terms in the Abhinavabhāratī:

rāgakāvya [a musical play] (NŚ.Vol. I, pp. 172, 174,181,182), nṛttakāvya

[a dance drama] (pp. 175, 177), rāgadarśanīya [a musical (play), to be

seen] (p.172), gīyamānarūpaka [a play to be sung] (pp. 171, 175, 176, 180)

and nṛtyapradhānarāgakāvya [a musical play presented principally

51 Raghavan, 1964-5, 1966-7, pp. 31-54; 1963, pp. 545-74.

MOVEMENT AND MIMESIS

177

through dance] (p.180). Dhanika uses the term nṛtyabheda for minor

dramatic presentations (DR. p. 8). Śāradātanaya refers to minor dramatic

presentations both as rūpakas [dramas] (Bh.P. p. 221) and nṛtyabhedas

(Bh.P. p.181, 255). Bhoja calls them padārthābhinayātmaka

preksyaprabandhas [a composition to be seen that expresses the meaning

of words] (Ṣr.P.Vol.II, pp. 466-69). Both Vāgbhaṭa and Hemacandra

refer to them as geya [to be sung] (K.Anu.H. pp.327-29;K.Anu.V. p.18).

Rāmacandra and Guṇacandra call them simply the other rūpakas [plays]

(ND. p. 190). Sāgaranandin lists the types without placing them under any

class (NLRK. pp. 113-14, 126-34) and Viśvanātha calls all of them

uparūpakas [minor dramas] (SD.6.6). The following list shows which

types appear in what sources. The types will then be taken individually

and their descriptions summarized.

MINOR DRAMATIC TYPES

Source

Types

Abhinavagupta

dombikā, bhāṇa, prasthāna, śidgaka, bhāṇikā,

preraṇa, rāmākrīḍa, hallīsaka, rāsaka and kāvya

(A.B. on NŚ. Vol. I, pp.171, 181-2).

Vātsyāyana

hallīsaka, nāṭyarāsaka, preksanaka, goṣṭhī

(Kāmasūtra, II.10. 25, p.154).

Kumārila

dvipadī, rāsaka (Tantravārtika, p. 279).

Bhāmaha

dvipadī, rāsaka, śamyā, skandhaka (KAlam. 1.

24).

Daṇḍin

lāsya, chalita, śamyā (KĀ.1. 39).

Agnipurāṇa

troṭaka, nāṭikā, saṭṭaka, śilpaka, karṇa,

durmallikā, prasthāna, bhāṇikā, bhāṇī, goṣṭhī,

hallīsaka, kāvya, śrigadina(ta), nāṭyarāsaka,

rāsaka, ullāpyaka, preṅkhaṇaka (AP. 338.2-4).

178

NṚTYA AND UPARŪPAKA

Dhanika

ḍombī, śṛṅgadita, bhāṇa, bhāṇī, prasthāna,

rāsaka, kāvya (DR. p.8).

Bhoja

śṛṅgadita, durmilitā, prasthāna, kāvya

(citrakāvya), bhāṇa (śuddha, citra, saṁkīrṇa),

bhāṇikā, goṣṭhī, hallīsaka, nartanaka,

prekṣanaka, rāsaka, nāṭyarāsaka.(Śr.P. Vol. II,

pp. 466-69).

Rāmacandra and Guṇacandra

saṭṭaka, śṛigadita, durmilitā,

prasthāna, goṣṭhī, hallīsaka, nartanaka,

prekṣanaka, rāsaka, nāṭyarāsaka, kāvya,

bhāṇaka, bhāṇikā (ND. 1959, pp.191-92).

Hemacandra

ḍombikā, bhāṇa, prasthāna, śiṅga, bhāṇikā,

preraṇa, rāmākrīḍa, hallīsaka, rāsaka, goṣthī,

śṛṅgadita (KAnu.H.p.327-29).

Vāgbhaṭa

ḍombikā, bhāṇa, prasthāna, bhāṇikā, preraṇa,

śiṅgaka, rāmākrīḍa, hallīsaka, śṛṅgadita, rāsaka,

goṣṭhī (KAnu. V. p.18).

Sāgaranandin

goṣṭhī, saṁlāpa, śilpaka, prasthāna, kāvya,

hallīśaka, śṛṅgadita, bhāṇikā, bhāṇī, durmallikā,

prekṣanaka, saṭṭaka, rāsaka, nāṭyarāsaka,

ullāpyaka (NLRK. pp.126-34: 3026-3216).

Śāradātanaya

toṭaka, nāṭikā, goṣṭhī, sallāpa, śilpaka, ḍombī,

śṛgadita, bhāṇa, bhāṇī, prasthāna, kāvya,

prekṣanaka, nāṭyarāsaka, rāsaka, ullopyaka,

hallīśa, durmallikā, kalpavallī, mallikā,

pārijātaka (Bh.P. p. 255).

Viśvanātha

nāṭikā, troṭaka, goṣṭhī, saṭṭaka, nāṭyarāsaka,

prasthānaka, ullāpya, kāvya, preṅkhaṇa, rāsaka,

saṁlāpaka, śṛṅgadita, śilpaka, vilāsikā, hallīśa,

durmallikā, prakaraṇī, bhāṇikā (SD. 6.6).

MOVEMENT AND MIMESIS

179

Śubhańkara nāṭikā, prekṣaṇa, toṭaka, śāṭaka, goṣṭhī, samlāpa, śilpa, bhāṇī, hallīśa, rāsaka, ullāpaka, śrīgadita, prasthāna, nāṭyarāsaka, durmallikā, lāsikā, kāvya (SDām. p. 79-80).

In performing most of the types listed above the style of presentation employed is the one known as kaiśikīvṛtti, which Bharata defines in the following verse:

या श्लक्षणनैपुण्यविशेषचित्रा

स्त्रीसंजुता या बहुनृतगीता ।

कामोपभोगप्रभवप्रचारा

तां कैशिकी वृत्तिमुदाहरन्ति ॥

(NŚ. 20. 54-55)

That which is graceful, varied with special costume and make-up, performed by [lit:added with] women, which [contains ] a lot of dance and song [and is] prominent in the enjoyment of passion is known as kaiśikīvṛtti.

This suggests that when a dramatic type calls for kaiśikīvṛtti it also requires singing and dancing. But the descriptions in the texts show that not all minor dramatic types could be called musical plays that used dancing, nor did they use all the features of kaiśikīvṛtti. The features of these types are summarized here from the sources; only those sources have been mentioned that either give the representative view or markedly differ from other sources.

DESCRIPTIONS OF MINOR DRAMATIC TYPES

1. Ullāpyaka: This term is variantly spelt ullāpaka, ullopyaka, ullāpya etc. It is first mentioned in the Agnipurāṇa as a minor drama. But, as Warder points out, it is understood in the NŚ as a kind of song (NŚ. Vol.

180

NṚTYA AND UPARŪPAKA

IV, 31.194, 221) and Abhinavagupta does not connect it with a musical

play (Warder, 1972, Vol. I, p.165). Sāgaranandin states that it includes

songs, the hāsya, śṛṅgāra and karuṇa rasas, a three-act structure and is

similar to śilpaka, an example being Devīmahādevam (NLRK. 3213-

16). Bh.P. has a definition similar to the NLRK, citing the

Gāndharvanirnaya, and gives two examples of the type,

Devīmahādevam, and Udāttakuñjaram (Bh.P. p.266). SDām. gives a

similar definition but emphasizes the need for a great number of songs

(p. 95). From this it would seem that by Śubhaṅkara's time the type had

evolved into a rāgakāvya.

2. Kalpavallī: This is described in the Bh.P. (p. 268) as a form that arouses

the śṛṅgāra and hāsya rasas, has an exalted hero and a supporting hero

(udāttanāyaka and pīthamarda upanāyaka). The situation it portrays is

that of the heroine waiting for the hero or going to meet the hero. It is

characterized by the use of dvipadīkhaṇḍa, rathyātāla, vāsakatāla,

three layas, ten lāsyas and sandhis. The example given of this form is

Māṇikyavallikā.

3. Karṇa: Agnipurāṇa (AP.338.3) is the only text to mention the name

giving no description.

4. Kāvya: Abhinavagupta quotes Kohala (NŚ.Vol. I, p.182) on rāgakāvya

or kāvya and treats it separately, distinguishing from other types of

minor dramas that are basically mimetic dance pieces (NŚ. Vol. I,

p.181). According to Abhinavagupta, a kāvya should be presented in a

variety of tempi to the accompaniment of rāgas and should have a well-

constructed plot, as exemplified by Rāghavavijaya or Mārīcavadha,

which are performed, respectively, to the ṭhakka and kakubha rāgas.

Dhanika includes kāvya in his list of nṛtyas, which suggests that he

associated it with music and dance although he does not define it.

Bhoja refers to kāvya and citrakāvya in his list of minor plays. He

differs from Abhinavagupta in stating that kāvya is set to one rāga but

citrakāvya to several. Bhoja also gives technical details regarding the

MOVEMENT AND MIMESIS

181

tāla and mātrā of the rāgas. The same information is repeated in the

Nāṭyadarpana (ND. 1959, p. 191). Sāgaranandin suggests more

technical details of tāla, vṛtti, rasa and sandhi in this minor play and

cites Utkanṭhitanāmdhavam as an example (NLRK.3151). The author

of the Bh.P. gives the longest description of kāvya. The description is

similar to that by Sāgaranandin but mentions in addition the use of

dvipadīkhaṇḍa and, occasionally, of lāsya, as seen in Gauḍavijaya.

Another work mentioned in the same context is Sugrīvakelana, which

exemplifies a variant of the type, one that has a brilliant hero, a young

heroine and employs joyful speeches (Bh.P. pp. 262-3). The author also

includes in his description of rāsaka two lines from Bhoja's definition

of kāvya (Bh.P. p. 265). Viśvanātha prescribes lāsya and śṛṅgārarasa

for kāvya, which he considers to be a type of one-act play in which

khaṇḍamātrā, dvipadī and bhagnatāla should be used but not the

ārabhaṭīvṛtti. An example is Yādavodaya (SD. 6.284-85).

Subhankara's description is similar to Viśvanātha's (SDām.p. 96).

5. Goṣṭhī: This does not appear in Abhinavagupta's account, nor in

Dhanika's. Bhoja was the first writer to include it in his list of minor

dramas. According to him, the purpose of goṣṭhī is to show the young

Kṛṣṇa sporting with cowherds and milkmaids (ṢṛP.Vol. II, p. 468). This

view is repeated in the ND (1959,p. 191). The author of the NLRK

calls goṣṭhī a one-act play that lacks garbha and avamarśasandhi and is

performed in the kaiśikīvṛtti (NLRK. 3026-27). Śāradātanaya's

description is quite elaborate: goṣṭhī is a one-act play, it is performed

in the kaiśikīvṛtti, and it celebrates the deeds of Kṛṣṇa, particularly his

killing of demons (Bh.P.p. 256). A similar definition is given by the SD,

which mentions the one-act structure, kaiśikīvṛtti and śṛṅgārarasa (SD.

6. 274-75). Kaiśikīvṛtti is mentioned also in the SDām (p. 93). All of

these accounts agree that these performances involve a number of

performers both male and female and employ numerous songs and

dances.

182

NṚTYA AND UPARŪPAKA

6. Ḍombī, ḍombikā: Abhinavagupta makes several references to this type

as a ṇtyakāvya (NṢ.Vol.I, pp. 166, 180, 181, to cite a few). According

to the ancient authorities he quotes, it is a dramatic dance composition

in four scenes, intended to please the king (NṢ.Vol.I, pp. 175,181,188).

Abhinavagupta gives two examples: Guṇamālā and Cūḍāmaṇi

(NṢ.Vol.I, pp.171,175; Vol. IV, p. 271). The Bh.P. (p. 257-58) repeats

AB as do Hemacandra and Vāgbhaṭa (KAnuH. pp.327-29; KAnuV.

p.18) and Dhanika mentions the type without providing much

information. Warder has a detailed discussion on ḍombī (Warder,

1972, Vol. I, pp.156-7).

7. Toṭaka, troṭaka: This is a variety of nāṭaka ( NLRK. 2766, BhP..238;

SD.6.273). Although this is not a musical piece, nor a dance type, and

although the writers do not prescribe dance and music for it, the

example they give of the form is Vikramorvasīya, which does include

dance and music. The Agnipurāṇa includes the form in its list of

dramatic types (AP. 338.2) and SDām. describes it as a primarily

verbal drama (SDām. p. 92).

8. Durmallikā, durmilitā: This is mentioned in AP (338. 2), Ṣṛ.P. (Vol. II,

p. 466) ND (p. 191), NLRK (3187), SD (6. 303) and SDām. (p. 69).

This is a humorous play in four acts which employs kaiśikivṛtti. The

story involves a 'stolen love' and a deceitful female messenger whose

function is interpreted by Raghavan as blackmail. According to

Raghavan this is a vulgar performance. No author has cited any

example. The reason, according to Warder, may be that this kind of

performance did not attract scholarly attention. Raghavan and Warder

discuss this form in detail (Raghavan, 1963, p. 547; Warder, 1972, Vol.

I, pp. 142-3).

9. Nartanaka: This seems to be a mimetic dance in which the dancer

mimes, through graceful movements, the meaning of the words of a

song. According to the Ṣṛṅgāraprakāśa, it has four varieties, namely,

śamyā, lāsya, chalika and dvipadī (Ṣṛ.P. Vol. II, p.468; Raghavan, 1963,

pp.555-56). The same description is repeated in the Nātyadarpana (1959, p.191). Raghavan calls it a nautch dance (Raghavan, 1963, p.

555).

10. Nāṭikā: This is similar to nāṭaka and employs songs and dances (A.B. on the NŚ. 1956,Vol.I, p. 436). Examples can be taken from all periods,

such as Pratijñāyaugandhārāyaṇa, Ratnāvalī(SD. 6.269-72, AP.338.2, Bh.P. p. 243-4, SDām. p. 91). A variant form of the term is nāṭī, which

appears in the list of twelve rūpakas in ND ( 1959, p. 7).

11. Nāṭyarāsaka: This is mentioned in the Agnipurāṇa and described in Śr.P, which is followed by ND. Śāradātanaya, Sāgaranandin, and

Viśvanātha treat it as a one-act play. According to Bhoja nāṭyarāsaka is also known as carcarī. Nāṭyarāsaka is described as a springtime

dance performed by a group of female dancers (Raghavan, 1963, pp. 562-67), requiring formation dancing or piṇḍibandhas. Warder points

out that Bhoja may have confused rāsaka and nāṭyarāsaka: "he inserts the description of the nāṭyarāsaka into the middle of that of the rāsaka,

but this may be the result of manuscript corruption" (Warder 1972, Vol. I, p.146). The term nāṭyarāsaka suggests some kind of dramatic

content but the description shows it as a dance form. Bh.P. has a long description (pp. 263-65) following Bhoja which includes formation

dancing or piṇḍībandhas. It also gives carcarī as an alternative name for the type. Rāmacandra and Gunacandra describe it as dance danced

by women to the rāga vasanta (ND.1959, p. 192). Sāgaranandin says that this form employs a number of lāsyāṅgas and a variety of rhythm

and tempo, as for instance in the composition Vilāsavatī(NLRK. 3210-12). Viśvanātha similarly suggests the use of ten lāsyāṅgas and a

variety of rhythm and tempo (SD. 6. 277-9), as in Narmavatī and Vilāsavatī. SDām. follows the definition of SD.(SDām. pp. 95-6).

12. Pārijātaka, pārijātalatā: This is a one-act play mentioned only by Śāradātanaya (Bh.P. p. 268) who says that it requires a number of

female dancers and the performance of daṇḍarāsaka, citing as

example of the type Gaṅgātarangikā. Warder thinks there may have

been some confusion between pārijātaka and daṇḍarāsaka.

13. Prakaraṇikā, prakaraṇī: This is a species of prakaraṇa (A.B. on the

NŚ.Vol. II, p. 436). No examples are given or found. ND., which is

followed by SD.( 6.306), says that it is similar to nāṭī but the hero

should be as in Prakaraṇa, including including the form in its list of

rūpakas ( ND. p. 7).

14. Prasthānaka: This type is characterized by descriptive gestures. It is

in two acts divided into four scenes. It mainly uses delicate

movements, with occasional vigorous passages, such as the gait of an

elephant, which stands for the idea of journeying abroad (NŚ. Vol. I,

pp. 166, 181). It is listed in the Agni Purāṇa and described by

Abhinavagupta, Dhanika, Bhoja, Rāmacandra and Gunacandra,

Hemacandra, Vāgbhaṭa (who like Hemacandra simply reproduces

Abhinavagupta), Sāgaranandin, Śāradātanaya, Viśvanātha and

Subhaṅkara. In their detailed descriptions Abhinavagupta and Bhoja

mention apasāra, that is, exit, as one of its features. ND. defines

apasāra as interludes between dances (p. 191). Bh.P. defines it as a

musical performance set to rhythm and tempo and performed in

kaiśikīvṛtti (p. 262). As example it mentions Śṛṅgāratilaka. SD.

suggests rhythm, tempo and both kaiśikīvṛtti and bhāratīvṛtti (6. 280-

81). SDām's definition is similar to SD. 's (SDām. p. 95). NLRK says

that it requires many rhythm and tempo arrangements (NLRK. 3147-

50). Both Raghavan and Warder have extensive discussions on this

type (Raghavan, 1963, pp. 548-49; Warder, 1972, Vol. I, pp. 160-61).

15. Prekṣaṇaka, preṅkhaṇaka: The term prekṣaṇaka appears in Bhāsa's

Bālacarita and, as Warder points out, may simply mean drama

(Warder, 1972, Vol. I, p. 139). The Kāmasūtra refers to it as a one-act

play (K.S. II.10.25). The AP mentions preṅkhaṇa. No similar term

appears in the A.B. but Bhoja, Rāmacandra and Guṇacandra,

Sāgaranandin, Śāradātanaya and Viśvanātha all list this type. Bhoja

says that spectacles like the immolation of Eros are characteristic of

the type, and this view is repeated in ND (1959, p. 191). SD states that

all the vṛttis are to be used in this type which it calls preṅkhaṇa (6. 286-

87). SD 's description follows that of Sāgaranandin (NLRK. 3192-7)

and both texts cite Bālivadha as an example. Bh.P. has a confused view

because it uses the account of nartanaka in Bhoja to define prekṣanaka

(p.263). SDām describes the type as a play that uses all vernaculars but

śaurasenī in the main (p. 92). It is discussed in detail by Raghavan and

Warder (Raghavan, 1963, p. 561; Warder, 1972, Vol. I, p. 144).

16. Preraṇa: This is a one-act play featuring comic action and characters.

Abhinavagupta describes it as full of laughter and riddles

(Abhinavagupta on the NŚ.1956,Vol. I, pp. 172, 181). Among writers

on alaṅkāra Abhinavagupta alone discusses it but in saṅgīta literature

preraṇa (or its variants, peraṇa or peraṇī) appears in almost every

work as a form dancing (SR. 7.46, 1316-25; NR. 7. 34-58; SSam. 7. 217-

22). The form is discussed in both Mankad and Warder (Mankad,

1936, pp.127-28; Warder, 1972, Vol. I, p.144). It will be discussed in

detail later in the present study in course of examining the dance forms

that developed from the time of the SR through the medieval period.

17. Bhāṇa, bhāṇaka: This term represents both a major and a minor

dramatic type. The term has caused considerable confusion in the

literature of dance and drama. Its importance to the understanding of

the evolution of dancing lies in the fact that though it is a type of drama

it employs bold, vigorous body movements and instrumental music. As

a minor dramatic type it first appears in the A.B. (NŚ. 1956, Vol, I,

.p.181) which distinguishes it from the bhāṇa that appears in the list of

major dramatic types. The major type is a satirical monologue but the

minor type is a presentation that calls for vigorous body movements

(NŚ.1956,Vol. I, p.181) and a great deal of instrumental music (NŚ

1956,Vol. I, p.166) with irregular beats. Bhoja has a long discussion on

this type (Raghavan, 1963, pp. 451-54) which he divides into three

186

NRTYA AND UPARŪPAKA

varieties, namely, śuddha, saṅkīrṇa and citra. According to him it

requires dancing. ND closely follows Bhoja and like Bhoja states that

it is difficult to perform (ND. p. 191). It is said to be of three varieties:

uddhata, lalita, and lalitoddhata. Bh.P.'s description of the minor type

known as bhāṇa is very long, running from page 258 to 262. It says that

lāsyāṅgas form part of bhāṇaka. Later in the text (p. 266),

Abhinavagupta's definition of the term is given in the middle of

discussing rāsaka. Abhinavagupta's definition is reproduced also by

Hemacandra and Vāgbhaṭa. Dhanika offers no definition. A variety of

bhāṇaka is known as nandimālī (BhP, p.261, ṢṛP. Vol. II, p. 467). Both

Raghavan and Warder have substantial discussions on bhāṇa

(Raghavan, 1963, pp. 551-54; Warder, 1972, Vol. I, pp.158-59). A

detailed study has been made by S. S. Janaki in her dissertation (1971).

18. Bhāṇikā: Bhāṇikā is mentioned in the Abhinavabhāratī, Agni Purāṇa,

Ṣṛṅgāraprakāśa, Nāṭyadarpaṇa, Kāvyānuśāsana (of Vāgbhaṭa), Nāṭakalakṣaṇaratnakośa and

Sāhityadarpana. Abhinavagupta refers to the enactment of bālakrīḍā

(presumably of Kṛṣṇa) in bhāṇikā (NŚ. 1956, Vol. I, p. 181). Bhoja says

that bhāṇikā is similar to bhāṇa but that the movements are delicate

(Ṣṛ.P.Vol II, p.467). Śāradātanaya considers lāsyāṅgas as necessary in

bhāṇikā as in bhāṇa and states that the minor dramatic type known as

bhāṇa has the features of bhāṇarūpaka, of which Viṇāvatī is an example (Bh.P. p. 262). NLRK prescribes both bhāraṭī and kaiśikīvṛtti

for bhāṇikā (3160-63). SD gives a similar account of the type and cites

Kāmadattā as an example (6. 308-13). Vāgbhaṭa and Hemacandra

follow Abhinavagupta, quoting him. Raghavan and Warder discuss it

in detail (Raghavan, 1963, pp. 551-5; Warder, 1972, Vol.I, pp.159-60).

19. Bhāṇī: This type is mentioned in the AP, the avaloka of Dhanika,

NLRK, and Bh.P. In the AP and NLRK both bhāṇī and bhāṇika are

discussed. AP offers no definition of bhāṇī but NLRK distinguishes it

from bhāṇikā by saying that bhāṇī contains lāsyāṅgas and arouses

śṛṅgārarasa, offering as example Viṇāvatī (NLRK. 3184-6).

Śāradātanaya lists bhāṇī as one of the nrtyabhedas (Bh.P. p. 255) but in describing it he equates it with bhāṇikā(p. 262). Subhaṅkara (SDām. p.94) follows NLRK and Bh.P.

20. Mallikā: This is mentioned only by Śāradātanaya (Bh.P. pp. 267-8) who identifies it as the same as maṇikulya, a mystery story. The description is unclear but seems similar to that of durmallikā. It includes a jester or parasite among its characters and is in two acts. Warder discusses it in some detail (Warder, 1972, Vol. I, pp.144-5).

21. Rāmākrīḍa: Abhinavagupta briefly discusses this type which is characterized by the description of the seasons (NŚ.Vol. I, p. 181). BhP., the Kāvyānuśāsana of Hemacandra (KAnu.H.p. 327-29) and the Kāvyānuśāsana of Vāgbhaṭa (KAnu.V. p. 18) mention this type of minor drama, Bh.P. repeating a line from its discussion on rāsaka (p. 266). No other text mentions rāmākrīḍa.

22. Rāsaka: This is mentioned in almost every early text. It is found in Abhinavagupta, Kumārila, Bhāmaha, Dhanika, Bhoja, Rāmacandra and Guṇacandra, Hemacandra, Vāgbhaṭa and appears in the AP, NLRK, BhP, SD. and SDām. As noted earlier in this chapter (pp. 159-60), it is treated both as a drama and as a dance in the literature. Abhinavagupta identifies it as a dramatic dance (NŚ. Vol. I. p. 181). Kumārila, Bhāmaha, Dhanika and the AP offer no definition. According to Bhoja hallīsaka becomes rāsa when it is performed to certain defined tālas. In his detailed discussion he states that piṇḍībandhas or group dances are a feature of this type and describes them (ŚrP.Vol. II, p. 468; SKA. p.125). Evidently, Bhoja sees it primarily as a dance, although he places it under padārthābhinayātmaka prekṣyaprabandhas. Rāmacandra follows Bhoja (ND.1959, p. 191). Hemacandra and Vāgbhaṭa quote Abhinavagupta’s view. The Harivaṃśa has references to rāsa (2.89.67; 93.24). NLRK’s definition implies the importance of verbal acting

188

NṚTYA AND UPARŪPAKA

(sakalabhāṣāvibhāṣāśobhitam) but does not include dance or music (3205-9). As example it cites Madanikākāmūka. Bh.P. gives a highly confusing description (pp. 265-66). Śāradātanaya reproduces

Abhinavagupta’s quotation from the cirantanas in the middle of his own definition. From the discussion in SD it seems that rāsaka is a minor drama characterized by verbal acting but employing singing and dancing as well (6. 288-90). SDām. follows SD but adds that the type is also known as the krīdārāsaka of the gopīs ( SDām. p. 95). Both Warder and Raghavan have long discussions on rāsaka that attempt to resolve the confusion in the texts (Raghavan, 1963, pp. 561-67; Warder, 1972, Vol. I, pp. 145, 148).

The dance called daṇḍarāsaka perhaps evolved out of śamyā, as Warder has suggested (Warder, 1972, Vol. I, p. 152). It is a group dance performed with sticks, as we find in Karpūramañjarī (4.10-11). This is a feature of pārijātaka as described in Bh.P. (p. 268), which suggests that daṇḍarāsaka and pārijātaka are one and the same. However, Bh.P. later (p. 297) equates rāsaka with latā, a variety of lāsya, and divides rāsaka into three classes, daṇḍarāsaka, maṇḍalarāsaka, and nāṭyarāsaka. A saṅgīta text, Saṅgītasamayasāra,

describes it as a dance that requires various tālavādyas and is performed with sticks (SSam.6. 240-45). NR. describes rāsaka primarily as a piṇḍībandha, (7. 84-97); nāṭyarāsaka as mime to a song (7. 100) and daṇḍarāsaka as a group dance performed with coloured sticks( 7.101-7). All three are described as deśīnrttas danced to music. The Nartananirnaya describes rāsa and daṇḍarāsa in the similar manner putting them under the anibandha category of deśī dancing (NN. 53a). This type of performance has survived into our times, especially in Gujarat (Raghavan, 1963, pp. 561-2).

23. Vāra: Bodhāyana (Bhagavadajjukīya, 1925, p. 4) mentions this but offers no definition. Bh.P. (p. 241) refers to a form of nāṭaka it calls nṛtyacāra (vāra being an alternative reading). Abhinavagupta

MOVEMENT AND MIMESIS

189

mentions nāṭyadhāra (NŚ.Vol. III, p.172), which, Warder suggests,

may be a scribal error. He conjectures that vāra was an ancient and

obsolete type of drama which was not recorded by later writers

(Warder, 1972,Vol. I, p.141). None of the available references is clear

enough to indicate the nature of vāra.

24. Vilāsikā: This appears only in the SD which describes it as employing

ten lāsyāngas and arousing śṛṅgārarasa. This type was evidently also

known as vināyikā (SD. 6. 301-2, p.106, comm.). In his discussion of

vilāsikā (Warder, 1972, Vol. I, pp.162-3) Warder derives the term

from lāsikā, which he considers to be an alternāte form of lāsya. He

further states that lāsikā, which appears in the Alaṅkārasaṅgraha of

Amṛtānanda, is the same as the bhāṇī of Sāgarandin. In SDām.

lāsikā has the same meaning as vilāsikā mentioned in SD (SDām. p.96).

25. Śilpaka: This term appears in AP, NLRK, BhP, SD and SDām.

NLRK says that it consists of four vṛttis and twenty-seven aṅgas

(NLRK 3029-46). BhP ( p. 257) and SD (6. 296-300) agree with

NLRK's description. All of them cite Kanakāvatimādhava as an

example. Śubhaṅkara follows them (SDām. p. 94).

26. Śṛṅgadita: This term appears in the AP and in Dhanika, Bhoja,

Rāmacandra, Hemacandra, Vāgbhaṭa, Sāradātanaya, Viśvanātha and

Śubhaṅkara, but the AP and Dhanika do not describe it. Bhoja says that

it depicts through songs and recitations the state of separation in love.

ND follows Bhoja (1959, p.191) as do SDām (p. 95) and NLRK except

that NLRK says that the performance involves mainly verbal acting

and does not call for songs, an example being Kṛdārasātala (NLRK.

3157-59). Bh.P. agrees that śṛṅgadita should mainly employ verbal

acting but recommends some singing and lalita movements (p. 258).

SD gives a definition similar to Bhoja's but adds that the presentation

should frequently use the word "śrī," as in Kṛdārasātala. Raghavan

compares śṛṅgadita with the Tamil kuruvañci and with the varṇam and

padam of Bharatanāṭyam (Raghavan, 1963, pp. 546-7).

190

NṚTYA AND UPARŪPAKA

27. Ṣidgaka, ṣiṅgaka: This type is termed ṣidgaka by Abhinavagupta (NŚ.Vol. I, p.181), and ṣiṅgaka by Hemacandra (KAnuH. p.327) and Vāgbhaṭa (KAnuV. p.18), but has the same definition in all three.

Abhinavagupta says that it is a musical play full of songs ("gīyamānarūpaka"-NŚ.Vol. I, p.166) and that although it is a delicate form, it makes extensive use of vigorous movements. He later quotes the cirantanas (NŚ.Vol. I, p.181) who describe it as a delicate form.

The type described as śṛṅgadita by Bhoja seems to be similar to ṣidgaka. Hemacandra refers to śṛṅgadita and ṣiṅgaka (a variant spelling of ṣidgaka) as different types but defines ṣiṅgaka similarly as Bhoja's śṛṅgadita. Abhinavagupta's definition of ṣidgaka is quoted in Bh.P. (p. 266) to form the definition of śilpaka, although a few pages earlier Bh.P. describes śilpaka as in NLRK (Bh.P. p. 257; NLRK 3029-46).

Raghavan compares ṣidgaka with śṛṅgadita (Raghavan, 1963, pp. 546-7). Warder has a long discussion (Warder, 1972, Vol I, pp. 161-62) on the possible connection between this and other types of minor drama.

28. Śāṭaka, saṭṭaka: This is mentioned in the AP, in Hemacandra, NLRK, BhP., SD and SDām. Hemacandra, who includes it in the class of rūpakas, says that it is performed in prākṛta and is similar to a nāṭikā (KAnuH. p.317). The NLRK recommends kaiśikīvṛtti for this type which it calls similar to nāṭikā, citing as example Karpūramañjarī (NLRK.3198-3204). Karpūramañjarī uses śāṭaka, the prākṛt form of the term. Śāradātanaya agrees that the type is similar to nāṭikā but does not include it in the list of minor dramas or recommend kaiśikīvṛtti (Bh.P. p. 244). Both NLRK and Śāradātanaya suggest the use of prākṛta. SD agrees with Bh.P. and cites Karpūramañjarī as an example ( 6. 276-7). The edited text of SDām. refers to śāṭaka (pp. 79, 92-3) as a type of minor drama and defines it in the same way as saṭṭaka, which is found in Bh.P. and SD. This text also gives (SDām. p. 93) saṭaka as a reading from the India Office Library manuscript, which suggests that the two terms are alternate spellings of the same term. The term is

MOVEMENT AND MIMESIS

191

listed by Bhoja as a rūpaka (Ṣr.P.Vol. II, p. 466) and is so defined by

him. In Karpūramañjarī (Act IV, 10-11) there is a description of the

dance called daṇḍarāsa which is introduced by the stage direction

"tataḥ praviśati carcarī." This suggests that while śāṭaka was not a

dance drama, it used dance and music. Except for NLRK no text

prescribes kaiśikīvrtti as such but the common view that the type

resembles nāṭikā implies the use of songs and dances as found in

Karpūramañjarī.

29. Sallāpa, sahlāpa: NLRK, Bh.P, SD and SDām. describe this as a

minor drama that does not use the kaiśikī and bhāratī vrttis. NLRK

suggests the sāttvatī and ārabhaṭī vrttis as essential to the type (NLRK.

3028). Bh.P. gives a similar definition at greater length (p. 256). SD (6.

291-92) agrees with NLRK and Bh.P., as does SDām. (p. 93). The

example cited by NLRK, SD and SDām. is Māyākāpālika.

30. Hallīśaka, hallīsaka: This is a group dance resembling rāsaka and is

mentioned by Vātsyyāyana (KS. II.10.2, p.154), by Bhāsa in the

Bālacarita (Act 3), and in AP (338.3), Harivaṃśa (89.68), A.B. (NŚ.

Vol. I, p. 181), Ṣr.P (Vol. II, p. 468), ND (1959, p. 191), Hemacandra

(KAnuH.p.327), Vāgbhaṭa (KAnuV. p.18), NLRK (3154-6), Bh.P. (pp.

266-7), SD (6. 307), and SDām (p. 94). Abhinavagupta describes it as a

dance and places it under musical or dance dramas. Ṣr.P. considers it to

be a minor dramatic type but defines it after A.B., which is quoted by

Hemacandra and Vāgbhaṭa. ND. reproduces Bhoja's account. NLRK

follows Abhinavagupta and Bhoja and cites Keliraivataka as an

example. The same example is given by BhP which treats hallīśaka at

some length, calling it a dramatic type that employs geyalāsya and

some of the technical features of drama. The technical features of

drama are mentioned also by SD, which gives a shorter account,

stating that the type calls for kaiśikīvrtti and the use of a variety of

rhythm and tempo. The example given is again Keliraivataka.

192

NṚTYA AND UPARŪPAKA

Raghavan (1963, pp. 554-5) and Warder (1972, Vol. I, pp.139,152)

have long discussions on the type.

These, then, are the thirty types of minor dramatic performances

which came to be known as uparūpakas. We find them in alaṅkāra

literature as well as in related works. Some of these types involve dance

and music, and these are found in the works on saṅgīta that begin to appear

around the 13th century. Of these, only those are relevant to the present

study that employ dancing, some being primarily dance forms while some

are dance dramas. Minor dramatic types may thus be classified into three

broad categories: first, those that use no music or dance; second, those that

use music and dance; third, those that are primarily dance dramas, or

musical plays, or dance performances. In the first category we may

include the following: durmallikā, preksanaka, mallikā, preraṇa,

śilpaka and samlāpa. The second category includes the following: nāṭikā,

trotaka, kalpavallī, prakarana and sattaka. The third category has two

subdivisions: musical plays and dance dramas. The following types of

musical plays are mentioned: kāvya, prasthānaka, rāmākrīḍa (this may be

included in the first type but its definition is not clear), ullāpyaka and

ṣidgaka. The other sub-category, that of dance drama, includes the

following: goṣṭhī, ḍombī, nartanaka, nāṭyarāsaka, pārijātaka, vāra,

vilāsikā, bhāṇaka, bhāṇikā, bhāṇī, rāsaka and hallīsaka. Of these, the

following may be considered as a separate group, for they are exclusively

dance compositions: goṣṭhī, ḍombī, rāsaka, nāṭyarāsaka and hallīsaka. To

this group belongs one of the few surviving forms from this period,

namely, rāsaka or rāsa, which is still practised in Gujarat and Manipur.

The use of the term uparūpaka thus reflects a gradually coalescing

view of those forms of performing arts which required body movements

52

A term very likely derived from this later came to denote a different presentation; see

p. 185 supra and pp. 232-33infra.

MOVEMENT AND MIMESIS

193

that were aesthetically pleasing as well as representational. In other

words, these forms possessed characteristics of both dance and drama,

though some were more dance than drama and some the other way about.

As our survey here shows, till the early medieval period there was a great

variety of such forms, not always clearly distinguished from one another

and even when so distinguished, not greatly differing from one another.

With the passage of time it became evident that these forms fell into two

broad divisions, both mimetic in nature but one achieving its effect

primarily by the representational functions of body movements, the other

by their form, one emphasizing the verbal representation of events and

character, the other appealing to the senses through physical action. Art

forms of the first kind were considered to belong to a sub-category of

drama proper, hence the term uparūpaka, while the other forms were

placed under nṛtya, or mimetic dance. Thus, the development of the two

broad categories marked a progressively clearer conception of the

characteristics of dance as well as drama. It also meant that the

classification of the performing arts was simplified, whereby the class

names used earlier, except for lāsya, became part of the terminology of

music rather than of drama.

The attempt to classify categories was, however, not restricted to the

use of the terms nṛtya and uparūpaka. A similar trend towards definitive

classification may be seen also in the attempt to distinguish the tradition of

dancing described first by Bharata from the later, more broad-based and

popular forms of the art. The next two chapters will examine how this

distinction came about and what it meant in terms of the content and style

of performance.

Chapter 6

BANDHA AND ANIBANDHA

During the period that started with Someśvara in the eleventh century and ended with Mahārāṇā Kumbhā in the fifteenth, that is, the second period of the evolution of the concept of dance, we find that the thought of these authors focused on the division of dance into two traditions, a division that was firmly entrenched by the end of this period. One tradition was termed mārga, and regarded as the central tradition established by Bharata, and the other deśī, the tradition which represented local and popular forms of dancing and which Bharata left out of his account. In this second period the most influential of all the authors was Sārṅgadeva who discussed deśī and continued to influence the authors in the third period of development.

In the third period, however, we find another development-a different kind of categorization of dancing. At least one author, Puṇḍarīka Viṭṭhala, draws our attention to another major development while describing dance traditions of different regions of his time.1 In his work on dancing and dramaturgy, the Nartananiṛṇaya, written in the sixteenth century, he does not merely divide dance into the two traditions, but gives a full account of the deśī tradition, presenting material he evidently collected from different regions, showing the varieties of dancing that existed in his time. Quite often these descriptions offer full identifications of different styles.

1Puṇḍarīka Viṭṭhala is better known in the field of music and although his work on dancing, the Nartananiṛnaya, has been known by name, neither a critical edition nor a critical study of this work has been done so far. An account of its content has been given by the present writer in her B. Litt. thesis and her book (Bose, 1964, 1970).

194

MOVEMENT AND MIMESIS

195

Information is given about which part of India the particular style came

from, what language was used in the accompanying songs, and, at times,

modes of presentation. By comparing his accounts of those styles with the

styles practised today, it is possible to establish links between these styles.

This text thus offers us a major breakthrough in understanding both the

evolution and the continuity of the art of dance because it enables us to

reconstruct the styles prevalent at a transitional period in the cultural

history of India.

One important contribution of the Nartananirnaya is the evidence we

may draw from it to establish firmly the time of the origin of two major

styles of India today, namely, Kathak and Odissi. There has always been

some controversy about the evolution of these two styles. Dance historians

in general are agreed that while the roots of Kathak go back to ancient

Hindu culture, its present form is derived from dancing styles imported by

Mughal rulers. There is no doubt that Kathak did absorb some Persian

influence, but the case for that influence is overstated. This can be easily

seen by comparing the detailed descriptions found in the Nartananirnaya

with the movements of Kathak. The style described in the Nartananirnaya

is, of course, not termed Kathak, a name that came into use much later, but

the descriptions clearly show it to be the same as what we know today as

Kathak.

The same historical perspective is lent by the Nartananirnaya to the

Odissi style. Odissi was in oblivion for a long time. Even during the

renaissance of the dance, when Bharatanatyam was rediscovered and

hailed as the true core of the authentic tradition of the art, Odissi remained

unknown to the general public and even to many dancers and dance

scholars. When eventually Odissi began to be performed in the early

1950s, its authenticity was called into question, some critics suggesting

that it had been concocted by some interested performing artists of

196

BANDHA AND ANIBANDHA

Orissa.2 But the Nartananirnaya provides incontrovertible proof that as a distinct style Odissi goes back at least four centuries, although it was not then known as Odissi, a name that came into use much later, just as we find in the case of Kathak. Pundarika's discussion of these and other styles show a high degree of correspondence with present-day styles. The Nartananirnaya thus marks a definite terminal point in the ancestry of today's classical Indian dancing.

The historic importance of the Nartananirnaya has remained unacknowledged for two reasons: first, Pundarika has been traditionally known as an expert in music rather than dancing; secondly and surprisingly, most of the other texts of this third period, although written after the Nartananirnaya, have followed the Sangitaratnākara in the main. As a result, the detailed information available in the Nartananirnaya has remained unexamined.3

Before we examine the technical details of the styles described in the Nartananirnaya, we must consider the categorization attempted in it. Throughout the history and evolution of dancing, its literature has been given to categorizing the art in many ways. Many of the categories used at one time or another seem of doubtful use since they overlap with others or are not distinct enough to be taken as individual categories. It is for this reason, no doubt, that most of such categories have disappeared from accepted usage and have no relevance to the art as it is practised today. However, the readiness of the authors of Sanskrit dance manuals to formulate categories indicates a sustained attempt to systematize the study of dancing by classifying its forms. One such attempt of particular historical significance is seen in the use of the terms bandha and

2Bose, N. "Odissi Dance today and its Exponents," Classical and Folk Dances of India, part, iv,'Odissi,' p. 51.

3A critical edition is under preparation by the present writer.

MOVEMENT AND MIMESIS

197

anibandha. Applied to dancing, these terms appear first in the

Nartananirnaya. The author describes the two categories thus:

कार्यं तत्र द्विधा नृत्तं बन्धकं चानिबन्धकम्

गत्यादिनियमैयुक्तं बन्धकं नृत्तमुख्यते

अनिबन्धं त्वनियमाद . . . ॥

(NN.39b)4

Nṛtta can be of two kinds: bandha and anibandha.

Bandhanṛtta is structured with gati etc. while

anibandha comes from following no rule.

How these categories came into existence must be considered in the

context of how dancing and its literature evolved. As we have seen in the

previous chapters, in the earliest works the categories are few; it is in the

later works that categories begin to multiply. The Nātyaśāstra of Bharata,

the first and most influential of the texts on dance, drama and music, does

not divide dancing into any categories. Bharata recorded only one kind of

dancing: nṛtta, the kind that later came to be known as mārga or classical.

He mentions the existence of popular, regional dance forms, forms later

categorized as deśī.5 Bharata obviously did not find them important

enough to be included in his treatise. For several centuries after Bharata,

manuals on music and dancing were virtual copies of the Nātyaśāstra and

described nothing that was substantially new. However, as time went on,

the manuals began to extend their accounts to include new forms and

techniques of dancing. The texts began to include the deśī style as an

individual, descriptive category around the seventh and eighth centuries

4Textual references are to the India Office Library Manuscript, London, unless

otherwise noted.

5NS.9.163-64.

198

BANDHA AND ANIBANDHA

A.D. The Brhaddesī by Matanga, written in the eighth century,6 was the

first text to deal with the desī style in the field of music.7 In the field of

dance too desī styles began to be mentioned around this time. And although

Someśvara, from the eleventh century, was the first writer to give us a

brief account of the desī style, it is not until we arrive at the

Sangītaratnākara by Sārngadeva that we find a systematic study of both

mārga and desī dance styles as separate categories.8 The

Sangītamakaranda, which also gives an account of one desī style, may

have been earlier but with a corrupt and undependably edited text we

hardly get a picture clear enough to identify that style as a separate desī

style. However, from the description that it gives it seems likely that the

style was a regional one that was practiced by prostitutes.9

A further subdivision occurred later when the terms bandha and

anibandha were applied to types of dancing. Parallel terms with similar

meanings but different forms-nibaddha and anibaddha-had already been

in use in music to denote two parts of a musical composition, or prabandha.

While both parts followed certain rules of structure and of development,

anibaddha was comparatively loose in its structure since it was free of the

regimen of tāla. By contrast, baddha was more rigorously constructed,

bound as it was by the constraining patterns of tāla. These musical terms

were first used in the Nātyasāstra :

निबद्ध चानिबद्धं च येन तेन द्विधा स्मृतम् ॥

(NS.32. 28)

Nibaddha and anibaddha can be of two kinds.

6Katz, 1983, p. 59.

7अवालगोपाले: क्षितिपालैर्निजेच्छया।

गीयते सानुरागेन स्वदेशेो देशिकन्यते ॥ (Brhaddesī. 13.)

8SR. 7. 26-7; 749-50; 1207; 1278.

9SMak.4.

MOVEMENT AND MIMESIS

199

Later the Brhaddesī and the Saṅgītaratnākara used them and the distinction was clearly drawn in both:

निबद्धश्चानिबद्धश्च मार्गोऽयं द्विविधो मतः।

आलापादि(?)निबद्धो यः स च मार्गः प्रकीर्तितः॥

(Brhaddesī. 14)

This mārga is said to be of two kinds, namely, nibaddha and anibaddha. That which is structured by ālāpa (?) etc.is called mārga.

बद्धं धात्वभिरूढेश्च निबद्धमभिधीयते।

आलप्तिबन्धहीनत्वादनिबद्धमितीरिता॥

(SR. 4. 5)

Nibaddha is known by being structured with dhātus and [other] aṅgas. Anibaddha is so called because it is not structured with ālapti.

The distinction between nibaddha and anibaddha clearly represents one of the important principles by which ways of composing music may be differentiated. It is to this categorization that we owe the term ālāpa, which still carries the connotation of free improvisation within the given framework of a rāga.

Almost 1500 years after these terms came into use in music, Puṇḍarīka Viṭṭhala’s work on dancing applied the terms bandha and anibandha to dancing,10 attempting to draw a distinction between categories of dancing that reflected the spirit of the distinction made in music, though the terms did not exactly correspond in form or meaning to those used in music.

Before examining Puṇḍarīka Viṭṭhala’s use of these terms it is necessary to note that the bulk of his work was in music. It is, therefore,

10Bose, 1970, p.5.

200

BANDHA AND ANIBANDHA

not surprising that when he tried to distinguish categories of dancing by

using as the differentiating principle the degree of improvisation

permissible, he should have borrowed terms from his primary expertise,

even though the parallel was not exact and the forms of the words were

different.11 It would seem that this is the sense in which this distinction

came to be understood after Puṇḍarīka, for the only substantial work on

dancing to be written after the Nartananirnaya, the Saṅgītadarpana by

Dāmodara, repeats the usage and reproduces the discussion of the terms as

in Puṇḍarīka's work.12

Puṇḍarīka uses the terms unambiguously. Bandhanṛttas are set pieces

with every movement in their structured sequences clearly specified for

the dancer. Given such detailed descriptions, a dancer trained in the

appropriate style of dancing could reconstruct a composition perfectly,

executing all the movements down to the minutest detail in precisely the

sequence prescribed. The bandhanṛttas are, then, fully structured

compositions that the dancer is required to reproduce unaltered, with no

variations, no straying from the set choreography. Herein lies the

significance of the term bandha, that is, a "closed" composition.

By contrast, anibandhanṛttas are flexible in both form and content

within broadly specified frameworks of aesthetic purpose. The

Nartananirnaya does not describe any of the anibandhanṛttas in sufficient

detail to allow us to reconstruct a full composition. What then was this

anibandhanṛtta? On the basis of the scanty descriptions in

Nartananirnaya, anibandhanṛttas seem to be short dance-sequences, using

which a dancer can choreograph her own piece. Thus, they have the same

11कार्य तत्र द्विधा नृतं बन्धकं चानिबन्धकम् ।(NN.40a).

निबद्धश्चानिबद्धश्च मार्गोऽयं द्विविधो मतः ।(Bṛhaddesī.14)

निबद्धमनिबद्धं च तद्‌द्वेधा निगदितं बुधैः ।(SR.4.4.).

12Dāmodara is believed to have been a court poet of Jahangir, which places him in the

17th century; see Krishnamachariar, 1974, p. 806.

MOVEMENT AND MIMESIS

201

function in the dancer's choreographic design as the karaṇas of the mārga

tradition. But their structural principle is entirely different from that of

karaṇas in that they are entirely flexible as to their components and

structure while karaṇas are of course rigidly set structures. The

Nartananirnaya does not give details about the movements constituting the

anibandhanṛttas but only discusses the movements in broad, general terms.

It would seem that the anibandhanṛttas were unlike any other dance

pieces described in the literature before the Nartananirnaya. Did they

then represent a new, hitherto unrecorded style of dancing that differed

from other post-mārga styles in being still uncoded and therefore

allowing the dancer a far greater measure of freedom? Was the author

trying to create for this new style a new category since it could not be

fitted into any other, inasmuch as even the relatively new, deśī styles

adhered to the accepted principle of prescriptive design and this new style

did not? Evidently, the author was recording two different classes of

dancing, the distinction between which demanded the use of two different

terms. This hypothesis, and perhaps only this would explain the use of the

terms bandha and anibandha and clarify their meaning as categories,

bandhanṛttas denoting dances for which there already were prescribed

rules, and anibandhanṛttas denoting dances for which there were none.

That this is more than mere speculation is suggested by the history of

dancing. We find that a new style of dance, still unnamed but later to be

known as Kathak, was indeed taking shape at this time, and that this style

placed a greater emphasis on creating dances within the general structure

of movements than on following a set regimen of technical details. This

new style offered a new approach to dancing and appeared on the dance

scene precisely at the time when the Nartananirnaya was written.

On its part the Nartananirnaya describes several entirely new dances,

all of them deśī and all previously unreported, although it follows the

Sangītaratnākara in describing some other deśī dances. When we look

closely at the technique of the dance described under the anibandha

202

BANDHA AND ANIBANDHA

category, we begin to see certain striking similarities with the technique

of Kathak. One cannot say that the style described in the Narananirnaya

matches Kathak in every detail, but one may certainly view that style as

the precursor to Kathak.

The historical context of the Nartananirnaya makes this connection

highly plausible. The text was part of the same cultural world of the

Mughal court that nurtured Kathak. Indeed, the colophon states that the

work was written by Pundarika Vittthala for the delectation of the

Emperor Akbar, the Mughal ruler :

अकबरनृपरुच्यार्थ भूलोके सरलसङ्गीतं ।

कृतमिदं बहुवरंेदं सुद्ददां हृदये सुखं भूयात् ॥

(NN. 53b)

In [this] world this simple sangita is created [with] a

lot of varieties in order to please the king Akbar. May

it please the heart of the good-hearted ones.

By Akbar's time the ideals and techniques of Persian art were firmly

entrenched in the cultural life of India, co-existing with varied native

traditions. Pundarika declares that along with the regional styles of

dancing he is describing the dance of the yavanas, in this case, the

Persians:

यत्नचेष्टाविरहितं तन्नृत्यं जक्कड़ी मतं ।

पारसीकै: पण्डितैस्तूद्ग्राहादिस्वरभाषया ।

यदि गीतं जक्कडीसंज्ञं यवनानामतिप्रियं ॥

(NN. 53a)

That dance is known as jakkadi which is devoid of

effort and action. The song sung by the experts from

Persia using udgraha, svaras etc. and vernacular is

known as jakkadi which is the favourite of the

yavanas.

MOVEMENT AND MIMESIS

203

Since modern scholars believe that Kathak is influenced by Persian dance-forms, Puṇḍarīka's statement provides a strong argument to trace the ancestry of Kathak to the style he describes. The Nartananirnaya seems, thus, to be the proper textual source for Kathak. This claim becomes stronger still on examining points of technique, as I shall shortly attempt.

Two arguments help to validate the hypothesis that Puṇḍarīka Viṭṭhala devised the anibandha category to classify the kinds of dances that eventually evolved as Kathak. First, what we now call Kathak represented in its early days a new principle of design, because as a style largely moulded by the Persian influence it was certainly outside the framework provided by Bharata. Thus, even though it could be viewed as a regional or deśī form, it had to be distinguished by being put into a special category.

Secondly, the similarities between the techniques of Kathak and anibandhanṛttas are most suggestive. Unlike other styles, Kathak does not prescribe in detail the movements it employs. This does not mean that Kathak does not use the movements that we find in other styles. It uses all the minor movements prescribed for the different parts of the body in the Nāṭyaśāstra just as the other styles do. But unlike the other styles Kathak does not have set compositions consisting of body movements, which are to be rigidly followed by the dancer. In Kathak, for instance, descriptions of dances do not identify particular constituent movements by name; rather, the descriptions identify general types of movements. The parallel with the anibandhanṛttas found in the Nartananirnaya is obvious. Even a brief comparison of the respective techniques and requirements of Kathak and anibandhanṛtta underscores the parallel.

The Nartananirnaya has four chapters, one each on rhythm, drum, song and dance. In the chapter on rhythm, when the author specifies the tālas appropriate to the dance he describes, the technical terms that he uses, sāadhanās in particular, match those used in Kathak today and listed by

204

BANDHA AND ANIBANDHA

Nirmala Joshi in her article, "Technical terms pertaining to Dance in

general and used in Kathak."13 These terms, which Joshi calls prāṇas, or

the life of tālas, are almost identical with the different sāadhanās

mentioned in the chapter on tāla in the Nartanairṇaya. For instance, in

Kathak, yati or the rhythmic arrrangement of the tempo is divided into five

categories, samā, srotagatā, gopucchikā, pipīlikā and mṛdaṅgī. The

Nartanairṇaya lists the same types of yatis similar in every detail,

although it includes a sixth type, kharjurikā. Another term, kuvāḍa, used in

Kathak to denote the climax of a complex rhythmic pattern is also found in

the Nartanairṇaya.

This similarity offers good reason to believe that the style described in

the Nartanairṇaya was something very much like Kathak, since it

required musical elements similar to those needed for Kathak. The

argument becomes even more persuasive when we examine the specifics

of the dance technique. But first let us briefly consider the typical

characteristics of Kathak as known today.

A major characteristic of Kathak is that although like other forms of

classical dancing it follows the same basic rules for the movements of the

individual parts of the body, it expects the dancer to find his or her

individual way to elaborate those movements within certain aesthetic

condition given. A good example may be taken from Maya Rao's article,

"The Hastas in Kathak": "In Kathak the body as a whole is visualised as

the prime medium of expression. . . . For instance, if the dancer intends to

represent the moon, not only will his hands show the Ardha-chandra

Hasta, but his body will also bend in an arch to suggest the idea of a

crescent moon."14 The same approach to elaborating and dramatizing

13Joshi, 1963."Technical terms pertaining to Dance in general and used in Kathak,"

Classical and Folk Dances of India, part. iii, 'Kathak,' pp. 8-9.

14Rao, 1963. "Hastas in Kathak," Classical and Folk Dances of India, part. iii, 'Kathak,'

p. 45.

MOVEMENT AND MIMESIS

205

basic movements is found in the Nartananirnaya. The description of a

dance called lāvaṇī includes an almost identical movement in which the

dancer bends her body from her waist in ardhacandra :

समपादे स्थितं पातं कटिन्यस्तार्धचन्द्रकम् ।

कटेःपरि तत् कायं भ्रामयेल्लावणी तदा ॥

(NN. 52b-53a)

[When] the performer stands in samapāda with her

waist bent in [the shape of] half-moon and she rotates

her body [lit : upper part from her waist], it is lāvaṇī.

The description of nāmāvalī, an anibandhanṛtya, strongly supports the

contention that in this kind of dancing only a general guideline was

provided for the dancer:

यथाभिनयसम्पन्नं विचित्रगतिसुन्दरम् ।

तीवटिग्रहभेदन लयतालसमन्वितम् ।

नामावलीनृत्तनिमदं नृत्येज्जनमनोहरम् ॥

(NN. 51b-52a)

Created with proper abhinaya, beautified with

various gaits and distinguishing between tīvaṭī and

graha [which are] filled with laya and tāla, this

nāmāvalīnṛtya, which is pleasing to people should be

danced.

This description, which gives the dancer broad, general directions

rather than detailed, specific instructions, is typical of Kathak, in which

the dancer is similarly guided. The striking feature of such a description is

its lack of detail; but that is precisely the feature that makes the dance

form flexible and permits the dancer the freedom to make creative

choices in building an individual choreography.

206

BANDHA AND ANIBANDHA

Maya Rao also mentions a composition by Shambhu Maharaj, called

"rang-manch ka tukra," which is performed to invoke the blessings of the

presiding deity of the stage. Rao finds striking resemblance between this

composition and the invocatory dance described in the Saṅgītadarpana. 15

As a matter of fact, the dance first appeared in the Nartananirnaya, from

which the Saṅgītadarpana has merely copied the description.

Two of the most distinctive movements of Kathak are chakkars and

tatkars. A chakkar is a rapidly spinning movement while a tatkar means to

stamp on the ground with one foot or both, and marking the rhythm with

ankle bells. Chakkar can be identified as the cakrabhramarī mentioned in

the Nartananirnaya, which describes it as a spinning movement:

चक्रभ्रमरिकाखण्डसुच्यर्धे चक्रवद् भ्रमात् ।

Cakrabhramarī [is performed] by spinning [lit: moving round] like a wheel in the middle of

khaṇḍasūcī.

(NN. 47b)

It is true that the bhramarīs were known long before the time of the

Nartananirnaya-Bharata refers to them-but they were not given the

prominence that they receive in the Nartananirnaya. Revolving

movements are of course integral to all dance styles, but in classical styles

other than Kathak the movements are never fast enough, nor sustained

enough to achieve the aesthetic form that a chakkar creates in Kathak. It

is the speed of revolution that sets it apart and it is precisely this element of

fast spinning, comparable to that of the pirouette, that we find in the

description of cakrabhramarīs in the Nartananirnaya.

In its discussion of revolving movements the Nartananirnaya also

describes tirapabhramarī:

15Ibid.

MOVEMENT AND MIMESIS

207

तिरपभ्रमरी तिर्यक् द्वावंची स्वस्तिकात् परम् ।

(NN. 47b)

[Revolving] obliquely with both the legs after crossing them is tirapabhramarī.

A similar movement termed as tirapa is found in Kathak as well.16

As for the tatkār, it clearly corresponds with the gharghara, of which details are given in both Nartananiraya and Sangītaratnākara. It was first mentioned in Sangītaratnākara as a feature of the deśī dance called peraṇī. A reference to it is also made in the Nrttaratnāvalī, which refers to this movement as a feature of the deśī dance named preraṇa but gives no detail. In the Nartananiraya it is a part of cindunṛtta which is identified as a dance from the south.

देशी द्वाविडदेशस्य चिन्दुरित्यभिधीयते ।

(NN. 49a)

The Saṅgītaratnākara gives the distinctive feature of gharghara as:

तत्र घर्घरिकावाद्ये बहुनिर्घर्षरो मतः ॥

(SR. 7. 1304)

Of these, gharghara is known to be the special practice in making the sound of ankle-bells.

The action is further clarified in theNartananirnaya :

यत्र किंकिणीकावाद्यैराहतिर्‌घर्षरो मतः । (NN. 50a)

16Vyas,1963, "The Background of Kathak" Classical and Folk Dances of India, part. iii, 'Kathak,' p. 6.

208

BANDHA AND ANIBANDHA

Where striking [the ground] to make [lit: with] the

sound of the ankle bells [is done] it is known as

gharghara.

Six varieties of ghargharas are described in the Saṅgītaratnākara and

the Nartananirnaya. These descriptions show that ghargharas closely

resemble the taṭkārs practised by Kathak dancers.17

Two more examples will strengthen the relationship between Kathak

and anibandhanṛtta. The Nartananirnaya abounds in descriptions of

anibandhanṛtta that fit Kathak. A particularly persuasive example is the

following:

यावनीभाषया युक्तं यत्र गीतं धृतानुच्चलम् ।

कल्लादिगजराद्युक्तं त्वाहारैर्न विभूषितम् ॥

विदग्धयानर्त्तनं नानालयत्रयविचित्रितम् ।

कोमलाक्रैैरयंदा नृतं भमर्योदिविराजितम् ॥

सशब्दा च क्रिया यत्न धुवशम्यादिभेदतः ।

यत्नचेष्टाविरहितं तन्नृत्यं जक्कड़ी मतम् ॥

पारसीकै: पणितैस्तुतूद्ग्राहादिस्वरभाषया ।

यद् गीतं जक्कडीसंज्ञं यवनानामतिप्रियं ॥

(NN.52b-53a)

Where the song is [sung by the dancer] in the language

of the yavana, holding her veil, [words] uttered with

kalla etc. and gajara etc.18 and beautified with

āhaṅga, the dance should be performed being adorned

with various three layas. When [this] dance is

17NN. 49b-50a. Kalyampurkar,1963, "The Techniques of Kathak" Classical and Folk

Dances of India, part. iii 'Kathak,' pp. 25-27.

18The technical terms are not clear here. They occur in no other text, nor has there been

any light shed on them by musicolgists.

MOVEMENT AND MIMESIS

209

performed with soft movements adorned by

bhramarīs, where the kriyā (keeping time with hands)

is done with sounded beat in accordance with the

difference between dhruva and samyā, that dance,

which is devoid of effort and action, is known as

jakkaḍī. The song sung by the experts from Persia

using udgrāha,svaras etc. and vernacular is known as

jakkaḍī which is the favourite of the yavanas.

This is an almost exact description of the ghungat gat, one of the best

known compositions in Kathak.19

The Nartananiṛnaya describes a certain anibandhanṛtta as follows:

कुञ्चिताङ्गुलिना यत्न प्रसृताङ्गुष्ठकेन चेत् ।

प्रसार्य जघ्घिकामं विचित्रदुतमाचरेत् ॥

घर्घरीभि: समायुक्तं तदैतत् कुडुपं मतम् ।

(NN. 52b)

Where [the dancer] contracts her toes [lit: one of her

toes] with the big toe extended, shakes her shank after

extending it with various quick [movements] and with

ghargharīs [that is, tinkling her ankle bells]it is known

as kuḍupa.

Precisely this action can be recognized today in Kathak when the dancer

beats a fast tattoo on the ground to create rhythmic sounds with her ankle-

bells.

The correspondence between Kathak and anibandhanṛtta is important

not only for discovering the roots of Kathak but also for understanding the

value that came to be attached to improvisation in medieval times. In

19Rao,1963, "The Hastas in Kathak" Classical and Folk Dances of India, part. iii,

'Kathak,' p. 47.

210

BANDHA AND ANIBANDHA

contrast with the prescriptive nature of the descriptions found in the

earlier texts, those in theNartananirnaya and its contemporaries allow the

dancer more structural flexibility while retaining the basic movements

described by Bharata and his successors. This difference in the approach

to choreography is most strikingly shown in the use of the contrasted

categories, bandha and anibandha. The descriptions of the

anibandhanṛttas give us a general sense of the choreography but leaves the

specific movements largely to the discretion of the dancer. In contrast,

Puṇdarīka Viṭṭhala describes the bandhanṛttas exhaustively, specifying

every movement required as well as the sequence of the movements.

Allowing-indeed encouraging-the dancer to choreograph parts of a

recital according to personal choice was a new approach to dancing.

There can be little doubt that the categories bandha and anibandha

mark a historical point in the evolution of Indian dancing at which

established traditions and new forms existed side by side. The bandha

category encompassed the styles that had been firmly defined and codified

by long usage. Other styles, still fluid and evolving, were placed within

the anibandha category. With the passing of time, as all styles lost their

unfamiliarity and became more firmly entrenched in the corpus of Indian

dancing, the distinction between the categories faded, eventually making

them obsolete. However, the term bandha is still in use in a very limited

way and later in this chapter we shall examine the dances to which it is

applied.

The term anibandhanṛtta has gone entirely out of use. Why it has

disappeared from the vocabulary of dancing while the term bandha still

has a limited use needs to be considered. Undoubtedly, the main reason is

historical, as explained above, but there may also be a reason related to the

very nature of dancing as an art form. We noted earlier that both in music

and dancing similar categories, anibaddha in music and anibandha in

dancing, were formulated to denote forms that allowed the artist wide

choices and architectonic individuality within broad guidelines. Such a

MOVEMENT AND MIMESIS

211

form is entirely appropriate for music, which permits-indeed demands-a

vast range of permutations of notes and improvised structures. That is why

the anibaddha category has survived in music as has the term itself, though

in popular use it is known as ālāpa.

But dancing is an art in which, once certain structural guidelines are

accepted, the dancer cannot have anything like the degree of freedom in

composing a dance that a musician has with a musical composition. A

dancer may still show her creativeness in a composition by combining

different sets of movements in a variety of ways or by exploring the

emotional content of the composition but not with the degree of freedom

that a singer or an instrumentalist enjoys. The concept of anibandha in

dance, therefore, can not be equated with anibaddha in music. The

creativity of a dancer is shown on an entirely different level. The

frameworks of music and dance are so different that a parallel

categorization is not possible. Hence the term anibandha became

redundant in the field of dance and we find the term neither in theory nor

in practice any longer.

When the term anibandha disappeared from the theory and practice of

dance, the need for its converse, the category bandha, became

meaningless. In most styles of dancing the term bandha is no longer found.

Nevertheless, there are two present-day styles in which this term is used.

These are Odissi and Manipuri. In Manipuri the term has very little

significance and according to one of the most eminent teachers of this

style, guru Bipin Singh, bandha is just another dance-sequence in the

Manipuri repertoire, which calls for greater technical virtuosity than

other sequences. Odissi is the only style today in which practising artists

still use the term to mean a separate category of dance sequences. These

follow the strictest possible rules of movements, sometimes producing

acrobatic postures.

When we compare the bandhanṛtta found in the Nartananiṛnaya with

the bandhanṛtta as practised in the Odissi style, we find striking

212

BANDHA AND ANIBANDHA

similarities. The bandhanṛtta of the Nartananirnaya requires

combinations created out of a base of sixteen karanas or short dance

sequences;20 the bandhanṛtta of Odissi requires seventeen.21

In Orissa the bandhanṛttas have been kept alive only by the goṭipuās

(goṭi=one, pua=putra) or boy-dancers attached to some of the temples,

whose numbers are rapidly dwindling. The few who are taught

bandhanṛtta start their extremely rigorous training at the age of seven and

perform till they are eighteen, after which their muscles no longer remain

flexible enough to be able to do such acrobatic movements. The goṭipuās

are dressed as girls and perform in religious festivals, though not inside the

temples. These dances are generally quite acrobatic. The most famous

Odissi guru of the present time, Kelucharan Mahapatra, is one of the very

few living artists trained as a goṭipua. When in the course of an interview

with the present author he was shown line drawings of the bandhanṛttas

described in the Nartananirnaya, Mr. Mahapatra confirmed that these

were indeed the sequences he was taught. According to him, these dances

are physically so difficult that girls are not trained in them and that they

are seldom performed even by men, the more acrobatic of the dances

being performed now mainly by the Chau dancers of Mayurbhanj in

Orissa. The difficulty inherent in these dances was further confirmed for

by another leading dancer of the Odissi style, Rani Karnāa,who said that

as a female she was not trained in the bandhanṛtta by her guru.

A comparison of the technique of the bandhanṛtta of the

Nartananirnaya with the bandhanṛtta of the Odissi style suggests a close

relationship between the two. In particular, there are remarkable

similarities in the standing postures prescribed in the Nartananirnaya with

the basic standing postures used in the Odissi style: chaukā and tribhaṅgī in

20NN.32a-33b.

21Khokar,1963, " Technique and Repertory," Classical and Folk Dances of India, part.

iii, 'Kathak,' pp. 28-31.

MOVEMENT AND MIMESIS

213

Odissi are comparable to vaiśākha-sthāna and agratalasañcara-pada as

described in the Nartananirnaya. Some of the acrobatic postures found in

the Nartananirnaya are still in use: danḍapakṣam, lalāṭatilakam and

niśumbhitam and several others are found both in Odissi and in Chau.

To the evidence taken from bandhanṛtta I may add three particularly

telling examples of correspondence between Odissi and the movements

described in the Nartananirnaya. The text describes the use of hand

gestures to express seven principal musical notes. Each note, according to

the author, is a correlative of a bird or an animal, which is represented by

hand gestures, as the following passage explains:

मयूरचातकछागक्रौञ्चकोकिलदर्दूरा:।

गजश्च सप्त षडजादीन् क्रमादुच्चारनत्यमी॥

(NN. 20b)22

Peacock, rainbird, goat, heron, cuckoo, frog and

elephant are the seven (notes starting with) ṣadja etc

that should be recited in order. [सा=मयूर: peacock;

रे=चातक: rainbird; गा=छाग: goat; मा=क्रौञ्च: heron;

पा=कोकिल: cuckoo; धा=दर्दूर: frog and नि=गज:

elephant].

Secondly, there is a dance called baṭunṛtta, a particularly difficult

dance in the repertory of Odissi, that involves both non-representational

dance and mimetic dance. In the Nartananirnaya we come across the

description of a dance called baṭu. Although it is a brief description, one

can see how difficult it is. It can not be a mere coincidence that the same

term should not only occur both in the Nartananirnaya and in Odissi but

22Reference to this verse is to the Ms. in the Asiatic Society Library Manuscript,

Calcutta.

214

BANDHA AND ANIBANDHA

designate dances involving similarly difficult poses. The brief description

in theNartananirnaya is as follows:

जानुभ्यां भूमिलग्नाभ्यां पद्भ्यां वा मण्डलाकृतिः ॥

नमपृष्ठं लताहस्तौ पात्रं भ्रमणमाचरेत् ।

तदासौ बटुरित्युक्तः सूर्यमण्डलवद् गति: ॥

(NN. 53a)

When the performer revolves touching the ground

either with both the knees or with both the legs

describing a circle [while her] back is bent

[backwards] with her hands in latā then it is known as

baṭu [and its] movement is like [moving] in the orbit

of the sun.

This description, insufficient as it is, corresponds to the pattern of

movements in the baṭu dance of Odissi. Further evidence is provided by

the following description of a bandhanṛtta in the Nartananirnaya, a

karananeri:

झम्पाताल: सगोपुच्छ: हस्तकोडप्यलपल्लव: ।

पार्श्वोर्ध्वजानुनौ दण्डपक्षं तलविलासितम् ॥

विद्युद्भान्तं ततश्चन्द्रावर्त्तनाम निशुम्भितम् ।

ललाटतिलकं पश्चाच्चाल्लतावृश्चिकसञ्ज्ञकम् ॥

नवाभिः करणैरभिः क्रमात् संय्यापसव्यतः ।

कृत्वालीढे स्थितिर्यत्र नेरिः करणपूर्वकः ॥

(NN. 43b-44a)

[The dancer dances] in jhampātāla in gopucchayati.

One of her hands is in alapallava. Pārśvajānu,

ūrdhvajānu, daṇḍapakṣa, talavilāsita, vidyudbhrānta,

candrāvarta, niṣumbhitā, lalaṭatilaka and latāvr̥ścika:

where these nine karaṇas are performed one after the

other facing the left and then the right in order and she

MOVEMENT AND MIMESIS

215

finishes off standing in the ālīḍha sthāna, it is

Karaṇaneri.

This sequence is one of the twelve ūrupas described in the

Nartananirnaya. Ūrupas are sequences formed with the karaṇas

prescribed for bandhanṭta and are danced to specified varieties of yati,

tāla and laya. Specific sthāna, cārī and hand gestures characterize them.

Using these twelve ūrupas a dancer can reconstruct a composition as

described in the Nartananirnaya, which will not be far from what we see

being performed by artists today. In Odissi we do find similar

compositions. Such close correspondences are now proving to be of

particular interest to many dancers and teachers who are trying to

reconstruct older dance forms by following the Sanskrit manuals. In

Tamilnadu and Orissa, eminent dancers and teachers such as Padma

Subrahmanyam and Kelucharan Mahapatra are drawing upon the śāstras

to choreograph their dance pieces. Dharu, a type of bandha dance from

Andhra described in the Nartananirnaya is being reconstructed by the

famous dancer Guru Chinnasatyam in Mysore. The term dharu is used by

Kuchipudi dancers of Andhra.

The categories known as bandha and anibandha therefore remain

relevant not only to the dance scholar but to the practising dancer. As

functional terms they may be of little use, since the distinction between

codified and uncodified dance forms has largely disappeared. But the

categorization attempted by Puṇḍarīka Viṭṭhala affords scholars valuable

knowledge about a vital stage in the evolution of dancing in India, a stage

which was marked by the infusion of new styles and techniques that

characterized the deśī tradition.

Chapter

7

THE

DEŚĪ

TRADITION

The

deśī

tradition

was

first

recorded

in

detail

by

Someśvara

in

his

Mānasollāsa.

As

a

separate,

distinct

tradition

deśī

is

mentioned

earlier

in

the

Daśarūpaka

but

no

details

are

given

there.

Earlier

still,

Bharata

recognized

in

his

Nātyaśāstra

the

existence

of

regional

variations

but

he

neither

described

them

nor

referred

to

them

as

belonging

to

a

separate

tradition

called

deśī.

1

In

the

entire

period

between

the

Nātyaśāstra

and

the

Mānasollāsa

we

come

across

only

one

other

manual

on

dancing,

the

Viṣṇudharmottara

Purāṇa,

and

this

too

does

not

record

or

refer

to

any

deśī

tradition.

Similarly,

Abhinavagupta

is

silent

about

the

deśī

tradition.

As

we

have

noted

earlier

in

the

fourth

and

fifth

chapters

of

the

present

study,

many

works

of

the

classical

period,

from

manuals

to

literary

compositions,

refer

to

a

number

of

mimetic

dances

but

no

detailed

description

of

a

dance

presentation

ever

appears

in

any

of

these

works.

This

picture

changes

around

the

eleventh

and

twelfth

centuries

when

scholars

begin

to

include

the

deśī

tradition

along

with

the

older

tradition

in

their

manuals

on

dancing.

With

the

growth

of

the

popularity

of

dancing

and

the

inclusion

of

the

deśī

as

a

tradition

in

the

early

medieval

literature

on

poetics

and

musicology,

that

is,

between

the

tenth

and

the

fifteenth

centuries,

a

number

of

manuals

appear

on

the

subject

when

art

was

flourishing

in

all

of

its

branches,

in

music,

dance,

sculpture,

architecture

and

painting.

This

interest

in

recording

the

later

styles

of

dancing

continued

through

the

succeeding

centuries

into

the

nineteenth.

1

NS.9.163-4

216

MOVEMENT AND MIMESIS

217

The description of the regional tradition differs from the mārga tradition in two ways: first, by putting its emphasis on the style of presentation rather than on the content of the composition, and second, by concentrating on the use of more acrobatic movements. As for the content, in structuring a dance piece the deśī tradition had very little to offer that was new. Mārga and deśī cannot, therefore, be judged to be exclusive of one another in terms of their components. They should rather be seen as different stylistic approaches that grew through time into separate traditions of the same basic art form. This relationship is clearly seen in the descriptions of the two traditions in the dance manuals of the medieval period.

In the first manual of this period, the Mānasollāsa (early 12th century), the deśī tradition is described only briefly, but a fuller picture appears a century later in the Sangītasamayasāra. An even more elaborate description of deśī appears still later, in the Sangītaratnākara, which was written in the late thirteenth century. The Nṛttaratnāvalī, which came a few years after the Sangītaratnākara and which deals only with dance, provides the most elaborate description of both the mārga or classical and the deśī or regional traditions of dancing. These four texts give us the first introduction to the deśī dance tradition. The deśī dance is constructed first by laying down the groundwork of dance compositions with smaller movements and gradually progressing to build up the whole composition by forming small units consisting of individual movements and moving on to the large units of a composition. In deśī, this basic method of constructing a composition did not change, nor-generally speaking-did the stock of body movements, but the larger units of movements included more varieties of smaller movements in the deśī tradition and these varieties of smaller movements were more acrobatic in general. Structurally, this is what changed in the appearance of a deśī dance, as we shall find in the four early works mentioned above.

218

THE DEŚĪ TRADITION

To begin with, these four texts will be examined to trace the progression in the content, in the styles of presentation, and in the treatment of the subject by the authors who wrote on deśī. With the exception of the Bharataṛṇava, the later texts of this period more or less follow the same pattern. Not all the manuals on deśī record a full dance presentation. The Mānasollāsa, for instance, does not give us any description of a composed dance piece. The Saṅgītasamayasāra is the first text to describe deśī dance pieces. Its descriptions of the dances include references to music, rhythm, tempo and styles of presentation but no detailed description of individual movements. This text introduces a new class of features of the presentation, calling them elements of deśī, which are absent in the Bharata tradition. The later texts also describe this new element in their description of the deśī tradition. The Saṅgītaratnākara describes dances in a similar manner with further details of the musical accompaniment. The Nṛttaratnāvalī's more elaborate descriptions have more information on the technical details of presentation but not on individual movements and it repeats the emphasis on the musical accompaniment along with the rhythm and tempo employed. The component movements of a composition, however, hardly ever appear in the descriptions of the deśī tradition found in this period of our study. The task of comparing the movements constituting the mārga and deśī styles is thus a difficult one.

This difficulty is further compounded by the fact that not all the details of the mārga style are known. By the early medieval period the tradition recorded by Bharata, was established and recognized as mārga, a term first used by Dhananjaya to distinguish between two traditions. Beside Bharata's account of the older tradition we have the only other extant record in the Viṣṇudharmottara Purāṇa. In contrast, regional styles, termed as deśī, have been recorded by a number of authors who give us a varied picture of the deśī styles. These authors also describe the mārga tradition but in doing so merely follow the Nāṭyaśāstra. The earliest record

MOVEMENT AND MIMESIS

219

of a dance presentation appears in the fourth chapter of the Nāṭyaśāstra

which forms the preliminary part of a dramatic presentation, that is, the

part in which the gods are invoked. It includes nṛtta and abhinaya.

Unfortunately, even Bharata does not give detailed descriptions of the

movements to be employed in the presentation, nor does he record any

specific rhythm and tempo to be used in this presentation. Nevertheless,

the basic difference between the descriptions of the presentations of the

two traditions, I believe, becomes apparent when we closely compare

them. Bharata describes the structure of the presentation but is silent

about the presentation technique or specific music. Also, he speaks only

briefly on the aesthetic aspects of a presentation. On the contrary,

discussing these seems to be a growing concern of the writers on deśī.

To understand Bharata’s view of dancing we shall first examine how

he describes the presentation in the fourth chapter of the Nāṭyaśāstra, a

performance meant for the offering of flowers, that is, puṣpāñjali, at the

beginning of a play in order to please and invoke the gods. This

presentation involves both abstract dance movements (that is, non-

mimetic movements) with drum accompaniment, and mimetic action set

to songs. Bharata describes no presentation that is constituted exclusively

of the abstract movements known as karaṇas and aṅgahāras. In describing

this preliminary presentation he does not specify the names of movements

to be employed. The only specific name given by Bharata is that of a

posture, vaiśākha sthāna, mentioned in this context as the opening posture

of this presentation.2 He does not specify any music for the presentation as

do the works on deśī while describing the dance presentations. Bharata

leaves the actual choreography of a dance piece as well as the

accompanying music to the imagination of a dancer or a choreographer.

The presentation is described in the following manner: a female dancer

performs nṛtta to the playing of drums, offers flowers, moves around,

2Vaiśākha sthāna happens to be the basic posture of the Odissi style of today.

220

THE DEŚI TRADITION

bows to the gods and mimes the meaning of a song without the accompaniment of drums. After this, she goes out and a group of dancers enter and perform various formations resembling the emblems of various gods, after which they make their exit. The main dancer enters again and repeats her performance of nṛtta and abhinaya and exits. The others again enter in the same manner as before and repeat their dance. Their performance is devoid of miming.3 All of this is repeated again. The entire presentation is a repetition of nṛtta and abhinaya by one female dancer followed by the performance of nṛtta by a group of female dancers.4 This presentation seems to be similar to what we see in the varṇam of the bhāratanāṭyam style of today except for the part played by the group of dancers.5 However, since the description of puṣpāñjali lacks specific instructions for the movements of the music, we do not know which movements constituted the nṛtta portion, nor do we have any idea of the specific music or rhythm used. All that we can say is that it seems to indicate that a scope for improvisation was permitted within the prescribed framework even in Bharata’s time but, as we shall see when we examine the deśī tradition, such improvisation was acceptable here in a more restricted sense than in the deśī tradition.

The analysis of the contents of the dance presentations described in the later texts seems to indicate that the dancers had greater freedom to improvise, because they were allowed to begin improvising at an earlier stage of composition than in the mārga tradition. Apart from this greater freedom, the emphasis on one other aspect of presentation gradually becomes more and more apparent. Not only do these authors of the early

3 paryastaka or abhinayaśūnya (non-representational) dance is recommended. NS. 4. 280; AB. 1956,Vol. I, p. 188.

4 NS. 4. 272-291.

5 Padma Subrahmanyam, the noted performer and scholar of the Bharatanāṭyam style, is experimenting with Bharatanāṭyam danced by groups of dancers.

MOVEMENT AND MIMESIS

221

medieval period often emphasize the mimetic aspect of the presentation,

they also describe the accompanying songs and music, often even

specifying the tāla. The overall presentation technique, especially its

aesthetic aspects, seems to be of greater concern for the writers on deśī.

This approach is different from the way Bharata describes dancing. When

we take a closer look at the specific dance pieces described in the later

texts it seems that those who developed regional variations based their

repertoire on the movements recorded in the Nātyaśāstra but combined the

movements in their own ways rather than in a set pattern when forming

dance units (or karaṇas) and dance-sequences (or aṅgahāras). Bharata

gave directions that guided the dancer to the stage of the formation of the

basic dance-units as well as the basic dance-sequences. His approach

permitted the dancer only the permutation and combination of a number of

the basic dance-sequences. Creativity was based on the use of such groups

of basic dance sequences. However, while describing the presentation

mentioned above, Bharata does not specify any dance unit, movement or

sequence except for the introductory posture of the performance. The

presentation is neither described as a dance piece nor is it described as a

mimed piece but as a combination of both. Nor does Bharata offer any

detailed guidelines to the aesthetic qualities of the presentation, which are

of far greater concern to the later authors of the early medieval period.

Their interest in these qualities is seen, for instance, in their use of a

variety of terms to describe this aspect of presentations, whereby they

attempt to specify the aestheic impact. Bharata does mention sauṣṭhava or

grace and pramāṇa or harmony in a presentation, but he devotes little time

to it. By contrast, the later authors emphasize the aesthetics of a

presentation and the greater number of terms they use testifies to the rise

of a more elaborate set of ideas in the field.6

6 See Bose, 1970, pp. 165-66.

222

THE DEŚĪ TRADITION

We may now attempt to examine these later texts chronologically to

see the evolution of dance-movements in the tradition as recorded by the

writers on deśī. The Mānasollāsa, the earliest text on deśī, shows us that it

follows the basic framework laid down by Bharata and this seems to be an

accepted pattern through the entire history of the writings on dancing.

Although Someśvara's description of the deśī tradition is comparatively

brief, his contribution in defining the parts of the body is important. To the

existing division of the body into aṅgas and the upaṅgas, he adds one more

division, the pratyaṅgas. We find that after this text appeared, all the

authors of the medieval period, except for Pārśvadeva in his

Saṅgītasamayasāra and Rāmacandra and Guṇacandra in their

Nātyadarpaṇa, began to describe the body parts under these three

divisions. However, this was only a reclassification: it did not materially

change the movements. The movements of the smaller body parts,

previously known as the upaṅgas, now came to be placed under two

categories, the upaṅgas and the pratyaṅgas, while the category of major

limbs, the aṅgas, remained unaltered. The descriptions of the smaller

units of movements by Someśvara show us that new movements were

added to the same class-name found in the Nāṭyaśāstra. but under deśī

sthāna, the later authors added more varieties to the tradition recorded by

Bharata.

Can we not say, then, that these later authors merely added to the

tradition of Bharata rather than branching out from it? In what way did

they differ then and why was the tradition they described termed deśī?

The answer is that the difference was on the emphasis: the deśī tradition

differed from the tradition of Bharata because the deśī dances involved

more leaping and acrobatic movements and the writers describing them

insisted on greater attention to aesthetic qualities. No new categories

were created for these acrobatic movements, although new-one might say

enhanced-varieties of sthānas, cārīs and even a few karaṇas of the deśī

MOVEMENT AND MIMESIS

223

type involving leaps and acrobatic movements appeared in the

Mānasollāsa. The writers on deśī dances remained content with the

established class-names for such movements, namely, sthānas, cārīs and

karaṇas.

The texts of our period seem to have been mainly interested in

recording the movements of the parts of the body and their smaller units in

course of describing the new regional varieties. The descriptions of the

deśī cārī, sthāna and maṇḍala occur in all the later texts but the basic

dance-units or karaṇas of the deśī variety are not. As for the aṅgahāras or

basic dance-sequences of the deśī variety, they are described only in the

Bharatārṇava, being absent in the other texts even when a composed

dance-piece is described with the accompanying music. However, this

approach of not describing the dance movements in a choreographed piece

is not much different from what we find in the description of the puṣpāñjali

of Bharata. Still, the consistent absence of descriptions of the entire set of

aṅgahāras in the deśī tradition tells us that the regional varieties may not

have used the aṅgahāras recorded in the Nāṭyaśāstra. Instead, the deśī

tradition grew in a different direction where the choreography depended

less on set units of movements taken from Bharata's tradition.

The interest in acrobatic compositions that characterized the deśī

tradition grew through time. We find that in later medieval times the

authors attempted to describe dance pieces that often required new

combinations of smaller movements to produce new varieties of dance-

units or karaṇas, often involving leaps and acrobatic movements. Such

dance-units are called deśī karaṇas, or quite often, utplutikaraṇas, a name

that indicates that they involve leaps. The Mānasollāsa, the

Saṅgītasamayasāra, the Saṅgītaratnākara and the Nṛttaratnāvalī, all

describe these deśī karaṇas. However, the Mānasollāsa describes only a

few, whereas the later texts record a great many more.

The distinctive features of the deśī varieties of movements involving

jumps and acrobatic movements become clear when the movements

described in these texts are examined. To begin with, we shall take a few

instances from the Mānasollāsa of Someśvara to describe them and

compare them with the tradition of Bharata. Someśvara describes the deśī

karaṇa named eṇapluta where the dancer jumps and performs sūcī in the

air (Mānas.16.4.1396-97). Again, in the deśī añcita karaṇa the dancer

stands in sama and then jumps straight up (Mānas.16.4.1384), and in the

deśī antarālaga karaṇa the dancer jumps while arching her back and

keeping her head between her shanks (Mānas. 16. 4. 1388). In the tradition

of Bharata there are some difficult and acrobatic karaṇas like

gaṅgāvataraṇa where the feet are raised with the soles turned up and the

hands turned down in tripatākā while the head is bent. The whole body

rests on the hands (NŚ. 4. 168-69). Or, we may take the vidyudbhrānta

karaṇa, where one foot is moved in an arc backward and then extended

and arched to touch the head and the dancer revolves. (NŚ. 4. 125-26).

Thus, some acrobatic karaṇas can be cited from the Nātyaśāstra but few

karaṇas in it involve such difficult movements. Besides, they seldom

involve leaps which seems to be the most important feature of the deśī. On

the other hand, the deśī karaṇas always contain acrobatic and leaping

movements. Even in the descriptions of the sthānas and cārīs we find that

the deśī varieties use more jumps and acrobatic movements compared to

those in Bharata's tradition.7

We must also note that the Mānasollāsa is concerned only with the

basic movements and a few deśī karaṇas, that is, the components of larger

dance units and not with dances in their entirety. Because its interest lies

in recording the smaller units of movement, it describes the deśī varieties

of sthāna, cārī and karaṇa. But large compositions of any kind using these

movements are not its concern.

It is not until the Saṅgītasamayasāra of the fourteenth century that we

find any description of a complete dance. This text not only describes

7 Ibid. chapters 5 and 6.

MOVEMENT AND MIMESIS

225

specific dance pieces but adds a number of new movements of the cārī, the

sthāna and the karanas of the deśī variety, all of which involve

complicated leaping movements.8 In the beginning of his chapter on

dancing, the author Pārśvadeva mentions two kinds of presentation, nṛtta

and nāṭya. He states that he is going to describe only āṅgīka or body

movements, a class of movements that is of particular relevance for nṛtta.

When he finishes describing these movements he proceeds to describe

modes of presentation and finally to fully composed dance pieces. Such

pieces he calls deśīṅṛtya and not deśīṅṛtta. The four complete

compositions which he describes, calling them deśīṅṛtyas, are peraṇa,

pekkhaṇa, guṇḍalī and daṇḍarāsa (SSam. 7.132; 7. 217-47). Peraṇa may be

the same as the preraṇa described by Abhinavagupta but it is hard to get a

full picture of latter from his cryptic description, and hence it is difficult

to be certain that the two were the same. It is likely that Abhinavagupta's

preraṇa was not a dance proper but it may well have developed as a dance

by the time when Pārśvadeva recorded it. Abhinavagupta quotes the

cirantanas on a type of minor drama involving dance and music which was

named preraṇa and which he describes as a farcical play, full of riddles.

The peraṇa of Pārśvādeva calls for five components, which are nṛtta,

kaivāra, gharghara, vāgaḍa and gīta. He describes nṛtta as consisting of

lāsya and tāṇḍava, which are based on rhythm and tempo. This definition

is different from previous definitions of nṛtta found in the Nāṭyaśāstra and

in the Viṣṇudharmottara Purāṇa. Nevertheless, the basic idea of nṛtta does

not seem to be different. Kaivāra is praising the king by praising his

ancestors. Gharghara is stamping the ground with the feet in ankle bells to

mark the rhythm. Vāgaḍa is the miming of farcical or ludicrous

characters.9 Finally, gīta, the last feature of peraṇa (the term gīta here is

used as a technical feature peculiar to preraṇa, very likely a set musical

8 SSam.7.126-92.

9 This feature resembles preraṇa described in the Abhinavabhāratī (NS. Vol. I, p.181).

226

THE DEŚI TRADITION

piece), denotes a song sung according to the rules of a pure or mixed rāga,

complete with ālāpa (SSam. 7.217.-22). The details of the instrumental

music to accompany are also given with the rhythm and time-beat

specified(SSam. 7.223-25).

The other two dances, pekkhaṇa and guṇḍalī, are the most commonly

described in all the texts dealing with deśī. They are described along with

accompanying songs and instrumental music, rhythm and time-beat

(SSam. 7.225-36). These songs can be of three types 10 and all of them

can be applied to the two above mentioned dances.

No song is prescribed for the last deśīnrtya described, called

daṇḍarāsa, but the instrumental music along with the rhythm and tempo

and the number of dancers dancing with sticks are mentioned (SSam.

7.240-47). The first three dance pieces described by Pārśvadeva call for

abstract movements mixed with mimetic movements. Pekkhaṇa and

gaunḍalī are described with accompanying vocal and instrumental music.

The fourth dance, daṇḍarāsa, requires no miming but is likely to draw

upon some Krṣṇa legend, as we gather from the later texts on dancing as

well as from other literature. Daṇḍarāsa and its connection with the Krṣṇa

story cycle has been discussed in a previous chapter while discussing the

minor dramatic types that are described mainly in the works on poetics. 11

However, the Sangītasamayasāra is silent about daṇḍarāsa's connection

with Krṣṇa. The association of Krṣṇa with daṇḍarāsa may have developed

later when stories of Krṣṇa became popular in the western regions of

India.

The Sangītasamayasāra adds a vital piece of information to the

description of the deśī tradition when it lists what it calls the nineteen

aṅgas or elements of deśī. These elements are not physical components but

10 These are śuddha (pure), citra (varied) and sālaga (based on pure, chāyālaga>sālaga)

SSam. 7. 229-34.

11 See pp. 187-88 supra.

MOVEMENT AND MIMESIS

227

modes of presentation particular to the deśī style. This emphasis upon the

mode of presentation shows the basic difference between the writers on

deśī and Bharata (and his immediate followers) in their respective

approaches to dancing. Even though Bharata describes body movements in

detail in the context of āṅgikābhinaya (that is, using the body in acting), he

never discusses the manner of performing the movements. By contrast,

the description of the deśī aṅgas in the Saṅgītasamayasaāra shows the

reader the qualities that characterize the movements.

The terms used by Pārśvadeva are mostly in the regional languages and

not in Sanskrit. Most of them refer to delicate modes. These nineteen

aṅgas of deśī, according to Pārśvadeva, are the following: mukharasa

(facial appearance created by make-up, garlands and dress), sauṣṭhava

(grace produced by proper and erect stance), lali (charm), bhāva

(expression), tūkalī(swinging movement in rhythm), anumāna (miming

while moving in specific gatis), pramāṇa (harmony), jhaṅkā (moving the

body, which is raised to the left and to the right), revā ( the head and the

corners of the eyes moved in emotion), surekhatva (dancing without

ludicrous movements), aṅga (performance of tāṇḍava), anaṅga

(performance not dependant on tāṇḍava), dhāla (expressing emotion by a

female dancer), dhillāyī(delicate and relaxed slow movement or a

standing relaxed posture), namani12 (bending down with ease even while

performing difficult movements), kittu (moving the arms and the breasts

delicately in rhythm), tarahara (quick movements of the breasts, the

motion reaching upto the arms), ullāsa (the body moved upwards

following the instrumental music) and finally sthāpana (the final stance

producing beauty, grace and creating a symmetry of lines with difficult

movements). Pārśvadeva remarks that both nāṭya and ṇṭta depend

entirely on sauṣṭhava, without which beauty cannot be created (SSam. 7.

199-200). He also states that delicate movements are dependent specially

12 Tavaṇi in the Trivandrum edition of the SSam. 6. 208.

228

THE DEŚI TRADITION

on two features, lali and bhāva, that is, charm and expression (SSam. 7.

208).

Bharata is not entirely silent on the manner of presentation but his

treatment does not emphasize it as a technical requirement. When he uses

the two terms, sauṣṭhava and pramāṇa (NŚ.10.94), he takes them as

general qualities of presentation rather than as the technical aspects of a

presentation. The closest successor to the Nāṭyaśāstra is the

Viṣnudharmottara Purāṇa, which does not mention qualities of

presentation, nor does Someśvara. Pārśvadeva, however, brings them to

the forefront when he calls them deśyaṅgas or features of deśī.

The same emphasis on presentation appears in the text that comes next,

the Saṅgītaratnākara by Śārṅgadeva. It does not use the term deśyaṅga but

describes ten deśī lāsyāṅgas, that is, features of deśī lāsya, namely,

delicate modes of presentation, sometimes abstract and sometimes

mimetic. These are different from the lāsyāṅgas of the Bharata tradition,

regarding which Śārṅgadeva is silent. Lāsyāṅgas of the Bharata tradition,

as we have mentioned earlier in chapter 4, are part of dramatic

presentations and as such are not relevant in the context of dance.13 These

lāsyāṅgas of Bharata are quite different from the nineteen deśī aṅgas

described in the Saṅgītasamayasāra, from the ten deśī lāsyāṅgas described

in the Saṅgītaratnākara, and from the forty-six deśī lāsyas in the

Nṛttaratnāvalī.

The diversity of the accounts of these movements shows that although

the deśī tradition depended upon the body movements mentioned by

Bharata and other early authors, in some cases the deśī tradition also added

to the stock by considerably elaborating upon the basic movements

described by Bharata. A case in point is the Saṅgītasamayasāra's

inclusion of five bhramarīs or revolving movements as a separate set of

movements, which is a new addition to the dance tradition. Bharata's

13

See pp. 145-46 supra.

MOVEMENT AND MIMESIS

229

tradition described a bhramarī merely as a variety of cārī, which is a revolving movement performed with one leg. It seems that from the time when Sangītasamayasāra recorded bhramarī as a separate set of movements, they gained major importance and became a common feature of the deśī tradition. However, the Sangītasamayasāra does not describe the five bhramarīs that its author mentions (SSam. 7.193-4). The Sangītaratnākara is the first text to describe the movements of bhramarīs in detail. It is apparent that in the styles from the south bhramarīs were important features, hence their being recorded in the Sangītaratnākara, in the Nṛttaratnāvalī and in the Nartananiṛnaya 14 In today's Indian styles, however, the styles that have retained bhramarīs are those from the north, Kathak and Manipuri, although these styles delineate bhramarīs in two different ways. Kathak has spinning movements while Manipuri has slower, revolving movements which involves covering larger space.15

Śārṅgadeva's account, again, shows how the deśī tradition was being expanded, though on the groundwork of Bharata's tradition. Besides describing the bhramarīs and deśī lāsyāṅgas, Śārṅgadeva also distinguishes between two traditions of presentation, one consisting of what he calls the śuddhapaddhati or pure method, and the other consisting of two deśī forms named peraṇi and gauṇḍalī. By the time Śārṅgadeva wrote his text, these two dances seem to have been established in the traditions of deśī and they remained part of the deśī technique from the beginning of the history of deśī dancing till the eighteenth century, as evident in the literature of dancing.16 Śārṅgadeva, like Someśvara and Pārśvadeva, describes the

14 SR. 7. 775-82; NR.5. 83-99; NN. 47b.

15 Chakkars or spinning movements are vital to the kathak style (Kalyanpurkar, 1963, p. 24) and revolving movements are vital to cāli, the basic dance of the Manipuri style(Chatterjee, 1978, p. 226).

16 Guṇḍalī and peraṇī appear in all the texts, starting with the Sangītasamayasāra and ending with the Saṅgītasārasaṅgraha. Evidently, in Andhra Guru Vempati

230

THE

DEŚĪ

TRADITION

movements

of

the

parts

of

the

body

following

Bharata.

But

he

adds

more

varieties

of

deśī

cārīs,

sthānas

and

karaṇas

to

the

existing

lists

in

the

two

former

writers

on

deśī.

He

records

nine

bhramarīs

and

includes

them

in

his

list

of

deśī

karaṇas.

Apart

from

making

these

additions,

Sārṅgadeva

demonstrates

a

strong

concern

for

the

overall

qualities

of

presentation,

which

is

expressed

by

his

interest

in

rekhā,

defined

by

him

in

the

following

lines:

शिरोनेत्रकरादीनामझानां मेलने सति ।

कायस्थितिर्मनोहारि रेखा प्रकीर्तिता ॥

(SR.

Rekhā

is

said

to

be

the

attractive

position

of

the

body

[when

there

is

a]

harmonious

combination

of

the

movements

of

parts

of

the

body

such

as

the

head,

the

eyes

and

the

hands.

According

to

Sārṅgadeva,

then,

rekhā

is

the

composite

image

of

the

lines

inscribed

on

the

mind's

eye

by

the

movements

of

the

body

at

any

given

moment.

In

addition

to

his

interest

in

rekhā

he

was

also

aware

of

the

need

for

yet

another

intangible

quality,

one

that

was

known

as

the

sauṣṭhava

of

a

presentation,

about

which

he

says:

कटी जानुसमा यत्न कूर्परांसशिरः समम् ।

उरः समुन्नतं सन्नं गात्रं तत्सौष्ठवं भवत् ॥

(SR.

Where

the

waist,

the

knee

are

kept

in

straight

[line],

the

elbows,

the

shoulders,

the

head

are

held

straight,

the

chest

is

raised

and

the

body

rests

in

sanna

[=

resting

in

a

natural