1. A Study of Dattilam A Treatise on The Sacred Music of Ancient India Mukund Lath

A STUDY OF DATTILAM

A Treatise on the Sacred Music of Ancient India

Mukund Lath

A STUDY OF DATTILAM

A STUDY OF DATTILAM

A Treatise on the Sacred Music of Ancient India

MUKUND LATH

IMPEX INDIA

NEW DELHI

© MUKUND LATH

January, 1978

Price : Rs 150.00

$ 30.00

PUBLISHED BY SITA RAM GOEL FOR IMPEX INDIA, 2/18 ANSARI ROAD, NEW DELHI-110 002

AND PRINTED BY R. K. PRINTERS, 80-D, KAMLA NAGAR, DELHI-110007. PHONE : 222638

ACKNOWLEDGEMENTS

I AM indebted to many discerning people in forming and expressing my ideas.

The work was originally submitted as a thesis for Ph.D to the Department of Sanskrit, University of Delhi. Dr. (Km) S. R. Kulashreshtha who was my guide and supervisor, has been of the utmost help. She offered fruitful suggestions at every step.

Much of the present work was done on the basis of notes taken in 1967-68 at the International Institute for Comparative Music Studies and Documentation, West Berlin, directed by the famed indologist and musicologist Mr. Alain Danielou. I was fortunate in having my friend Mr. Vivek Dutta as my colleague at the Institute. Mr. Dutta creatively helped in the process of sorting and sifting relevant material, solving problems pertaining to Dattilam and in clarifying many of the major ideas presented. Without Mr. Dutta's perceptive collaboration, the work may not have been possible. Mere thanking him cannot be acknowledgement enough for his help.

I thank the Berlin Institute (whose headquarters are now in Venice) and its director for having placed its well-stocked collection of manuscript copies of saṅgīta texts and other valuable works at my disposal.

I express my gratitude to my gurus in music, Sri Ramesh Chakravarty, Pandit Maniram and Pandit Jasraj for having introduced me to the living traditions of the art of Indian music .

Thankful acknowledgement is due to Sri Ram Swarup, Sri Hari Prasad Lohia, Sri Sitaram Goel and Dr. Govind Chandra Pande (Professor of History and Indian Culture, University of Rajasthan, Jaipur) for having read significant portions of the work and offered helpful suggestions. Dr. Satyavrata, Head of the Department of Sanskrit, Delhi University, was kind enough to take time out of his busy schedule and go through the work.

A number of valuable points emerged in discussions with Dr. Premlata Sharma at the Hindu University, Varanasi, where she is the Head of the Department of Musicology. I am grateful to her for the keenness with which she examined the work and her perceptive suggestions.

Goelji further deserves my thanks for having undertaken the publication of the work. He has contributed much to the final shape in which the book appears—not only as to format but also as to contents—for he is a fine scholar and writer as well as a publisher. He has spared the author days of painful proof-reading by taking on himself a large share of the taxing job.

I am grateful to my friend Vinay Jain for having helped in the long, dull task of index-making with cheerful patience and perseverance.

ABBREVIATIONS

A. B.

The Abhinava Bhāratī on the Nāṭyaśāstra, Gaekwad Oriental Series Edition.

Ait. Br.

Aitareya Brāhmaṇa.

A.V.

Atharva Veda.

B. B.

The Bharata Bhāṣya of Nānyadeva (or Nānyabhūpāla), Chapters I-V edited by Chaitanya P. Desai. Published by Indira Kala Sangita Vidyalaya.

B. B. (I)

The Bharata Bhāṣya of Nānyadeva, copy of original ms. no. 312 belonging to the Bhandarkar Institute, Poona, at the International Institute for Comparative Music Studies and Documentation, West Berlin. We refer to chapters only, as verse numbers are uncertain.

Br̥.

The Br̥haddeśī of Matanga Muni edited by K. Sambasiva Sastri, Trivandrum Sanskrit Series.

Bh. Sang. Iti

Bhāratīya Sangīt Kā Itihās (Hindi) by Dr. Sarat Chandra Paranjape.

Bh. Sang. Iti (B)

Bhāratīya Sangīter Itihās (Bengali) by Swami Prajnananand. Published in two volumes by Ramakrishna Vedanta Math.

Datt.

The Dattilam of Dattilamuni, Trivandrum Sanskrit Series, 1930.

G.O.S.

Gaekwad Oriental Series.

IICMSD

International Institute for Comparative Music Studies and Documentation, West Berlin, Directed by Mr. Alain Danielou.

Hist. of Ind. Lit.

History of Indian Literature by Maurice Winternitz.

J. M. A.

Journal of the Music Academy, Madras.

Kalā

The Commentary Kalānidhi on the Sangītaratnākara by Kallinātha. We refer to the Adyar edition of the text.

MBh.

The Mahābhārata. We refer to the Chitraśala edition, Poona, published with Nilakaṇṭha’s commentary.

N. S.

The Nāṭyaśāstra of Bharata. We refer to the G.O.S. edition published in four volumes with the Abhinava Bhāratī commentary unless otherwise stated.

Puranic Record.

Puranic Records on Hindu Rites and Customs by Dr. R. C. Hazra.

Ram.

The Vālmīki Rāmāyaṇa, Gita Press, Gorakhpur.

R. Kau

The Rasakaumudi of Śrikanṭha edited by Dr. A. N. Jani published in the G.O.S.

R.V.

Rgveda. Numerals refer to maṇḍala, sūkta, mantra.

S.Br.

Śatapatha Brāhmaṇa

S.C.

The Sangītacintāmaṇi of Vemabhūpāla. We refer to the ms. copy no. 189 in the collection of HCMSD. The number and location of the original ms. is not indicated. The verse order is cerrupt. We refer only to chapters and sections.

S,R.

The Sangītaratnākara of Śārṅgadeva. We refer to the Adyar edition published in four volumes.

S.Raj

The Sangītarāja of Rāṇā Kumbha. We refer to the Banaras Hindu University edition, edited by Dr. Premlata Sharma.

S.S.S.

Sangīta Samaya Sāra of Pārśvadeva, Trivandrum Sanskrit Series.

Sudhā.

The Sudhākara commentary on Sangītaratnākara by Simhabhūpāla, Adyar edition.

T. ed.

Trivandrum edition.

TABLE OF TRANSLITERATIONS

Devanāgarī

Roman

Devanāgarī

Roman

अ

आ

इ

ई

उ

ऊ

ऋ

ॠ

ऌ

ॡ

ए

ऐ

ओ

औ

ं (anusvār)

ः (visarg)

क

ख

ग

घ

ङ

च

छ

ज

झ

ञ

ट

ठ

ड

ढ

ण

त

थ

द

ध

न

प

फ

ब

भ

म

य

र

ल

व

श

ष

स

ह

To

Ma

who has devoted her life to the

promotion of Arts and Culture

PREFACE

Authoritative studies of ancient Indian music are rare. Textual studies are rarer still. Yet such studies are needed if an authentic picture of ancient Indian music is to emerge. It is necessary that we study the important texts, relate them to other texts of that and other periods and try to analyse and sift forms, concepts, theories and ideas with a view to interpret and comparatively understand the changing and evolving traditions of music.

In a very fundamental and general way, musical culture in India has maintained a vital continuing link with the past. In form and spirit, our music remains true to the seeds sown in a very remote age. Witness, for example, the significant fact that the melodic features and formal principles (called the ten jāti-characteristics) which governed the early musical forms still govern our own rāgas in their salient features. A similar link with the past is to be found not only in music but also in other art forms with a living tradition.

At the same time, dynamic transformations or rather transitions have been as much a part of Indian musical history as its essential living continuity. This is bound to be so in any living tradition which has not become stagnant. Old forms in music which held sway for centuries, have given place to new. The ancient doctrinal scheme of interpreting musical structures has been transformed in many basic ways.

To form an idea of these changes we must turn primarily to the testimony of the texts. Not many ancient texts have survived. One of the earliest and perhaps one of the most eminent is the Dattilam, attributed to Dattila Muni.

The importance of Dattilam was realised as soon as it came to light and was published in 1930. But doubts still continue to abound regarding its status and indeed its very authenticity and originality as well as its completeness.

It has been our task, in this dissertation, to show through evidence both internal and external that the text is a work complete in itself and is both authentic and original. To this end we have examined the testimony, the logical plan and structure of the text itself which reveals it to be a self-contained fully-formed treatise and not a fragment or a partial work as alleged. We have also examined the text vis a-vis the chapters on music in the Nātyasāstra (of which the work has been alleged to be a derivative) and attempted to present data, which establishes the originality of Dattila and the Dattilam. We have also presented the testimony of much of the significant saṅgīta literature from the Brhaddeśī of Mataṅga (Circa 7th Century A.D.) onward, which in references to and quotations from the Dattilam attests to the independent views, original stance and authentic stature of Dattila as an ancient teacher. We have collected together a string of passages quoted from the Dattilam in subsequent texts and have endeavoured to sift and review this data critically. It has helped us to reveal the authenticity of Dattilam.

The whole testimony shows that Dattila was at least as ancient as Bharata and that the Dattilam is almost certainly his authentic creation.

Our study, further, tries to show that the impression of its being a fragmentary and incomplete treatise arises due to its confinement to only one area of ancient Indian music. It deals with a specific and well-defined corpus of musical forms which consisted of only a part, though a very significant part, of the entire body of ancient musical forms. This corpus was called ‘gāndharva’ and the Dattilam is avowedly about gāndharva. In our opinion, its author had deliberately and knowingly limited himself to this form.

Confusion arises regarding Dattilam because of the use of the term gāndharva by the author for defining its subject-matter. The word in ordinary usage has for centuries meant music in general or, in other words, all music. Yet the Dattilam does not deal with all contemporary music. We have reasoned at length that the term ‘gāndharva’ was also used in a restricted sense for a particular form of music which was ancient and sacrosanct and was governed by strict rules and regulations. Dattilam, we have argued, deals with this form.

The independent stature of gāndharva has been recognised by many scholars; but confusion remains as to its exact forms and extent. Scholars would assign to it many forms other than those expounded in Dattilam. The muddle is due largely to the fact that our present day notions regarding gāndharva are essentially derived from Śārṅgadeva and his successors who equate gāndharva with mārga. Swami Prajñānānanda, for example, has often spoken of gāndharva in his works. In Rāgā o Rūpa (Bengali), following the cue of Śārṅgadeva and his commentators, he speaks of gāndharva as a form of sacred music and equates it with mārga (see pp. 94-100). In his introduction to Saṅgīta Sāra Saṅgraha1 (pp. 13-15) he again speaks of gāndharva. In this introductory essay, he lists the very topics which Bharata in the 28th chapter of Nāṭyaśāstra names as definitive of gāndharva. This is a hint pointing at the direction that we have taken in our present study. But Swami Prajñānanda again equates gāndharva with mārga.

A more pointed hint at the direction we have taken is to be found in Dr. Premlata Sharma’s introduction to Sahasarasa (Hindi). She speaks of the distinction between gāndharva and gāna made by Abhinava and quotes passages from Abhinava, as well as Bharata, on the matter, citing briefly a few of the major traits that distinguished gāndharva from gāna, especially the different treatment of words in the two.2 But she, too, seems to imply some equation between gāndharva and mārga.3 She has, moreover, merely touched the periphery of the matter.

Our own discussion of gāndharva is detailed. We have attempted to study the matter comprehensively on the basis of the thorough and masterly analysis of gāndharva made by Abhinava (in his commentary the Abhinava Bhāratī on the

1 Published by Ramakrishna Vedanta Math, Calcutta, 1956.

2 Sahasarasa, Introduction, pp. 111-117. The work is published by the Saṅgīta Natak Akademy, Delhi.

3 Ibid., Introduction, p. 111 fn.

Nāṭyaśāstra). Abhinava's treatment vividly brings out the distinctness of gāndharva as a special corpus of musical forms.

The Nāṭyaśāstra in its chapters on music (collectively called the geyādhikāra) has also expounded gāndharva as does the Dattilam. But the Nāṭyaśāstra expounds other musical forms too. And though there are explicit hints here that point at the distinctness of gāndharva, these are easily overlooked without a study of Abhinava, who carefully sifts the gāndharva from the non-gāndharva. In the process, Abhinava reveals the scope, characteristic features, and the distinctive purpose and motivation of the gāndharva form.

Abhinava's analysis shows that gāndharva in a restricted sense was no more and no less than the body of music defined and expounded in the first three chapters of the geyādhikāra in the Nāṭyaśāstra (chs. 28-31), which gives the same extent of gāndharva and confines it to the same boundary as does the Dattilam.

Dattilam is thus revealed to be an independent treatise on a specific corpus of ancient Indian music.

Abhinava's valuable cue regarding the nature and extent of gāndharva has acted as our basic guide in judging the subject-matter of the Dattilam. We have then tried to show that a distinct body of manuals defining and expounding gāndharva was in existence before Bharata and Dattila and the tradition must have continued after these two authorities. These manuals are unfortunately lost now. But traces, in the form of names and references are found in Dattilam and to a greater extent in later saṅgīta literature, where quotations are sometimes found from ancient authorities other than Bharata and Dattila regarding forms which were the field of gāndharva.

Though early manuals of gāndharva, besides the Dattilam and the section on gāndharva in the Nāṭyaśāstra, have been mostly lost, awareness of gāndharva as a specific form can yet be gleaned from significant pre-Mataṅga (i. e., pre 7th Century A.D.) literary testimony. This testimony is recorded in some of the earliest Purāṇas, where the subject of music has been discoursed upon. It is also recorded in the Yājñavalkya Smṛti. We have reviewed this testimony, too.

Much valuable material regarding gāndharva is also found in later post-Abhinava saṅgīta literature. In fact, any understanding of gāndharva forms is bound to lean greatly upon this later literature which is happily very thorough and extensive in its treatment of many aspects of ancient music. Without the help of this literature, much would remain obscure. Paradoxically, it is here that confusion sets in regarding gāndharva and its extent. This confusion begins, notably, with the Sangītaratnākara, and continues in later texts all of which have cumulatively contributed to it. We have analysed the nature of this confusion, its reason and its source. The term gāndharva has in later literature been equated with mārga—a term with a comparatively larger denotation. This has been the source of muddle regarding gāndharva. We have argued that mārga was not the same as gāndharva. Our analysis also aims to show that even though Śārṅgadeva has confounded gāndharva with mārga, he yet evinces a tacit awareness of the distinction between the two,

The importance of our analysis of gāndharva is, however, not only one of categorisation, classification and arrangement of musical forms. These are necessary for any analysis and they have been undertaken to show that Dattilam is a complete work and it completely (though concisely) describes the forms and extent of a specific corpus of music—the gāndharva.

But there is a deeper significance underlying our study of gāndharva. Gāndharva was not only a specific musical form, but also the form-supreme. Gāndharva was said to be born of sāma-music and was as sacrosanct as sāma. Like the older form, it was said to be ‘revealed’ (and therefore called anādi, i.e. existing from time immemorial); it was as well-defined and controlled as a body of Vedic ritual to which it was likened. Though in its origin, it might have been born of sāma, yet it contained much that was not in sāma. It emerges, in fact, as the first, the primal body of known music to which we can relate our own present-day musical forms. Indeed, gāndharva has been called the progenitor of all subsequent musical forms. The process of unfoldment and creative change which was initiated with gāndharva in our musical culture has continued over the centuries and is still alive in current music. Old links remain. Much that is distinctive in all post-gāndharva music is, in essence, existent in gāndharva.

The importance of gāndharva is thus two-fold.

Firstly, it gives us an example of a body of music which was pervaded by a transcendental significance not only in spirit but in every detail of its melodic and rhythmic (tāla) content. Every movement in gāndharva, whether of melody or of tāla, was sacrosanct and had to be rendered as prescribed by specific injunctions which spelled out each particular detail. Recording these prescriptions was the function of authoritative manuals like the Dattilam. Gāndharva was thus akin to religiously prescribed and liturgically regulated musical forms such as were known in ancient Egypt and other ancient cultures. It reflected the same tendency which has been preserved to a great extent in the Gregorian Chant of Europe. It was a body of music the like of which we no longer know in India.

We call our music ‘spiritual’ as opposed to Western music which is said to be ‘profane’. Such appellations, however, reveal nothing about the nature of the music—neither its spirit nor its form. Though vaguely termed ‘spiritual’ in general, the same musical forms are today used for both profane and spiritual purposes. The same rāga can be sung to a devotional hymn and to a profane erotic lyric, without any formal ‘musical’ difference being noticeable. But apart from the lyrical content and of course, a general psychic attitude—which somewhat seeps into and formally moulds the music to some extent—there are no structural musical determinates separating a specific body of contemporary music as spiritual. It is a common occurrence that the same ṭhumrī-form sung to excessively amorous words is often utilised for a devotional song without any change of musical form. Gāndharva, however, was marked as ‘spiritual’ not only in its lyric but even in its musical structures down to every single detail. It was devoutly and diligently maintained as a distinct form. Being thus unique, it is worthy of close study.

( xlii )

Secondly, a study of gāndharva has also a great historical importance.

It can help in interpreting the history of musical forms in India and in understanding

and placing such forms in a kind of relative chronology. Gāndharva, historically

speaking, occupied a crucial position. It arose from sāma and in turn gave rise

to those later forms which have come down to the present through transfor-

mations and transitions. To know gāndharva will thus help us to understand later

musical forms in a historical prespective and will also provide a key to their links

with tradition as well as a cue to the nature of change.

Gāndharva thus stands as a doorway between the now extinct sāma and

subsequent musical forms. To know and distinguish gāndharva can provide many

handy clues to a historian of music. We have pursued a few of these but only to a

limited extent. We have not fully entered into historical implications as that would

have been beyond the scope of our work.

Having analysed gāndharva as a distinct body of music, we have then

presented the text of Dattilam with a translation and exposition. Since Dattilam

is a work on gāndharva per se, the latter part of our work is in effect a study of the

forms of gāndharva : the musical scheme to which they were related and the rules and

injunctions which governed them. Dattilam is a methodological work and our study

here follows it in all its contours.

In a short chapter, following the above, we have collected together some

material relating to Dattila, including legends about him. We have also collected

at this place some material covering Nārada, Kohala and Viśākhila, the ācāryas

to whom Dattila makes references.

In interpreting Dattilam, we have often taken recourse to other texts, without

which Dattilam is too recondite to be understood. We have compared Dattila's

injuctions with those of others, especially Bharata and have discussed their agreement

and also their divergence.

In our entire study, we have relied greatly upon Abhinava Gupta's

masterly exposition of the theory and doctrine of music and his analysis of details

pertaining to musical forms. The Abhinava Bhāratī has been a well studied text

in so far as it concerns poetics, the problem of rasa and literary aesthetic experience.

But its equally masterly study of music has been largely neglected and remains an

unexplored field.

There is much that is obtuse in the Abhinava Bhārati itself. Yet it greatly

repays study and helps in clarifying many problems of ancient music. We hope

our essay will arouse further interest in Abhinava's body of comments on music.

More in-depth study of these will, we are sure, clarify many points where we have

stumbled and will reveal details in areas still remaining unexplored.

A Hindi translation of Dattilam by Kalinda was published from Hathras a

few years ago. The author has added some comments by way of short notes to the

translation. This translation is rather literal, while the subject calls for a technical

treatment t. It does not, therefore, make much musical sense.

Another work on Dattilam has been published from the Netherlands. The work

is entitled Dattilam : A Compendium of Ancient Indian Music. The author is Emmie Te

Nijenhuis. This is certainly a close and erudite study of the text. (It became known to us

when we had already worked out our major ideas). The reader will find that on some

essential matters our conclusions differ. Nijenhuis believes that the Dattilam is an

abridgement of a more extended earlier work, and the original as well as the abridged

work were modelled on the Nāṭyaśāstra which belongs to a considerably earlier

period. Nijenhuis also believes that gāndharva is not a specific and distinct corpus of

music. Part of our own main task has been to establish a contrary thesis. This

difference in basic conclusions pervades our treatment of minor points as well.

It would have taken up too much space in an already lengthy work had we spelled

out our stand vis-a-vis Nijenhuis at every step. We have, however, indicated our

differences on a few sample points of detail concerning textual meaning and

interpretation in Part III of our work.

MUKUND LATH

January, 1978

Centre for Jain Studies

University of Rajasthan

Jaipur.

Acknowledgements

Abbreviations

Table of Transliterations

Preface

PART I

TEXTUAL CONSIDERATIONS :

AUTHENTICITY OF DATTILAM

Chapter

1. The Originality of Dattilam

1

2. Testimony from Quotations

36

PART II

GĀNDHARVA : ITS NATURE AND EXTENT

3. The Two-fold Meaning of Gāndharva

61

4. A General Introduction to Gāndharva

81

5. Gāndharva and Gāna

91

6. Gāndharva and Bharata

130

7. Some Minor Gāndharva Forms

135

8. The Genesis of Gāndharva

145

9. Gāndharva and Post-Dattila Writers

152

10. Gāndharva and Post-Abhinava Theorists

161

PART III

THE TEXT WITH EXPOSITION

Introductory Remarks

191

Preamble

192

Section 1 : On Svara

195

Section 2 : On Tāla

311

PART IV

11. Dattila : What We Know of Him

431

Appendix A : The Date of Dattila

440

Appendix B : The Dual Meaning of the Word ‘Nāṭaka’

445

Appendix C : The Use of Drum in Gāndharva

451

A Select Bibliography

453

Index

459

PART I

TEXTUAL CONSIDERATIONS:

AUTHENCITY OF DATTILAM

CHAPTER I

ORIGINALITY OF DATTILAM

Dattila has always been held as a great ācārya among students, scholars and musicians versed in the traditions of saṅgīta literature. He has been hailed as a compeer of Bharata and other ancient sages by numerous authorities on music over the centuries. Later writers have used the epithet ‘muni’ with his name as with Bharata and Nārada. Like other ancient Masters of music, he has passed into legend as a sage.

Dattila’s work, the Dattilam was, however, lost to modern scholars until K. Sambasiva Sastri chanced upon a copy and published it in the Trivandrum Sanskrit Series in 1930. Unfortunately, only a single manuscript of the work was available and this, too, was “found almost neglected and left out to the care of insects” as Sambasiva Sastri puts it (Datt. Preface p. 3). No other manuscript of the work has since been found and the text stands on the basis of a single manuscript alone. The readings, fortunately, are coherent and only a few words here and there have been lost due to the tattered condition of the manuscript.

Dattilam is a very short treatise consisting of only 244 verses or kārikās and is written in a very terse and often obtuse language. It is natural that around this important ancient text—especially because there is such a dearth of manuscript material—many controversies have sprung up and its authenticity is questioned.

The original editor, K. Sambasiva Sastri, believed the work to be by Dattila. Examining the evidence then available to him, he found that Mataṅga in his Brhaddeśī and Kṣīrasvāmin in his commentary, Amarakośodghāṭana on the Amara-kośa, have quoted kārikās from Dattilam. These are found to contain the same order of succession as they have in the extant Dattilam (samudhṛtāṃ dattilakārikānāṃ paurvāparyābhyāvamajahatināṃ tathaiva sanniveśadarśanam). Deducing the condition of the whole from a significant portion thereof, according to the maxim of sthālipulāka-nyāya, Sastri concluded that the work was authentic.1

Subsequent scholars have, however, questioned this conclusion. The shortness of the treatise has led many to believe that it is incomplete or at best a later abridgement of a part of a much larger ancient treatise. The original treatise, some posit, must have dealt comprehensively with all the multiple aspects of ancient

1 अत्र केचिदिमं ग्रन्थम् "अङरोद दतिलः; शास्व्रं गीतं दत्तिलसंज्ञितम् ।" इति वचनात् स्वादन्यकृतो दत्तिलशास्त्रसंग्रह इति संंशयिनुमलम् । परन्तु स्थालीपुलाकन्यायेन परीक्षणानामसंशयमिमे क्षीरस्वाम्याचार्येण च समुद्धृतानां दत्तिलकाङिकार्णां पौर्वापर्यभावमजहतां तद्वेव सन्निवेशादर्शन्तं विचक्षणानामसंशयिमिदं श्रद्धमेव दत्तिलं प्रत्याययात् ।

—Datt. Sanskrit preface, p. 3.

See also English preface, p. 2.

2 A Study of Dattilam

music and probably also theatre. The extant Dattilam patently does not. It deals only with one small part of ancient music.

Dattila has been named as a son of Bharata. This has been understood by some as evidence that he was a teacher in the tradition of the Nāṭyaśāstra. It is concluded, therefore, that he must have modelled his work upon that of Bharata and dealt comprehensively with music, dance and drama. Dr. Raghavan says that "Dattilam published now in the Trivandrum Series is only a very late fragmentary selection or condensation of the early original and big work which must have, like other early works, dealt with dance and dramaturgy. It must have been big. The Trivandrum text of Dattila is very poorly small even as regards music. It has no section on drama and dance. There is no denying the fact that Dattila's work treated of Nātya also".1

Other scholars have arrived at similar views. Swami Prajñananand in his History of Indian Music (in Bengali) comments : "it is certain that he (Dattila) had written a large work dealing comprehensively with dance, music and drama and the present Dattilam is a condensed form of the saṅgīta aspect of the large compendium".2 Dr. Premlata Sharma opines: "Dattilam, a work published in Anantasayana Sanskrit Series No. 102, is very small and briefly deals with a few topics of gīta alone. It appears to be an abridged version of a portion of gīta of the original work which must have dealt with gīta, vādya, nṛtya and nāṭya, but is lost to us."3 Dr. Dharmavati Srivastava also concludes: "the work written exclusively on saṅgīta by Dattilamuni does not appear to be a complete treatise. It seems that the text is an abridged form of a much larger work by him."4

E. Wiesma Te. Nijenhuis in her recent work on Dattilam analyses the matter on the basis of textual data and strengthens a similar conclusion. Nijenhuis says : "the limited scope of this treatise and the conciseness of its style create the impression that it was written as an abridgement of a more extended work. The fact that the author himself announces in verse 44, that he will not discuss vīṇā-playing lest he make the treatise too extensive ('for fear of making this treatise too lengthy') supports this view. We might suppose that Dattila wrote a larger work on music, which also contained a chapter on rāgas, bhāṣās etc. Since he is referred to in the Saṅgītarāja of Rāṇā Kumbha 2, 2, 1, 93 and 2, 2, 3, 2 as an authority on bhāṣās and Saṅgītarāja 2, 1, 2, 37 in connection with the rāgas. The original work of Dattila may also have dealt with dramaturgy. According to K. Sambasiva Sastri, preface to Dattilam page 2f, two verses of Dattila referring to this subject are quoted by Sarvānanda in his commentary Tīkāsarvasva on the Amarakośa.

"A very long quotation on the other hand which appears in Bṛhaddeśī of Mataṅga, 201-230 ab, 243 cd-248 and 250 cd, introduced by the words : 'tathā cāha bharata', 'Bharata said.........' a reading which does not occur in the Bombay and

1 The Journal of the Music Academy, Madras, Vol. III, p.18.

2 Bh. Sang. Iti. (Bengali), Swami Prajnananand, Vol. II, p.360.

3 S. Raj. -ed. Dr. (Km.) Premlata Sharma, Vol. I, p. 663.

4 Prācīna Bhārata men Saṅgīta (Hindi), Dr. Dharmavati Srivastava, p. 222.

Baroda editions, but which may exist in some unedited manuscript of the Nāṭyaśāstra, suggests that Dattila borrowed some material from Bharata’s work. Not only this quotation in connection with the jātis, but also the general structure, that is to say, the way in which the subjects are arranged and the order in which they are discussed, reminds me of Bharata’s Nāṭyaśāstra.

“In my opinion we may assume not only that the Dattilam is closely connected with Bharata’s Nāṭyaśāstra but even suppose that this small treatise and quite possibly also the longer original work, which may have contained chapters on rāgas, instrumental music (vādya), especially viṇā-playing (cf. Dattilam 44), and dramaturgy was modelled on the Nāṭyaśāstra.”1

Ancient texts on music are few. Tradition speaks of many teachers but most authorities exist for us as mere names or at best as few lines of stray quotations in later works. Apart from the Nāṭyaśāstra of which many manuscripts exist—and at least one masterly and comprehensive commentary is known2—the only other work was the Nāradī Śikṣā on sāmagāna till K. Sambasiva Sastri discovered a single manuscript of Dattilam and two incomplete manuscripts of the Bṛhaddeśī of Mataṅga—one extremely worn out and the other very fragmentary.3 Till mere manuscripts are discovered and other texts in the field are available, much will remain as mere conjecture regarding the nature, inter-relation and place in musical history of texts like the Dattilam.

However, we think that there are sufficient grounds to believe that the Dattilam, as extant, is a work complete in itself and is found in its original form ; that it is of a fairly early date and represents a tradition independent of the Nāṭyaśāstra in many respects, though parallel to it. We shall in the following pages attempt to establish these points.

The conciseness of Dattilam alone should not lead us to believe that the work is an abridgement. The great sūtrakāras and kārikākāras of ancient Indian literature bear ample testimony that brevity is, indeed, the soul of wit and only a great master can be truly brief. It is, certainly, unfortunate for us that the Dattilam is so concise, since what we demand from an ancient work on the art and the science of ancient music is elucidation and not brevity, however intellectually commendable the latter might be—so scanty is our knowledge of the subject. Yet, brevity was much esteemed during ancient times and only great ācāryas were deemed as fit practitioners of this lofty art.

Concise, pithy works also served a great practical purpose during the early period both for students and teachers, since through them one could keep in mind

1 Dattilam—A Compendium of Ancient Indian Music by Wiesma Te Nijenhuis, Introduction, pp. 2-3.

2 The Abhinava Bhāratī of Abhinava Gupta (10th Cent. A.D.).

3 Bṛhaddeśī of Mataṅga, ed. K. Sambasiva Sastri, 1928, Preface, p. 2.

4 A Study of Dattilam

the outlines of a subject in a precise yet authoritative and comprehensive form. Regarding details, one could refer to the more extensive works, many of which were then certainly current.

An examination in some detail of the internal evidence of the Dattilam itself and the available external evidence from other works will, we believe, reveal the originality and authenticity of Dattilam.

The Dattilam is a technical or a śāstric treatise. It is a manual dealing logically and scientifically with a single subject. In many ancient works of this nature, where the purpose was an organised theoretical delineation of a specific subject, it was customary to name and list the topics to be expounded and explained in the text. This initial enumeration of topics was technically termed uddeśa. It precisely outlined the subject under consideration. Abhinava Gupta commenting on the twenty-eighth chapter of the Nāṭyasāstra, where a list of topics on music is contained, gives a definition of uddeśa. Uddeśa, he says, is a collection [of topics] from which a knowledge [of the whole] can be had in a concise form—“saṅkṣiprāvagamo yatah iti samavāya uddeśaḥ”1 (A.B. on N.S. 28, 13-15). To take a well known ancient example, the first sūtra of the Nyāyasūtra of Gautama enumerates sixteen padārthas (entities); these covered the entire subject matter of nyāya-philosophy.2 This list was technically known as the padārthoddeśa. Subsequent sūtras of Gautama define, elucidate and discuss these padārthas.

The uddeśa has been defined by Jayanta Bhaṭṭa and other philosophers as a list containing a group of terms or paribhäṣās which outlined the entire subject and the contours of the formal treatment to follow.3 One could know the plan and scope of the entire subject under discussion in a treatise from its uddeśa. It was somewhat parallel to what in modern works appears as a list of contents.

The Dattilam, too, contains meticulously and methodically grouped uddeśas. It has two uddeśas—one pertaining to svara or the melodic aspect of music and sequently another pertaining to tāla or broadly the rhythmic aspect of music. Like a true śāstric treatise, Dattilam follows a precise plan in dealing with its subject-matter. After the customary salutation to the deity (vighnavighātāya viśiṣṭadevatā nama-skriyā), Dattila broaches the object of his essay. He expressly declares his intention to be concise and says that he shall briefly delineate the essentials of the science of gāndharva4.

1 तत् वैणतानां स्वराणां प्रकारानुद्दिशति स्वरा ग्रामावित्यादि । संकषिप्यावगमो यतः इति समवाय उद्देश । -A.B. on N.S. 28, 13-15. also नाट्यस्य नाट्यविधयार्हस्य संग्रहं संकिप्य गृहीतुनेतति तमुद्देशं बद्धयामि । -A.B. on N.S. 6, 9.

2 Nyāyasūtra 1.

3 नामधेयेन पदार्थानिमित्तो हि: Nyāyamañjarī of Jayanta Bhaṭṭa, p. 11 (Kasi Sanskrit Series No. 106). Jayanta, further in the passage, says : तत्रोदे शः: प्राथमकमवच्यं क्रतव्यं: अनुद्दिष्टस्य लक्षणपरिक्षानुपपत्ते: Lakṣaṇa was the exposition of topics listed in uddeśa. In philosophical works a parikṣā or examination of the exposition was also undertaken ; this in a work like Dattilam, which is a technical manual, was unnecessary.

4 गान्धर्वंगास्त्रसंक्षेप: सारतोऽर्थ मयोच्यते । -Datt. 1.

Gāndharva he defines as consisting of four elements : svara, tāla, pada and avadhāna. It is, he says, a group of notes (svara sañghātah) well measured in tāla (tālena sumitah) and set to words (padasthah) ; these three, he adds, should be availed with due intentness (avadhāna).1 Of these four elements he deliberately chooses only svara and tāla for detailed exposition. He gives reasons for this choice. The nature of padā, or words, he says, should be learnt from common usage and a study of grammar etc.; hence, he observes, there was no need for him to expound this in a work on music. Avadhāna, too, according to him needed no exposition for it was commonly understood and denoted a state of mind in which the intellect and other faculties were properly concentrated2. It was, he implied, the inner composure or attentiveness demanded by any serious human activity and was thus not limited to the performance of gāndharva alone. Therefore, not considering it necessary to go into details regarding pada and avadhāna—especially since he wanted to be brief—Dattila announces that he will take up the other two viz. svara and tāla for exposition (dvayamanyadato vācyamiba sañkṣepamicatā. Datt. 5). He deals first with svara (tatra svaragatam pūrvam. Datt. 5).

Throughout his work, Dattila follows a methodical rationale in treating his subject and consequently his choice of svara before tāla, as the aspect of gāndharva to be taken up first, was not arbitrary. Details of svara, he says, are being expounded first because svara is the object to be measured (meyatvat) — “tatra svaragatam pūrvam meyatvādupadiśyate” (Datt 5).

Music is a temporal (kālātmaka) art—it extends over time and is conceived in terms of a temporal relationship within its various elements : just as painting, sculpture and architecture are spatial (deśātmaka) arts which extend over space in various ways. In music the svara aspect is rendered in certain patterns, which are often moulded to the framework of certain set and structured rhythmic measures. In Indian music, these are collectively termed tāla. Tāla sets and maintains the pattern in time over which svara extends. Svara has thus been called by Dattila “that which is to be measured (meya)”. The measure was tāla. Hence in defining gāndharva, Dattila had called it a group of svaras (svarasañghātah) properly measured by tāla (tālena sumitah).3 For this reason, Dattila takes up svara before tāla, for a measure has significance only in terms of the object to be measured.

Music is a science as well as an art. The morphology of music has to be understood in terms of a conceptual fabric with its own paribhāṣās or technical terms which are used as tools of analysis and delineation. This, broadly, is the theory or

1 पदस्थस्वरसंघातस्तालेन सुमितस्तथा । प्रयुक्तस्त्वावधानेन गान्धर्वंमधीगतंते ॥ —Datt. 3.

2 लोकाद् विद्यालुपदा (हीनं नोहू) शब्दास्त्रानुप्रहार्त् । प्रसिद्धमवधानं तु सम्यक्कबुद्धयादियोजनम् ।। —Datt. 4.

3 Abhinava also calls gāndharva.—तालप्रमितस्वरप्रधान —A.B. on N.S. 33, 1.

6

A Study of Dattilam

śāstra of music. It stands in the same relation with musical forms as does grammar

with language. Gāndharva śāstra (as Dattila calls his subject-matter) dealt with

the theory or science of ancient music (of a particular aspect or corpus of it as we

shall see) formulated during a certain period. Its theoretical vocabulary consisted of

well-defined terms and ideas which had become quite well set by the time Dattila and

Bharata, who also deals with gāndharava, wrote their descriptions.

The elements and considerations which governed svara-formulations in

gāndharva ; the nature of musical sounds and how they arise ; the notes and pitches

on which they should be fixed ; the octaves and the different ways of constructing

them (as in the two grāmas) ; the range and gamut of svara-arrangements, patterns

and structures-all were comprehended and denoted through certain given pāribhāsika

terms and these together were listed in the svaroddeśa. Dattila’s svaroddeśa

follows his short preamble and is contained in two verses consisting of a list of 12

topics which together describe the svara-aspect of gāndharva ( . śrutayotha svarā

gramau...eṣa svaragatoddeśah. Datt. 5-7). He proceeds to expound them with

the words : “I shall now authentically relate their nature” (saṅkṣepenātha nirṇayah.

Datt. 7). The exposition is completed in precisely a 100 verses as he later asserts :

"iti ślokas̄aten̄amuktaṁ svaragataṁ sphuṭam" (Datt. 109).

We notice that the exposition of the topics follows the same order in which

they have been listed. The order itself is not haphazard but seems to follow a

preconceived rational plan-one topic logically leading to the other.

The section on svara is followed by that on tāla (atha tālaṁ pravaksyāmi.

Datt. 109). Here, too, after a few words regarding the nature of tāla in general,

we have tāloddeśa where a list of topics constituting the tāla aspect of gāndharva is

given (tatra jñeyāḥ kalāḥ pātāḥ......ityuddeśaḥ padārthānāṁ jñeyastālagato budhaiḥ.

Datt. 110-112). Again the description and explanations follow an ordered pattern

as in the foregoing section and the work ends only after all the listed topics have been

dealt with.

Dattilam thus presents a well-knit whole. The author follows a thoroughly

organised plan, almost architectural in its layout. He completely covers the entire

span of the subject he sets out to deal with and not a single topic is left out.

A comparison in depth between corresponding matter in the Nātyaśāstra

and Dattilam also reveals the independent stance of Dattila.

The Nātyaśāstra is a large compendium dealing with many arts and sciences.

One of these arts is gāndharva, which is accorded a full śāstric treatment. Bharata

treats of gāndharva in the final section of the Nātyaśāstra, traditionally called the

geyādhikāra, which begins at the 28th chapter of the text. In his treatment of

gāndharva Bharata lists almost an identical set of topics as Dattila. However, the

disposition of topics in the initial uddeśa list follows quite a different order.

Moreover, the entire topics of gāndharva, pertaining to svara, tāla (and also pada as

Bharata has a padoddeśa) are grouped at one place (N.S. 28, 13-20).

We do not have a complete description of gāndharva from any other early authority except Bharata and Dattila. Later authors have, however, spoken of and quoted some ācāryas besides these two. Evidence points out that the art of gāndharva conformed to strict rules of theory and practice. Its śāstra, too, followed a well-trodden path conforming to norms. Basic differences between authorities expounding the forms and statutes of gāndharva were absent ; yet differences on points of structural detail, and conceptual nuances of theory seem to have abounded.

A comparison between Dattilam and Nātyaśāstra reveals such differences. Further, authentic testimony regarding the nature of these variant views can be gleaned from later saṅgīta-literature. Among later authorities Abhinava Gupta’s discussion of gāndharva is the most detailed, scholarly and informed. In his commentary on gāndharva in the Nātyaśāstra, Abhinava introduces opinions of many other theorists besides Bharata. Among these, Dattila figures prominently. Abhinava has noted many points of disagreement as to doctrine and formal details between Bharata, Dattila and others on matters of varying importance. So have other authorities like Śārṅgadeva, Rāṇā Kumbha and others. Let us examine these differences, (with the help of Abhinava and the others), and we shall see that in the light of these, Dattila stands out as an independent ācārya on gāndharva with an acknowledged original stature and standpoint of his own.1

In fact, a basic semantic and doctrinal difference is to be found in the very definition of gandharva as given by Bharata, Dattila and by Viśākila–another authority whom Abhinava quotes and often refers to. Abhinava has expounded the definitions of these three ācāryas and discussed their merits (A.B. on N.S. 28, 11-12).

Bharata defines gāndharva twice.

(a) “gāndharvamiti tajñeyaṃ svaratālapadātmakam : gāndharva is known as comprising svara, pada and tāla” (N.S. 8, 8).

(b) The second definition is a slightly rephrased version of the first : “gāndharvaṃ trividhaṃ vidyātsvaratālapadātmakam’” (N.S. 28, 11).

The use of the word ‘trividham’ in the second definition was not considered as very happy by many ancient theorists for it suggests that svara, tāla and pada are the three modes or varieties of gāndharva, whereas, in truth, gāndharva was formed by a fusion or conjunction of the three. Abhinava in a attempt to vindicate Bharata, refers to an ancient tradition regarding this point and explains ‘vidhā’ in ‘trividha’, not as ‘mode’ or ‘manner’ but as vidhāna or ‘regulation’. The word vidhā, he says, here denotes a ‘dependence’ (āśritavācī vidhā śabdaḥ); in other words svara, pada and tāla are the three elements on which gāndharva depends. He then quotes Viśākhilācārya who had unambiguously defined gāndharva as “svarapadatālasamavāye

1. K.C. Pandey in his Indian Aesthetics (p. 538), which forms the first volume of his magnum opus, the duology called Comparative Aesthetics, opines : “Dattila seems to have been a contemporary of Bharata”—mainly on the basis of the numerous references found in the Abhinava Bhāratī from the Dattilam and the great authority accorded to Dattila by Abhinava.

tu gāndharvam : gāndharva is an inseparable concomitance (samavāya) of svara, pada and tāla".1

Dattila's definition of gāndharva, besides implying a togetherness of svara, pada and tāla also suggests an heirarchical relationship between these elements, because he says "padasthasvarasaṅghātastālena sumitaḥ" : i.e. "gāndharva is a group of notes set to words and well-measured by tāla". Svara, here, has definitely been assigned the central position, since it is a group of svaras—set to words and sung to tāla—that forms gandharva. Abhinava has, indeed, remarked that in defining gāndharva, it is not enough to state that gāndharva is a conglomeration of svara, tāla and pada because this implies an equal status for all three, whereas in actuality a hierarchy (pradhānagunupabhäva) exists; svara is the dominant element, tāla second in importance : its function is to give an equipoise and equilibrium (sāmya) to svara; pada follows last in the hierarchy. Bharata unlike Dattila has not explicitly set forth this hierarchy. Though, according to Abhinava, he has implied it in listing the three elements of gāndharva serially as svara, tāla and pada (svaratālapadātmakam); svara being named first because it is foremost in importance, tāla following it and pada coming last. In thus interpreting Bharata, Abhinava, in support of his contention, approvingly cites the line from Dattila which contains the definition of gāndharva.2

A definition in a śāstric work should, ideally, be able to comprehend the object defined in all its possible aspects. Śastrakāras, therefore, are very careful regarding their choice of words in a definition. Of the three available ancient definitions of gāndharva that we have cited, the one by Dattila is, in its wording, the most happy and exact. It explicitly implies both the necessary togetherness or fusion existing between the various aspects of gāndharva as well as their hierarchical inter-relationship. Bharata's definition—insplte of Abhinava's vindication—is obviously ambiguous on both counts while that of Viśākila leaves out the hierarchical aspect.

Dattila's definition of gāndharva includes avadhāna as an element in its formation :

"padasthasvarasaṅghātastālena sumitastathā prayuktaśca vadhānena gāndharvamabhidhīyate" (Datt. 3).

Bharata and Viśākhila, evidently, considered avadhāna as superfluous. Abhinava, after appraising the relative importance of the elements, svara, pada, tāla and their inter-relations, remarks : "avadhāna is not manifest as an element deserving to be included as a facet (of gāndharva) like svara, pada, tāla. It should not, therefore, be

1 गानधर्व विविघं विधादिति । त्रिष्यं: स्वरादिघ्य विधादिति । त्रिष्यं: स्वरादिम्घ विधा विधानं यस्य । आश्रितवाची विधाशब्द इति विरसतां। न त्वत्त विधाशब्द: प्रकारार्थ एवं सति पूर्वकपुष्पगतान्धर्ववहारो भवेत् । न चैवं, तथा च विशाखिलाचार्य: स्वरपदतारसमवायो तु गानधर्वम् ।

—A.B. on N.S. 28, 11-12.

2 इति कर्तव्योपयोगिन हि तत्वत मिथोभाष: प्रधानगुणभावेनैव भवति । तत् पूर्वनिपातादेव प्राधान्य सूच्यति । तेन स्वरा: प्रधानं, तारो नामास्ति तस्मैयनोपकारक:, "तल प्रतिष्ठाकरण" इति तार एवं तदाह । ततोपि दूरं पदं पदशब्देनैव तस्याधारतालाः । तदुक्त दत्तिलाचार्यै—"पदस्य: स्वरासघातस्तालेन सुमित" इति।

—A.B. on N.S. 28, 11-12.

Originality of Dattilam

9

included in that group."1 For, he adds : "avadhāna means concentration (yogarūpam) and is not pertinent in this context."2

Simhabhūpala, the famous commentator on Sangitaratnākara, quotes a definition of gāndharva attributed to Dakṣaprajāpati, alorg with that of Dattila.

Dakṣaprajāpati not only included avadhāna as a part of gāndharva but considered it to be the determining factor. "It is avadhāna," he says, "which forms gāndharva;

svara, pada and others come after it. Without avadhāna the [other] three elements cannot be accomplished."3

Dattila, on this point, appears to have taken a stand which lay in-between the one taken by Bharata and Viśākhila and that of Dakṣaprajāpati.

He considered avadhāna a part of gāndharva, but, evidently, not an essential part, for he states that avadhāna was a commonly understood term; he did not think it important enough to be discussed or expounded.

Coming now to the list of topics or the uddeśa, we notice that the order of arrangement in Dattila is quite different from that noted by Bharata.

This might appear as a superficial matter and a mere whim on the part of the two authorities.

Such was not the case. The order of terms in a uddeśa was expected to follow a rational plan.

Commentators like Abhinava—who was certainly not the only one to have done so—took pains to justify the order presented in the Nātyaśāstra and to defend it against different uddeśakramas, that is, the order of arrangement accepted by other authorities like Dattila.

Let us compare the serial order of topics in the uddeśa relating to svara in Bharata and Dattila.

In Bharata we have :

1. svara

2. grāma

3. mūrchanā

4. tāna

5. sthāna

6. vṛtti

7. śuṣka

8. sādhāraṇa

9. varṇa

10. alankāra

11. dhātu

12. śruti

13. jāti'

(N.S. 28, 13-14)

Dattila, on the other hand, has :

1. śruti

2. svara

3. grāma

4. mūrchanā

5. tāna

6. sthāna

7. vṛtti

8. śuṣka

9. sādhāraṇa

10. jāti

11. varṇa

12. alankāra

(Datt. 5-6)

न हि स्वरतालपदतुल्यतया इदं भागनिविष्टमवधातंवग्धानं भाति, येन समूहमध्ये गण्येत ।—A.B. on N.S. 28, 11-12.

अवधानं योगरूपं तच्चात्र नोपयोगि ।—ibid.

दत्तिलदक्षपजापत्यादयो वग्धानं गान्धर्वाङ्गलेनैकायः । यदाह दत्तिलः—‘पदस्थः स्वरसङ्घातस्तालेन सुमितस्तथा । प्रयुक्तश्चावधाननेन गान्धर्वमसिधीरितम् । इति । दत्तिलपक्षे तिरपि—“अवधानानि गान्धर्वं पश्चात्स्वरपदादयः । अवधानं हि-तिरेकेण त्रिविधं नोपपद्यते ॥”——Sudhā on S.R. 1, 3. 15-16.

The G.O.S. ed. reads ‘yatayāḥ’—clearly a faulty reading, as yati was a topic in tāla and is listed and expounded as such. We accept the reading ‘jātayaḥ’ which is appropriate here and noted as a variant reading.

10 A Study of Dattilam

We notice that there is quite a discrepancy in the order. We do not know of the details of the uddeśa-order noted by other authorities. Abhinava, however, remarks that Viśākhilācārya in his uddeśa places śruti after grāma : “viśākhilācryo grāmānantaraṃ śrutiruddiśati sma” (A.B. on N.S. 28, 22).

The position assigned to sruti vis-a-vis svara in the uddeśa seems to have been a matter of great doctrinal importance in ancient times.

Dattila has assigned śruti the first position, both in his uddeśa and in the exposition that follows. But in Bharata's list śruti is almost the last topic to be named. This discrepancy is not accidental. It is indicative of a major controversy in ancient times regarding the comparative status of svara and śruti. Dattila, expressedly attributes to śruti a dominant position. He believed that svaras arose from the śrutis. The śrutis according to him were twenty-two specific sounds in the octave (dhvaniviśeṣāḥ), which the human ear could discriminate as distinct and separate (śravaṇācchrutisamjñitāḥ. Datt. 9). The śrutis in ancient times were twenty- two graded pitches into which the octave was divided. It was believed that these included all the musically significant microtones which could be differentiated one from the other.1

Although within an octave twenty-two microtones were believed to be cognizable, not all of these were usable in gāndharva. Gāndharva forms were based on svara and not śruti, and its octave had only seven svaras (plus two intermediary or sādhāraṇa svaras which were not accorded in gāndharva the full status of svaras). In Dattila's view, certain specific srutis out of the range of the twenty-two in an octave, had been selected and given the status of svara (Datt. 9-11). This made the svaras dependent upon śrutis, for śrutis came first and svaras were chosen out of them.

The grāma poses a question, for in different grāmas different śrutis became svaras. The gāndharva seven-tone octave had a basic twofold division on the

Fig. 1.

1 Cf. प्राक्तनस्य ध्वनेर्वैलक्षण्यं यावता हीनेनाधिकेन या तौवमदात्मना रूपेण लक्ष्यते सा श्रुतिरिति यावत् । यच्चापि परमाण्वाद्यप्युक्रान्तराविकर्षो वा भवेद ध्वनिविशेषस्वस्थापि नासौ गृह्यत इत्य पायते ।

—A.B. on N.S. 28, 27-28.

basis of the somewhat different śrutis which, in each, attained the state of svaras. The two differing octaves were called the ṣadja-grāma and madhyama-grāma. The above circular maṇdalas with twenty-two śruti-demarcations will illustrate the differing śrutis on which the seven notes were established in the two grāmas :

In each of these grāmas, as the maṇdalas show, only certain specific śrutis, bearing a fixed śruti-interval with each other, became svaras. Thus, although somewhat different śrutis in one or two cases became svaras on the basis of the grāma division, the principle that certain fixed śrutis attained svara hood remained unchanged. (See also topic ‘grāma’, Part III).

Bharata, contrary to Dattila, evidently believed that the śrutis were subservient to svaras. The very first topic he names and expounds is the topic svara. The nature of śruti and the number of śrutis in the svaras of the two grāmas are recounted after svara. His method of demonstrating the śruti-intervals between svaras, clearly indicates that he considered svara as self-revelatory and śrutis as secondary demarcations made within svaras. In outlining his method, he says that we first tune two vīṇās on the same seven svaras of the ṣadja-grāma: the correct pitches of these svaras were evidently to be arrived at by a musician through his ear alone as is commonly done in tuning instruments even today. So self-evident, in Bharata’s view, were the positions of the accepted svaras in the octave that he did not consider it necessary to describe how they were to be arrived at. After identical svaras were tuned on two vīṇās, the position of the śrutis could be arrived at by an elaborate process of slightly lowering one of the two vīṇās through a number of steps. These steps helped in illustrating the number of śrutis contained in each of the svaras of the two grāmas. For, Bharata believed that the śrutis were revealed through the svaras (dvāviṁśati śrutayaḥ svaramaṇḍalasādhitāḥ. N.S. 28, 23). Viśākhila, too, evidently, gave svaras precedence over śrutis, because in his uddeśa śruti has been mentioned by Abhinava as following grāma and does not occur as the first topic. Grāma itself must have followed svara, for on no count can it be considered as preceding svara. Although Dattila had consulted Viśākhila’s work on gāndharva (Datt. 177), yet he had an independent stand on the question of the relative importance of śruti and svara.

Abhinava vindicates the order followed by Bharata. He says (obviously pointing towards authorities like Dattila in whose uddeśa śruti precedes svara) : “Some have in this regard concluded that since svaras are revealed through the succession of śrutis (śrutikramābhivyañgyāḥ), they should therefore be named first. To this criticism the answer is : such would indeed have been the case had the śrutis, which are defined as definite sounds (dhvaninādasaṁjñitāḥ), themselves become svaras at fixed śruti intervals. But this is not so, and, indeed, it is on the basis of svara that the high or low position of the śrutis themselves is recognised.”1

Many commentators on Bharata, too, appear to have criticised him for not having placed śruti before svara. They, like Dattila, believed that śrutis were the basis

1 अन्य केचित् चोद्यं प्रति समाधानदर्शयति—श्रुतिक्रमाभिव्य्॑ज्ञ याः स्वरा इति श्रुतय एवादौ वाच्याः इति चोद्यम्, तदोत्तरं—भवेदेवं, यदि ध्वनिनादसंज्ञिताः श्रुतय एव नियतश्रुयन्तरत्वेन गृह्यमाणा स्वरा (इति)। न हि श्रुतयो ह्य उच्चनीचतया

अपि स्वराश्रया एव प्रतियन्ते !—A.B. on N.S. 28, 21.

of svaras and had voiced this objection : “śrutyapekṣinaḥ svarāḥ, tatkimakāṇḍe tā

uddiṣṭāḥ—svaras depend upon śrutis, why then have they been misplaced in the

uddeśa?” (A.B. on N.S. 28, 16-20).

Abhinava has gone to great lengths in justifying the primacy of svara over

śruti. He has adduced many arguments of a fundamental importance to music in

defence of svara. He reminds us of the fact that musical notes are not arbitrarily

determined intervals but are produced through an inexorable accoustic law of

harmonics. Thus a true note can be recognised and experienced as such by anyone

with a musical sensibility. Putting this in the philosophic language of his age, he

remarks : “svaras are self luminous (svayaṃ......rājante) and are cognised as such

in all musical forms. So inexorably effective are they by their very nature that they

irresistably stamp their impress on the mind of the listener who experiences their

aesthetic import spontaneously (hṛdyatādivasaṭ svatāmākṣipantạḥ, svaviṣaye abhidhā-

naṃ kurvataḥ svarā ityuktāḥ) ; on listening to a rightly sounded svara, the psyche

of a listener is forcibly shaken away from his normal everyday consciousness and

ascends to a new aesthetically moving state of being (cittavṛttimadhyasthatārūpasvā-

sthyāvastāparityājanenopatāpayantạḥ).”1 This is not true of the śrutis because these

depend in their very position upon svaras : “śrutayo uccanicatayā api svarāśrayā eva

pratīyante” (A.B. on N.S. 28, 21).

Yet, though Abhinava considered śruti as subsidiary to svara, he did feel that

the position of śruti so far down the Nāṭyaśāstra list (as no. 12) was not defensible.

He, therefore, concludes that there is no point in endeavouring to justify Bharata’s

entire uddeśa-order : “tannātra kramasamarthanena prayāsitavyaḥ”.,2 Viśākhaḷa had

followed a more rational and defensible plan in placing śrutis immediately after

grāma, since the form of each grāma depended upon the distinct śruti-arrangements

of the seven svaras constituting it.3

On examining the detailed exposition of topics that follows the uddeśa in the

Nāṭyaśāstra, we notice discrepancies between their serial position in the uddeśa and in

1 तेन च्वद्स्वभावां चित्तवृत्तिमध्यस्थतारूपस्स्वास्थ्यवस्थापरित्याजनेनोपतापयन््तो हृदयतावशात् स्वतामाक्षिपन््तः स्वविषये

अभिधानं कुर्वंन्तः स्वरा इत्युक्ताः | वर्णसांयादिपुं निह्नितिमाह स्वयं स्वेच्छेव जातिरागभामेदयुं राजन् इति स्वराः |

—A.B. on N.S. 28, 21

2 We give here the whole relevant passage which has many points of interest :

श्रुत्यपेक्षिणः स्वराः तत्रिमकाऽऽण्ठे तु उदीष्ट इति चोद्यशिल्वं यत्नयुक्तं शारीररागादिष्येव तु द्वारंस्य ते स्वरा इति

तद्संगतप्रायम् | एलावता तु द्वारंस्य नासति श्रुयत्पयोग इत्युक्तं तु | तत्र ताभिहिताद्भिः ग्रामविभागाय ता इति

चेत्तत्स्वित्र एवोदेक्ष्य इत्यनुमतं , तन्नात्र कमसमर्थनेन प्रयस्सितव्यं |

—A.B. on N.S. 28, 16-20.

This may be rendered as : “Some raise the question—‘svaras depend upon śrutis, why then have

the śrutis been misplaced in the uddeśa ?’ They, then, answer this question by arguing that

'cnly in relation to the śārira vinā (i.e. singing) are the svaras dependent upon [śrutis] but not

in the vinā made of wood’. This does not stand to reason, for the implication is that in the

wooden vinā, śrutis do not serve any purpose. Why, then, at all include the śrutis in the

uddeśa ? If it be answered that ‘the śrutis help to differentiate’ the grāmas, then they should

occupy a place adjacent to grāma in the uddeśa. Enough, however, of this--there is no point

in endeavouring further to justify the uddeśa-order.'

3 ग्रामविभागार्थमेव श्रुतिकीर्तनं तद्ग्रत्संवन्धादिविवेचनं कृत्वं तद्गतेपुं रागेपु नैक स्वराणां नियतोद्भवांदिकमन्या| शारीरसाम्यं

भवतीत दर्शन्तुमित्ति श्रुतयः स्वरेम्योऽनतरमुक्ताः , एतदाश्येयेविच् विशाखिलाचार्यो ग्रामानन्तरं श्रुतिसद्विष्यति स्म |

—A.B. on N.S. 28, 21.

the lakṣaṇa or exposition. Bharata does not expound the topics in the sequence in which he lists them. Though, śruti is named twelfth in the uddeśa list, it occurs as third following grāma in the lakṣaṇa (N.S. 28, 24) ; sādhāraṇa, eighth in the uddeśa, follows sthāna which is the fifth topic to be expounded (N.S. 28, 34) ; jāti which is at the end of the uddeśa list, follows sādhāraṇa (N.S. 28, 37). Besides this jumbling up of the uddeśa-order in expounding the topics we also find that gīti, a topic included in the tāloddeśa (N.S. 28, 19), is expounded in the svara section (N.S. 28, 44) along with svara topics, without any justification being offered.

Dattila on the other hand follows an orderly course in his exposition, keeping strictly to the uddeśa-krama. At points he even justifies the order he accepts, by hinting at the dependence of a topic which follows upon the one which precedes. We have already noted that he has based svaras upon śrutis. He states so expressedly. After describing śruti, which in his exposition is the first topic, he describes svara, stating : "of these (śrutis) some are selected and sung in all the gītis. Among these śrutis those that attain the state of a note (svaratva) are esteemed as the seven svaras, beginning with ṣaḍja."1 Again, having described svara and grāma, Dattila deduces the mūrchanās and their serial numbers as a natural consequence from these (Datt. 21). Similarly, in describing tāna, he shows their dependence on mūrchanās, the preceding topic : "these fore-mentioned mūrchanās", he says, "[when formed] with five notes and with six notes have been declared by the wise as the tānas".2 Describing alaṅkāras3 after varṇas, he notes : "the alaṅkaras should be known as based upon the varṇas".

Other disagreements, too, are observable between Bharata and Dattila. These are more basic, as they pertain to the nature of the exposition itself. Let us examine some of these.

1. Ancient music recognised two grāmas on the basis of two different arrangements of śruti-intervals accorded to the seven svaras of an octave. Each of these grāmas could result in seven mūrchanās which were the seven notes of the octave in a serially ascending order ; each new mūrchanā beginning on a new and successively lower note. These mūrchanās were numbered serially and each had a distinct denomination. The first mūrchanā of the ṣaḍja grāma was : sa ri ga ma pa dha ni, and was called uttaramaṇdrā. The second was : ni sa ri ga ma pa dha, and was called rajanī. The third was : dha ni sa ri ga ma pa, and was called uttarāyatā.

1 इति ध्वनौ (वि)शेषवास्ते श्रवणाच्छॆ तिसंज्ञिता: । तेम्य: कांचिदुपदाय गीयन्ते सर्वगीतिषु ।। —Datt. 9-10.

2 पञ्चस्वरा: षट्स्वराश्च मूर्छना याः प्रकीर्तिताः । (तनाना च ताना) चतुर्यीतिस्तु ता एवान्यैरुदाहृता: ॥ —Datt. 30.

3 वर्णान्वयास्तु विज्ञेया अलङ्कारास्त्वयोदय: । —Datt. 91.

Thus beginning with each of the seven notes there were seven mūrchanās in the ṣadja-grāma.

Similarly, seven mūrchanās were formed in the madhyama-grāma also: these, too, had their specified serial order and denomination. The first mūrchanā in this grāma began on madhyama : ma pa dha ni sa ri ga and was called sauvirī. The second began on ga, a note lower : ga ma pa dha ni sa ri ; this was called hariṇāśvā. Observation will readily reveal that given any mūrchanā of any particular grāma, say, dha ni sa ri ga ma pa, of the ṣadja-grāma, one could easily tell its serial number by locating the position of ‘sa’ or ṣadja in it. The mūrchanā noted here is thus the third mūrchanā of the ṣadja grāma, the uttarāyatā, as ‘sa’ occupies the third place. The same applied to madhyama-grāma. To take an illustration : the mūrchanā, ni sa ri ga ma pa dha of the madhyama-grāma is easily recognisable as the fifth mūrchanā by noting that ‘ma’ or madhyama occupies the fifth serial position in it. In this manner any mūrchanā of either grāma—provided we know the grāma to which it belongs—can be placed serially, simply by locating the serial position of the notes ṣadja and madhyama.

Dattila has delineated this process concisely in a single kārikā. He also draws an important conclusion from the process : “ṣadja and madhyama by their different serially displaced positions in the two grāmas give the serial number of the mūrchanās in them; and it is for this very reason that the two grāmas are called ṣadja-grāma and madhyama-grāma”.1

Bharata gives no such rule-of-the-thumb method for arriving at the serial position of a mūrchanā in the two grāmas, nor does he provide any reason as to why the two grāmas have been given the names they bear. Abhinava, quoting Dattila’s kārikā (translated above), dismisses this process as insignificant and of no practical use. “Dattilācārya”, he remarks, “has described how the mūrchanās can be numbered as the first and so on beginning with ṣadja and madhyama, but, this [process] is of almost no value because the counting [of mūrehanās] as the first and so on serves no purpose anywhere and is of mere bookish importance (āgamamātrasāratva) ; moreover, the importance given to serial numbering in this process tends to make one mūrchanā more important than the other merely on the basis of number, implying an interdependence, which, in fact, does not exist.”2

Though Abhinava thus dismisses Dattila’s process as pointless, later authors eagerly adopted it. Sārṅgadeva—slightly rephrasing Dattila’s kārikā—describes the process, almost claiming it as his own.3 Śrīkaṇṭha, in his Rasakaumudī, follows Sārṅgadeva (R. Kau. 1,71). Other writers like Puṇḍarīka Viṭṭhala and Raghunātha-

स्वरी यावतिस्थो स्यातां ग्रामयो: षड्जमध्यमौ । मूर्छना तावतिस्थयेच तदग्रामावत एव तौ ॥

-Datt. 21.

दत्तिलाचार्यस्तु मूर्छनानां पद्जस्वरप्राधान्येन च प्रथमादिव्यपदेशलाभात पद्जजमध्यमयोव्यपदेशहेतुत्वमाह—स्वरो यावतिस्थो

स्यातां ग्रामयो; षड्जमध्यमौ । मूर्छना तावतिस्थयेच तदग्रामावत एव तौ । इति । एतद् फलमुप्रायं प्रथमादिव्यपदेशलाभात

कुतश्चिदन्यनपयोगादागममात्रसारत्वात् । गणनादिलव्बाच प्राधान्यादितरतरेष्वयापत्त: ।

-A.B. on N.S. 28, 24.

यस्यां यावतिस्थो पद्जजमध्यमौ ग्रामयो: कमालू । मूर्छना तावतिस्थयेब सा न:श्रृङ्खल कीर्तिता ॥

-S.R. 1, 4, 18. Niḥśaṅka was a title of honour accorded to Sārṅgadeva.

prasāda also mention the process,1 Rāṇā Kumbha, however, criticises Dattila on this

point, evidently following Abhinava's argument ( S. Raj. 2, 1, 1, 304-307).

2. There was an ancient decree concerning the dropping of notes in the two

grāmas when forming tānas from murchanās and in rendering the melodic structures

of jātis, ṣāḍava (hexatonic) or auḍuvita (pentatonic).

Dattila has said : “pañcama in the madhyama-grāma, dhaivata in the ṣadja-

grāma and madhyama in both the grāmas (sarvatraiva) should be known as indispens-

able [notes]” (Datt. 20).

Bharata expressedly mentions the indispensability of only the madhyama :

“anāśitvātmadhyamasya” (N.S. 28, 33 ; also 28, 65). He is silent about the other notes.

Abhinava in this context quotes Dattila's kārikā, Datt. 20, and remarks : “some

theoreticians here believe that [Bharata] muni was not in favour of the view expressed

by Dattila and similar thinkers ; he considered only madhyama as indispensable”.2

In another passage Abhinava reiterates this statement, noting that Viśākhila, too, like

Dattila, believed in the indispensability of dhaivata and pañcama in the two grāmas

respectively.3

3. A kūṭa-tāna has been defined as a mūrchanā in which the serial order of

notes is disarrayed (Datt. 38). For example : ni dha ma sa ga ri pa ; dha sa ma ni ga

pa ri etc. Such tānas can be both pūrṇa—containing all seven notes—and apūrṇa—

containing fewer than seven notes. Dattila gives the number of all possible pūrṇa

kūṭatānas as 5033.4 He also gives a method for arriving at their number. This was much

the same as the process of naṣṭoddiṣṭa, akin to the mathematical method of computing

permutations, which later authors took such care and pains to expound.5 In Bharata

we find no indication of this process. Abhinava quotes a long passage describing it

but he has taken it from another authority (A.B, on N.S. 28, 34). He gives a reason

for its absence in the Nāṭyaśāstra : “these (kūṭatānas) are not described here because

[Bharata] muni has included them all (sarvamgarbhīkṛtya) in a general statement

where he says that ‘there is a [great] variety in tāna and mūrchanā for the pleasure of

the performer and the listener’.6

1 यस्यां यावतिः पड्जस्तावतिभिर्विनियोगः मुखेना :-

Ragamālā of Puṇḍarīka Viṭṭhala I, 30 ; ms. No. 1985 in Oriental Institute, Baroda. The

same line with many scribal errors is quoted (at1, 43) in a work ascribed to Raghunāthapra-

sāda. The colophon calls the work Bhāratisvarasamvāda ; ms. no. B66431/D10669, Tanjcre

Library.

We consulted copies of the above mss. at the library of I.I.C.M.S.D., West Berlin.

2 अन्ये दत्तिलादिमतं मुनेर्नेष्टं मध्यममित्याचक्षते ।

-A.B. on N.S. 28, 34.

3 अन्ये तु धैवतपञ्चमयोगप्रविभागेन यदनाशितं विशाखिलाचार्यामतिदूषितभ्रुवंत नानुमन्यंते मुनेर्मध्यममेवाविनाशि मत

इत्याहुः ।

-A.B. on N.S. 28, 65-65.

4 पूर्णी: पंचसहस्राणि त्रयस्त्रिंशद्वाच्चयसंयया ।

• Datt. 39.

Note that the total number of possible permutations is 5040. Dattila, however, does not include

the seven mūrchanās, as they are not kūṭatānas but have the seven notes in a serial order.

5 See S.R. 1, 4 and S. Raj. 2, 1, 1. etc.

6 ते चेह न दर्शिता इति सर्व गभीर्यः मुनिः साधारणं वचनमाह “प्रयोक्‍तॄन् श्रोतॄनुगाथं च मुखंनाताननाताल्वमिति ।

-A.B. on N.S. 28, 34.

16 A Study of Dattilam

This deduction on the part of Abhinava is, however, quite far-fetched. To assume that merely by using the term ‘variety’ (nānātva) to qualify mūrchanās and tānas, Bharata implied the entire process of reckoning the number of kūṭatānas, does not stand to reason.

There seems to have been a definite textual tradition of describing the process of reckoning kūṭatānas since quite early times. Abhinava, as we have noted, quotes verses from an unnamed ancient authority. The Vrtti on Brhaddeśī, too, describes the process in detail, basing its account upon Datt. 39-40. It quotes these Dattila kārikās and expounds them in detail (Vrtti to Br. 117).

Later authors like Nānyabhūpāla in Bharatabhāṣya(B.B. 4,68-70), Śārṅgadeva in his Saṅgītaratnākara,1 Rāṇā Kumbha in his Saṅgītarāja (S. Raj, 2, 1, 1, 495), Śrikaṇṭha in his Rasakaumudī (R. Kau. 1,74-80), Śubhaṅkara in his Saṅgītadāmodara (ch. 3) and numerous other authors have all deliberated upon the process to a greater or lesser extent. This evidence is very suggestive. It implies that if not Dattila himself, then the tradition of theorists to which he belonged initiated this process of enumeration, and in doing this they included in their works a theoretical aspect of gāndharva missing from Bharata’s account.

Bharata’s uddeśa-list contains thirteen topics whereas Dattila has only twelve. The topic dhātu is absent from the Dattilam. Dhātu was a term relating to the technique of playing instruments, especially the vīṇā. Bharata describes it as : “dhātavo jñeyā vāditrakaraṇāśrayāḥ” (N.S. 29, 50). Vāditra denoted both fingers and external devices such as plectrums used for making strokes on the vīṇā (vāditramaṅgulikon-ādi. A B. on above). Karaṇa (literally, the act of making, doing, producing, effecting etc.), stood for the technique of making various strokes on the vīṇā or the methods of playing it (karaṇam prayatnaḥ. A.B. ibid). Bharata, evidently, considered dhātu an important topic and has described it at length (N.S. 29, 50-70).

It is curious, then, to find that Dattila makes no mention at all of dhātu. In fact, in regard to the entire matter connected with the details of vīṇā and other instrumental playing Dattila refers us to other ācāryas who had expounded the subject in detail. In Datt. 44 he says : “In apprehension of making the treatise lengthy, I shall refrain from expounding the characteristics of vīṇā and other instrumental [playing] in [various] vrttis as laid down by ācāryas.”

What could be the reason for Dattila’s exclusion of dhātu ? On no other point, in spite of his desire for brevity, does he refer us totally to the works of others. It seems probable that he considered vīṇā and instrumental techniques as of subsidiary importance in describing the scheme of gāndharva. His intention, avowedly, was to deal only with the essentials : “sāratoyam mayocyate”, he says in the very first verse. Dhātu embraced a matter concerning instrumental techniques alone, and matters of technique in his view do not seem to have been of much substance regarding essentials, that is, the formal and conceptual structure of gāndharva. Dattila engaged himself in delineating the principles, codes, patterns, mores and other fundamental matters underlying and governing the forms constituting gāndharva. He left

1 S.R. 1, 4, 32-36 ; note also the commentary of Siṁhabhūpāla who quotes Datt. 39.

out dhātu as irrelevant. Being a logical and painstaking theorist, he entirely omitted

the term dhātu from his uddeśa-list as he did not intend to take the topic up for

exposition.

Bharata, on the contrary, lays a great and basic stress upon the vīṇā. Not

only does he describe its techniques at length, he also arranges the topics in his

svaroddeśa in relation to the vīṇā. We find in his uddeśa a double arrangement of

the svara topics. He says :

dvyadhiṣṭhanāḥ svarā vaiṇāḥ śarīrāśca prakīrtitāḥ

eteṣāṃ sampravakṣyāmi vidhānaṃ lakṣaṇānvitam (N.S. 28, 12).

"Svara is described as having a twofold basis : relating to the vīṇā and to the

voice (śārīrāḥ). These I shall expound, along with the precepts [governing them]."

This statement is followed by two verses listing the thirteen topics contained

in his svaroddeśa. Bharata concludes the list with the remark : "this group [of topics]

pertaining to svara is considered as connected with the wooden vīṇā".1 A few of

these very topics are then listed again with the announcement, "here is the saṅgraha

(i.e.: uddeśa) relating to the śārīra vīṇā",2 that is, to the svaras as produced through

the human body. The śārīra list is much shorter and contains only seven of the

thirteen topics. These in succession are :

1. svara 5. sthāna

2. grāma 6. jāti

3. alaṅkāra 7. sādhāraṇa

4. varṇa

Abhinava comments that "vīṇā-playing is of primary importance and therefore

topics pertaining to it have been listed first : vaiṇānāṃ prādhanyātpūrvamuddeśaḥ"

(A.B. on N.S. 28, 11-12). He also gives reasons for the exclusion of certain topics

from the śārīra list. His remarks are interesting. "Tāna", he says, "though

renderable through the śārīra vīṇā, should not be thus rendered as the effect is not

happy. When for sake of practising, tāna is rendered through the voice, then, too,

this can be accomplished only with the help of the vīṇā, which is capable of

producing exactly the required notes-neither more nor less than necessary. Thus,

one who cannot produce the notes through his voice, due to being unable to

comprehend a particular note in his mind, can easily render these notes on the vīṇā".3

1 स्वर ग्रामो मूर्छनाश्च तानः; स्थानानि वृत्तयः। श्रुयोन यतयश्चैव नित्यं स्वरगतात्सकः।। दारव्या समवायस्तु वीणायां समुदाहृतः।।

—N.S. 28. 13-14.

2 स्वर ग्रामावलङ्कारो वर्णाः स्थानानि जातयः।। साधारणे च शारीरेण वीणायामेव सङ्ग्रहः।।

—ibid. 28, 15.

(regarding śārīra vīṇā as denoting the human voice see Part III).

3 तानं तु झारीयं सम्भवेदपि न प्रयोगाह्नमुखाबहलवात। अभ्यासार्थं च तत्सयोगेनैव निपुणैः स्वरव्यजनवशेनैव तत्सिद्धेः।।

बत एव शारीरस्वरप्रयोगसक्तोऽपि दृढप्रयोगरहितो हतस्वरविभोजः। श्वसनोति वैणस्वरान् प्रयोक्तुं।।

—A. B. on N. S. 28, 13-15.

Giving a reason for omitting śruti from the śārīra list, he says : “śruti can be demonstrated only on the vīṇā, because śrutis are revealed through sāraṇā (a method of tuning) which in turn depends upon loosening and tightening of strings : śrutayaśca vīnāyamevopayoginyah sāraṇāyāstantryutkarṣaṇāpakarṣaṇasya tanmū-latvācca.” (A.B. on N.S. 28, 13-15).

Regarding vrtti, dhātu and suṣka--also omitted from the sārirra-list--Abhinava remarks : “the vrttis form an helpful base for the dhātus (which, we have seen, was a term specific to the vīṇā) and the suṣkas ; and these latter two can be pleasantly rendered only on the vīṇā : vrttayaśca dhātūnām śuṣkāsya vā nugrāhak-ādhārah dhātuśuṣkaprayogaśca vīnāyāmeva sukhāvah.” (A.B. on N.S. 28, 13-15).

Dattila makes no division of the topics pertaining to svara into the two compartments of vīṇā and song. He gives only a single list of topics pertaining to svara without entering into an allocation on the basis of instruments or the voice.

Brevity certainly must have been one reason for disregarding a two-fold list. Another, a more potent reason, seems to have been that he was basically interested in the delineation of musical forms and not in techniques of playing. These techniques had only a practical importance in rendering musical forms.

Abhinava in giving reasons why Bharata allocates only a few topics to the voice in contrast to the vīṇā, says : “the vīṇā is appropriate for both practising and rendering all the various modes of svara, for it has a naturally pleasing tone : vīṇādhi-nam samastasya svaragatasya prakārasya abhyāse prayoge copayogitvam tatra saṅkrāntasya cārutvāt.” (A.B. on N.S. 28, 13-15).

These remarks show that Bharata’s two-fold distribution of the svaroddeśa was based upon a practical rather than a theoretical consideration.

We do not know what relative importance Dattila placed upon the vīṇā in comparison to the voice when it came to practical demonstrations.

His treatment of gāndharva is too brief for us to indulge in any valid speculation on this point.

His description of the location of śrutis in the human body and the vīṇā does not, apparently, make any distinction between the two as means for producing the śrutis, since he describes the śrutis as distributed in the same way in both the human and vīṇā frames (Datt. 8-9).

But later in the text we find him referring to experts1 and their instrumental tuning (sāraṇā) in connection with the determination of mūr-chanās, which suggests that the mūrchanās, according to him, could be better demonstrated on the vīṇā (Datt. 29).

But paucity of evidence bars us from any fruitful conjecture.

Yet it is apparent that Dattila as a theorist did not accord vīṇā the great importance in his scheme which Bharata evidently did.

1. इत्येतामुच्यते: प्रोक्ता: सारणायाश्चैव वैनिकै: ।

The text here has ‘vaiśika’, literally, ‘expert in the manners of courtesans (veśa)’, hence a sophisticated person, an expert. A more happy reading could be ‘vainika’–‘vīṇā-player,’ a term more befitting the context.

Originality of Dattilam

19

Let us now compare the exposition of alaṅkāras in the Nāṭyaśāstra and the Dattilam. Wo notice that their number is different in the two texts. Dattila recounts thirteen alaṅkāras (alaṅkārāstrayodaśa. Datt. 99), whereas the number is much greater in Bharata (N.S. 28, 19-26).

This is suggestive. An increase in alaṅkāras usually points towards greater elaboration and complexity and is often considered a proof of a later stage in the history of an art-form. This is testified by the accumulating number of alaṅkāras during the later development of Sanskrit poetry and poetics. True, this fact alone is not a strong enough basis for concluding the historical priority of Dattilam over the Nāṭyaśāstra. But it does show that Dattila was voicing a very early tradition.

More interesting is the fact that Bharata makes a double classification of alaṅkāras. Having described alaṅkāras in gāndharva,1 he warns that “naite sarve dhruvāviṣṭa” (N.S. 29, 26).

Dhruvās were songs sung during the stage-acting of ancient plays and they formed an essential part of the plot. The poetic content of a dhruvā song was its mainstay and a clear enunciation of the words sung was of the greatest importance. Incorporation of melodic figures was secondary and subservient to the primary consideration of communicating the meaning of words. Certain alaṅkāras could so split the letters of a word that its meaning could by entirely lost to the audience. Such alaṅkāras were to be avoided when singing dhruvās.2

Bharata has enjoined : “too great a splitting of letters is not desirable in dhruvās, for then they cannot be properly rendered”.3 In singing certain alaṅkāras, a syllable within a word was dragged for so long that when the next syllable was pronounced, the link between the two was lost to the listner.4 This Bharata calls : śrotṛvarṇaprakarṣaṇa (N.S. 29, 26). This was undesirable in dhruvā sinces it interfered with the

1 The description of gāndharva-alaṅkāras in the text closes with the words : सप्तरूपगता ज‌‍ेऽप्य अलङ्कारा बहुध‌‍ैरिसे । N.S. 29 26.

Saptarūpa was another name for the seven gītakas of gāndharva described in the tāla-section. Here the word is representative of gāndharva as opposed to dhruvā : remark Nānyadeva's comments on the above line from N.S. :

त‌‍ल्ल‌ सप्तरूपशब्द‌‍ः सकलगान्धर्वजात्युपलक्षणस्थ‌‍थया हि सप्तरूपगता जो‌‍े‌वेदिति नियम‌‍ान्देव ध्रुवास्वबल‌‍द्धारनिपेध‌‍ः सिद्ध एव करोति, तन्विप्रसङ्ग च न नियोज्यो ग‌ ध्रुवास्वि‌वति ह्रस्वदीर्घप्रकल्पनादिव्यतिरेकदेवापलक्षणं कारणमिति ।

—B.B. 5, 17.

2 Cf. त‌‍ल्ल‌ जास्वस‌्मेदि त‌ त्व प्राहुःयम‌ । वर्णा जालङ्कारबलिन‌ प्रकल्प्यम‌नाधृतप्रतिपत्तिविकर्षित‌‍े‌ष्वतीत‌प्वरे‌षवनुसन्धानम‌-

सम्भावि ।

—A.B. on N.S. 29, 26.

3 न हि वर्णप्रकर्षस्तु ध्रुवाणां सिदरिष्यते ।

—N.S. 29, 27.

4 Abhinava has indicated this in his comments quoted in fn 2. Ancient poeticians and philosophers have made interesting reflections on the role of time-lapse in pronouncing different words (padas) forming a sentence. Without āsatti or a properly close interval of utterance, two padas in a sentence could not become related into a single unitary whole in the mind of a hearer : thus if we pronounce 'Rāma' right now and 'is coming' a few hours later, the sentence will not make sense. The same principle of āsatti clearly extends to individual syllables forming single padas within a sentence.. For āsatti see Sāhityadarpana of Visvanātha, 2, 1, and other standard works.

20 A Study of Dattilam

goal of dramatic display : the arousal of rasa (rasaniṣpatti) in the spectator, mainly

through the words of the play-wright. Abhinava remarks that the aesthetic process

consisting in the presentation of vibhāva etc. leading to sādharanīkaraṇa and the

resulting relish of rasa, could only be initiated if meanings could be communicated.1

Hence the omission of alaṅkāras unfit for this purpose.

Dattila was dealing with gāndharva alone. Its application to drama and the

resulting adaptations to be made were outside his field. He, therefore, makes no

classification of the alaṅkāras on the basis of their suitability to dramatic songs.

Speaking of gīti—an exposition of which follows alaṅkāra—Bharata again

relates the topic to dhruvā. Gītis, he enjoins, were not to be formed within dhruvā

songs; they were to be restricted to gāndharva. His injunction is : “The gītis should

be known as inapplicable to dhruvā and are to be employed only in gāndharva by the

experts who render songs.”2 Dattila here makes no mention of dhruvā. In fact, he

does not speak of it in any connection.

Related to this point is another significant omission in Dattilam. Bharata, after

describing the jātis (which were elaborate svara-structures in gāndharva somewhat

parallel to the modern rāgas), speaks of their relation to rasas (moods).3

Let us take some examples :

The two jātis named ṣadjodiçyavatī and ṣadjamadhyā are enjoined in the

rasas, hāsya and śṛṅgāra (laughter and love); ṣāḍjī and ārṣabhī are promulgated

for the rasas, vīra (heroic), raudra (fierce), and adbhuta (wonderous); naiṣadī and

ṣadjakaiśikī are laid down for the karuṇa rasa (the sentiment of compassion). Similar

injunctions are given for other jātis on the basis of the notes that predominate in them.

The guiding rule was: musical forms where ṣadja or ṛṣabha were profusely used were

to be employed in karuṇa; those with an abundance of madhyama and pañcama were

suitable for śṛṅgāra and hāsya; and forms with an exuberance of dhaivata were to be

used in vībhatsa (repulsive) and bhayānaka (terrible) (N.S. 29, 1-12).

Dattila makes no mention of any ulterior application or utility of the jātis or

their suitability for certain rasas. His concern is only with delineating the formal

elements of jātis. Even if, for the sake of argument, we were to posit that in

writing his manual of gāndharva he had the Nāṭyaśāstra before him as the model

from which he made his own brief digest, yet it must be admitted that he has done so

with a remarkable sense of selectivity. We have remarked upon the organic and

orderly grasp he displays of his subject-matter. Another proof, in his work, of a great

unity of conception and excecution is the preclusion of elements extraneous to the

scope of the topics he was describing. His exposition covers all the topics that he

1 अयंवद्रो(?)पि तु सामाजिकानां तत्प्रतीतिसिद्धये प्रकृतिमेदो विभावादिस्यस्ततः: साधारणीभूतस्तयावगतो

rasasamaptyupāyogi प्रभवतीति यावत् ।

—A.B. on N.S. 29, 29.

2 एतास्तु गीतयो जेया ध्रुवायों विनिव्हे हि । गान्धर्व एव योज्यास्तु नियं गानप्रयोकॄभिः ॥

—N.S. 29, 49.

3 एवमेतो वधंनेनेयो दशलक्षणा: । यथा यरिमन् रसे याक्च गदतो मे निबोधत ॥

—N.S. 28, 141.

lists as forming gāndharva and each one of them is satisfactorily defined and described; what he leaves out are matters not strictly pertaining to gāndharva.

It was not a desire for brevity alone which restrained Dattila from dwelling upon the rasa aspect of jātis. The reason was more fundamental and concerns the very nature of the two texts, the Dattilam and the Nātyaśāstra and their approach towards gāndharva. Indications of this can be discovered in the latter text in the light of Abhinava's interpretation.

Bharata, in concluding his description of the rasas compatible to the various jātis, states : “thus the wise know the jāti as associated with nāṭya (drama).” (N.S. 29, 13). Abhinava's comments are worth dwelling upon. He raises the basic question as to why at all the rasa aspect of jāti was elaborated upon by Bharata. He asks: “after suggesting that jāti have no ulterior use, why does [Bharata] lay down their connection with rasas, as this aspect is not at all included in the uddeśa?” Abhinava then suggests an explanation : “it is to answer this (above) criticism that [Bharata] says : “evametā budhairjñeyā jāatayo nāṭyaśamśrayāḥ” (thus the wise know the jāti as associated with nāṭya): the purport is that these jāti are only thus (evam) applicable in the nāṭya, not on their own but by way of their utility in dhruvā singing”.1

Dhruvās, we have observed, formed an inherent part of ancient theatre which incorporated many operatic elements in its presentation. Bharata significantly says of his theatre that “gīta vāadya, and nāṭya, which depend upon differet factors, should be so interwoven as to make a unitary whole”. An expressive phrase he uses is that, gīta, vāadya and nāṭya should be like the alāta-cakra : a burning log which when skilfully rotated in a circle made a single unbroken band of fire.

Abhinava's comments are revealing. He first points out the difference between gīta, vāadya and nāṭya. He argues that these arts depend upon quite different factors in their formation (yasmāadvividhāśrayam) : that is, gīta depends upon the human voice, vāadya upon musical intruments and nāṭya upon gestures, mime and such like. Their rendering, too, consequently, requires quite distinct skills. So different, he says, are these arts from each other that their perception requires different human sense-organs (bhinnendriyagrāahyavidhakriyārūpam) : gīta and vāadya are perceived through the ear while nāṭya is perceived mainly through the eye. To combine gīta, vāadya and nāṭya in such a way that a spectator may experience the whole as a single presentation requires great artistic skill and finesse (tasmādyatnenāsyāikatā tatsampādyā yenāikabuddhiviṣayatā sāāmājikasya gacchet). Just as the alāta (the burning torch), which consists of a single flame, appears like a complete circle of unbroken fire through the skill and effort of an acrobat, similarly a theatrical presentation should be exhibited as a single unitary whole.2

1 जातीनां नानोपयोगं कवयतां रसादियोगे जातीनामनु दृष्ट एव किमिति कवित इत्याशङ्क्य शमयति । “एवमेता बुद्धीर्ं ज्ञेया जातयो नाट्यसंश्रयाः” । इति । इत्थमेता नाट्योपयोगिन्यो ध्रुवागानोपयोगद्वयोदितर्थः । —A.B. on N.S. 29, 13.

2 यसमादधिविधाश्रयम् भिन्नेन्द्रियग्राह्यविधक्रियारूपं, तस्माद्यतनेनास्यैवकता तत्सम्पाद्या, येनैकबुद्धिविषयता सामाजिकस्य गच्छेत् । अलाततेजःकणो हि न कश्चिद् युगपदनेकदेशसम्बन्धी, लाघवयतेन तु तयातथा साम्यमापादितम् । एवं प्रयोक्ता अपि …………………………… —A.B. on N.S. 28, 7.

Bharata, for this reason, treats of music with a constant eye towards the theatric whole of which music constituted only a part. Being an independent art, music needed a separate and independent delineation; yet Bharata constantly keeps relating music to theatre, wherever such a relation is due.

The Nātyaśāstra is a work on theatre with sections not only on music but on many other ancient arts and sciences. Bharata deals with architecture, dance and mime as well as ritual and prosody at quite some length. But whatever he treats of is directly or indirectly related to contemporary theatre and a proper presentation of plays. Thus architecture finds a place in his work in describing auditoriums and how to construct them; dance he describes in detail for it played a very basic role in ancient dramas. Similarly, the other arts he describes have a bearing upon theatre, his fundamental theme.

In describing gāndharva, and its jātis too, he speaks of their affinity with the different rasas (sentiments) portrayed on the stage. True, such a description entailed a deviation from a strict adherence to the topics constituting gāndharva and their proper scope, yet it concerned something important from the viewpoint of theatre; therefore it was included.

Dattila had no such ulterior purpose in mind. He treats of gāndharva as gāndharva alone and scrupulously keeps within the confines of his subject.

The exposition of svara-topics in Dattilam is followed by that of tāla, the second aspect of gāndharva that Dattila expounds. Before describing tāla in detail, Dattila in a short sentence defines the relevance and significance of tāla in gāndharva. “Tāla” he says, “results in sāmya (‘equipoise’, ‘equanimity’, ‘balance’) and only through sāmya can gāndharva be truly rendered: “tālāt sāmyaṃ bhavet sāmyādiḥa siddhiḥ” (Datt. 110). We have remarked that Dattila considered svara in gāndharva as the element which needed measure (meya). Tāla had the function of according measure to svara (Datt. 3). Patterns created by svara extend over time and tāla in Dattila’s period (as it still does in Indian music) constituted the time-measures which rhythmically demarcated, organised and distributed these patterns. Tāla-measures accompanied svara-patterns and gave them a balance and an equipoise of properly distributed and arranged structural formations. By giving them a strict measure, they restrained svara-structures from being loosely formed and being distorted out of proper shape and in this manner imparted sāmya to gāndharva. (for further discussion on sāṃya see ch. 5).

Having defined tāla and its function in gāndharva, Dattila sets forth the tāloddeśa ; the list of paribhāṣās or terms relating to and comprehending the entire span of tāla. Let us compare his tāloddeśa with that of Bharata.

Bharata has a list of 21 topics : “ekaviṃśatvidhaṃ jñeyaṃ tālagataṃ budhaiḥ” (N.S. 28, 20). The topics are successively named as :

1. dhruva 5. prāveśana

2. āvāpa 6. śāmyā

3. niṣkrāma 7. tāla

4. vikṣepa 8. sannipāta

Originality of Dattilam 23

9. parivarta 15. yati

10. vastu 16. laya

11. mātrā 17. gīti

12. prakaraṇa 18. avayava

13. aṅga 19. mārga

14. vidārī 20. pādabhāga1 (N.S. 28, 18, 20)

15. pāṇi

The list in Dattilam is :

1. kalā 8. aṅga

2. pāta 9. laya

3. pādabhāga 10. pāṇi

4. mātrā 11. yati

5. parivarta 12. prakaraṇa

6. vastu 13. avayava

7. vidārī 14. gīti2 (Datt. 110-112)

8. mārga

We notice quite a discrepancy between the two lists. In the first place, the difference in the number of topics is quite conspicuous.

The first eight terms in Bharata’s list : dhruva, āvāpa, niṣkrāma, vikṣepa, praveśana, samyā, tāla and sannipāta were names of various beats used to form tāla. They are absent from Dattila’s list, which names kalā and pāta as the first two terms. These two, in turn, are missing from Bharata’s list. However, kalā in Dattila is a generic term and it denotes all the beats which Bharata has listed singly.

Yet there is a difference : the number of beats themselves in Dattila’s account are only seven; he does not name dhruva as one of the beats. The others are the same as in Bharata. The exclusion of dhruva by Dattila indicates a major disagreement between him and Bharata. Bharata names dhruva as the very first term in his tāloddeśa (N.S. 28, 18). It occurs foremost in his list, because—as Abhinava comments—it was, according to him, the fundamental beat : “tatra dhruvatātsyā mūlabhūtaṁ saṃjñādhyāyoddeśe prātham yāddarśitam” (A.B. on N.S. 31, 38).

In Dattilam we do not find any mention of the dhruva beat. Vemabhūpāla, author of the Saṅgītacintāmaṇi, has noted the absence of dhruva in Dattila. He states : “some like Dattila do not accept the dhruva beat. According to them kalā (time) in tāla is demarcated through only seven beats.”3

1 N.S. 28, 19 reads ‘pādamārgah’. Dattilam always reads the word as pādmārgah—Datt. 110, 131, 132 etc. Bharata, too, gives the word as pādabhāga in the exposition of the topic at N.S. 31, 52. Later authors also read the word as padabhaga.

2 There are lacunae in the text here. However, the missing topics can be inferred from the expositional section and a comparison with other texts.

3 अत्र केषुचिद्दत्तिलादिषु (दय:) ध्रुवापदं न मन्वते । तन्मते तु कला सत्ता ताभिस्तालमितिर्मिता ॥

—S.C. ch. 1 in the section on vādya.

The copy of the ms. in the IICMSD consulted by us is incomplete. No indication is given about the source of the original ms. The copy is no. 189 of the IICMSD ms. collection. The work is copied in two parts and begins with :

अथ वैममहोपालो लक्ष्यलक्षणकौविद:। लक्षण तत्प्रवादस्य प्रतिपादयति स्कुटम् ॥

it ends abruptly with :

अत्र कतिचिदुक्त्वन् च मत्तलीतरत्नं तथा । प्रस्तुततारांनुसारेण नाट्ये चाप्युप…॥

These beats—seven according to Dattila, eight according to Bharata—were divided in two classes: nihśabda and saśabda. Āvāpa, niṣkrāma, vikṣepa and praveśa were the unsounded (nihśabda) beats (N.S. 31, 30); śamyā, tāla and sannipāta—plus dhruva according to Bharata—were the sounded (saśabda) beats proper and presumably consisted of various strokes, perhaps on instruments of the cymbal type.1 They have been called pāta by Dattila; the term is second in his uddeśa. Though the function of all beats—whether sounded or unsounded—was to measure time in tāla and so they were collectively called kalā, yet sounded beats were quite distinct from the unsounded; therefore the term pāta is used by Dattila to classify them separately (Datt. 120, 121). Bharata, too, uses the term pāta to denote the sounded beats (N.S. 31, 36) but he does not include the term in his uddeśa.

Bharata, unlike Dattila, does not use the term kalā to denote beats in general. But he does use the term in expounding tāla in gāndharva and it stands in his work for a very important concept.

Kalā in Bharata denotes a specific unit of time which formed the basis for measuring rhythm in tāla (N.S. 31, 2-4). Dattila also uses kalā in this additional sense (Datt. 114). By kalā, Dattila denotes both the beats which demarcated the time-measures in tāla and the measure or unit of time which was thus demarcated. The beats and the time-units that they measured, were, according to Dattila's conception, two sides of the same coin which he calls kalā. Bharata, on the other hand, uses kalā to denote units of time only and not the beats which formed them. Abhinava's comments reflect the notion, clearly implicit in Dattila, according to which kalā was defined both as the time to be measured and the specific beats which measured time.

At one point Abhinava describes kalā as the unit of time in tāla : “time demarcated through beats like āvāpa etc. is called kalā : kālaḥ paricchhidyate āvapādikriyayā, sā kalā” (A.B. on N.S. 31, 1). Concomittently, —as he remarks at another place—the beats which define and demarcate time-units may also be called kalā. (and Abhinava here significantly echoes Dattila's classification of beats as seven) : “the word kalā denotes the seven beats in tāla, niṣkrāma etc. : kalā śabdena ca saptavidhā tālakalā niṣkrāmādirucyate” (A.B. on N.S. 5, 5-7).

The ancient beat system had many elements quite alien to modern pract ce. Notions parallel in Hindustani musical usage to the ancient sounded and unsounded beats are the khālī and bharī. The modern khālī, however, is not a totally unsounded beat. In percussion-playing the difference between khālī and bharī is a difference only of accent and stress—the khālī having a weaker sound in comparison to bharī : but neither khālī nor bharī is really quite unsounded (nihśabda). Only when tāla is indicated with a hand or a palm movement alone in order to keep time in the absence of percussion-playing or to demarcate the structure of specific tāla-cycles, is the khālī commonly formed as totally unsounded. But the hand-gesture for the khālī too, in

1 Bharata mentions ghaṇa, metallic instruments, and connects them with the beats and time-units in gāndharva-tāla in the very first verse of his chapter on tāla :

वाच्यं तु यदूद्यतं प्रोक्तं कालापातलयानुवर्तम् । कालस्तस्य प्रमाणं हि द्विजोऽयं' तालयोगतः ॥

—(N.S. 31, 1)

truth, is just a convenient gesture signifying a lack of stress rather than total soundlessness and has no mandatory or prescribed hand-movement.

The sounded beats or pātas of the ancient gāndharva can, in principle, be understood from available descriptions : one was formed with the left hand (this was tāla), the other with the right hand (śamyā), the third with both hands simultaneously (sannipāta) ; and dhruva apparently was an ever-accompanying rythmic snapping of the fingers before striking any beat.

Apart from the sounded beats, four quite distinct unsounded beats also existed. This fact is difficult to comprehend, since silence can be shown by a single indication or, indeed, through an absence of any indication at all.

What, then, was the purpose of the four silent beats ? Scrutinising the descriptions, we curiously mark that these unsounded beats were formed with very specific mudrā-like movements of the palm and fingers : āvāpa was formed by putting the palm face upwards and drawing in the fingers ; niṣkrāma was formed with the palm facing downwards and stretching out the fingers; and when, along with the āvāpa gesture, the hand was also tossed to the right, the movement was called vikṣepa; praveśa was drawing in of the fingers stretched during niṣkrāma. Such elaborate gestures were certainly not for merely indicating or marking time during tāla movements, for which purpose these complex movements are obviously quite pointless. The reason for forming them was that beats in gāndharva had, besides the musical, a definite ritualistic significance which aimed at transcendental merit or adṛṣṭa.

Abhinava points out: “although to indicate time, whether fast or slow, any method can, in truth, be adopted, but a ritualistic rule is followed because the aim is to attain transcendental merit : tatra ciraśīghratādau kriyā yadyapi yā kācidapyupāyastathā niyamādṛṣṭaphalasiddhaye viśiṣṭahastasthāṅgulakriyivaipopayogini” (A.B. on N.S. 31, 372).

This makes the absence of the dhruvapāta from the Dattilam worthy of notice: since its absence denotes a divergence of not a merely technical nature but concerns a fundamental matter of ritual and spiritual merit. One cogent explanation could be that Dattila belonged to a school or tradition of theorists differing on this matter with the tradition represented in the Nātyaśāstra: such divergences, we know, were not uncommon even between two śākhās regarding the same sacrificial ritual. Another reason could be—as suggested in the Abhinava Bhāratī—that dhruva is, in truth, implicitly included by Dattila though not expressedly expounded. Quoting Datt. 199 —with a reading varying from that in the available text—Abhinava says that Dattila here has the dhruva in mind.

If Abhinava is right, then the inference would be that Dattila took dhruva so much for granted that he did not feel the need for

1 यद्गान्धर्वे कृतचित्त परिच्छेद्ये परिच्छेदविभेदस्य साम्प्रेडपि श्राकुञ्चनास्थान्तमतमा काॅचिदेव क्रिया करिष्मिचदेव अज्ञोऽयादिप्रदेशे वर्त्माना परिच्छेदहेतुचेनात्रैव तद्वैवृत्सिद्धे(:)....

—A. B. on N. S. 31, 1.

2 ध्रुवतव यत्त गच्छति तदध्रुवे(वमि)रादि व बद्धन्ति । दत्तिलाचार्येणापि 'उपोहनत्वादितस्याः (न च तन्न्राचः) स्मृतं ताल(चार्य) द्विज्जितम्' (Datt. 199) इत्युक्तम् ।

—A.B. on N.S. 31, 38.

See also S.R. 5, 141 and Kalā.

either naming or expounding it. But howsoever we might explain his neglect of dhruva, Dattila clearly did not attach the same importance to it as Bharata.

Apart from kalā and pāta which are absent in Bharata’s uddeśa, the other topics are common between him and Dattila; though again we notice a divergence in their serial order. Dattila in his exposition of tāla topics, as in those of svara, follows a logical step-by-step manner of delineation. His lakṣaṇa or exposition maintains the same order as his uddeśa. He, evidently, had arranged the topics in uddeśa after due deliberation and forethought. The exposition in the Nātyaśāstra is disorderly. The topics are not expounded in the order that they have been listed : kalā, not even listed in the uddeśa, is the first term that is discussed, followed by laya which is sixteenth in the list. Laya is followed by mārga, which is nineteenth,3 and is in turn followed by a description of the process of forming pādabhāgas without, however, a prior definition or exposition of pādabhāgas being made (N.S. 31, 3-29). A description of sounded and unsounded beats, enumerated at the very beginning of the uddeśa, comes next. Then follows pādabhāga again (here the pādabhāga is defined). After pādabhāga is the mātrā. After mātrā, vardhamāna, also called vardhamānaka, is described in great detail (N.S. 31, 54-188).

Vardhamānaka was a complex tāla-structure built of many smaller tāla units.2 Songs were sung in accompaniment to it and were, evidently, composed to the jātiś described in the svara section. On occasion, as in the pūrva-ranga, that is, the prologue of the ancient theatre, vardhamānaka was also accompanied by a specific dance (N.S. 4, 271-291). Dattila describes vardhamānaka after prakaraṇa and, evidently, as an appendix to it. Prakaraṇa was a name given to the seven gītakas of gāndharva which were also collectively called the saptarūpa. These gītakas, too, were complex tāla-structures. The simpler tāla-components which the gītakas were composed of, are described by Dattila before he describes gītakas themselves. The vardhamānaka was formed on principles similar to the gītakas, and with like components. It seems to have been composed largely of patterns taken from of the gītakas, but arranged in a different and perhaps an even more complex formation. Abhinava notes that “Dattila and others have expounded it (the vardhamānaka) after having first desciibed the gītakas: since between the two exists a difference that exists between the parts and the whole : dattilditobāhistu gītakalakṣaṇāntaraṃ tallakṣaṇamuktamavayavayāvayavibhe-datvāditi.” (A.B. on N.S. 31, 55). Indeed, clarity in describing a complex structure lies in describing simpler structures first.

What, then, is the reason for the abrupt and out-of-context description of the vardhamānaka in the Nātyāśāstra, without logically leading up to it as does Dattila ? Apart from the fact that Bharata does not really appear to have taken pains to impart an organic build-up to his subject-matter in delineating gāndharva, Abhinava suggests another reason. He recalls Bharata’s statement in chapter five of the Nātyaś-

1 This is according to the reading followed by Abhinava. According to another reading mārga followed kalā as the second topic in the exposition.

—N.S. 31, 3-4.

2 For details see ‘vardhamānaka’, Part III.

āstra to the effect that vardhamānaka was a necessary ingredient of the pūrvaranga in ancient theatre and that without vardhamānaka the ritualistic tāṇḍava dance in the pūrvaranga remained incomplete, and argues that the relevance of vardhamānaka to the theatric context is the reason why Bharata describes it before its true place.1 Bharata was writing a manual mainly for theatre and, therefore, vardhamānaka drew his attention more urgently than other facets of tāla in gāndharva. Contrary to Bharata, Dattila was describing vardhamānaka from the point of view of gāndharva alone. He does not even hint at any theatric use that vardhamānaka could be put to. He accorded vardhamānaka its proper place in his exposition : describing it after the components on which it was based.

Vardhamānaka in Bharata's exposition is followed by aṅga (N.S. 31, 190); after which prakaraṇa, that is, the gītakas are described (N.S. 31, 200). In Bharata's uddeśa, prakaraṇa precedes the aṅga : “prakaraṇāṅgāni vidārī yatayo. . .” (N.S. 28, 19). But this order is not kept in the exposition where aṅga precedes prakaraṇa. Justifying the precedence of aṅga in the exposition, Abhinava argues : “although in the uddeśa-list aṅga, vidārī etc. are named after prakaraṇa as they are of use in it; still, in the lakṣaṇa-section, propriety really demands that the whole (aṅgi) i.e. the prakaraṇa should be delineated after the parts (aṅgāḥ), i.e. components like, vidarī etc., have been described. Therefore its description (meaning, that of the component aṅga) is given before [that of prakaraṇa]—yadyapi coddeśavidhau ‘vidārī yatayo layah’ iti prakaraṇopayogāt tadantaramaṅgānāṁ vidāryādeścoddeśaḥ tathāpi laksanāvasare’ṅganirūpaṇaṃ pūrvamaṅginīrūpaṇam vaktavyamiti proktaṃ tannirūpaṇam.” (A.B. on N.S. 31, 190, 191).

In classifying aṅgas, Nāṭyaśāstra again differs with the Dattilam. Bharata recounts three kinds of aṅgas—vivadha, ekaka and vṛtta (N.S. 31, 190). Dattila—whose statement is quoted by Abhinava in this context2—recounts four : ekaka, avagādha, pravṛtta and vividha. Dattila does not name vṛtta as an aṅga ; instead he has pravṛtta and avagādha, which in the Nāṭyaśāstra are named as two sub-classes of the vṛtta.3

Prakaraṇa, the topic under which name gītakas were described, was the central topic in tāla. The other topics were helpful in formulating the prakaraṇa and were

1 ननु नृत्योपयोगे प्राधान्याद् यदृते कौन लक्षण तद् वर्धमानस्यैव प्रथमलक्षणं वाच्यम् । —A.B. on N.S. 31, 69. also : तत् ‘ताण्डवे (व) यत्न यत्स्यात्’ (भ.नाट्य 4, 123) इति वर्धमानप्रयोगस्य प्राधान्यात् तदपयोगिनाम्/rittalakṣaṇaṃ pūrvamird vakṣyamīti —ibid. 31, 55.

2 तथा च दत्तिलाचार्येणोक्तम्—वक्ष्यमाणमुख्याद् द्विको यं तच्चतुर्विधम् (Datt. 143) इति —A.B. on N.S. 31, 190-191. Rāṇā Kumbha notes the differences in the two classifications—those of Dattila and of Bharata—explicity : गीताङ्गानि विधा चान चतुर्था दत्तिलोऽब्रवीत् —S. Raj 2, 4, 1, 38.

3 प्रवृत्तमवगाढ च द्विविधं वृत्तमुख्यले । —N.S. 31, 195.

28 A Study of Dattilam

subservient to it. In the Dattilam, prakaraṇa is the twelfth topic in tāla. The topics preceding prakaraṇa include elements with which the structure of the gītakas was built. Dattila, after describing kalā, pāta, pādabhāga, mātrā, parivarta and other elements necessary for rendering or formulating the prakaraṇa, takes up the seven gītakas one by one. Their serial order in the Dattilam is :

1. madraka 5. oveṇaka

2. aparāntaka 6. rowindaka

3. ullopyaka 7. uttara

4. prakari

(Datt. 160-221)

The order followed in the Nāṭyaśāstra, in naming the gītaka, is different :

1. madraka 5. oveṇaka

2. ullopyaka 6. rowindaka

3. aparāntaka 7. uttara

4. prakari

(N.S. 31, 200-201)

In the Nāṭyaśāstra these gītakas are at first not described separately but taken up all at once. Their differences and correspondences are pointed out on the basis of similar or dissimilar ‘aṅgas’ formed in them (N.S. 31, 200-229 ; for ‘aṅga’ see ch. 9). Then the gītakas are taken up one by one and further details of each gītaka are described singly and separately. They are described in the order in which gītakas are listed but with one deviation : aparāntaka precedes the ullopyaka instead of following it as in the list.

The gītakas (the tāla-structures) in gāndharva appear to have been time-honoured forms. They followed strict codes of formation like their recipro-cal svara-forms, the jāti s. Descriptions of gītakas in different ancient texts—reports of, and quotations from which are available—are all essentially the same. This is, largely and in all fundamental respects, true of the Dattilam and the Nāṭyaśāstra too. But in spite of a correspondence regarding over-all patterns and primary structures, many differences regarding details exist. Abhinava, with the thoroughness of an erudite commentator, records many differences occurring between Bharata and other authorities. In recounting these divergent views, he refers to the authorities differing with Bharata as ‘some’ ‘others’ etc. : ‘anye’, ‘apare’, ‘kecit’ . 1 He also speaks of ‘ācāryas

1 अन्ये तु वस्तुनिबद्धं पृथक् समासु पृथग्भूतामेव पटकलान् तालिकामुखैः

—Ā.B. on N.S. 31, 251.

अन्ये तु एकैकशोभेदवात् त्रिस्वरत्वाद् विविधक्रमयुक्तमित्याहुः

—ibid. 31, 261.

एषा च माला दृश्यते कलमात्रकृतालः पञ्चमात्तालस्यो कार्यति चेचित । यद्यैषं तत्त्वतः स्थितंस्य वैनतिक मन्यते

—ibid. 31, 282.

अत्र केचिदाहुः । कदाचिद् पूर्वोक्तः कदाचित् पञ्चिमसाधः ।

अन्ये तु मापघातस्य पट्कलामु तैरेबान्तरेरम्याः पटकलाः पञ्चपाणिगतमूलतोनादिविभागेन गेयाः ।

अपरे त्वाहुः । सङ्क्षिप्तवृत्तेकलं न पञ्चपाणिना कार्ये:

—ibid. 31, 286.

अन्ये तु पञ्चपाणिप्रकृतं णानुवर्तयन्ति

—ibid. 31, 295.

अन्ये त्वेकमेव शीर्येक तथा कार्य यथा

—ibid. 31, 312. and etc.

other than Bharata' and their views (ācāryāntaramata). A.B. on N.S. 31, 296-298) ;

but of these ācāryas he mentions only Dattila and Viśākhila by name. Viśākhila is

named only once : in the context of the prakarana (A.B. on N.S. 31, 252). But

Dattila is often spoken of and discussed ; quotations from him,—all traceable to the

available Dattilam—abound (for details see chart, ch. 2). Abhinava mentions a

number of points on which the Dattilam differs with the Nāṭyaśāstra. Some of these

we are pointing out in short : the details, along with further instances, will emerge

when the gītakas are described at length (Part III).

(i) In the gītaka, aparāntaka, Dattila (Datt. 172-173) has described a

certain beat-formation in its pādabhāgas (groups of beats and time-units

into which tāla-structures were analysed), that quite differs from the

account in Bharata.1

(ii) In ovenaka, Dattila's description concerning a matter of tāla-formation

again disagrees with Bharata.2

(iii) In rovindaka a similar disparity is noted.3

(iv) In uttara, some commentators on the Nāṭyaśāstra, following Dattila, had

interpreted a certain formation in accordance with his alien description.4

The entire sangīta literature of the medieval period bases itself fundamentally

upon the Nāṭyaśāstra in expounding ancient music. This work had acquired an

almost canonic authority in the field of music as in that of several other arts. But

other ācāryas are also spoken of and their views are often quoted in support of

Bharata. Differing views, too, are noted with due deference.

Nānyadeva, author of the Bharatabhäṣya, flourished somewhat after Abhinava.

His work is a free exposition of the geyādhikāra (section on music) of the Nāṭya-

śāstra. In his discussion of topics constituting gāndharva, he leans greatly upon

the Dattilam, and quotes many passages and lines from it : all to be found in the

present text. Like Abhinava, he too, records a few cases of divergence between

Bharata and Dattila in this context.

1 दत्तिलादयस्तु द्वितीयं पादविभागं आननिवेशयः चतुर्थं ग्रानिविग्रुपमाचक्षते। “आष्टमी विंशिका चैव द्रे शा᳚म्ये पारिकीर्यिंते

द्वादशाष्टादशौ तालवेकविधानस्य च ॥” (Datt. 172-173) इति बहुलः ।

—A.B. on N.S. 31, 251.

2 अन्ये तु पञ्चपाणिप्रग्रहं नानुवर्तयन्ति । तनाद्यपादद्वितीयार्धे तालमुपपात् प्राहुः । आनिबिता आस्ताविणा तानिविष्टं उपपातः ।

ततुकं दत्तिलाचार्यण—उपपातद्वितीयस्तु तालः के᳚शचिदुदाह,त᳚ः (Datt. 208) इति । ‘द्वितीय’ इहान्यपादद्वितीयोऽर्थ-

निविष्ट इत्यर्थः ।

—Ibid, 31, 295.

3 आचार्यैरमतानुसारेण व्याख्या चैते यत्न टीकाकारण कलिप्तं तालोर्ध्वोज्झते चाङ्ग्यासां शा᳚म्यार्घ्यै᳚न्ते न तु᳚र᳚स्स्यां तालोर्ध्वोज्झते

पञ्चम्यामिति तदुपेक्ष्यमेव । तथा हि दत्तिलाचार्यः—“तथा᳚र्द्धे᳚नते च मालानां तालः शा᳚म्ये यथाक्रम᳚म᳚” ।

—A.B. on N S. 31, 296-298.

4 अन्ये तु दनिलाशनुसारिणो व्याचक्षते । शाखाप्रतिष्ठाख्यमात्रस्तः पृष्ठग्रन्थः सन्निवेश्यते ।

— ibid. 31, 312.

Later authors like Śārṅgadeva and Rāṇā Kumbha (author of the encyclopaedic Saṅgitarāja), had also studied and deliberated upon the Dattilam. Śārṅgadeva often refers to the views of Dattila and notes instances where Dattila's descriptions clash with those of Bharata. In his section on the gītakas, Śārṅgadeva speaks of many views and opinions expressed by authorities other than Bharata but recounts only two such authorities by name : Dattila and Viśākhila.1 Of these, Dattila figures more prominently.2 Śārṅgadeva has often closely followed Abhinava, especially in matters relating to ancient music, but his remarks on various points of difference between the Nātyaśāstra and the Dattilam testify that he had, indeed, made an independent study of the Dattilam ; this he, in fact, claims (S.R. 1, 1, 15-20). He has often compared views of Dattila and Bharata on points where Abhinava is silent.

Kumbha wrote much later than Śārṅgadeva, whose work he had studied thoroughly, along with all the commentaries written upon it.3 Kumbha's is the last extant work on music which treats the subject comprehensively in all its multiple facets, both ancient and contemporary. The work is based to a great extent upon a study of many ancient texts in the original, as Kumbha himself points out (S. Raj. 2, 4, 1, 187). Dattila4 and Viśākhila5 are often spoken of and among these Dattila's name stands out more prominently.

We note below pertinent references made by Nānyadeva, Śārṅgadeva and Rāṇā Kumbha to some decrees in the Dattilam concerning matters of tāla where they found this text enjoining forms different from those in the Nātyaśāstra. The present Dattilam testifies to the truth of these observations and shows that the above authors had this very work in mind :

(i) In the gītaka called madraka, Dattila has enjoined a component called the dvaigeyaka (Datt. 168), but the Nātyaśāstra does not mention it. Kumbha has recorded that an injunction regarding the formation of dvaigeyaka occurs in Dattila's account.6

(ii) In aparāntaka, a certain tāla-formation has been described differently by Bharata and Dattila. Nānyadeva, like Abhinava, has noted Dattila's view7 (Datt. 172A-173B); Śārṅgadeva also records this view but does not

1 Viśākhila is mentioned at : S.R. 5, 90 ; 5, 111.

2 Dattila named at : S.R. 5, 74 ; 5,100 ; 5, 119; 5, 141.

3 यः श्रुत्वा भारतं चरितार्थभिलषैर्माध्यस्यते रत्नाकरम् . . .

– S. Raj 1, 1, 39.

4 Dattila is mentioned at : S. Raj 2,4, 1, 29 ; 2, 4, 1, 31 ; 2, 4, 1, 38 ; 2,4, 1, 54 ; 2, 4, 1, 97 ; 2, 4, 1, 124 ; 2, 4, 1, 161.

5 Viśākhila is named at: S. Raj 2, 4, 1, 58 ; 2, 4, 1, 103 ; 2, 4, 1, 188 ; 2, 4, 1, 248.

6 दृशं नेयकादयो विविधवस्तुत्विकल्पनं वरतुनः । प्राङ् नेये गीतविधौ कायौ द्व(द्रु)तिलादिमितादिह ॥

– S. Raj 2, 4, 1, 97.

7 तथा च दत्तिलाचार्यः-वंशटमी विशातिका चैव दृशं शम्ये परिकीर्तिते ।

द्वादशाष्टदशोक्ता (ताली) एकविंशतिस्थयव च ॥ इति

– B.B. (I) ch. VIII section on aparāntaka.

explicitly name its propounder and refers to it as the view of 'others' : 'anye'.1

(iii) The arrangement of aṅgas (a component) in the aparāntaka has not been decreed by Bharata : he is silent on the point. Abhinava, his commentator, too, makes no specifications. Dattila, however, has enjoined : "the aṅgas in aparāntaka should be the vidha and ekaka" (Datt. 179A). Noting Bharata's silence in this matter, Śārṅgadeva says : "Bharata has enjoined no rule regarding aṅga in aparāntaka, but Dattila names ekaka and vidha as the aṅgas to be applied".2 Kumbha has construed Bharata's silence to indicate an absence of any rule : "Bharata", he says, "lays down the application of aṅga in this gītaka according to the choice or desire of the performers ; but Dattila has here spoken of the two aṅgas, vidha and ekaka".3

(iv) In the gītaka, ullopyaka, Bharata describes a certain part as being rendered first in duple time (N.S. 31, 261), but according to Dattila it was first formed in triple time (Datt. 186). Abhinava has not commented upon this disagreement, but Śārṅgadeva and Kumbha have noted it.4

(v) Vidārī was a method in gāndharva of dividing songs and their accompanying tāla into parts or sections. Bharata enumerates the total number of vidārīs that could be formed in a gītaka (N.S. 31, 197-198 and A.B.). Dattila has no rule regarding the number of vidārīs to be formed. This difference was noted by Śārṅgadeva and Kumbha : "Bharata" formed while Dattila mentions no such rule."5 Kumbha, too, notes Dattila's silence, but interprets it to mean that according to him any number of vidārīs could be formed : "etāḥ punarasaṅkhyatā dattiladyairudahṛitāh." (S. Raj. 2, 4, 1, 31).

1 मतान्तप्रामाण्यादृष्ट्यैस्तदमतद्वादपोषणम् ।

—S.R. 5.104.

See also the two commentaries. Abhinava on N.S. 31, 251, notes this view, ascribing it to Dattila.

2 एकै विवृद्धा च गीता द्वे इत्यौद्‌बिल्यवद्‌भवत् । गीता द्वे नियमे कार्ये इत्यननुमतेऽपि भगवान्मुनिः ॥

—S.R. 5, 100.

3 अनाचार्यमते तावद्‌गुणकल्पनमिच्छया । अन्यद्‌रयमभापृष्टो दत्तिलो विविधैकके ॥

—S. Raj 2, 4, 1, 124.

4 Śārṅgadeva say : युग्मोक्तः प्रथमस्तेऽपां, द्वयथ इत्याह दत्तिलः—S.R. 5,118. Note Kalā: युग्मोक्तः प्रथम इति भरतस्य मतान्तरेति शङ्‌ख; । दत्तिलस्तु द्वयस्थः प्रथम इत्याह । तयोः प्रथमः श्रुततुरस्खविनिर्मितः । द्वयस्थोक्तः: प्रथमस्तेऽपां दत्तिलोक्ते: प्रतीयते ॥ Kumbha states : तेषा

(चतुरस्य is the same as युग्मम)

5 संख्यानियममनेतां नावोच्‌ददतिलः । । भगवान्भरतस्त्वस्या संख्यानियमममन्यधात् ।

—S.R. 5, 73-74. See also Kalā and Sudhā.

Before proceeding, we, note further points of disagreement between Dattila and Bharat to strengthen our argument:

1. After delineating the gītakas and the modes or mores of their formations, Bharata incorporates a long passage on a matter which is again extraneous to gāndharva. He expounds the lāsya form of dance and the technique of rendering it (N.S. 31, 330-367). He says: “I have already spoken to you of lāsya; now I shall describe its lakṣaṇa and its application (prayoga) in due order”.1 Gāndharva was explicitly defined as containing three elements : svara, pāda and tāla ; nṛtta or nṛtya did not form a part of it by definition. Yet Bharata describes lāsya as part of his exposition of gāndharva because its musical accompaniment was, evidently, formed with elements of gāndharva. Strictly and logically speaking, lāsya had no place in an exposition of gāndharva, and Dattila, indeed, makes no mention of it. Lāsya is not included within the scope of any of the topics listed under gāndharva or even remotely suggested by them. Bharata’s chief interest, however, lay in theatre of which lāsya formed an integral part.

Abhinava, following Bharata, classifies dramas into two main classes, the uddhata drama and the sukumāra drama: “idam dvvidham hi nāṭyam, uddhatam sukhumāram ca” (A.B. on N.S. 31,332). In uddhata plays, the masculine–virile and robust–sentiments predominated while in sukumāra plays the ruling ethos was one of femmine tenderness and softness. Accordingly, in the pūrvaraṅga of the uddhata drama the dance corresponded with the dominating ethos : it was rendered with the vigorous Śiva-born tāṇḍava. In sukumāra dramas, on the other hand, the typical dance for the pūrvaraṅga was lāsya, a dance rendered with delicate movements, said to be created by Devī, the consort of Śiva.

Quoting an injunction made by Bharata in the fourth chapter, Abhinava comments : “vīra (heroic) and śṛṅgāra (love) are the two principle emotions that govern the actions of dramatic heroes in their pursuit of puruṣārtha : achievement of the desired aim. In dramas where vīra predominates, the pūrvaraṅga should be rendered in the samuddhata (i.e. uddhata) mode, consisting of movements created by Śiva: vīraśṛṅgārayoreva nāyakagatayoraśeṣapuruṣārthasiddhiṣu vyāpārāt tatra prathame samudyata(ddha)taḥ pūrvaraṅgo māheśvarairaṅgārairuddhatairiti” (A. B. on N.S.31, 332). In dramas where śṛṅgāra governed the action, the pūrvaraṅga, as Bharata had earlier indicated, was to be rendered as sukumāra with delicate lāsya motions: “śṛṅgārapradhāne tu nāṭye sukumāra eva pūrvaraṅgo yadi hi prāyeṇa pañcame lāsyam sūcitam” (A.B. on N.S. 31, 332).2

1 लास्यमित्यव यत्पूर्व मया बः परिकोचतितम् । लक्षणं तस्य वक्ष्यामि प्रयोगं च यथाक्रमम् ॥

• N.S. 31, 330.

2 The indication in Bharata to which Abhinava refers, is found towards the end of the fourth chapter of N.S. where Bharata describes two types of dances, the uddhata dance rendered with Māheśvara gestures (i.e., the tāṇḍava) and the lalita (i.e. sukumāra) dance rendered with gestures created by Devī :

यतु शृङ्गारसम्बद्धं गानं स्त्रीपुंसाश्रयम् । देवीकृतै रलितै रेवस्तत्प्रयोगोजयेत् ॥

यतु शृङ्गारसम्बद्धं गानं स्त्रीपुंराश्रयम् । देवीकृतै रलितै रेवस्तत्प्रयोगोजयेत् ॥

• N. S. 4, 311-312. See also ibid 302-303.

The word ‘pañcame’ occuring in the A.B. text, seems a scribal error ; it should be ‘caturthe’.

Tāṇḍava was danced chiefly to ihe accompaniment of vardhamānaka : “vardhamānamathāpīha tāṇḍavaṃ yatra yujyate” (N.S. 5, 13). The dance-movements forming tāṇḍava, as performed during the pūrvaraṅga, have been described by Bharata in the fourth chapter of the Nātyasāstra (N.S. 4, 269-291). The musical formations of the vardhamāna (its tāla and pada structures) are described in the tāla section of gāndharva. The lāsya as danced in the pūrvaraṅga had not been described earlier, either in its dance-form or its accompanying patterns of tāla. Therefore, after having described the gītakas which formed the basis of tāla in lāsya, Bharata proceeds to describe this dance-form in detail.1

2. The exposition of lāsya in the Nātyaśāstra is followed by some general observations on tāla and then come the three topics: laya, yati and pāṇi.2 Laya is here taken up for a second time; it had already been dealt with in the beginning of the tāla-lakṣaṇa (N.S. 31, 5-6). Abhinava gives a reason for taking up the topic again: “laya”, he says, “had been classified into its various modes earlier; here in this latter-end of the tāla-section, it is being defined and expounded”.3 However justifiable the reason for this repetition, it does evince an haphazard arrangement of ideas, happily missing from the Dattilam.

3. After laya and yati, Bharata describes pāṇi. Pāṇi constituted three different ways of syncopation between the svara-forms (in song or on the vīṇa) and the accompanying tāla.

The exposition of pāṇi in Bharata and Dattila is almost identical. But Bharata does not define pāṇi while Dattila gives a characteristically short definition : pāṇi and tāla, he declares, are the same : “tatra tālaṃ [ca] pāṇiṃ ca prāhurekam” (Datt. 153). Abhinava quotes this definition4 (as he also does Dattila’s exposition of pāṇi5) and discusses its implications. He notes the absence of any definition of pāṇi6

1 Abhinava in commenting on N. S. 31, 322, recalls a previous verse relating to tāṇḍava (N. S. 5, 13), and says that tāṇḍava which is connected with uddhata drama, has been described earlier; lāsya the sukumāra form is being described now in the 31st chapter : गीतानामिति इलोके (भ.ना. ५, १३) । “इहं समुद्धते गीतकं वर्धमानं तत्र सुकुमारामिति दर्शयितुं 'मिहे' त्युक्तम् । तेन साधारणं चरितमारचायोः: निर्देश्योऽन्यद्वय च वर्धमानादप्रयोगेण सूत्ररक्षार्थ पठ्यते निरुपितम् । तदुपयोगी चाऽत्रार्थे तत्र तमिरूपेण निर्गिर्‌फः । अथुना सुकुमारोपयोगिन् दर्शयितुं ग्रन्थान्तरमिति ।

—A. B. on N. S. 31, 332.

2 अथभूता हि तालस्य यतिपाणिलया: स्मृता: . . .

—N. S. 31, 368-369. The exposition of these three follows in 31, 370-377.

3 उत्तो विभागो लयस्य लक्षण (न्) ह्वान्न (स्वनन्) ।

—A. B. on N. S. 31, 370-377. The elassification of laya is noted in N. S. 31, 5 तयो लयास्तु वृत्तो या द्रुतमध्याविलम्बिता: ।

A proper definition occurs at N. S. 31, 370-371 : छन्दोऽक्षरपदां हि समस्तव यत् प्रगीतमिततं कलाकालान्तरङ्कुल: स लयो मान (नाम) संज्ञित: ॥

4 तदाह दत्तिल: “तत्र ताल (च) पाणि च प्राहुरेकमिति”। (Datt. 153)

—A. B. on N. S. 31, 372.

5 यथाह दत्तिलाचार्यः: । सम् बोप्ति वा तस्य यद्राघ: सस्वितं भवेत् । वाच्यं पदनि पाणो वा तदेवममिधीयते । (Datt. 153-154)

—ibid. 31, 374.

pāṇi in Bharata and gives a reason for this, which from his remarks, can be inferred to be somewhat as follows:

Tāla, says Abhinava, consists of a well-defined interval of time and is manifested through the formation of various sounded and unsounded beats; its function is to properly divide and demarcate the total measure of a song.1 In gāndharva, we have seen, tāla was rendered through beats like āvāpa etc. Pāṇi (literally, ‘the hand’, with which the beats were formed) in its pāribhāṣic sense denoted the various ways in which the demarcation of time in tāla could be related to the accompanying svara formations. Hence pāṇi was identical with tāla, because both terms, in the last analysis, denoted the actions which manifested tāla.

It is with these considerations in mind that Abhinava explains the absence of a definition of pāṇi in the Nāṭyaśāstra. He argues: pāṇi—as Dattila has observed—is the same as tāla, for it denotes all the specific acts or actions (kriyā) which demarcate and define the tempo to which tāla itself is formulated. Such actions also define the measure in a melody rendered on instruments such as the vīṇā. Bharata has elsewhere described various hand-movements relating to gāndharva tāla and to vīṇā and other instruments. This is why he did not consider it necessary to include either a general or a particular definition of pāṇi in his exposition.2

Kallinātha, the commentator on Śārṅgadeva, too, perhaps with Dattila’s definition of pāṇi in mind, equates pāṇi with tāla : “atra paṇiśabdena tāla lakṣyate tadvyaapareṇa tasyābhivyaktatvāt : the word pāṇi here indicates tāla, because the functions of pāṇi taken together constitute tāla in its entirety” (Kalā on S. R. 5, 50-52).

4. Bharata closes his description of gāndharva with the topic pāṇi and initiates a new section dealing with dhruvā.3 Looking back, we notice that some more topics in tāla (besides those already mentioned), which have been named in the uddeśa, have not been properly defined or described by Bharata. Vastu and parivarta (topics 9 and 10 in Bharata’s uddeśa), were evidently important elements of tāla and Bharata, indeed, makes repeated use of these terms.4 But nowhere does he define them or describe their lakṣaṇa, their nature and function. Dattila, on the contrary, takes these topics up in their due order and expounds them in his neatly terse and concise manner.

1 Cf. तालेन भावना श्म्यादिसङ्केतवादिनः भावनक्रियाविशेषयोनिमन् सति यस्तालः परिञ्छिद्यतात्मकालखण्डे क्रियारूपो द्रव्यात्मा स एवं गीतक्रियाप्रमाणपरिछेदोपायः । —ibid. 31, 11.

2 तदाह दत्तिलः । तन्न तालं (च) पाणि च प्राहुरेकमिति । तदगतया क्रिययातरमपि वीणाप्रभृतिरवादनादि स्वीक्रियते । तेन विशिष्टक्रियापरिच्छेदो यदवच्छिन्नो लयस्ताल इति गान्धर्वीयताललक्षणादेव्हि क्रियाया: प्रदेशान्तरे लक्षणादिव्ह न पाणेः सामान्चलक्षणं कृतं विशेषलक्षणं वा । —A. B. on N. S. 31, 372.

3 The last line of Chapter 31 reads : अथ उद्धं प्रवक्ष्यामि ध्रुवाणां च कलपनम् । —N. S. 31, 378.

4 Vastu : N. S. 31 ; 127, 189. 190, 191, 202, 229, 240, 248, 275, 279, 280 etc. Parivarta : N. S 31 ; 98, 265, 269, 335, 336, 337 etc.

Šārṅgadeva and Vemabhūpāla (author of the Saṅgītacintāmaṇi), not finding an appropriate discussion in Bharata, drew upon the Dattilam in expounding the topic parivarta.1 Other authors like Nānyadeva and Kumbha seem to have been influenced by Dattila in their delineation of vastu.2

5. Another topic Bharata lists but does not expound is avayava (18th in Bharata's uddeśa). Dattila considered avayava an important aspect of gāndharva. A gītaka could be rendered as short or long by forming a permissibly lesser or greater number of the parts (avayavas) ordained in it. These were two distinct ways of rendering gītakas. They have been indicated by Dattila after prakaraṇa (Datt. 223).

6. In addition to the fact that exposition of some topics is missing from the Nāṭyaśāstra, others have been misplaced and described out of the uddeśa context : gīti listed as a topic in tāla, is described in the section on svara (N. S. 44-49) and vidārī, another tāla topic is taken up in the section on dhruvā (N. S. 32, 17-18).

Our survey here is not exhaustive. But we hope that it is ample enough to indicate the significant position of Dattilam as an independent text in saṅgīta tradition. It was known and studied as an authoritative text propounding an independent view, often divergent from the more esteemed Nāṭyaśāstra, on many topics. Other renowned ancient teachers like Viśākila—and to lesser extent Kambala, Durgaśakti etc.—were also studied by later authors, but it was the Dattilam which figures more prominently wherever the subject comes within the scope and range of the topics this concise treatise has delineated.

We have dwelt upon features where Dattila diverges from Bharata. They are tokens of his originality. We shall discuss still others when we study the text in detail. Matter, where the two authorities were in agreement, has also been cited by later writers because the Dattilam was often referred to in support of and as complimentary to Bharata's contention ; this, too, we will dwell upon. For the present, let us study the quotations and incorporations abounding in other texts from Dattilam, for proofs of its authenticity.

1 S.R. 5. 43 ; S.C. ch. 1 in the section on vādya, defines parivarta as : पादभागस्य मान्या: तालस्य सकलस्य वा । आवर्त्तनंतु यद्गीते परिवर्त्तः स उच्यते ॥ Compare this with Datt. 138.

2 S. Raj 2, 4, 1, 19-20 ; B. B. (I) Ch VIII. For details see 'vastu', Part III.

CHAPTER II

TESTIMONY CONCERNING DATTILAM

FROM QUOTATIONS

We have observed that many saṅgīta-writers evince a direct knowledge of Dattila's opinions. As late an authority as Kumbha claims to have read his works. Siṃhabhūpāla appears to speak from direct knowledge when, justifying the contemporary need for a work like Śārṅgadeva's Saṅgītaratnākara, he says : “the ancient saṅgītaśāstras of Bharata, Dattila, Kohala and others are full of mysteries difficult to unravel. Śārṅgadeva has therefore written his Saṅgītaratnākara, incorporating the quintessence of saṅgīta in a single treatise for the convenience of modern readers.”1 Śārṅgadeva, himself names Dattila among prominent ancient writers 'the ocean of whose opinions' he had 'churned' in order to produce his own compendium.2

To take other passing examples, Pārśvadeva, author of the Saṅgīta-Samaya-Sāra, who was perhaps a contemporary of Śārṅgadeva, speaks of Dattila as an authority on tāla : “śrī someśvaradattilaprabhṛtibhiḥstālaśvarūpaṃ purā proktam sarvajagadhitāya...” (S.S.S. 9,2). Pārśvadeva, has, in fact, incorporated a few ślokas from Dattilam into his work in describing certain tāla-topics, as we shall see. Even a very late author, Raghu­nāthabhūpa, writing in the 18th century, claims, in his Saṅgītasudhō, to have 'repeatedly deliberated' upon the writings of Dattila, among other authorities : “mataṅgaśārdūlakadattilādyaih praṇītaśāstrāṇi muhurvicārya.”3 The testimony of saṅgīta literature, from the Bṛhaddesī (c. 7th century A.D.) to the 18th century, shows that Dattila was a much honoured authority.4

1 The full passage reads :

भरतदत्तिलकोहलादिप्रणीतानि संगीतशास्त्राणि भूतत्वर्तिभिरवलोक्य प्रगेनु रबोधरहर्यान्नीतिमत्स्व परमाचार्यनिकः जिनिर-किरणाभरणचरणपङ्क्तिप्रवणनमानसः श्राङ्गदेवः संगीतसंग्रहार्थमेवं संगीतसारं लोकोपकाराय चिकीर्षुर्निर्जात्यमुहरिसमास्वादये विषयप्रकृष्टद्वारा प्रभञ्जनानुमितस्मृतिसमुन्नतिमत्सुतिविहितं विशुद्धदेवतानमस्काररूपं मज्जन्माचारतन्मप-सञ्ज्ञानमित्येयप्रयोजनसङ्गाञ्छकेनोल्लोकैर्न कद्यति ।

—Sudhā on S. R. 1, 1, 1.

2 S. R. 1,1, 15-21, esp :

अत्राग्रहोपसंग्रहैस्तैरपि मतप्रमाणोद्भवम् । निर्मध्यं श्रीशाङ्गदेवः सारोदारमभिव्यधात् ॥

3 Quoted by K. Sambasiva Sastri; Datt. Sanskrit preface p. 2; he also quotes another verse from the Saṅgītasudhā where Dattila is named along with ancient authorities of repute :

शक्तिश्च शास्त्रं लकोहलौ च समीरगणः कायपमास्ती च । रम्भाजु नोनारदस्तुष्टं च पोतनरदिद्रुवासुरगुरुद्विलासच ॥

4 Besides those named above many other works mention Dattila with reverence. To take a few examples :

(i) Sudhākalāśa, author of the Saṅgītopaniṣadsāroddhāra (composed in 1350 A. D.), claims to have studied Dattila :

शीलातमजाकोहलदत्तिलादिमेरुत्पन्नैश्च भोजादिमुखैः प्रणीतान् ।

सद्ग्रन्थसारांशन परिलेख्य समयं वितत्य स्वानुभवं च कीर्तित

—Saṅgītopaniṣadsāroddhāra, 1, 4.

(Continued on next page)

Testimony Concerning Dattilam from Quotations

37

Not only was Dattila studied with due respect, he with also copiously quoted by major writers. His Dattilam is a work of only 243½ verses or 487 lines; of these almost 200 lines occur in other texts : 92 lines occur with an explicit acknowledgment of Dattila as the source ; 10 lines are quoted in contexts which makes Dattila the obvious, though implicit, source, for they occur in works where Dattila is often quoted by name and are placed in proximity to acknowledged quotations. Other 97 lines have been borrowed from the Dattilam without naming Dattila and are incorporated in the body of their texts by later authors. 12 more lines occur as quotations where an authority other than Dattila (namely Bharata) is mentioned as the source.

It is worth noting that when a string of lines from the Dattilam on a single topic is reproduced in any later text, it usually occurs in the same serial order as found in the Dattilam. And, significantly enough, we do not find any material or opinion or quotation ascribed to Dattila, concerning any of the topics described in the Dattilam, which is not to be found in the extant text.

Before deliberating upon the testimony of these quotations—acknowledged or unacknowledged—we are presenting it in the form of a chart for the sake of clarity and easy reference. The chart follows the serial verse-order of the Dattilam: each line can thus be graphically placed as to its occurrence and also as to whether it is acknowledged or unacknowledged or occurs in both the ways, in different texts or at different junctures in the same text. Many lines, we shall see, are quoted by more than one authority. Some lines are unacknowledged as Dattila’s by one or more writers but another text makes a specific mention naming Dattila as the source, thus according a strong basis for their authenticity.

(Coutinued from previous page)

(ii) Govinda in his Sangrahacūḍāmani, bows to Dattila among others in introducing his work (see 1, 6).

(iii) Jagadekamalla, a predecessor of Sārṅgadeva, mentions Dattila in his Sangītacudāmani (1, 59) in connection with yati :

लयानां यतिः सम्पक् कविता दत्तिलादिभिः ।

समा स्तोत्रह्राद्या च गोपुच्छा चोत्तमा तिथ्या ।।

Comparing this verse with Datt. 154B-155A, reveals that Jagadakamalla had indeed consulted Dattilam on this point.

38 A Study of Dattilam

THE CHART

Showing Dattilam in Quotations

Dattila Kārikā Works in which quoted. Ascribed to Dattila Works in which quoted. Not ascribed to Dattila

A : first line

B : second line

1 2 3

3. (a) A.B. on N.S. 31,11. (quotes 3A). Sudhā on S.R. 1,4, 15-16. S. Raj. 1,2,3,48.

4. (a) Kśirasvāmin in his Amarakośodghāṭana on the Amarakośa (nāṭyavarga). (b) B. B. (I) ch. 6. Nāṭyacūdāmaṇi of Somanārya (copy of ms. at I.I.C.M.S.D. No 12998 at the Gaekwad Oriental Institute Library).

11B. B.B. 3, 55 (quotes parts of 11B). It is wrongly ascribed to Nārada ('tathā ca nāradah'). The editor of B.B. has corrected 'nāradah' to 'dattilaḥ'.

12. } (a) A.B. on N.S. 28, 21 (quotes 12A).

13. } (b) Sudhā on S.R. 1, 4, 15-16 quotes all five lines in the same order.

14A. }

16. } (a) B.B. 3, 152 quotes Datt. 16. (b) B.B. (I) ch. 7, also quotes Datt. 16. (c) Sudhā on S.R. 1,3,51 quotes all four lines in the same order.

17. }

Testimony Concerning Dattilam from Quotations 39

1 2 3

18. } (a) B.B. 5, 129-130. (b) B.B. (I) ch. 7 (quotes 18A at two separate places). (c) Sudhā on S.R. 1, 3, 46-56 (quotes 18B).

19. (a) A.B. on N.S. 28, 33. (b) Sudhā on S.R. 1, 4, 6. (c) B.B. 4, 58.

20. (a) A.B. on N.S. 28, 24. (b) S. Raj 2, 1, 1, 304-305. (c) Sudhā on S.R. 1, 4, 18-26 R. Kau. 1, 71 is the same as Datt. 21 with a slight modification. The point dealt in the verse does not occur in the N.S. and is, in fact, criticised as ‘phalgu-prāya’ by Abhinava (A.B. on N. S. 28, 24). It is thus likely that Śrikaṇāha, author of R. Kau, had bodily lifted the verse from Dattilam after slightly rephrasing it.

21. } 23A. } Vāyupurāṇa 86, 18 (a three-line verse) is the same as Datt. 22-23A ; these three lines occur in the same order as in Datt.

22. } 25A. } Vāyupurāṇa 86, 16-17A is the same as 24-25A ; these three lines again occur in the same order.

25B. Sudhā on S.R. 1, 4, 18-26.

26. } Sudhā on S.R. 1, 4, 15 quotes 27. } these six lines in the same order. 28. }

27. (a) Brṛ. Vṛtti on 117. (b) Kalā on S.R. 1, 4, 17. (c) Sudhā on S.R. 1, 4, 29-31.

1 2 3

37. (a) Br̥. Vṛtti on 117.
    

(b) A.B. on N.S. 28, 34.

39. (a) Br̥. Vṛtti on 117.
    

(b) Sudhā on S.R. 1, 4, 32-36 (quotes it within a larger quotation of a passage from Br̥. Vṛtti on 117).

40. Br̥. Vṛtti on 117.
    

41. A.B. on N.S. 29, 70-71.
    

42. (a) B.B. 3, 151.
    

(b) B.B. (I) ch. 7.

48. B.B. (I) ch. 7.
    

52. } Br̥ Vṛtti on 187, quotes these 12 lines, ascribing them to Bharata.

55. } Br̥ 195B-197A, incorporates these four lines within text.

57A. (a) B.B. (I) ch. 6.

(b) B.B. (I) ch. 7.

58A. B.B. (I) ch. 7.

58B. B.B. (I) ch. 7.

1

60A. (a) A.B. on N.S. 28, 73 quotes part of line 60A.

61. (b) B. B. (I) ch. 6 quotes 60A with 61B.

(c) B. B. (I) ch. 7 quotes 60A. Br. Vṛtti on 251 quotes Datt. 61A.

62. (d) B. B. (I) ch. 7 quotes 61B-64. Order of lines as in Datt. Br. 251 same as Datt. 62.

63. (e) B. B. (I) ch. 6. Br. 201-230A is the same as Datt. 63-91A. These 58 lines occur in the same order as in Datt.

64. &

65. (f) ibid (quotes 62 twice).

67. (g) B. B. (I) ch. 7 quotes 62 with 48.

70. (h) B. B. (I) ch. 6 quotes 65 (Dattilam here is quoted as an authoritative text supporting N. S. 28, 98, which is quoted as ‘sūtra’).

78. (i) B. B. (I) ch. 6, quotes 66 as authoritative text supporting N. S. 28, 120-122.

1 2 3

80. (j) B. B. (I) ch. 6 quotes 67,

81. again as parallel authoritative text supporting N. S. 28,

82. 102-103.

86. (k) B. B. (I) ch. 6 quotes 78 as

87. parallel text suporting N. S.

88. 28, 120-122 which is quoted

89. as 'sūtra'.

90. (l) B. B. (I) ch. 6 quotes 80-81

91A. and 83B as parallel text supporting N. S. 28,125B-126.

91B.

93. Br̥. 244-248 same as Datt. 91B-95,
    

96A. Br̥. 250B same as Datt. 96A.

96B. A. B. on N. S. 28,140.

97A. B. B. (I) ch. 7.

97B.

99. S. S. S. 1, 42-46 same as Datt.
    

100. 97B-102A.

102A.

1 2 3

113B.

114A.

118B. S. S. S. 7, 6B-10A same as these

119.    lines.
     

121A.

116B. S. S. S. 7, 14B-15A same as

117A. Datt. 116B-117A.

130B. A. B. on N. S. 31, 40.

134B. A. B. on N. S. 31, 51.

135B. (a) A. B. on N. S. 4, 315 quotes

136. there in the same order.

(b) A. B. on N. S. 31, 40-50

again quotes these three

lines in the same order.

143B. A. B. on N. S. 31, 190-191.

147B. A. B. on N. S. 31, 196-197.

148A.

148B. B. B. (I) ch. 8 quotes Datt. with-

149. out explicitly naming Dattila.

150. Order of these five lines same

as in Datt. Context shows that

Dattila is evidently meant as

the source, though not avowed

as such.

153A. (a) A. B. on N. S. 31, 26 quotes

153A, which is cited again at

A. B. on N. S. 31, 372.

154B. (b) A. B. on N. S. 31, 374 quotes

153A-154B.

1 2 3

159. A. B. on N. S. 31,316 names the source as 'anye' which, evidently, seems here to refer to Dattila.

165A. A. B. on N. S. 31, 240.

172B. 173A. } (a) A. B. on N. S. 31, 251. (b) B. B. (I) ch. 8

174A. B. B. (I) ch. 8 quotes in part.

175A. B. B. (I) ch. 8 quotes in part (in the same passage where Datt. 174A is partially quoted).

197B. B. B. (I) ch. 8 A. B. on N. S. 31, 280 (quotes part of the line).

198A. B. B. (I) ch. 8.

199A. B. B. (I) ch. 8.

199B. (a) B. B. (I) ch. 8. (b) A. B. on N. S. 31, 280.

201B. 202A. } (a) A. B. on N. S. 31, 281-282. (b) B. B. (I) ch. 8 quotes 201A.

203A. A. B. on N. S. 31, 207.

205A. S. R. 5, 150A almost same as Datt. 205A.

207B. A. B. on N. S. 31, 294, quotes the line 207B saying it occurs in 'tantrāntare' (i.e. another text).

208A. (a) A. B. on N. S. 31, 294. (b) B. B. (I) ch. 8.

211B. A. B. on N. S. 31, 296-298.

240B. A. B. on N. S. 31,39.

(For fuller details of the text of these quotations and the musical context to which they pertain, see Part III where the text of Dattilam has been discussed).

Testimony Concerning Dattilam from Quotations 45

In making this chart we have been able to consult only the major works of saṅgīta literature that have been published. Of these the Bharatabhāṣya of Nānyadeva has been published only partly by the University of Music at Khairagarh, and contains only chapters 1-5. Fortunately, we were able to consult the rest of the work in manuscript form. This, as the chart shows, proved fruitful. The Saṅgītarāja often refers to Dattila and quotes him. A part of this lengthy treatise is still unpublished, viz. the vādyabhāga, where some topics of gāndharva tāla have surely been described and this may be found to contain further reference to Dattila. Unfortunately, in this case, we were not able to avail of any copy of the necessary manuscript. Other unpublished texts, too, may contain references to Dattila but they have, so far, remained beyond our reach.

The chart bears ample testimony to the fact that the Dattilam, as we know it, has been known in the same form since at least the days of Mataṅga and his Brhaddesī (ca. 7th cent. A.D.), where two long passages occur with the same verse order as in the Dattilam

Earlier incorporations occur in the Vāyupurāṇa. The Vāpupurāṇa is perhaps the oldest of the extant Purāṇas.1 Portions of it have been dated to the third century A.D. This, however, cannot be asserted of the entire available Purāṇic text, which may, like the other Puranas, have undergone many changes, editions, re-editions and insertions of new material over the centuries. The definite period when its two chapters on music, the 86th and the 87th, were written is unkonwn, but they certainly belong to an early age. The exposition of music in the Vāyupurāṇa comprises of elements borrowed from established works on music. Many passages must have been lifted bodily from existing works of renown. Sources are rot cited since a Purāṇa presents itself as a supreme canon in no need to consult other works. We do not know of the different sources from which the Vāyupurāṇa drew its material on music, but the Dattilam was certainly one, since six lines from it occur in the Vāyupurāṇa, and they occur, moreover, in the same verse-order as in the Dattilam.

The Brhaddesī contains many verses in common with the Dattilam. Two such long passages are introduced with the words: “tathā cāha bharataḥ”. This has led some scholars like E. Wiesma-Te. Nijenhuis to conclude that Dattila borrowed some material from Bharata's work. However, none of these passages can actually be traced to any of the extant recensions of the Nāṭyaśāstra.2 Nijenhuis argues that they 'may exist in some unedited manuscript' of the work.

This argument cannot really be settled till more recensions of the Nāṭyaśāstra are forthcoming; but we can, meanwhile, examine the matter from other angles and considerations. One reason which leads Nijenhuis to believe that the Dattilam is based upon the Nātyaśāstra is 'the general structure that is to say, the way in which

1 Studies in the Puranic Records—R.C. Hazra, p. 13.

2 We should, in this context, keep in mind the observation made by M.R. Kavi who edited the G.O.S. edition of the N.S after consulting numerous manuscripts and variant readings. He was of the opinion that basically different recensions of texts like the N.S. are, by the very nature of their subject-matter—which is scientific rather than legendary or mythical—quite unlikely. See his preface to the second edition of N.S., Vol. 1, p. 11.

the subjects are arranged and the order in which they are discussed'; this she says : "reminds me of Nātyaśāstra". This judgment regarding the Dattilam seems to have been current since M. Krishnamachariar wrote his encyclopaedic History of Classical Sanskrit Literature (first published in 1937, only a few years after the Dattilam was published). Krishnamachariar remarks : "Dattila was a devout follower of Bharata and never differed from him in his expositions, so that, when later writers wanted anything to support Bharata they invariably invoked Dattila."1 Our examination has shown that this is definitly not so, and that beyond an agreement as to the general scope of the subject-matter and its topics, the structure, the layout and the intent behind the plan of the two works, Dattilam and Nātyaśāstra, is quite different. The similarity of treatment that does occur, is explained by the sameness of the subject-matter.

We have noted divergences regarding detailed delineation of specific topics in the two works : a fact which has been recognised at least since the days of Abhinava, and since Abhinava himself was writing within a long tradition of commentators, it is quite likely that his predecessors, too, had commented upon the divergence to be found between the Nātyaśāstra and other ancient treatises dealing with the same topics, among which the Dattilam must surely have figured. Abhinava, in fact (commenting on the gīta named uttara), makes mention of writers who had interpreted the details of gītaka-formation given by Bharata in accordance with the alien views propounded by Dattila and others : "anye tu dattilādyanusāriṇo vyācakṣate, śākhāpratiśākhāt-māntaḥ prthagbhūtaḥ sanniyataḥ." (A.B. on N.S. 31, 312).

It would here repay to examine the Brhaddesī evidence. The Brhaddesī contains many verses from the Dattilam and seemingly ascribes them to Bharata. The extant Brhaddesī is an incomplete work. It ends with the description of prabandhas; with the author's promise that he will next take up a discussion on vādya: "idānīṃ kathayiśyāmi vādyasya nirṇayo yathā." (Br. 511).

The number of verses contained in the Brhaddesī are 511, and their order throughout the extant text continues unbroken inspite of chapter-endings or section-endings. The colophon at the end states : "iti mataṅgamuniviracitabṛhaddeśyāṃ prabandhādhyāyaḥ saṣṭhaḥ." This colophon leads us to deduce that the available portion of the work contains six adhyāyas or chapters. But, actually, besides the sixth adhyāya, noted in the colophon, at which the available texts ends, we find only two other adhyāya-endings. The fifth adhyāya ends after verse 375 with : "iti mataṅgamuniviracitabṛhaddeśyāṃ bhāṣālakṣaṇādhyāya nāma pañcamah samāptaḥ."

Another adhyāya-ending occurs at page 133 of the published text. It is obviously not an adhyāya-ending in the Brhaddesī, but in a long quotation from another work which is incorporated in the Brhaddesī at this juncture. The colophon reads : "sarvāgamasamhitāyāṃ yaṣṭikapramukhabhāṣālakṣaṇādhyāyaḥ caturthaḥ." Clearly it is a chapter-ending within a work called Sarvāgamasamhitā which was a compilation of opinions regarding musical forms like bhāṣās, ascribed to Yāṣṭika and others

1 See p. 822, para 957 of this work.

(yāṣṭikapramukha). A long portion of this Samhitā occurs within the Brhadeśī, and contains no less than 175 verses, which unlike those in the actual Brhadeśī text, are unnumbered.

There are other pauses in the Brhadeśī scheme besides these three, but none of them are adhyāya-endings. The only other major ending is at page 104 and is a lakṣaṇa-ending, or in other words, the ending of a topic or section within a chapter :

"iti matangamuniniviraciatabr̥haddeśyāṁ rāgalakṣaṇaṁ tṛtīyam." After this lakṣaṇa-ending, the delineation of bhāṣās begins with : "atha bhāṣālakṣaṇam", and then after verse 366 we have : "iti bhāṣālakṣaṇam." Still other endings may be observed, but these too are quite minor sectional endings occuring after the elucidation of specific topics or sections :after the verse 15 we have:"iti deśīlakṣaṇam"; and only three verses preceding this, that is after verse 12 occurs, "iti deśyottipattiprakaraṇam". Thus—not counting the ending included in the long interpolation from the Sarvāgamasamhitā — we have eight endings in all. Of these there are only two major adhyāya-endings at the end of the fifth and sixth adhyāyas. The other endings are clearly minor topic-endings or at best section or prakaraṇa endings which occur within adhyāyas.

We do not know how the adhyāyas were really distributed in the original work. Scribes who had no knowledge of the subject and were not careful enough, may be responsible for the present state of affairs. We notice, in fact, many instances of scribal error (lekhaka-pramāda-doṣa) in the Brhadeśī. The learned editor of the text has cited some of these in his footnotes:

(a) before verse 64 some lines describing auḍuva etc., are obviously missing;1

(b) after verse 99 again some lines have clearly been dropped;2

(c) verse 173 is misplaced and should have preceded verse 171;3

(d) another lekhaka-doṣa has been noticed on page 136.4

1 The editor notes :

इति ब्रोङ्बादिनिरूपणपरं; कियाद्विचद् ग्रन्थप्राप्तः; सम्भाव्यते ।

—fn. by Sambasiva Sastri, Br̥. p. 18.

2 The editor notes :

इह पूर्वप्रदर्शितमतमध्यमप्राममुख्यत्नामनुसरणे गान्धारनानिर्देशानन्तरम् ऋषभनामादर्शानन्ताद् आदर्शो तद्वाचकस्य ग्रन्थस्य सम्भाव्यते ।

—fn. on Br̥. p. 24.

3 The editor notes :

अस्याऽनस्यमिदं पचं ‘दक्षिणो अतीतः’ उत्यातः पूर्वं निवेशनीयमिति भाति ।

—ibid. p. 49.

4 The editor puts a string of notations (svaralipi) and a verse within brackets as they are obviously out of context; he remarks :

वलयान्तर्गतं वाक्यजातं लेखकप्रमादत् प्रक्षिप्तमिति भाति ।

—fn. on Br̥. p. 136.

The editor also points to passages obviously dropped by scribes on pp. 137 and 138. Again, in the section on prabandha, he notes many missing verses :

बल प्रकरण उद्धिष्टेषु केषाञ्चिल्लक्षणादर्शन् , लसितेषु च केषाञ्चिदनुद्देशः; उद्देशलक्षणवाक्येषु नामभेदश्च वर्तते ।

—fn. on Br̥. p. 142.

Besides these errors noted by K. Sambasiva Sastri, we note others, too. For example, the Vṛtti following 27A introduces two verses with: “tathā cāha caturah”. No authority known as Catura is known in saṅgīta literature which otherwise abounds in names of ancient authors. The same two verses have, however, been quoted by Kallinātha in his commentary on Saṅgītaratnākara, 5, 3, 10-16. He ascribes them to Tumburu—‘tathā cāha tumburuh’. This ascription is quite plausible. It also appears likely that Kallinātha had borrowed this quotation from the Bṛhaddeśī itself, which work he knew well, and since no other quotation from Tumburu is to be found in Kallinatha on the subject of music.

Besides scribal errors and other irregularities noted above, the Bṛhaddeśī also contains many major lacunae:

(a) Verses 201-251 describe the eighteen jātis of gāndharva. This description is followed by a long prose Vṛtti (Bṛ. pp. 69-74). Verses 252-276 describe the jātis a second time; this time in a different language and style. This second description is interspersed with prose passages of the Vṛtti. Also, the description this time is incomplete, since after eight jātis have been described, a sudden break occurs in the text at verse 276, and a new verse, entirely out of context, expounding the etymology of the word ‘bharata’, is abruptly introduced (verse 277).

(b) The very next verse, 278, is again abrupt and has no continuity or pūrvāparasambandha with the preceding matter. Verse 278 formulates a question regarding the nature of rāgas and the etymology of the word rāga.2 The answer to the question follows, and is introduced with the words ‘Mataṅga said’—‘mataṅga uvāca’. Quite a number of verses have obviously been dropped betweed verse 276 and 277 and again between verse 277 and 278.

(c) Another significant break in the continuity of the Bṛhaddeśī occurs after verse 366A. Heretofore, the work was in the form of a Purāṇa-like dialogue, between Nārada and Mataṅga: Nārada asking questions, Mataṅga answering them (Bṛ. 2, 4, etc.). But after verse 366A, the two characters of the dialogue suddenly change, and without any prior indication, Kaśyapa becomes the questioner and Yāṣṭika, the answerer. A long passage follows (pages 104-141 of the printed text) containing 175 verses which, unlike the verses of the Bṛhaddeśī text proper, are unnumbered. This passage is the interpolation from the Sarvagamasamhitā discussed above. It breaks off as abruptly as it was introduced and the

1 Verse 276 forms part of the description of the jāti nandayanti; it reads : हृद्यान(का) मुच्छना ताल: पूर्ववद् दिगुणा कलाः | विनियोगो श्रुतिस्थाने प्रयत्नप्रयत्नेषु भवेत् ॥ This is followed by the Vṛtti which gives notations of the jāti-structure. Nandayanti should have been followed by further description of other jātis, 10 of which had yet to be described. But verse 277 reads : भं ज्योतिस्तदत्रो हंसस्थमात् भरतं विदु: | तद्भ(वं) भरतज्ञानं तद्मवा भरती शुभा ॥

2 Verse 278 has no discernible relation with the previous verse; it reads : किमुच्यते रागबन्धेन कि वा रागस्य लक्षणम् । व्युत्पत्तिलक्षणं तस्य यथावद् वक्तुमर्हसि ॥

verse numbers of the Brhaddeśí are resumed at page 141 with verse 366B. But in spite of the unbroken numerical order in the verses, a wide gap in the subject-matter remains obvious. This gap had probably occured at an early date in the history of the available manuscript and some scribe-editor obviously filled it up with similar material borrowed from the Sarvāgamasamhitā. The gap-filling, however, was done in an inadept manner without any attempt at smoothing the joints and putting up at least a facade of textual continuity by numbering the borrowed verses in continuance with the original text and by obliterating or changing the colophon of the interpolated passage.

(d) Another break occurs after verse 368A. Mataṅga promises that he will next deal with deśí rāgas: “ataḥ paraṃ pravakṣyāmi deśírā-gakadambakam”. Testimony reveals that this line was originally followed by further explanatory lines regarding the nature of the contents to follow. Siṃhabhūpāla quotes these lines introducing them with the words: “adhunāprasiddhānāṃ rāgāṅgādīnāṃ lakṣaṇamuktaṃ mataṅ-gena”. His quotation reads:

ataḥ param pravakṣyāmi deśírāgakadambakam lakṣyalakṣaṇasamyuktaṃ trividhaṃ cāpi samyutaṃ rāgāṅgaṃ caiva bhāṣāṅgaṃ kriyāṅgaṃ ca tṛt́īyakam pratyekaṃ lakṣaṇam caiṣāṃ pravakṣyāmyanupūrvaśaḥ uktānāṃ grāmarāgāṇāṃ chāyāmātraṃ bhajanti hi gitajñaiḥ kathitāḥ sarve rāgāṅgāstena hetunā bhāṣāchāyāśrītai yena jāyante sadṛśāḥ kila bhāṣāṅgāstena kathyante gayakaiḥ stutitatparaiḥ śokotsākaruṇādidīpakātmakriyāditah jāyante ca yato nāma kriyāṅgāstenahatuā”

(Sudhā on S.R. 2, 2, 9-10; part of passage also quoted by Kalā, giving it as Mataṅga's: Kalā on S.R. 2, 2, 1).

The quotation clearly shows that a large section describing deśí rāgas, bhāṣ-āṅgas etc., existed in the original text. As the text exists, verse 368A is followed abruptly by seven verses describing only six rāgas. Then, after verse 375, follows a colophon which states that the fifth chapter on bhāṣās ends here. The new chapter, the sixth, takes up a new subject, namely, prabandha.

Obviously, the extant material on deśí-rāgas is incomplete and fragmentary. Only six rāgas are expounded where actually there were many more, as is attested to by Sārṅgadeva, Rānā Kumbha and others, who based their own accounts of rāgas and other related forms on authorities like Mataṅga. Siṃhabhūpāla, after quoting the passage from Mataṅga cited above, in fact, comments that rāgas are altogether 268 in number.

There is also not much room for thinking that Siṃhabhūpāla was not quoting from Bṛhaddeśí but from another text which may also have been ascribed

to Matañga, since verses relating to definitions of bhāṣas, grāma-rāgas, gītis, etc. that

he quotes from Matañga, are, indeed, traceable to the Bṛhaddeśī.1

The 511 verses comprising the Bṛhaddeśī are interspersed with passages of an

explanatory Vṛtti. The Vṛtti is mainly in prose but it also contains a few verses,

quoted from renovend authorities. We do not know whether the author of the Vṛtti

was the same as the author of the main body of verses. The two are unlikely to have

been identical. It is true that it was not uncommon for authors of śāstric works to

compose their own Vṛtti, but the Bṛhaddeśī assumes a Puranic tone and sets forth the text

as a dialoge between two ṛsis of authority in a canonic style. A simultaneously

composed and published Vṛtti of the nature of a gloss or explanatory notes would

have been incongruous with this tone. It is, therefore, quite likely that the Vṛtti

was composed only after the work became established in authority.

A clear distinction is observed in the Bṛhaddeśī text between the original

verses and the Vṛtti upon them. Verses quoted or incorporated from other texts in

the Bṛhaddeśī have been numbered but those occurring in the Vṛtti are unnumbered

and stand out as quite separate from the original Bṛhaddeśī verses.

This distinction appears to have escaped Nijenhuis's notice. The long passage

of about sixty lines which the Bṛhaddeśī has in common with the Dattilam begins at

verse 201 (Br.). Preceding this is the Vṛtti on verses 199-200, where Bharata is quoted

with the words 'tathā cāha bharataḥ'. The quoted verse is in fact a statement clearly

traceable to Bharata: it is N.S. 28, 72. This verse from the Nāṭyaśāstra is unnumbered

like other verses quoted in the Vṛtti. The verse is then followed by a short line in

prose : “daśavidhajātilakṣaṇamiti”—indicating that the Vṛttikāra herewith ends his

explanation or notes regarding jātilakṣaṇas. Then the main Bṛhaddeśī text is resumed

and the verses are again numbered (as 201 etc).

It thus cannot be said that the numbered verses following the verse quoted by

the Vṛttikāra from the Nāṭyaśāstra have been ascribed to Bharata by Matañga or even

by the Vṛttikāra, since the Vṛtti ends with a single verse from the Nāṭyaśāstra.2

Verses 201-230 of the Bṛhaddeśī are the same as Dattilam 63-91 : they form

part of the Bṛhaddeśī and are not acknowledged as borrowed from any author.

From the testimony of the Bṛhaddeśī alone these verses cannot unambiguously be

ascribed to Dattila; but Nānyadeva who was evidently very familiar with the

Dattilam—as his frequent quotations from this work testify—clearly ascribes many of

these verses to Dattila, as can be seen from the chart. Nānyadeva's citations also reveal

that Dattila's verses in this context were quite distinct from those of Bharata. He

quotes Dattila's lines in support of the descriptions in the Nāṭyaśāstra (referred to as

sūtra), as forming a parallel and yet a distinct testimony from a reputed authority.

The long passage describing the jātis, thus appears evidently to have been borrowed

from the Dattilam by Matañga.

1 Sudhā on S.R. 2, 1, 2-5 quotes 5 verses from Matañga with the words : “matañgenāpyuktam”;

these are nos. 293-297 in Br. The same passage in Sudhā also quotes Br. 285-286. Again

Sudhā on S.R. 1, 6-7 quotes two verses from Matañga introducing them with: “lakṣaṇasamuccayaścokto matañgena”; these are verses 302-303 of the Br.

2 K. Sambasiva Sastri has indeed enclosed this single verse between inverted commas to distin-

guish it from the verses that follow.

Besides these, a few other stray lines have also been borrowed from Dattila and incorporated in the body of the Brhaddeśī (as distinct from its Vṛtti). The Brhaddeśī has similarly incorporated verses from the Nāṭyaśāstra also, without acknowledging their author : Br. 189-195A corresponds exactly with N.S. 28, 50, 51, 52, 53, 57, and 54. These incorporations have not been alleged to prove anything against the authenticity of the original text.

The author of the Vṛtti on the Brhaddeśī also quotes Dattila, but he does so with due acknowledgment (see chart). A group of six verses from the Dattilam (49-54), however, has been quoted in the Vṛtti (on Br. 197) with the source mentioned as Bharata: “tathā cāha bharataḥ”. The quotation includes a total of eight verses : the first two are, indeed, from the Nāṭyaśāstra ; they are N.S. 28, 46-47 ; but the next six are found only in the Dattilam—and in the same order as quoted. Is this proof enough of Dattilam's textual indebtedness to the Nāṭyaśāstra? We think not. Firstly, as we have observed, the available Brhaddeśī is full of lacunae, scribal errors, irregularities and textual confusions. It is, therefore, not improbable that a line or lines are missing from the Vṛtti at this point where Dattila's name might have been cited by the Vṛttikāra. Secondly, in view of the many convincing evidences that we have adduced regarding Dattila's independent stature as an ancient authority of repute, and of the authenticity of the Dattilam text itself, it is unlikely that Dattila would have bodily lifted portions of another work in order to pass them off as his own. It is, we think, quite imaginable that with the discovery and study of more works on saṅgīta, these verses, too, will be found in citations unambiguously ascribed to Dattila: as we have seen in the case of the long passage from the Dattilam occurring unacknowledged in the Brhaddeśī, but proved to be unquestionably from the Dattilam through Nānyadeva's citations.

These remarks are not made to assert that Dattila was quite free from indebtedness to others. In fact he explicitly acknowledges his debt to his predecessors. In concluding his work, he humbly states that what he has written “is a mere pointer towards the system (mata) propounded by preceding teachers” (pūrvācāryamatasyaitad diñmātraṃ samudāhṛtam. Datt. 242) and invites the ‘virtuous’ readers to study their views and make amendments in his work, if necessary (Datt. 243).

Throughout his work Dattila often refers to preceding ācāryas (Datt. 66, 80, 197) as authorities on whom his exposition is based.1 Yet his indebtedness to his

1 यद्वृत्तियुक्तमचार्यैः वीणावाद्यादिलक्षणम् —Datt, 44. नित्यं पूर्णस्वरं चैमाचार्याः परिकीतिताः —Datt. 69. तृतीयमाहुराचार्याः परिकल्पनमाप्तिकम् —Datt. 169. धैवत्यां गुरुभिः प्रोक्तावनुस्वरूपमउचते —Datt. 80. ताला गुरुभिराख्याताः षोडशादशाष्टमा; —Datt. 197

predecessors appears to have been of a general nature : it was the kind of debt any author writing on a specific subject owes to the authorities who have preceded him. Dattila refers to only three of his predecessors by name: Nārada (Datt. 2, 31), Viśākḥila (Datt. 177) and Kohala (Datt. 128); the name of Bharata is conspicuously absent. Nārada is mentioned twice : the first reference, in the second verse, appears to be to Nārada, the Purāṇic figure who brought gāndharva from heaven to the earth; but the second reference (in Datt. 31) is certainly to Nārada, a very early theorist on music.1 Kohala and Viśākḥila are mentioned in connection with matters of formal detail pertaining to tāla. These two predecessors of Dattila had apparently written works on gāndharva and had become established authorities by the time the Dattilam was composed.

Bharata's Nāṭyaśāstra has been the most honoured work on dramaturgy, poetics, nṛtya, music and others related arts, over the centuries. Many commentaries of great originality and immense scholarship were written on it by numerous luminaries during the classical epoch. The work had almost come to occupy the position of a sacred text. It acclaims itself as an additional Veda, created out of the established four, by Brahmā Himself (N.S.1, 11-20), for it is related that Brahmā taught the Nāṭyaveda to Bharata, who epitomised it in the Nāṭyaśāstra. Bharata then taught it to his hundred sons: Kohala and Dattila are among the first four to be listed : "sāndilyaṃ caiva vatsaṃ ca kohalaṃ dattilaṃ tathā." (N.S.1, 26). Significantly enough these four sons—and these four alone—are again mentioned by name as prominent ācāryas in the last chapter,2 the name of Dattila, in this latter case, occuring in a corrupt form, as Dhūrtitā.3 The Nāṭyaśāstra has adopted a canonical and scriptural tone in teaching the subject of nāṭya and related arts, and hardly refers to preceding authorities except those who were so ancient as to have become mythical personalities. Yet individual works on the arts and the sciences that have been delineated in the Nāṭyaśāstra must certainly have existed prior to it. These must have included manuals and codes on gāndharva which were certainly studied by Bharata as they were by Dattila.

The hundred authorities mentioned in the Nāṭyaśāstra as sons of Bharata are relegated to the future in the Purāṇic fashion. These sons of Bharata, it is said, will write on nāṭya, music and related arts in order to popularise them;

1 Nārada, Dattila says, was responsible for the nomenclature of the 84 auḍuva and ṣaḍava tānas : अष्टाष्टोन्मादिनामानस्तु उक्ता नारादिभिः । देवाराधनयोगेन तत्पुण्योत्पादक यतः ॥ —Datt. 31. See also Dattilam—A Compendium of Ancient Indian Music by E.T. Nijenhuis, p. 70, where this author also makes a similar observation. The first Nārada, Nijenhuis thinks, was mythical and the second historical.

2 कोहलादिभिरेव तु विशिष्टगुणविद्यागुरुर्तिः । मत्स्येधर्मक्रियायुक्तः कतिच्चत्कालमवस्थितः ॥ एतच्छास्त्रं प्रणीतं हि नराणां बुद्धिवर्धनम् । — N S. 37, 24-25

3 Dattila has been called Dantila by Kunḍbha and Dāmodaragupta. Dhūrtitā also appears to be a corrupt form of his name. Dhūrtitā is another. The reference clearly seems to be to Dattila, because of the four recurring names; the other three are unquestionably common; the fourth, presumably, is also the same.

Dattila will be one. But the authorities named must have flourished prior to the Nātyaśāstra where they are named. This would place the Dattilam—where Bharata is nowhere mentioned—in a period predating the more honoured Nātyaśāstra.

However, the first and last chapters of Nātyaśāstra can be taken as later interpolations (a possibility also supported by tradition : Abhinava, though he comments upon 37 chapters, says at two places that Bharata's work consists of 36 chapters).1 Still the indebtedness of Dattila to Bharata cannot be established upon known evidence. Had Dattila drawn directly upon Bharata, he would surely have mentioned his name, for Dattila does not appear to have shy in the matter of acknowledging previous authorities. Nor in the face of the immense, almost superhuman, authority assigned to Bharata, is it likely that Dattila has in fact referred to him but has done so without explicitly naming him; that, in other words, Bharata is one of the prominent gurus or pūrvācāryas hailed by Dattila, though not named.

On the question of vīṇā and other instrumental techniques, Dattila has referred us totally to preceding ācāryas. This has been interpreted as a reference mainly to Bharata who has, in fact, in his Nātyaśāstra, described vīṇā, flute, and other instrumental techniques in detail. Dattila, however, has in this case referred us not to a single unnamed ācārya, but as elsewhere, to a plurality of them “yadvṛttișūktamāc-āryaiḥ vīnāvādyādilakṣaṇam. . . .” (Datt. 44) indicating a whole body of literature on the subject and not a single work alone. Even if we interpret this reference as being to a specific author or authors and their works, inference should lead us to Viśākḥila and Kohala whose names do occur in the Dattilam. We know that these two authorities also wrote on gāndharva: references to their opinions and quotations from their works on many gāndharva topics are to be found in works on music. Viśākhila's treatise, evidently, contained an elucidation of instrumental technique. A line from Viśākhila regarding instrumental technique has, indeed, been quoted by Abhinava while commenting on vīṇā-playing. It is quoted as a 'pramāṇa' or supporting evidence from a work parallel to the Nātyaśāstra : a 'śāstrāntara', on the subject.2 Abhinava also quotes Viśākhila regarding the flute ('vaṃśa'. A.B. on N.S. 30, 3).

Another hypothesis may, however, be posited: Bharata and Dattila, we may conjecture, both drew upon a body of common traditional material handed down from antiquity, and adapting it to their own purpose composed their works either contemporaneously or, perhaps, Dattila immediately followed Bharata during a period when the latter's authority had not yet become a byword.3 The two authorities also seem

1 पटत्रियाकं भरतमुनिमतं विवृण्वन् . . . —A.B. introductory verse 2.

Also : मध्येपद्र पटत्रिवाद्यध्वायाम. . . . —A.B. on N.S. 1, 6.

2 Abhinava in commenting upon a certain tāla formation during vīṇā-playing (outlined in N.S. 29, 91-92), criticises a certain view. In support of his own stand he quotes Viśākhila : न च शास्त्रान्तरप्रामाण्यादपि द्विकलादिविभागाश्रय (यत्स्वं) युक्तमेव। तादृशमयोगेऽन न स्यादप्रमाणश्च . . . ति विशाखिलाचार्य । —A.B. on N.S. 29, 91-92.

3 In the light of tradition, stretching over centuries, where Dattila appears as a luminary of lesser note than Bharata, one hesitates in assigning to him a period antedating the latter.

54 A Study of Dattilam

to have followed different theorical traditions in some important matters : this would explain the dissimilarities existing between the expositions in the Nātyaśāstra and the Dattilam.1

The Dattilam, as evidence clearly shows, has been held in great repute among authors on saṅgīta: it is not, therefore, suprising that commentaries, too, were written upon it, especially since the work was so terse and compressed. A commentary called the Prayogastabaka has been mentioned by Simhabhūpala, who quotes a passage from it.2 This citation has been noted by many modern scholars like M. Krishnamachariar, V. Raghavana and, following them, Prem lata Sharma3 and others.

Yet another reference appears to have escaped the notice of scholars; it is to be found in Abhinava. Discussing the gītaka named ovenaka, Abhinava quotes Dattila's view on a certain point, where it differed from Bharata (A.B. on N.S. 31, 295). Introducing, in this context, a line from the Dattilam (Datt. 208A), he gives his own explanation of it. He then cites the explanation of this line as given by 'dattile vivaraṇakṛtah' and criticises it as 'confused' (bhrāntāḥ).4 We notice that the word 'vivaraṇakṛt' is in the plural. It is difficult—in the absence of further reference—to decide whether Abhinava is naming a single commentator with an honorific plural—a commentator of repute, who wrote a ṭīkā called Vivaraṇa on Dattilam—or if he is pointing to a series of gloss or vivaraṇa writers (vivaraṇakṛtah) who had interpreted Datt. 208A in a certain way. The language used by Abhinava appears to support the latter hypothesis since 'bhrāntāḥ' seems to be too strong an adjective to use for a person who was being honoured by an address in the plural.

Another interesting deduction follows from Abhinava's phrase “dattile vivaraṇakṛtah”. The last line from Dattilam reads: “akarod dattilaḥ śāstraṁ gītaṁ dattilasam-jñitam : Dattila composed the śastra known by the name of Dattila” (which word as a lone pada becomes Dattilam). This line appears to have been originally a part of the colophon which a later scribe introduced into the text. The name Dattilam, which the present text bears, appears to be quite old as attested by this line, and this is why Abhinava, referring to commentators, says ‘dattile vivaraṇakṛtah’. The saptamī in the pada 'dattile' refers not to the author for that would in this case be ungramma-

1 Dr. Paranjpe places Dattila in the 7th century A.D. For a critical examination of his arguments see Appendix A.

2 Quoting Datt. 12-14a Simhabhūpāla comments :

“पइजलवेन पडिसंभावेन गीतीतं: परिकलिप्तो बुद्धया व्यवस्थापितो य: विवृतं चैतदग्र्योसरतकाव्यां दत्तिलटीकायाम् — “पइजलवेन पडिसंभावेन गीतीतं: परिकलिप्तो बुद्धया व्यवस्थापितो य: कविच्चद् धरानिविशेष: पइजालवेये गीते भवेत्समाद् धरानिविशेषयोहुद्र्रं तुतीय: स्यादुपष्म:” इति -Sudhā on S.R. 1,4,15-16.

3 History of Classical Sanskrit Literature: M. Krishnamachariar, para 957. Later Saṅgīta Literature : V. Raghavan, Saṅgīta Nāṭaka Academy, Bulletin No. 17. Saṅgitarāja ed. Premalata Sharma, p. 663.

4 एतद्बुद्ध्वा द दत्तिले विवरणकृतो भ्रान्तास्तलं विशेष व्याघ्यातवन्तः ।

—A.B. on N.S. 31, 295.

Testimony Concerning Dattilam from Quotations

55

tical, but to ‘Dattila’ (i.e. Dattilam) the treatise. Simhabhūpāla, too, in quoting the ṭikā Prayogastabaka says: ‘vivṛtaṃ catitatprayogastabakākhyāyāṃ dattilaṭikāyām...’. The phrase ‘dattilaṭikāyām’ also, evidently, refers to a ṭikā on the work called ‘Dattila’ (i e. Dattilam) and not to a person. This suggests that the text we have and know as the Dattilam has been current not only in the same form but also by the same name at least since the days of Abhinava (10th Century A.D.).

We have remarked that no quotation from Dattila on the group of topics discussed in the Dattilam (as constituting gāndharva), which cannot be traced to the Dattilam, appears anywhere. In the face of this fact, the shortness of the treatise appears, evidently, to have been intended by the original author himself and is not due to a later abridgement.

Moreover, Dattila evinces such ingenuity in his pursuit of brevity that it is difficult to call the composition a precis made by a lesser mind, incapable of original deliberation and writing at a later period when gāndharva practise and theory had started becoming a closed book, a vague matter of the gradually receding past.

In order to avoid long detailed descriptions regarding morphological patterns and type-bound regulated structures, Dattila composes some short, precise, almost mathematical formulas which dispense with much lengthy detail. These bring to mind a similar tendency in other śāstras of ancient times. It reminds one, for example, somewhat of the formulas so meticulously perfected by Pāṇini in describing the grammatical structure and morphology of the Sanskrit language. Pāṇini had, with the help of his formulas, reduced a lengthy grammatical discourse to a few pithy, short sūtras. Dattila evinces a similar tendency to some extent. We are noting two instances of his formula-making from the Dattilam :

(a) The gītakas described in the tāla-section had very complex beat-formations with variegated patterns of sounded and unsounded beats. Bharata, and later authors following him, have noted these beats and their patterns in full detail over long passages. Dattila, with a sūtra-like attitude towards brevity, constructs short rules or formulas on the basis of which the beats can be inferred and arranged.1

(b) Mudrā-like finger-indications formed a necessary part of the silent beats in gāndharva. The arrangement was complex, because, depending upon the context, the same finger-mudrā could be employed for different unsounded beats or different finger-mudrās could be required for the same unsounded beat. Bharata gives details (N.S. 31, 41-51). Dattila characteristically, sums the details up in a short formula (Datt. 135A-136). Abhinava quoting Dattila’s terse description on this point introduces it with the words : “dattilācāryeṇa tu saṅkṣipyoktametat” (A.B. on N.S. 4,

1 Details will emerge in Part III.

315). K. Sambasiva Sastri had noticed Abhinava's characterisation of Dattila's delineation at this point as a 'terse description'. He thought that this testified to Abhinava's awareness of the Dattilam's brevity and was a point in favour of the authenticity of the present text.1 We agree.

The Dattilam, as we have it, thus appears from all evidences to be a complete work, authentic and orignal; it has been known as such and in the same form over a long period. Some scholars have called it 'incomplete' but this is a hasty judgment since we have seen that its exposition covers the entire range of topics it promises to discuss. These topics are collectively described in Dattilam as constituting gāndharva and we have observed that the Nātyaśāstra also expounds an almost identical range of topics where, too, they are grouped together as forming gāndharva. The two texts do not disagree as to the range and scope of gāndharva, though on points of arrangement, emphasis and sundry matters of detail, there is quite a difference of opinion showing that Dattila had a stance of his own.

Besides gāndharva, Dattila also seems to have written on other subjects connected with music and drama. K. Sambasiva Sastri found two verses ascribed to Dattila in a commentary on Amarakośa, the Ṭīkāsarvasva, by Sarvānanda. This Ṭīkākāra, commenting on the meaning of the word 'nirvahana', introduces two verses with the words: 'tathā cābha dattilaḥ'.2 The verses are about sandhis in ancient plays, among which nirvahana was one. This quotation indicates—as Sambasiva Sastri had inferred—that Dattila may have composed another work dealing with plays or daśarūpaka, (Datt. Sanskrit preface, p. 4). Krishnamachariar also cites Dattila's views on various types of dramas and their classification (History of Classical Sanskrit Literature p. 544). He does not, however, mention his textual source. It is likely, thus, that Dattila had written not only on gāndharva but also on drama. Krishnamachariar cites Dattila in connection with poetics also (ibid. p. 714), but again he does not refer to any work by Dattila on the subject.

It is likely that he found reference to Dattila in works of Tantra. He says that works like the Yāmalāṣṭakatantram, Kāśyapatantram and Uddiśamahātantrodaya deal with a variety of arts including poetics, music and dance. These Tantras mention views of many early authors among whom Dattila also figures (ibid. p. 841).

1 नाट्यवेदादिवृत्तान्ताभिनवभारत्यां श्रीमदभिनवगुप्ताचार्यै:——"दत्तिलाचार्येण हु सदृक्ष*व्युत्क्रमेतद् इति प्रस्तुत्य" आच- द्वितीयमध्यायान्तात् पदाभागाद् विदुः । क्रमा निष्ठालामिकयुक्तमध्यमवदेवीकुण्डात् ॥ अयमुन्मध्ययोनिः स्वायं इति दत्तिलकारिकार्णां प्रामाणिकत्वं दृढीकुर्वते ।

The *points to the editor's footnote, which reads : 'इदं गान्धर्वशास्त्रसङ्क्षेपं' इति दत्तिलाचार्यप्रतिज्ञागते संक्षेपमनुसमारयन् अवश्यमेव प्रकृतो ग्रन्थो दत्तिलप्रणीत इति स्पष्ट्यति । —Datt. Sanskrit preface, p. 3.

2 श्रीमत्स्वनिन्दविरचिते टीकासर्वस्वे नाट्यवर्गो 'निर्वहणाद्यचर्यविचरणसमये 'तथा च दत्तिल:' इत्यारभ्य :—— मुखं प्रतिमुखं चैव गर्भो विमर्श एव च । तथा निर्वहणं चैव नाटकस्य पञ्च सन्धयः ॥ मुखाद्यैः सुपनीतान्ते यत्नु निर्वं हुति तद् तत् ॥

—Datt. Sanskrit preface, p. 3-4.

Kumbha has mentioned Dattila's views on musical forms like vibhāṣās and bhāṣāṅgas,1 though he does not quote Dattila's exact words. Kumbha claims a direct knowledge of Dattila's writings (dṛṣṭvā dattiladurgaśaktibhaṇitīḥ... S. Raj 1,1,1,36), and hence it is likely that Dattila had written not only on gāndharva but also on more popular derivative forms.

Although no authentic work of Dattila apart from Dattilam is now extant, yet some small works have spuriously adopted the name of Dattila in order to bask in his authority. But they are of an obviously late date. One such work is the Gāndharvaveda which is a collection of stray verses, 149 in number, that form no cognisable whole. Its author remains unnamed, but the work lifts many verses from the Dattilam to give itself an air of authority.2 Another work, the Rāgasāgara, is set forth as a dialogue between Dattila and Nārada, in order to borrow their glory; it is also, therefore, subtitled the Nārada-Dattila-Saṃvāda.3 The nature of the work and its subject-matter—which contains the very late rāga-ragini-putra theory of classification—proclaims it to be a recent work. Similar other works also exist.

We have seen that the Dattilam forms a neat and compact whole and it follows an intrinsically logical plan to its conclusion. Though short, it is by no means a partial or fragmentary work, for within its short span the author covers the entire range of topics (uddeśa) he expressedly sets out to expound. Its brevity is intentional and the methodical compression of its subject-matter evinces proof of the author's grasp of his subject.

We have also seen that quotations ascribed to Dattila that deal with gāndharva are numerous; they generally observe an identical verse order as in the present Dattilam, and can all be traced to it. No verses ascribed to Dattila and relating to any of the gandharva topics are to be found anywhere (in the works of repute known till date), which are missing from the Dattilam. Thus both by 'anvaya, (positive evidence) and 'vyatireka' (absence of detrimental evidence) the authenticity of the Dattilam can be taken as established.

Other roundabout and involved explanations of the evidences we have adduced in the foregoing discussions can possibly be thought of, but let us keep in mind the ancient precept that in making hypotheses the more straight forward explanation is preferable.

1 विभाषादिषु चान न योऽपि दत्तिलोदितम् । —S. Raj 2, 2, 1, 93. भाषाङ्ग दत्तिलोक्तेतु रीयमुत गीतकोहलादिभिर्गृहीत..... —ibid. 2. 2, 3, 2.

2 A copy of the Gāndharvaveda is in the collection of the library of the IICMSD, West Berlin. The original ms. is No. 8095 of the collection of the Asiatic Society of Bengal.

3 The first colophon reads : इति श्री रागसागरे नारददत्तिलसंवादे रागविमर्शको नाम प्रथमस्तरङ्कः । Similar colophons occur at the end of the other two tarangas of the work. Copy of the ms. in the collection of IICMSD. Original is ms. No. 15015 of the Madras Govt. MSS Library.

58 A Study of Dattilam

tion (lāghava) should be preferred to the tortuous one (gaurava). We can thus, for

the present, reasonably assert that the extant Dattilam is the work of Dattila himself

tiil compelling data to the contrary is forthcoming.

Dattila's early date is evinced by his independent views and traditionally well

known divergences from the Nāṭyaśāstra. His mention in the first and last chapters

of the Nāṭyaśāstra as a prominent authority—added to his own significant silence

about Bharata—shows that he perhaps composed his work at about the same period

when the main bulk of the Nāṭyaśāstra was being formulated.

Yet a basic problem still looms large if the above conclusions were to be

granted and the treatise Dattilam were to be accepted as a complete work and not

fragmentary; how is it that it deals with only a portion or aspect of ancient music ?

A work treating avowedly of gāndharva—the ancient word for music—should have

dealt with the subject in all its multiple forms and aspects. And we know from

many reliable sources that ancient music, indeed, covered a field much wider than is

envisaged in the Dattilam. Other ancient authors deal with bhāṣāṅgas, vibhāṣās,

rāgas, rāgāṅgas and a host of other forms which were part of music and hence of

gāndharva but of which this self-styled gāndharva-śāstra1 does not even speak.

An answer to this question can only be made through a proper study of the

term gāndharva and its denotations. To this we shall now turn.

1 गान्धर्वविज्ञानसङ्ग्रहः सारतोऽर्थ मयोच्यते

—Datt. 1.

PART II

GĀNDHARVA : ITS NATURE AND EXTENT

CHAPTER III

THE TWO-FOLD MEANING OF GĀNDHARVA

The term gāndharva (from ‘gandharva’ : deities delighting in music) had in ancient times a dual denotation concerning music.1 It was used both in a broad and restricted sense. In its larger connotation the word denoted all music; it was synonymous in this sense with the English word ‘music’. In a restricted sense, as we shall show, it denoted only a fixed body of musical forms constituting a very special part of the entire range of ancient forms. Gāndharva was in this sense a corpus of music firmly stabilized through meticulously defined rules and regulations formulated as decrees and injunctions which were imperative and binding. The Dattilam is a manual of this special gāndharva body of music; so are those chapters of the Nātyaśāstra which have a content parallel to the Dattilam and which deal with gāndharva and its established range of topics.

Many discerning scholars and historians of Indian music have long considered gāndharva to have denoted a special kind of music. But ideas regarding the nature, extent and scope of gāndharva have remained vague and confused. Notions regarding gāndharva have depended mostly upon Śārṅgadeva and his commentators—especially Kallinātha—and subsequent authorities. These later authors, too, like Śārṅgadeva are mixed up and muddled on the point. Evidently they inherited the legacy of confusion from Śārṅgadeva. Remarks concerning gāndharva in Śārṅgadeva and post-Śārṅgadeva saṅgīta-literature are, moreover, too summary to form the basis for any proper understanding of the matter.

Fortunately, Abhinava's commentary on the Nātyaśāstra contains an extensive analysis of gāndharva: its nature and span, and its distinct spirit and form as differentiated from other musical structures. Abhinava's remarks have not been accorded the attention they deserve. We propose to study gāndharva in the light of Abhinava's many valuable comments scattered throughout the relevant portions of the Abhinava Bhāratī and summed up in one long illuminating passage at the beginning of the 33rd chapter of the Nātyāśāstra. Abhinava will prove extremely helpful

1 Like many Sanskrit words ‘gandharva’ has in fact a host of meanings. It denotes ‘a horse’, ‘a cuckoo’, ‘a kind of deer’, ‘a celestial being’ and so forth (see lexicons); ‘gāndharva’ (that which relates to gandharva), too, could be used in all these senses. We are, however, concerned with the word and its usage only in the context of music. Grammatically gāndharva is formed as : गन्धर्वो देवता अस्य अण् ।

—Vācaspatyam.

Nārada in his Nāradi Śikṣā gives an etymology of the kind popular in ancient times ; he analyses the word in a manner to suit a denotation signifying ‘music’ through every syllable : गेति गीयते विदु: गेति धातु: प्रवादयमिति, वाग्यस्य संज्ञे ति, रेफस्तु वाग्यसोपलक्षणं काकुप्रवादनमिति वा

—Quoted in A. B. on N.S. 28, 9–10.

in understanding the ‘gāndharva’ Dattila had in mind. His analysis will, we think,

fully answer the criticism that the Dattilam deals with its subject incompletely. This

will, certainly, help clarify ideas regarding a great chapter of ancient Indian music.

Abhinava has stated that both in the śāstra of music and in ordinary usage,

the word gāndharva, like the word nāṭaka, had a dual meaning. The word, he

says, is used both in a general and a special restricted sense1.

Numerous instances of gāndharva used in a general sense can be found in

old texts. And before taking up the special meaning of gāndharva and its implica-

tions it would be interesting to take up the history of the word gāndharva and its

broad usage as ‘music’ in general. This will give a perspective to our discussion.

Surprisingly, gāndharva as meaning ‘music’ is not to be found in the earliest

classical lexicon, the Amarakośa2 (3rd Century A.D.). Yet the word can be observed

to have been in use as a general term for the art and science of music since at least

the epic times.

Vālmīki, describing the musicianship of Lava and Kuśa, calls them : gāndharva-

tatvajñau.3 The Mahābhārata often uses gāndharva to mean music and even refers to

gāndharva-veda and gāndharva-śāstra.4 ‘Gāndharva’ is the word used to denote music

even in the far-off Kharoṣṭī inscriptions of the Central Asian Indian community.5

Buddhist Jātaka stories use the word gāndharva (gandharba) as meaning music (e.g.

Guttila Jātaka).

The word ‘gandharva’ is as old as the Ṛgveda, where it is used quite a

number of times.6 It seems to have been an ancient Indo-European word and some

relate it to the Greek ‘kentaurós’ (‘centaur’, a class of demi-gods, half men, half

1. इहायं गान्धर्ववेदो लोके शास्त्रे च द्विविधो नाटकवदवत् । सामान्यविशेषणाभ्यां स्यात् प्रयो​गदर्शनात् ।

—A.B. on N.S. 33, 1.

For a dual use of the word nāṭaka in a manner analogous to gāndharva, see Appendix B.

2. One meaning of the word ‘gāndharva’ is stated to be ‘a celestial singer’ (गान्धर्वो दिव्यपायने — ऋ,

१४२) which according to later commentators could also mean singers in general. This is the

nearest that the Amarakośa comes to denoting music through the word ‘gāndharva’.

3. तौ तु गान्धर्वतत्त्वज्ञौ शस्तमेधन्तकौविदौ । श्रातरी स्वरसंपन्नौ गन्धर्वाविव राघवौ ॥

—Ram. 1, 6, 10.

4. Nārada is described as fond of wars and of gāndharva : युद्धगान्धर्व सेवी च —M.Bh. Vana. 5, 9.

Arjuna learnt the gāndharva-veda : Vana 91, 14-15. Nārada knew gāndharva : गान्धर्वे नारदो वेद:

Śānti 210. 20. The King should know gāndarvaśāstra : गान्धर्वशास्त्रे च कला परिरक्ष्या नराधिप

M. Bh. Anuśāsana 104, 49.

5. See the article entitled ‘Indian Culture in the Kharoṣṭi Documents from Chinese Turkestan’ by

R.C. Agarwal in India’s Contribution to World Thought and Culture, p. 278. Documents no.

514 and 565 contain the words ‘gandharva’ and ‘gamdharva’ (for Sanskrit gāndharva) used in

the sense of music.

6. R.V. 1, 22, 14; 1, 163, 2; 3, 38, 6; 8, 0, 11; 77, 5, 9; 83. 4; 9, 85, 12; 9, 86, 36; 10, 10, 4; 10, 85, 40;

10, 85, 41; 8, 123, 4; 10; 123, 7; 10, 136, 6; 10, 139, 4; 10, 139, 5; 10, 139, 6; 10, 177, 2.

horses).1 Like many other words, gandharva has a checkered early history. In the Rgveda it has no connection with music or with deities who delight in music and dance. In fact ‘gandharva’ in the earliest Vedic age rarely denotes a group of gods, it is rather an epithet denoting singular Beings : ‘gandharva’ is the heavenly God called Viśvāvasu (R.V. 10, 85, 21 and 22 ; 10, 139, 4 and 5) and Vāyukeśa (R.V. 3, 38, 6). His duty is to guard the heavenly Soma (R.V. 10, 83, 4; 10, 85, 12) which the gods had obtained through him (R.V. 1, 22, 14). Gandharva dwells in the sky and in the heavenly waters and is also once identified with the Soma (R.V. 10, 86, 36) ; he may have been considered the tutelary deity of the Moon (soma). He is also the God who knows and makes known divine truths and heavenly secrets (R.V. 10, 139, 5 and 6). He is the parent of the first pair of human beings, Yama and Yamī (R.V. 10, 10, 4).

There is one attribute of this Rgvedic ‘Gandharva’ which is somewhat related to the gandharvas of later conception (who were very handsome and very fond of women as well as music) : the ‘Gandharva’, Viśvāvasu, is said to have a mystical power over women ; he is said to be one of their divine Masters before they are married (R.V. 10, 85, 21 and 22 ; again 40 and 41). There is a word similar to ‘Gandharva’ in the Rgveda namely, ‘Gandharvī’, which may have contained in seed-form the later connection of gandharvas with song and music. ‘Gandharvī’ occurs only once (R.V. 10, 11, 2). ‘Gandharvī’, like ‘Gandharva’ is connected with divine waters (apyā ca yoṣāṇā) ; she is Speech who pleases the mind in the resounding (nāda) hymns of psalm singers.2

It was only during the later Vedic period that the word ‘gandharva’ was generally used to denote a class of deities. Gandharvas as a class have the same general characteristics as the one ‘Gandharva’ of Rgveda ; they live in the sky (S’Br. 14) and guard the Soma (S’Br. 3, Ait. Br. 1, 27). They follow after women wanting to possess them (S’Br. 3). Their wives are the apsarās, together with whom they are invoked by worshippers (A.V. 7, 109, 5). They are also connected with Speech and are said to have revealed the Vedas to Vāk (S’Br. 3 ; Par. Gr. 2, 12, 2) and are thus the preceptors of the ṛṣis (S’Br. 9).

Some facets of their later characteristics as deities of music have started emerging. We see that the apsarās are bracketed with gandharvas in the Atharvaveda: in phrases like ‘gandharvāpsaras’. Apsarās were joyous deities fond of dancing (yāyatiḥ parinṛtyatyādatādan......ānandinīṃ pramodinīmapsarāṃ tāmiha huve. A.V. 4, 38, 3-4).

During the epic age the gandharvas became definitely established as tutelary deities of music, dance and related arts. Viśvāvasu, the ancient Rgvedic figure,

1 In his Sanskrit-English Dictionary, Monier Williams observes : “L. Centaurus—gk. kentauros : a centaur; which some compare with skt. gandharva, a demi-god”. See also : Etymological Dictionary of the English Language by W.W. Skeat, under the word ‘centaur’.

2 रप्सरोगन्धर्वीणां च योषणा नदस्य नादे परि पातु मे मनः । —R.V. 10, 11, 2. Cf. Veṅkaṭamādhava’s comments : वदतु अमिनगणान् गन्धर्वस्य स्त्री भारती बन्तरिघ्या सरस्वती च । स्तोभिः स्तोभे परिपातु च मम मनः भारती सरस्वती च ।

became transformed in the Mahābhārata to a celestial musician, expert on the vīṇā.

He performed during king Dilīpa's sacrifice : “avādayattatra vīṇām madhye viśvāvasu

svayam” (M.Bh. Śānti, 29, 75 also Droṇa 61, 7) and he stung at the sacrificial dīkṣā of

Jamadagni (M. Bh. Vana. 90, 18).

It was also during this period that the term gāndharva became current as

denoting music. Before this period, not even in the Upaniṣads, do we find a mention

of gāndharva.1 The terms for common or ‘laukika’ music in the Upaniṣads are gīta

(song) and vāditra (instrumental music) ; also used in conjunction as gītavāditra to

denote music in all its aspects.2 Vedic sāma music was referred to as sāma and

udgītha.3

During the epic age, the word gāndharva like the modern word ‘music’

denoted music of any and every variety except Vedic music which was called sāma-

gāna4. Among actual forms we know, for instance, that gāndharva included bardic

music : the music to which epics like the Rāmāyaṇa were sung.

Music was inseparably connected with the Rāmāyaṇa right from its inception.

Valmīki, when he composed his famous verse “mā niṣāda pratiṣṭhām tvamagamaḥ….”

had sung it with an appropriate tempo5 to the accompaniment of the lyre or the

lute. His disciples had sung the verse after him : “tasya śiṣyāstatalḥ sarve jaguḥ

ślokamimam punaḥ” (Ram. 1, 2, 39). Later, having composed the Rāmāyaṇa, he

taught Lava and Kuśa, the twin sons of Rāma, how to sing it to the accompaniment

of string instruments (tantrīlayasamanvitam). The twins were endowed with a sweet

voice and a perfect sense of tone (bhrātarau svarasampannau……madhurasvara-

bhāṣinau : Ram. 1, 4, 10). They are called experts in the art of gāndharva (tau tu

gāndharvattvajñau sthānamūrchanakovidau : Ram. 1, 4, 10). When the Rāmāyaṇa

was sung before Rāma, the audience included pandits versed in gāndharva i.e. music

in all its aspects of svara and tāla6. Twenty sargas of the epic were sung on each day

1 Dr. Paranjape thus says that the only Upaniṣad that uses the word gāndharva for music is the

Sitopaniṣad, which is a late work. Bh. Sang. Iti. p. 106.

2 अथ यदि गीतवादित्रलोकामो भवति सङ्कुलादेवैस्य गीतवादित्रे समुत्पत्तिष्यतेन गीतवादित्रलोकेन सम्पन्नो महीयते ।

—Bḥadāraṇyaka 8, 2, 8.

3 पुरुषस्य वाग्रसो वाचः ऋचः: साम रसः साम्न उद्गीथो रसः।

—Chāndogya 3, 1, 2. Also :

यर्क तत्साम तस्माद्ग्राणनादग्रसाद गायति यत्साम उद्गीयरसतस्माद्ग्राणननुदगायति

—Chāndogya 1, 3, 4.

4 Cf. Bh. Sang. Iti. p. 137.

5 Ram. 1, 2, 14–18, esp :

पदबद्धौकसरसमस्तनृत्यसमान्वितः । शोकार्तस्य प्रवृत्तो मे श्लोको भवतु नान्यथा ॥

6 तां स शुश्राव काकुत्स्थः पूर्वाचार्यैःविनिर्मिताम् । अपूर्वीं पाठशज्जातिं च गेयेन समलंकृताम् ॥

प्रमाणैर्बहुविःबद्धां तन्त्रीलयसमन्विताम् । वाल्मीकिं राघवः श्रुत्वा कीर्तयिल्परमभवत् ॥

अथ कनन्तरे राजा समाहूय महामुनीन् । पारिषदांश्च नरव्याघ्रः पणिडतान् नैरमान्स्थया ॥

पौराणिका श्वदविदे ये वृद्धाश्च द्विजातयः । स्वराणां लक्षणज्ञाश्च उत्तुकान द्रिजसततमान् ॥

लक्षणज्ञाश्च मार्गाणां नैयामिक्रश्च विभीषपत् । पादाक्षरसमासाङ्गश्लान्वः सु परिनिष्ठितान् ॥

कला मातृा विशेपज्ञान.......... Ram. 7, 94, 2-7

(divase viṁśatiḥ sargā geyāḥ, Ram. 7, 93, 11) and its rendering—which Vālmīki calls a 'musical recitation', (cf. 'pāṭhyajāti geyena samalaṅkṛtā', Ram. 7, 93, 2)—was extended over a long period.

Other bardic compositions and recitative songs besides the Rāmāyaṇa must have been known. These must have been composed to both folk and sophisticated forms of gāndharva or music. Besides bardic ballads, a host of other musical forms must have been current, for music was an intimate part of life and culture during the epic period. Men were born, married and cremated to the sound of music; every function, sacred or profane, had its appropriate musical setting.

Music as performed in concerts, as a developed art-form with a tradition of teachers, experts, performers and discerning, connoisseurs also seems to have been current during the period when the Rāmāyaṇa was composed. We hear of 'pūrva-cāryas' (preceding teachers) signifying a long standing tradition of regular teaching ; of men with a discerning knowledge of svara (svarāṇām lakṣaṇajñāḥ); of experts with a specialised knowledge of musical time (kalāmātrā viśeṣajñāḥ); and of musicians well-versed in the techniques of their art (lakṣaṇajñāḥ gāndharvāḥ).1

Gāndharva or music was a standard branch9 of knowledge. As a science it was termed the gāndharva-veda. Arjuna had learnt the science of gāndharva (from the son of Viśvāvasu named Citrasena) : “kaunteyo gāndharvaṁ vedamāptavān” (M. Bh. Vana 91, 15)2 ; and had taught it to Uttarā, daughter of king Virāṭa, and her companions (M. Bh. Virāṭa, 11, 8-13). King Virāṭa had a 'nartana-śālā', a dance-hall (scene of the killing of Kīcaka by Bhīma, M. Bh. Virāṭa, 22), where his daughter and other young women of the court could learn dancing (and certainly also music).

Rāja-sabhās (big halls where courts used to be held) flourished, with gatherings of courtiers and eminent citizens around the king. On occasion, they also served as auditoriums where eminent performers of gāndharva performed for the king and his court (a tradition which continued well into our own century). In the sabhā, constructed by Maya for Yudhiṣṭhira, celestial gandharvas, experts in all aspects of gāndharva, used to perform for the pleasure of those present (M. Bh. Sabhā, 4, 36-41).

The kings of the epic period and their courtiers were often learned connoisseurs of gāndharva, for this was one of the sciences included in their curriculum. It is enjoined in the Anuśāsana-parva that the king should know music and other arts (gāndharva-śāstraṁ ca kalāḥ parijñeyā narādhipa. M. Bh. Anuśāsana, 104, 49).

However, like many generic terms of common usage, gāndharva denoted a group of related ideas, not fixed but loosely shifting. Though usually it stood for

1 7, 94, 2-7.

2 Indra had asked Arjuna to learn from Citrasena, the gandharva king, all the secret techniques of music that the latter knew :

नृत्य गीतं च कौतेय चित्रसेनादसुयुधि । वादितं देवऋषिभिः नृत्योके यन्न विद्यते ॥

.......................गीतवादित्रनृत्येषु भूय एवाभिदर्श ह ।

.......................गान्धर्वन्तु नृत्य वादितं चोपालभध्वान् ।

स शिक्षितो नृत्यगुणाननेकान वादित्रगीतार्थ गूणांश्च सर्वान् ।

—M. Bh. Vana, 44, 96-11.

non-sacred music in all its vocal and instrumental forms, yet its meaning could sometimes also include Vedic music (sāma) and even dance.1 We also notice usages where the term is used for instrumental music alone,2 or for only songs.3

During the period following the age of the two great epics, the word gāndharva remained current and the art-form it denoted continued to flourish. Gāndharva was enriched by the creation of new forms. The Harivamśa, traditionally known as the appendix (khila) to the Mahābhārata, describes in great detail the festivities and watersports of Kṛṣna, Balarāma and their companions and consorts on a beach near Dvārakā and speaks of a new form of music created by Kṛṣna for the occasion. This new form was named chālikya-gāndharva and was a composite form comprising song, instrumental music as well as dance (Hari. 89, 66-83).

An inscriptional evidence also survives from around this period of the use of the word gāndharva for music. The Hathigumpha inscription of Kharavela (2nd Century B.C.) calls the king ‘an expert in the science of music’ : gāndharva-veda-budho’ (Epigraphica Indica, Vol. XX, see line 5 of the Kharavela inscription, edited by Jayaswala and R.D. Bannerjee).

In early Buddhist literature, too, the word gāndharva (in its Pali form as ‘gandhabba’) is used for music in general. The Jātakas often refer to gandhabba as an established śilpa—a science or skill.4 The Milinda Pañha recounts gandhabba among the vidyās which the educated were trained in. Early Buddhist literature even tells of a faculty of music in the ancient university of Benaras, which was “presided over by an expert who was ‘the chief of his kind in all India’.”5 At the famous ancient academic centre of Takṣaśilā, too, gāndharva was, evidently, one of the 18 śilpas which, according to the Jātakas, were taught there.

Like the Mahābhārata, the Jātakas testify to a wide variety of gāndharva or music. We hear of sophisticated art-forms requiring a long and arduous training in which masters used to compete for superiority in the presence of large audiences with the king as the presiding judge6 (this practice, too, was not unknown in India till the very recent past). We hear of festive gāndharva, on the occasion of joyous processions (Mahāvessantara, verses 726-727); of the light-hearted erotic gāndharva of singing girls who used to perform for the pleasure of kings (Khāntivādī Jātaka) ; of the lone gāndharva played and sung as a seranade by estranged lovers (Kusa Jātaka) and even of gambling songs (Vidhura Jātaka).

1 वैष्णावसूत्रु तनयादिते नृत्यं च साम च । वादितं च यथान्यायं प्रत्यविन्दव्याकविधि ॥ एवं कृतास्त्वः कौन्तेयो गान्धर्व वेदमास्थवान । —M.B. Vana, 91, 14-15.

2 गीतगान्धर्वचर्चोपेतश्चतुरोणवति:स्वरै: । —ibid. Anuśāsana, 160, 64.

3 गान्धर्वमत्तुलं नृत्यं वादितं चोपबृंहणम् । —ibid. Vana, 44, 10.

4. See Guttila Jātaka, the story of a great teacher and expert in gāndharva; Kusa Jātaka where Kusa makes gāndharva or music by singing and accompanying himself on the vinā in order to attract his estranged wife.

5 Ancient Indian Education : Dr. Radhakumud Mukherjee, p. 490,

6 Guttila Jātaka.

The word gāndharva continued in usage as a generic term meaning music throughout the Gupta and later classical periods well into medieval times. Bhāsa in his Pratijñāyaugandharāyanam (3rd Cent. A. D.) often uses the word gāndharva in this sense : Vāsavadattā who was intent upon acquiring skill in music (asyā gāndharve' bhilāṣah' : Act. 2) was taught by king Udayana, a famous musician (gāndharvavit-takah', Act 2) and an expert on vīṇā who was proud of his hereditary skill in the art of gāndharva (darpayatyenaṃ dāyādyāgato gāndharvo vedah : Act 2). In the play Avimāraka, Bhāsa's forlorn hero expresses envy for the happy man who with his beloved by his side, makes and enjoys gāndharva (aye gāndharvadhvaniriva śrūyate. ko nu khalvayaṃ sarvakālasukhī puruṣaḥ kāntayā saha gāndharvamanubhavati : Avimāraka, Act 3).

Cārudatta, the gentle hero of Śūdraka's Mrcchakatikam (4th Century A.D.) was a great music-lover ; we hear of him spending the night listening to the sweet gāndharva of his friend, Rebhila, who sang to the accompaniment of the vīṇā1 (kāpi velāryacārudattasya gāndharvaṃ śrotum gatasya : Act 3).

The Bhāṇa Padmaprābhrtakam, also ascribed to Śūdraka,2 describes its love-lorn hero as suffering a heart-burn from the very things he used to delight in : the spring, the moon, garlands of flower, perfumes and also gāndharva or music (ramyai-ścandravasantamalyāracanāgāndharvagandhādibhiryayāireva pramukhāgataiḥ sa ramatetaireya santāpyate, verse 7).

The Viṣṇupurāṇa, which Dr. Hazra places in the 3rd Century A.D. (Pu. Re. p. 22; Dr. Baladeva Upadhyaya places it in the 2nd Century B. C.: Purāṇa Vimarśa, p. 545) recounts gāndharva as one of the 18 vidyās.3

Inscriptional evidence of the use of the word gāndharva to denote music in general is also not lacking from the early centuries of the Christian era. The Junagadh inscription of Mahākṣatrapa Rudradāman (1st Cent. A. D.) speaks of his great proficiency and fame in logic and gāndharva.4 Samudragupta (4th Cent. A. D.), the well-known Gupta king, was very fond of music and the coins that he minted show him as playing the vīṇā ; his inscriptional eulogy speaks of his extreme sophistication in the art of gāndharva. His skill could put to shame ancient ṛṣis like Tumburu and Nārada.5

1 Of Rebhila, Cārudatta says : tāṃ tasya svarasamkramam mudgurir; bhilāṣṭaṃ ca tantrīsvanam, vaṇṇanāmapi mūḍhānantaragatāṃ tāreṇ vīrāme mūḍam | helāsamyitam punarva lajitaṃ rāgadr̥ddhaircitam, yatsarvaṃ virete'pi gītasamaye gacchāmi śr̥ṇvanniva || —Mrcch. 3, 5.

2 See the Introduction to the Hindi edition entitled Srigāra-Hāt of the Caturbhānī (a collection of four Bhāṇas belonging to the Gupta age of which Padmaprābhrtakam is one) by Dr. Motichandra.

3 आयुर्वेदो धनुर्वेदो गान्धर्वश्चैव ते त्रयः । अर्थवास्त्रं चतुर्थं तु विद्या हस्तादिवाचिकाः ताः ॥ —Viṣṇupurāṇa 3, 6, 29.

4 गान्धर्वन्यायविद्यानां पारंगाधारण विज्ञाने प्रयागवाप्तविगुल कीर्तिना...... महाक्षत्रपेण रुद्रदाम्ना..... —Epigraphica Indica, Vol. VIII, p. 44.

5 निशितविदग्धमतिगान्धर्ववदलतैः श्रीहस्तविद्याप्रतिगुरू तुंबरू नारदौ... —Gupta Inscriptions, Fleet. No. I, p. 8.

Jaina canonical literature acquired its present written form during the 5th or early 6th Century A.D.1 In this body of texts we often find the word ‘gāndharva’ used as meaning music; the term sometimes also embraces dance. The words ‘gandha-vva-śālā’ and ‘gandhavva-ghara’, for example, were names of audience-chambers where music was taught and practised or performed.2

The Vasudeva-Hindī is a rightly famous work of early Jain ākhyāna literature (5th Century A.D.).3 It is a treasure of social and cultural information concerning the Gupta age, an era when people were passionately fond of music or gāndharva: experts were accorded great respect and all persons of education and culture were expected to be proficient in the art. One of the Hindī stories tells of the Vidyādhara Kāmonmatta (Kamummatta) who had abducted sixteen young girls from the homes of princes, rich merchants and tradesmen (rājā, setthī, ibbha, sāttavāhā). The girls were beautiful, well-educated and cultured. Two were expert musicians. One, known as Śrī (Sirī), was an expert vocalist, proficient in the singing of gāndharva : gāndharvagītakuśalā’ (gandhavvagīya-kusalā); the second, Vijavasenā (Vijayasenā) was an expert composer of music: she was ‘gāndharva-racanakusalā’ (gandhavvara-yanakusalā).4 The Gandharvadattā Lambh relates the story of Gāndharvadattā, daughter of the merchant Cārudatta. This beautiful damsel was excessively fond of music (gāndharva) and was a great expert in the art : gāndharvavedapāramgatā (gandhavvedapāramgayā)’. She had vowed to marry the man who will defeat her in gāndharva. Such was her beauty that every eligible person in the town where she lived frantically began practising gāndharva in the hope of winning her for himself (tūse rūvamohiyā māhṇā khattiyā vaisā gandhavve rattā). Vasudeva, hero of the story, under the assumed name of Khandila, asked a towns-man if expert teachers of gāndharva were available in the town (atthi iham uvajjhāyā gandhavva pāragā ?), for he too wanted to learn music (gāndhavvaṃ sikkhejjä).5 Vasudeva was only feigning ignorance of music but was actully an expert musician. He succeeds in winning the princess for his bride by playing the viṣṇu-gītaka (a composition unknown to his fair compitetor) on the vīṇā.

1 Hist. of Ind. Lit., Winternitz, Vol. II. p. 432.

2 For early Prakrit Jain references to gāndharva, see Paiasaddamahāṇṇava under ‘gandharva’, p. 284.

3 “The Āvaśyaka Cūrṇī mentions Vasudeva Hindi thrice. This shows that this work was composed before A.D. 600.” Haribhadra Ke Katha-Sāhitya Kā Ālocanātmaka Pariśīlana —Dr. Nemicandra Sastri, p. 37.

4 Vasudeva-Hindī : Dhamilla-cariyam p. 68, published by the Jain Atmananda Sabha, Bhav-nagar, Gujarat.

5 चासत्तसिदिणो धूया गंधव्वदत्ता परमगुब्बती गंधव्ववेद-पारगया। सो वि इदम्हो वेसमणस्समाणो। तं तिसे रुवमोहिया माहणकत्तिया-वड्ढसा गंधव्वे रत्ता। तं च जो जिणइ सिक्खिउं तस्स भज्जा होइहति पुण्णभोगिणो। मासे मासे गते च णओोगं देवु विडसाणं पुरओ। कलहो च समुदओ आसी। पुणो माणेण भविस्सति। मया ‘विसिणा-वहुविहसा’ गमेयज्जणा पुच्छलामि ताव ण-भो अण्णे इहं उवचओ गंधव्वपारगा। कहिय च से मया णामं खइलितोति गोयमो पुण गोत्तेण, गंधव्व’ सिक्खेज्जा"1

—Vasudeva Hindī p. 126–127.

In another section of the Vasudeva-Hindi, named the Sāmā Vijayā Lambho, we find a similar condition placed for winning the hand of the two princesses, Śyāmā and Vijayā. These two beautiful sisters were experts in both gāndharva and nātya (or nṛtya): gandhavve naṭṭe ya pariṇitthiyāo. Any eligible young man who could surpass them in these branches of learning could win their hand (jo ṇe āgameṇa visesijjā so ṇe bhaṭṭā ti). Vasudeva, the hero, this time under the assumed name of Gotama, defeats these princesses with his greater skill and knowledge of the intricacies of music and theatre (or dance) and receives them as brides, along with half the kingdom over which their father, the king Jitaśatru, ruled.1

Another famous work of the 5th Century, where the word gāndharva is often used as denoting music, is the Bṛhatsamhitā of Vārāhamihira. Gāndharva is mentioned as an art along with writing (lekhya), mathematics (ganita) and handling of weapons; musicians are called gāndharvika (for details see appendix ‘ri’, Bh. Sang. Iti. p. 521).

The Bhāgavatapurāṇa, which according to Dr. Hazra and Dr. Baladeva Upadhyaya, cannot be placed later than the 6th Century A.D.,2 often uses the word gāndharva to mean music. Gāndharva is recounted as an established branch of learning along with architecture, medicine and archery.3 A legend relates that the ten sons of king Prācīnabarhis, called the Pracetasas (pracetasah), set out to perform meditations and tapas near the sea. There they heard divine gāndharva accompanied by percussion instruments like mṛdaṅga and paṇava.4

King Revata, father of Revatī (later married to Balarāma, the elder brother of Kṛṣṇa) went to Brahmā, in order to seek his advice regarding a suitable groom for his daughter. Revata was distracted by the beautiful music being made in Brahma-loka : “āvartamāne gāndharve sthito” labdhakṣaṇaḥ kṣaṇam”. (Bhāg. 9, 3, 30).

It is also related, in a beautiful allegory, that King Pṛthu ‘milked’ the earth of her treasures which consisted of herbs and agricultural produce. Manu acted as the calf and helped in the milking (Bhāg. 4, 18). Following Pṛthu, the gods, the demons and the gandharvas all ‘milked’ the earth and the milk they extracted was the object which they most coveted : the gods, with Indra as the calf, extracted ‘soma’ as well as moral qualities like virility, inner strength and power as the ‘milk’ (vīryamoja balam), in a vessel of gold (hiraṇmayena pātreṇa). The demons, with Prahlāda as their calf, extracted wines and liquor (surāsavam). The gandharvas and apsarās, with Viś-vāvasu as the calf, extracted gāndharva or music along with beauty (saubhagam) and

1 इह रायगो जियसत्तूस दुवे धुयाओ—सामा विजया च, सुवंसणिहरो गंधव्वेण य परिणिट्ठियाओ विर्णिस्सयंवराओ। जो णे आगमेण विसेसिज्जा सो णे भट्टत्ति 1. ......ताओ च मया गंधव्वनट्टसमयमणुण्णाओ वि नट्टटे गिए ण विसेसिआओ । त तों ट्ठहेउ रन्ना सोहण्ण दिन्ण तस्सि पाणिण गाहिओ मि बिहिणा, अहं च रजजेसु णिसिट्ठ । —Vasud vā Hindi. p. 121.

2 Puranic Records by Dr. R.C. Hazra, p. 55; Purāṇa Vimarsa (Hindi) by Baladeva Upadhyaya, p. 548.

3 आयुव्वेदे धनुव्वेदे गान्धव्वे वेदमात्तण्णः । स्थापत्यं चासृजद वेदम्.... —Bhāg. 3, 12, 38.

4 तत्र गान्धर्वमाकर्ण्य दिव्यमार्गमनोहरेम । विसिस्मय राजपुत्ते स्ते मृदङ्गपणवादयसं ॥ —Bhāg. 4, 24, 23.

sweetness (madhu) in a vessel appropriately consisting of lotuses : “gandharvāpsaraso’

dhukṣan pātre padmamaye payah/ vatsaṃ viśvāvasuṃ kṛtva gāndharvaṃ madhu

saubhagam”. (Bhāg. 4, 18, 17). Other beings similarly extracted objects of their own

desires.

Another famous text of the 6th Century, the Śiśupālavadha of Māgha1 uses

gāndharva as meaning music or song in a verse with a double-entendre where the word

means both music and horses.2

Bāṇa, the renowned poet of King Harṣa’s court, flourished in the 7th Century.

From his own accounts we know that in his youth he was not quite respectable.

Among the companions of his young wandering days was a teacher of music—

gāndharvopādhyāyaḥ—named Dardurāka (Harṣacarita, ch. 1). In Kādambarī,

Candrapīḍa is taught many arts and sciences : carpentry (vāstuvidyā),

medicine (āyurveda), dentistry (dantavyāpāra), handling mechanical devices (yantra

prayoga) etc; these sciences include gāndharva or music as propounded by Nārada

and others (nāradiyaprabhṛti gāndharvavedaviśeṣaḥ : Kādambarī, p. 168, Nirnaya Sagara

ed.). Further in the story, Bāṇa describes Mahāśvetā playing on the vīṇā with

an ivory plectrum (danta koṇa) held in her right hand :

“she looked like the science

of gāndharva incarnate” (pratyakṣamiva gāndharvavidyām : Kādambarī, p. 283).

Listening to her music, Candrāpīḍa was completely overwhelmed; he thought that the

lady producing such music was certainly of divine origin for who has heard such

subtle gāndharva tones being produced by mere humans—“kuṭiśca martyāloke

sambhūtiravamvidhānāṃ gāndharvadhvaniviśeṣāṇām” (Kādambarī, p. 287).

The Kuvalayamālā of Udyotanāsūri is a kathā in Prakrit and was composed

during the 8th Century A.D.3 Gāndharva here is often used to denote music. When

a son is born to queen Priyaṅguśyāmā, king Dṛḍhavarman orders for joyful festivities

to be held : “the atmosphere becomes filled with the tinkling music (gāndharva) of the

jewelleries worn by intoxicated women: tāra tāra raṇantehiṃ kāñcīkalāvehiṃ tākiṅkiṇī-

tāla-mālā-ravāraddha-gandhavva-pūranta-saddam disā-mandalāṃ”. (Kuvalayamālā,

p. 18). Here we have a very looseʾuse of the word gāndharva to denote any pleasant

or ‘musical’ sound. The word is also used for the science of music which was one of

the disciplines taught to princes along with painting, poetry, theatre etc :

“ālekkhaṃ

ṇaṭṭaṃ joisaṃ ca gaṇiyaṃ guṇā ya rayaṇānaṃ/vāgara yaṇaṃ veya-sui gandhavvaṃ

gandha-jutti ya” (ibid. p. 22). Another passage speaks of the beautiful sounds of

1

Prof. Jacobi has argued that Māgha cannot be “placed later than about the middle of the Sixth

Century” : quoted in the preface to the Nirnaya Sāgara edition of the Siśupālavadha. ff. p. 4-5.

2

नानाविद्याविश्कृतसप्तस्वर; सहितस्वर्या चपलवीरुन्ध्रयः |

गान्धर्वभूमिविशिष्टतया समन्ततः स सामवेदस्य दधो बलोदधि: ॥—Siśu. 12, 11.

Mallinātha comments :

सहस्रशाखास्वात् सहितस्वर्या, गन्धर्वा एवं गान्धर्वा अंशा.....सामवेदौषि बहुविधाविष्कृतबहुध्वरसंतरादिसमोत्यतस्वरः।।

सहस्रशाखास्वात् सहितस्वत्ना, गान्धर्वं गानबहुत्वाच्चय इति ।

3

Haribhadra ke Prākṛta Sāhitya kā Ālocanātmaka Pariśīlana, Dr. Nemicandra Sastri, p. 61.

Also Kuvalayamālā, ed. Dr. A.N. Upadhe, Vol. 1, preface, p. 2.

gāndharva accompanied with tāla and string instruments : “gandhavva-tāla-tantī-samvaliya-milanta-mahura-saddenam.” (p. 43).

Another work of the 8th Century, the Kuṭṭanīmatam of Dāmodara Gupta, mentions the science of gāndharva, and names the semi-mythical exponent of the art, Nārada, son of Brahmā, when describing the great musical skill of a certain character (gāndharve kamalajānmanastanayah : Kuṭṭanīmatam, 875). Like Māgha, in a double entendre, Dāmodara uses gāndharva as meaning both a horse and a musician when describing the city Vārāṇasi : “turuṣkaseneva bahula-gandharvā : the city contained many gāndharvas (musicians) as the Turk army contains many gāndharvas (horses).” (ibid. 14)

Śīlāṅka's Cauppannamahāpurisacariyam, a Jain Prakrit work of the 9th Century, also mentions gāndharva as one of the essential sciences,1 taught to princes and others who were nobly born.

During the 10th Century, in the latter part of which Abhinava began his vocation as a writer, gāndharva was quite a common word for music. The Bṛhatkathā-kośa of Hariṣeṇa, a collection of stories written in Jain hybrid Sanskrit and composed in the middle of the 10th Century (see Introduction to this work by Dr. A.N. Upadye, p. 117-122) often uses the term gāndharva for music. Here, we again come across the story of the daughter of Cārudatta (the girl is here named Gāndharvasenā) who is won in a musical contest by Vasudeva through his superior skill in playing gāndharva on the vīṇā : “sughoṣavīṇayā tatra kurvan gāndharvamuttamaṃ/jitva gāndharvasenāṃ ca vasudevo'vatiṣṭhate.” (Bṛhatkathākośa, 93, 296). Another story with a similar theme is that of Gāndharvadattā, daughter of king Gandharvadatta. This princess, like Gāndharvasenā, had vowed to marry a man superior to her in gāndharva: “gāndharvavidhinā yo hi kumārīṃ māṃ vijesyate/tasya bhāryā bhaviṣyāmi saṃgaro'yam mama sphuṭam” (ibid. 114, 6). She defeated many a suitor by her great skill in gāndharva (narāḥ sarve gāndharveṇa nirākṛtāḥ). Then an expert in gāndharva named Pāñcāla came to compete with her. So excellent was he in the art of gāndharva that there was hardly a chance of his losing to Gāndharvadattā. The girl realised this fact, but did not want to marry Pāñcāla because he was extremely ugly. She was at the same time fascinated by the intoxicating magic of his music. Being in two minds about the man and unable to come to a decision, the poor girl committed suicide.

Another author of the 10-11th Century, Dhanapāla, almost a contemporary of Abhinava, also uses gāndharva in his Tilakamañjarī. Describing the evening activities of a king he says : “reclining on an ornate bed he spent his time listening to the music of flutes and vīṇās played by famous ācāryas of gāndharva so that he, too, could learn the art” (vidrumadāruparyaṅkamadhiśayānastatkālesevāgatairgītaśāstrapararijñānāya dūrārūdhagarvairgandharvikopādhyāyaiḥ saha veṇuvīṇāvādyasya vinodena dinaśeṣamanayat : Tilakamañjarī, Nirnaya Sagara ed., p. 70).

1 Cauppannamhāpuriscariyam, p. 38, names an author on gāndharva named Citraratha whose text formed the basis of teaching the art : gandhavvaṃ cittaraheṇam. We hear of no such author in the extant standard texts on music though he seems to have been well-known among Jain circles of the 9th Cent. as testified by Śīlāṅka.

The word gāndharva was, in fact, used to denote music in general till quite a late period. Its continuous usage in this sense is attested by its inclusion in late lexicons, such as the Śabdaratnākara of Vāmana Bhaṭṭa Bāṇa (14th Century A.D.). Bhaṭṭa Bāṇa includes gāndharva in the list of important śāstras and defines it as the 'science which governs songs' : "syādgāndharvaṃ gītiśāsanam" (line 1844); again in his list of synonyms he equates gāndharva with song or gīta : "samāne gītagāndharve" (line 1887). A very late lexicon, the Rasakalpataru of Viśvanātha (17th Century A.D.), also equates gāndharva with song : "gītirgītaṃ gānaṃ geyaṃ gāndharvamapyasmin" (line 1628).

We observe that the word gāndharva denoted music in general as well as its science or art. Its range of meaning was, in fact, almost the same as that of the English word 'music'.

Abhinava thus suggests that in a general sense gāndharva could connote any sound that might strike one as pleasant and not only the organised sounds of music as a fine art : the songs of children, the rustic singing of shepherds and even the songs of cranes and herons could all be loosely grouped as gāndharva.1

Bharata, too, uses gāndharva as denoting music in general : as in the passage where gāndharva is included among factors that could arouse the feeling of love.2 Dattila, also, must have been aware of such a meaning of the term, though his work is too confined to provide an instance.

By the period of Bharata and Dattila, gāndharva or music had evolved a large variety of forms. The historical limits of this period cannot be ascertained for sure, even as to the exact century, but it is generally said to fall somewhere between the 3rd century B.C. and the 3rd century A.D.

But along with its wide connotation, the term gāndharva had also, by this time, acquired a limited meaning as a particular body of music. What Bharata and Dattila have expounded under the name of gāndharva consititutes not the entire span of music known to them but a restricted range of forms.

The melodic or svara structures of this gāndharva were limited to the 18 jātiṣ. The tāla structures consisted of the seven major gītakas (described under the topic prakaraṇa) and some minor ones (such as pāṇikā).

Gāndharva in its restricted sense certainly did not exhaust the entire gamut of existing musical forms. Bharata himself, after having fully described the characte-ristics of forms he designates as gāndharva and having completely dealt with all the topics he lists under gāndharva, goes on to describe another distinct family of musical

1 न होताबदेव गांधर्वस्य लक्षणं बालगोपालतासरबलाकादिगीतैषडपि गान्धर्वापते: ।

—A.B. on N.S. 33, 1. Abhinava makes this remark in answer to the possible objecion that the special restricted meaning of gāndharva needed no exposition, as it was evident enough. This, Abhinava purports to say, was not so, because the word gāndharva was capable of being used in a very loose sense indeed, as meaning any pleasing sound whatever, even the song of a bird. Only analysis, he implies, could bring out its restricted meaning clearly.

2 ऋतुमाल्यालङ्कारेः प्रियजनगान्धर्वकाव्यसेवाभिः । उपवनगमनविहारः श्रृङ्गाररसः समुद्भवति ॥

—N.S. 6, 47.

forms, the dhruvā, i.e. dramatic songs, in a long chapter (the 32nd, containing 437 kārikās). He even differentiates these dhruvās from gāndharva (as we have seen) on matters like alaṅkāra-formation and the use of ‘gīti’. These dhruvās, as the evidence of early writers like Kaśyapa testifies, were sung to melodic forms like grāma-rāgas, bhāṣās and rāgas all of which were quite distinct from gāndharva jātis. Kaśyapa has named many of these and has noted the occasions for their use on the stage.1 These forms were evidently current during Bharata’s age also.

Dattila in his Dattilam has not described any form outside the range of what he calls gāndharva, which in his account covers the same limited range of topics and musical forms as in Bharata. Yet he definitely hints at other forms. He mentions hybrid (saṅkara) forms : these were created by compounding two or more gāndharva jātis. He speaks of the general maxim by which these saṅkara forms could be traced to its parent jāti (Datt. 96). In fact, Dattila makes a general statement regarding all existing musical forms annoucing that they were all ‘present’ in the jātis, that is to say, they could all be traced to the jātis : “tasmād yad gīyate kiñcit tatsarvam jātiṣu sthitam” (Datt. 97). The hybrids, though not specifically named, evidentiy, included grāma-rāgas, bhāṣās, rāgas and related non-gāndharvic forms.

Bharata and Dattilla were evidently not the only figures who described gāndharva in a limited sense : there were many others, of whom the works of Viśākhila and Kohala seem to have been known directly by the Vrttikāra ’on Brhaddesī and by Abhinava and others as well as by Dattila himself. Evidence points to a whole tradition wherein this special body of music was described and preserved under the name of gāndharva. All its peculiarities and features—structural and doctrinal—were summed up with a more or less fixed set of terms listed in uddeśas. These in turn were classified on the basis of savara and tāla.3 Bharata and Dattila are, in

1 Kaśyapa recounts many grāma-rāgas, bhāṣās and rāgas along with specific situations in the dramatic plot where they were most suited. The names of some melodies are familjar even today, though their musical form must have undergone great transformations. To take a few examples, we meet with, gauri, baṅgāli, gurjari, gāndhāri, mālava kaiśika, bhinnaṣadja, kaku-bha and so on. See the passage from Kaśyapa quoted in A.B. on N.S. 29. 13.

2 Bharata himself speaks of grāma-rāgas in connection with percussion accompaniment of dhruvās. The context is interesting. In the 34th chapter of the Nātyasāstra, Bharata deals with percussion instruments, their technique and use in theatre. He describes the placing of the orchestra (kutapa-vinyāsa), pointing out the area on (or near) the stage where the various musicians should be seated and how they should be arranged (N.S. 34, 216). He proceeds to give directions regarding the tuning of drums. The drummers after being properly seated, he says, should tighten or loosen the leather straps (vadhra = ‘baddhi’ in Hindi) of their instruments to get right sounds ‘śithilāñcitavadhrastanitēṣu: N.S. 34, 217). Abhinava explains that the purpose of adjusting leather straps was to get the right svara or pitch. The mrdaṅga, Bharata further says. was then to be properly pasted (this was done with dough as it is to this day in North India) and tuned to the required grām-rāga before being rested on the thighs of the player : “yathāgrāmarāgāmārjanaliptēṣumṛdaṅgēṣurudvayāpiḍane tēṣu . . .”(N. S. 34 217).

3 We are not including ‘pada’ here, because, as we can gather from the Dattilam (Datt. 4) and observe in the Nātyasāstra, ‘pada’ was a minor factor. Bharata takes the pada as denoting ‘ordinary words’, and they were as such governed by laws of grammar, usage, prosody etc.; these laws governed all words and were not limited to gāndharva. Bharata recounts the topics governing pada as follows :

व्यञ्जनानि स्वराश्च वर्णा: सन्धयोध विग्रकत: । नामाद्यतोपसर्गान्तच निपातास्तद्विदिता कृत: छन्दोविधिर्द्विरलङ्कारा जैेय: पदगतो विधि: । —N.S. 28, 16-17.

This may be rendered as:

(Contd. on next page)

truth, two links—albeit the only extant links—in a long chain of ācāryas who were

bearers of the tradition of gāndharva. Bharata and Dattila, indeed, appear to be late

ācāryas in the tradition, for by their time all the formal and conceptual facets of

gāndharva had already acquired set patterns. Divergences between Dattila's and

Bharata's exposition of gāndharva are certainly to be found, but they hinge on details

or on matters of theoretical outlook—the main body of gāndharva as a whole is

accepted as an established entity.

Abhinava has devoted much attention to gāndharva in its restricted sense. He

has dwelt upon the peculiarities of gāndharva music both as to spirit and form

and has analytically distinguished it from other musical forms. His testimony is of

the greatest importance, because he was not only one of the most discerning men that

medieval India has known in the field of the fine arts, but evidence reveals that he

had a deep and direct knowledge of the matter that he was analysing.

Abhinava Gupta (10th Cent. A.D.) was one of the most remarkable thinkers

of India. His vision had a true philosopher's range. He has dwelt illuminatingly

over the philosophical subtleties and depths of Śiva pratyabhijñā darşana. He has

also opened up new vistas in the field of aesthetics and literary criticism. He had a

genius for analysis, rare in the history of thought. His contributions to philosophy

and literary aesthetics are well known and have fruitfully occupied many a learned

scholar.

It is not so well-known, however, that Abhinava was also a great musicologist.

He has analysed and discussed the intricate terminology of music with the sure grasp

of a master. Abhinava has commented on the entire Nātyaşaşstra, a feat possible only

for a versatile scholar, because the Nātyaşaşstra is an epitome of many arts and

sciences: from aesthetics to architecture and music to mine.1 Although Abhinava had

before him the examples of some very competent ṭīkās on the Nātyaşaşstra2—some of

which he mentions and refers to—yet his own contribution in this field was so

original as to have far-reaching influence over the centuries. The geyādhikāra of the

Nātyaşaşstra (chs. 28-36) is a very technical section dealing with music. It deals with

(Contd. fram previous page)

"The consonants, the vowels, the alphabet, the suffixes, the nouns, the verbs, the prefixes, the

nipātas, taddhitas and kṭantas, rules of metre, figures of speech—such are the laws governing

pada."

Yet, words as they entered into structural relations with melody and tāla were certainly in the

domain of music and Bharata and Dattila as gāndharva-şaştrākaras have described quite a few

'musical' aspects of the pada element in gāndharva. We shall come across these aspects later

in the essay.

1 न तक्शानं न तच्छिवा न सा विद्या न सा कला । न स योगो न तत्कर्मं यन्नाट्यहिसमिन्न दृश्यते ।

—N.S. 1, 116.

2 For example those of Mātrgupta (5th Cent. A.D.), Sri Harṣa (7th Cent. A.D.), Udbhata (8th

Cent.A.D.), Lollaṭa (9th Cent. A.D.), Sañkuka (9th Cent. A.D.) etc.

matter needing a very specialised knowledge of the subject. Abhinava's commentary on this section is as masterful as elsewhere.

At places his meaning is obscure, but the reason is not the shoddiness of the commentary but the fact that the subject dealt with is now a remote one. Technical terms are meaningful only within a living technique, and when, as in this case, the art and its technique have undergone a transformation over the centuries, it is difficult to recapitulate the significance of terms as they were originally used. Moreover, the available edition of the Abhinava Bhāratī on the geyaadhikāra has, many faulty and fragmentary readings at places; so much so, that discovering a coherent meaning sometimes becomes a very exacting and frustrating task. The fault, surely, lies with the perplexing nature of the subject-matter itself and the best editing efforts may remain largely futile.1 Yet, in spite of these difficulties, this commentary proves to be one of the most illuminating documents for our understanding of ancient Indian music. No medieval author on music can compare with Abhinava in clarity of analysis and range of thought. Furthermore, Abhinava is relatively close in point of time to the ancients. After the Bṛhaddeśī, his Abhinava Bhāratī is the first extant work on music.

Abhinava's reflections show compelling evidences of his direct knowledge of the dramatic and musical forms he was discussing and commenting upon. It is likely that he was himself a proficient musician. A contemporary pen-portrait by one of his pupils, Madhurāja Yogin, shows him as playing on the Nādavīnā, while dictating some matter to his disciples.2 K. C. Pandey, who has made an admirably detailed study of Abhinava and his works, makes the following observations: “Abhinava Gupta, as a matter of principle, writes on those subjects only of which he has personal experience. This is true not only in the sphere of philosophy but also in that of art. Thus, in the case of dramatic presentations, he writes on the basis of the presentation of dramas by great actors (Mahānata) who were his contemporaries and to whom he refers in the Abhinava Bhāratī.”3 K.C. Pandey proceeds to produce references from the Abhinava Bhāratī which show that Abhinava in commenting upon gīti-kāvyas and their stage presentation, was speaking from personal knowledge (Abhinava Gupta, K.C. Pandey, p. 115-116).

Abhinava flourished during a period when ancient art-forms and the doctrines underlying them were living things. Though new forms in music and the other arts were certainly being created—old forms were very much alive and were performed by artistes versed in tradition. These forms must have undergone changes, but by and large, they maintained their orginal outlines and formal characteristics. Authori-

1 Unless, of course, more manuscripts are discovered. The present edition is made from two sets of manuscripts both full of ‘scribal errors, broken pieces, moth-eaten leaves and other natural decays’ (see preface to N.S. Vol. 1, second edition, p. 20). So faulty are the readings in the original mss. that Ramakrishnakavi went to the extent of remarking : “a scholar friend of mine is probably justified in saying that even if Abhinavagupta descended from heaven and saw this ms. he could not easily restore his original reading” (ibid. p. 63).

2 Abhinavagupta, K.C. Pandey, p. 20-22. The author also quotes relevant stanzas from this pen-portrait in his Appendix, p. 738.

3 ibid. p. 115.

tative manuals like the Nāṭyaśāstra and the Dattilam were studied and discussed by

ācāryas and were taught to groups of eager śiṣyas. These śiṣyas must have included

actual performers, for communication between performers and theoreticians seems to

have been unbroken. The ācāryas in their theoretical delineations referred to actual

practise and performers must have referred to ācāryas in order to authenticate, and if

necessary, correct what they practised ; for only by this reciprocity must ancient forms

have been kept so alive that even relatively late ācāryas like Abhinava could

evoke the evidence of 'actual renderings' (lakṣya) in proving a point.

Abhinava's teacher in the Nāṭyaśāstra was Bhaṭṭa Tota, whose teachings and

discussions form the basis of Abhinava's comments and ideas recorded in the

Abhinava Bhāratī. Abhinava says that he composed the work because of an eager

demand from certain persons (arthijana) for information regarding Tota's ideas.1

Tota was himself a representative of a long line of teachers from whom he

had learnt the Nāṭyaśāstra. Abhinava was another luminary in this tradition and

similarily acted as its vehicle. The persons (arthijana) who had prompted him to pen

down his store of practical and doctrinal knowledge regarding the *Nāṭyaśāstra

must have included performers, for who else can be more eager than a performer

regarding authentic pronouncements concerning his art ?2 Abhinava's pronouncements

are all the more authentic and authoritative because of his wide and intimate know-

ledge of actual practice as current in Kashmir, which during his period was the hub

of learning and culture, and was termed by impressed contemporaries as the 'land of

Sarasvatī'.

As in the rest of his work, many illustrations are strewn over the geyādhikāra-

section of the Abhinava Bhāratī which show Abhinava's direct and personal

acquaintance with his subject. We jot down some instances :

(i) Commenting upon the range to which certains jātiṣ were to be played or

sung in the tāra octave, he criticises a certain view saying that this

view does not agree with actual practice (mahān lakṣyavirodhaḥ). He

then records the svara-prayoga, in the tāra, of three jatis: ṣāḍjī, ārṣabhī

and niṣādavatī, with the words "this is how they are actually rendered ;

dṛśyate ca lakṣye".3

1 सद्यःप्रसूतवदनोदितनयवैदग्ध्यवार्यंभिजनैःछलितसिद्धिहेतोः ।

माहेश्वराभिनवगुप्तपदप्रतीतः संस्कृत्यभृतिविदधिना विशेषदीकरोति ॥

—A.B. introductory verse 4.

2 According to Acarya Visvesvara, these 'arthijanas' belonged to South India where the only

existing mss. of the A.B. have been discovered and where the tradition of Bharata's nāṭya

are still alive. His speculation is, however, supported by too meagre a thread of evidence.

—Abhinava Bhāratī ke tīn adhyāya, p. 41.

3 अथ तारमाह । पञ्चस्वरपरा तारगतिरिति । ननु यच्च गस्वरादारोहक्रमेण तावदारोहणीं यावत्तत् पञ्चमसंख्यस्य स्वरो न

तु तत्रपरिमिति नियमः । (एतन्न) तदा महान् लश्ष्यविरोधः । पड्ज्यां हि गड्जोषे पञ्चमस्वरारात्तमेवारोहणीया स्व्यात् ।

दृश्यते च लश्ष्ये पड्यावां—ननु घा घानिसनिधघनिषपघनिषपा इति । आर्षभ्यां च छैवतान्तः स्यात् । दृश्यते च निषादवतां न

घरनिमगमम् इति । निषादवतां च पञ्चमालातमारोहणं दृश्यते पापारिणा इति । एवमन्यतापि विचार्यम् ।

—A.B. on N.S. 28, 70 ; again in the same passage :

एत एवैति, तल्लश्ष्येण (स्थाप्यते) । तथाहि । निषादवत्यां निषादेऽपि तारपड्जौ दृश्यत एवाञ्जन साधनेन इति ।

(ii) Jātis, he says, are basically patterns of svara. He adds that rules regarding the use of specific tālas though enjoined were not always observed in current practice as could be seen in actual performances.1

(iii) Speaking about the use of notes in the rāga-forms, born of the jātis, he says that the ṭakkarāga and the mālavakaiśika both use the same ṣadja with the same number of śrutis but still the note sounds different in the two rāgas (because of the changed context). A similar experience, he continues, can be had on hearing the rāgas vegasvara and saindhava, where the same ṣadja acquires yet new expressions. In fact, he concludes, on hearing the note ṣadja we are immediately able to recognise whether ṭakkarāga, saindhava, vegasvara or mālavakaiśika is being sung.2

(iv) The sthāyī varṇa in music was a generic term for patterns restricted to a single note. But, paradoxically enough, in rendering some alaṅkāras (melodic graces) of the sthāyī varṇa, all seven notes of the octave appear to have been utilised. To dispel the apparent contradiction, Abhinava, remarks: “in an alaṅkāra of the sthāyī varṇa the introduction of other notes occurs in such a way that they do not hinder the experience of hearing one single note being rendered continuously: madhye tu svarānta-rānupraveśo bhavannapi na sthāyivarṇadhiyaṃ bādhate” (A.B. on N.S. 28, 20).

Alaṅkāras like preṅkhita, hindu, sannivṛttapravṛtta etc., were subtle alaṅkāras. According to Abhinava it was difficult to analyse them and separate them from the musical context in which they were rendered (yatra vihitāstatvena bhānti). Like many subtle musical movements they could only be experienced and not described in words. Abhinava therefore, says that his own descriptions of these alaṅkāras are mere tokens through which a listener might be able to recognise the alaṅkāras when he hears them sung. Other alaṅkāras like recita, kampita and kuhara were so ethereal that they could not be indicated. These alaṅkāras were formed by special trills of the three-śruti notes (ri and dha in the ṣadja-grāma, ri and pa in madhyama-grāma) in different octaves. Their subtlety, says Abhinava, is the reason why they have neither been indicated through a token nor described.3

1 tālavṛttamanudrūtvā sāmyenāyāṃtanuccalitaguptādin loka iti vācikamārgāḥ jātyādiṣu prasiddham iti tu vayam chacchaliguptādi na tad vikalpaḥ | 3tuṣaklādānuṣāreṇa kalapratibandhiḥ: kartavyodapi loke na prasiddhaḥ | svarabhāgāgrahāṇyāt |

—A.B. on N.S. 28, 95-97.

2 tulye ca śruti(tye)kaṭa(k)eṅyādyoṣo mālava(k)kaśike paḍjo bhāti | Ṭakkarāge tvanyadrūṣṭaḥ | tathāpi vegasvarasaṅghavadane(n) bhinnāvāvabhāti ca | karṇyaḥ paḍjadhanināma(n) eva(k) śrutisthākarāgādyaṅge saṅghavo vegasvar iti mataḥ |

—ibid. 29, 8.

3 yathāpi caṅku(ṅ)ḍukāraḥ: panvar(yatvā) hitāstatvena bhāti(smit) tathāpi pratītasya gītamayapraviṣṭasyaṃabhijñānadānārthaṃ likhitāḥ | reḍhitakampitakuharāstu śirovallāḥ:kaṇṭhaniviṣṭasya viśruteḥ: svarasya kamalutvā ityaṃbhijñānenāpi darśayiṣyati muṣakā iti n vilikhitā iti |

—A.B. on N.S. 31, 21-22.

(v) We have noted that in dhruvās—ancient threatrical songs—alaṅkāras which might split a word or pada beyond recognition were not used. They were, however, used in gāndharva and Abhinava, had, evidently, heard them being rendered, for he says : “we find in actual performances of jāti-gītas that words are split to any desired extent : laksye'pi ca jā tigītā dau yatheṣṭaṃ prakarṣaṇamupalabdhameva” (A. B. on N. S. 29, 26).

(vi) Vidārī in ancient times was a division of musical compositions or songs into parts or phases in order to impart charm and variety to the structure and to break their non-stop continuity which could only result in monotony. Abhinava, evidently, speaks from personal experience when he says : “in songs devoid of the diversity rendered in them through vidārī formations, the resulting music is no different from a lesson being monotonously read out from a book: vidārīvaicitryaśunye hi gītavyavahāre svrālāpaḥ pustakavācanādāviva” (A.B. on N.S. 31, 200).

It is clear that Abhinava speaks from experience in matters of music. His testimony can thus be of gratest importance in our understanding of gāndharva as described by Bharata and Dattila.

Abhinava analyses the musical forms, pertinent in connection with the Nāṭyaśāstra, into two basic categories : gāndharva and gīna.

Abhinava realises that the term gāndharva could be confusing. The definition of gāndharva, as consisting of savra, tāla and pada, that Bharata gives at the very outset of the geyādhikāra, is so broad that it could give rise to the notion that the term denoted music in general. One could thus misconstrue that the entire geyādhikāra dealt with gāndharva alone. To avoid this confusion, Abhinava reminds his readers that the word gāndharva also had a definite restricted denotation.1 He points out the extent of the geyādhikāra text which deals specifically with gāndharva in the restricted sense. Gāndharva, he says, denotes the specific body of music described by Bharata in the first four chapters of the geyādhikāra, i.e. chapters 28 to 31.2

This observation is of the greatest interest in attempting to understand the subject-matter of Dattilam, because we have seen that the Dattilam and the first four chapters of the geyādhikāra deal with an identical subject-matter.

Abhinava further says that topics relating to the svara aspect of gāndharva have been delineated by Bharata in the three chapters 28 to 30 and tāla is expounded

1 'svaratalalātmakaṁ gāndharvamiti' geyādhikārāramḍhaṃ eva muninā saṅgītānāṁ ca tat-svarūpādirkaṁ lakṣaṇamuktaṁ 'idānīṁ tan-nāma-vyutpattir loka eva śāstre ca drṣṭā nāṭakābhyadhāvat. sāmannyaviseṣa-bhedenāsya prayogadarśanāt' —A.B. on N.S. 33, 1.

2 nanvai' gāndharvasya ca lakṣaṇamuktaṃ tad-dhyāyacchubhṭyāpeva muninā ibid—. 33,1.

in chapter 31, where the nature of pada in gāndharva is also indicated : “gāndharve” pi tathaiva ca svaravidhiradhyāyatrayeṇa (i.e. chs. 28-30) darśitaḥ tālavidhiradhyā-yena ca (ch. 32) tatraiva padavidhirāpi darśitā” (A.B. on N.S. 32, 27). This, in fact, precisely defines that portion of the Nāṭyaśāstra where the specific topics listed as gāndharva are named and delineated.

Abhinava has used the word gāna, too, in a pāribhāṣic sense. Gāna ordinarily means ‘song’—any song. Bharataḥ has himself used the word very often in a general sense; and has made repeated use of it even within the chapters on gāndharva.1 Abhinava gives a peculiar meaning to the word ‘gāna’ when contrasting it with gāndharva. It is not very clear if by ‘gāna’ he denoted all non-gāndharvic forms current during his period. But by ‘gāna’ he certainly meant all musical forms utilized in dhruvā singing and in the music which was played in the background during the staging of ancient drama. We will not be wrong in taking ‘gana’ in Abhinava to stand for such non-gāndharvic musical forms which have been described or indirectly assumed in the geyaadhikāra and which were employed in the presentation of theatrical plot. ‘Gāna’ used musical forms like the grāma-rāgas, bhāṣās, nāgas etc., moulding them to the structure and ethos suited to dramatical plots.

Abhinava’s sources for the term ‘gāna’ seem to be ancient texts on music where dhruvā was also referred to as simply gāna. In an important passage at the beginning of the sixth chapter of the Nāṭyaśāstra, Bharata gives a short resumé of the subjects he proposes to deal with. Among elements connected with music he recounts the five accepted varieties of dhruvā songs and calls them ‘gāna’ (‘gānaṃ pañcavidham’ N.S. 6, 29-30).2 Here the word ‘gāna’ clearly does not mean song in general but denotes dramatic songs.

‘Gāna’ has been used in this sense also in a passage ascribed to the ancient authority Kaśyapa. This passages is quoted in the Vṛtti on Bṛhaddesí. Kaśyapa was an authority on musical forms employed in the dhruvās and has been often named in this capacity along with Kohala and others, by Abhinava. Abhinava also quotes a passage from Kaśyapa where many grāma-rāgas have been traced to their parent jāti-forms and these grāma-rāgas are said to be employed in dhruvā-ganā.3

1 To take a few examples :

चतुर्विधवद्मेतेषां विलयं गानयोक्तृभिः:

—N.S. 28, 22.

गान्धर्व एवं योज्यास्तु नित्यं गानप्रयोक्तृभिः:

—N.S. 29, 49.

तत्रवृत्तं रसे गानं कार्य गेये प्रयोक्तृभिः:

—N.S. 29, 9.

मध्यमपञ्चमसंयुक्ता गानं श्रृङ्गारहास्ययोः:

—N.S. 29, 10. etc.

2 प्रवेशकाद्यपनिष्कान्तप्रस्तावादिकमनुसृत्य तत्र गानं पञ्चविधं जायं ध्रुवायोगसमन्वितम् ।

—N.S. 6, 29-30.

3 The passage occurs at the end of the A.B. on N.S. ch. 28. The last verse of the quoted paasage reads :

एषा विभाषा गीतो तु ग्रामरागाश्रिते मते । ध्रुवागानोपयोगाय तदतद्गोत्रिसम्वकम् ।

In the passage quoted in the Vṛtti on Bṛhaddesī, Kaśyapa decrees that grāma-rāgas should be sung to the five established dhruvās which, like Bharata in N.S. 6, 29-30, he terms ‘gāna’.1

The sense in which Abhinava uses gāna is perhaps best illustrated through another reference from Bharata. In the 28th chapter of the Nātyaśāstra, the delineation of gāndharva begins with verse 8 which contains a definition of the term gāndharva. The verses preceding this describe various types of instruments and different instrumentalists needed for presenting a theatrical show. Then in verse 7 Bharata makes a general observation: “gāna, vādya and nāṭya should be so skillfully combined as to make a unitary whole”.2 The word gāna here evidently refers to theatrical music in general—music of the variety which was an inherent part of the unfoldment of ancient dramatic plots. This is indeed the sense in which Abhinava, too, uses the word.

1 तथा चाह काश्यप:-

ग्रामरागा: प्रयोक्तव्या विधिवद् दशरूपके । प्रवेशालोपनिष्क्रान्तं प्रासादिकमयान्तरम् ॥

गानं पृथक्विधं यत् तद् रागैरेभिः प्रयोजयेत् ।

—Vṛtti on Br 364A.

2 एवं गानं च वाद्यं च नाट्यं च विविधाश्रयम् । अभिनातचक्रप्रतिमं कर्तव्यं नाट्यैकस्नुभि:

—N.S. 28, 7.

We have discussed this verse earlier.

CHAPTER IV

A GENERAL INTRODUCTION TO GĀNDHARVA

Before going into details regarding gāndharva and its distinction with gāna, we would do well to form a general idea, in the light of Abhinava, of what gāndharva was. Gāndharva was a specific body of music having a well-defined form in all three aspects of svara, tāla and pada ; Abhinava calls it : 'svara-pada-tālaviśeṣātmakam' (A.B. on N.S. 33,1). Only certain specific musical movements of svara and tāla were permitted in gāndharva. When sung, gāndharva was rendered with certain traditionally set songs.1 Its forms were believed to have been time-immemorial and were maintained intact through an unbroken 'guru-śiṣya-param-parā'—the teacher handing them down to the disciple and so on : “nāradādigurusan-tānānuyāyīti gāndharvasya prabhavah” (A.B. on N.S. 28, 10). Bharata's and Dattila's treatises are, indeed, manuals of gāndharva composed within this param-parā.

Gāndharva was, in its rigidity and timelessness, believed to be somewhat akin to the Vedas and it was, indeed, considered to be quite as immutable and sacred. Regarding the 18 jātis of gāndharva, Bharata says that they were established by Brahmā himself: 'jātayo' sthādāśetayevam brahmaṇābihitam purā' (N.S. 28, 39). Abhinava comments that their establishment by Brahmā himself shows them to be revealed forms which permitted no change or mutation : 'brahmaṇeti āptāgamasya-na-nyathābhāvamāha'.

The nrtta (such as tāṇḍava), too, when rendered as part of the purvaranga was considered inviolable and sacrosanct like a prescribed Vedic ritual. Its accompanying music was gāndharva. Abhinava says of them that "in these (purvaranga) renderings, both music and the dance-gestures (gītamaṅgam ca) are fixed and have their own predetermined forms (sva-pratiṣṭhitam). Their tempos, rhythmic patterns (yati) and other characteristic formal structures should be rendered exactly as enjoined. Like Vedic mantras these should never be disfigured : 'iha tu gītamaṅgaṃ ca dvaya-mapi svapratiṣṭhitam tathā hi yasya yadrṣṭaṃ layayatisvarūpādikaṃ nirūpitaṃ tanna viparyeti mantradivat' (A.B. on N.S. 4, 248-259). Abhinava strengthens his point by further exhortations. He says that in nrtta when Bharata enjoins a particular movement being in fast tempo with an accompanying gesture called apaviddha (vide N.S. 4, 407), then this is exactly how the aṅgahāra has to be rendered, no mutation being permissible. Similarly, in the case of the gītakas of gāndharva, musical forms had to be performed just as described: when in a gītaka it was ordained

1 This is what Abhinava denotes here by padaviśeṣātmaka. Though the general linguistic nature of pada was governed by grammatical and other laws common to all padas, yet the actual form of the padas, or the particular lyrics sung in gāndharva constituted a specific body of hymns in praise of Śiva. Bharata has noted some of these, as we shall see later.

that a certain rhythm-cum-tempo formation should be made in the middle of a specific structure, such as in the gītaka called aparāntaka where the decree was 'madhye srotogatalayah' (cf. Datt. 175), then it was imperative that the formation had to be made exactly as decreed without any transgression or violation : ["tatha hi 'apaviddham drutam caiva' (N.S. 4, 207) ityaṅgahāre, gītakādāvapi madhye srotogā-layādirityādikam nānyathā kriyate"]; (A.B. on N.S. 4, 248-259].

Gāndharva was not only immutable like Vedic mantras, its performance was not unlike the performance of'a Vedic sacrifice. It was a special kind of sacrifice in honour of the gods. It was a yāga that was free of the immense expenses involved in Vedic sacrifices, and was especially pleasing to Lord Śiva : "dhananinirapeksam cedam devānām yajanam, yathā purāñayogādibhyo'dhikā prītirgāndharvācchañkarasyeti" (A.B. on N.S. 28, 9). Like Vedic sacrifices, gāndharva was governed by strict commands regarding dos and donts. Every movement in it had to strictly follow the injunctions or vidhis laid down for it. These injunctions were preserved in the manuals on gāndharva (such as the Dattilam). Movements forbidden or not prescribed were to be avoided: "gāndharvavedoditavidhiprabandhapradhānatvādidam gāndharvam" (A.B. on N.S. 3.3, 1); "gāndharvaśastram ca pramanamasyeti gāndharvam" (A.B. on N S. 28, 8). For this very reason Abhinava has called gāndharva 'pravṛttinivṛt-tipradhāna,' governed by observances regarding what is applicable and what is non-applicable.

A properly performed Vedic sacrifice was said to result in adṛṣṭa (or apūrva, a concept synonymous with adṛṣṭa).1 The same belief existed regarding gāndharva.

It would be worthwhile here to understand the concept of adṛṣṭa so as to be able to relate it properly to gāndharva and to understand a vital aspect of the spirit geverning gāndharva. Apūrva or adṛṣṭa is a concept elaborated by the Mimāṁsakas and is intrinsically connected with the liturgical spirit behind the performance of Vedic rites and sacrifices.

The goal of Vedic sacrifices is the attainment of heaven, hence such ancient injunctions as: "yāgena svargam kuryāt : one should attain heaven through sacrifices" and "svargakāmo yajeta : one desirous of heaven should perform sacrifices".

It was, however, argued by many opponents of Vedic sacrifices that a result (phala or kārya) in principle immediately follows a cause (kāraṇa) and when the cause is absent the result or phala also is non-existent. Now, a sacrifice is a destructible temporal activity with a definite termination and we know that at its end heaven does not necessarily result; the person performing a sacrifice continues to dwell in this very unheavenly world for years after. And if heaven were to result years after the death of the yajamāna, the performer, how can a yāga or sacrifice be called its cause, for the yāga had become non-existent when it terminated years before.2 A

1 अपूर्व कमंज न्या स्वर्गादिफलदास्यावित्तः स्वर्गादिफलस्य प्राप्तत्तनी सुखसामग्र्यो वा । इदमेव अदृष्ट नामः see 'apūrva' in the Mimā ṃsā Kośa.

2 cf. 'स्वर्गकामो यजेत' इत्यादि समयाध्याहृतत्वात्स्वर्गसाधनत्वं यागं विधिप्रत्येत बोधयते तच्च न सम्भवति, स्वर्गादि: कालान्तरभावित्वात आभु विनाशिनो यागादेष्टदानोऽसत्वेन कारणत्वासम्भवत्वात -see "adṛṣṭa", Vācaspatyam, p. 116.

cause which has been non-existent for years cannot result in a phala now; how then can one say that sacrifices cause heaven : katham vinaśvareṇa phalaṃ kartavyam ?

(Mīmāṃsā Kośa on 'adrṣṭa').

In reply to this objection, it was argued by the proponents of Vedic karma, especially the Mīmāṃsakas, that yāgas generate a certain transcendental power or śakti which remains even when a yāga is over and has become non-existent. It is this śakti, accruing to the performer of a yāga, which at the appropriate time results in heaven. This śakti was named apūrva (=adṛṣṭa): “apūrvaṃ nāma yāgāvāntaravyā-pāraḥ, kim vā, yāgajanyā kācickaktiḥ kartṛsamavāyini” (Jaiminīyanyāyamālāvistaraḥ, 2, 1, 2 quoted under 'apūrva' in the Mīmāṃsā Kośa).1

A total Vedic yāga is made up of many smaller or greater rites. Each small rite has to be performed strictly according to injunctions, for only then can the complete yāga be rendered faultless as a single entity or 'whole'. The apūrva resulting from the yāga as a whole is named the parama-apūrva, while that resulting from a rite within a yāga is the avāntara-apūrva. The parama-apūrva is the totality of avāntara-apūrvas.2

Thus every single rite down to its minutest detail had to be performed according to vidhi handed down by authorities. Mādhavācārya in his Sarvadarśana-saṅgraha gives an interesting example from the sacrifice called darśapūrṇamāsa. In this sacrifice husking of rice was a part of the rites performed. Now, rice particles can be separated from the chaff in many ways, even by using one's nails. However, it was enjoined that rice in this case should be husked through avaghāta, i.e., by bruising it in a wooden pestle with a mortar of the same material.3 This was an imperative vidhi (“yathā nakhavidalanādinā taṇḍulaniṣpattisambhavat pākṣiko 'vaghato vaśyaṃ kartavya iti vidhinā niyamyate,” Sarvadarśana-Saṅgraha, Anandas-rama Series, p. 89). Only husked rice resulting from avaghāta could be used in certain rites to be performed during the darśapūrṇamāsa sacrifice and only thus could the

1 also cf.

अपूर्व पुनरस्र यत आरम्म शिष्यते — ‘स्वर्गकामो यजेत’ इति ।

इतरथा हि विधानं अनर्थकं स्यात्, म्रि षात्पवत यागस्य

—Śabara Bhāṣya, 2, 1, 2, 5 (p. 390) as quoted in Mīmāṃsā Kośa under 'apūrva'.

also cf.

शतं विधिवोद्ध तं यागादेरिष्टसाधनस्व फलकालपर्यंनतस्यापिसवजनस्याध्यापारं विनानुपपत्तिमनुपपत्तिज्ञानेन द्वाराभूतमपपवं कल्पयते

—these are lines in continuation of those given in fn. 1, p. 191, from the Vācaspatyam.

2 The Pūrva-Mīmāṃsakas, in fact, classified apūrva into three types—the parama-apūrva, the aṅga-apūrva and the kālika-apūrva. The first was the apūrva of a yāga as a whole and the other two were sub-apūrvas and depended upon the size and importance of particular rites to which they accrued :

तच्चापूर्वं त्रिविधं प्रधानापूर्वमड.गापूर्वं कालिकापूर्वं चेति मीमांसका: । तत्र दर्शपौर्णमासाद्यपूवंं प्रधानापूर्वं तदेव परमापूर्वं,

प्रयाजादि चड.गजन्यापूर्वमड.गापूर्वं तद्वतस्तर क्रियाकूटजन्यमपवं कालिकापूर्वं

— see Vācaspatyam under apūrva.

3 See the Sanskrit-English Dictionary, Monier Williams, under 'avaghāta'.

avāntara apūrva be acquired which helped in forming the parama-apūrva: “avaghāt-anispannaireva taṁdulaiḥ piṣṭapurodāsādikaraṇe ‘vāntarāpūrvadvārā darśapārṇamā-sau paramāpurvamutpādayato nāparathā ato’ pūrvamavaghātāsya niyamahetuḥ” (ibid.).

Just as in a sacrifice, so in gāndharva, only a proper performance led to adṛṣṭa. Gāndharva was in fact a gānayajña—a sacrifice with musical movements taking the place of the usual rites. As in a yāga, each specific musical movement (which was here analogous to a rite) was to be performed strictly according to injunctions; the true adherence to the enjoined rules led to adṛṣṭa (“gāndharve niyamamadrṣṭasiddhaye”, A.B. on N.S. 28, 35).

We have already seen in the last chapter that the beats in gāndharva served an adṛṣṭa end.1 The same was true of every single element of gāndharva. Let us take another instance: Abhinava points out that in the jātiś (the svara-structures enjoined in gandharva) the use of graha (the initial note) and other structural factors had to be made in a certain specified and distinct manner because only then could adṛṣṭa accrue: “grahādi-yojanam tu vaiśiṣṭyāvaham gāndharve tu vaiśiṣṭadrṣṭasiddhaya” (A.B. on N.S. 28, 24).

Here is another telling example. In yāgas smaller avāntara-apūrvás gave rise to a big totality of adṛṣṭa. It followed from this that the larger a sacrifice and greater the number of smaller rites in it, the greater was the resulting adṛṣṭa. A similar principle applied in gītakas (complex tāla-structures ordained in gāndharva). A gītaka could be rendered in a short or compressed form as well as in a lengthened and spread-out form, known respectively as kulaka and bhedyaka (chedyaka in Datt.).

These two types had three further sub-forms. Altogether, including the subclasses, the number of ways in which a gītaka could be rendered was quite sizeable (see kulaka and chedyaka, ch. IV). In some modes or renderings relatively fewer aṅgas or components were formed in comparison to lengthier renderings which included all possible aṅgas. Each of these possible modes of rendering a gītaka was equally governed by the rules or vidhis of gāndharva and each resulted in adṛṣṭa but there was a simple maxim governing the extent adṛṣṭa each could result in : the greater the number of components formed in a gītaka, the greater the resulting adṛṣṭa—“te ca sarva evopabhedā apyadṛṣṭopayoginaḥ, aṅgabhūyastvaṁ tu phalabhūyastvāyeti nyāyaḥ” (A.B. on N.S. 31, 324).

This is not to say that gāndharva had no dṛṣṭa-phala or an immediately pleasing effect; a dṛṣṭa-phala was natural to it by virtue of the fact that gāndharva, what distinguished it from other musical forms was its adṛṣṭa value too, was music. What distinguished it from other musical forms was its adṛṣṭa value and a strict formal regulation to that end. Thus Abhinava says: “gāndharva is a special body of musical forms regulated by strict laws of application and has both

1 also, cf.

गान्धर्व त्वदृष्टप्रधानतया शम्यादितालगणननियानियमिका लिङ्गैर्‌नवैष्यँ्‌वामवदधिणकरैर्‌विमानादभ्युदयैर्‌यति नियता एवंविगादयो द्रष्टव्या:

—A.B. on N.S. 31, 30-32.

an immediate and an adṛṣṭa goal…. "svatalāpadaviśeṣātmakam pravṛttinivṛtti-pradhāna dṛṣṭādṛṣṭaphala.. (A.B. on N.S. 33, 1).

As the adṛṣṭa motive clearly reveals, gāndharva was a musical form seeped through and through with a transcendental ritualistic purpose. No such musical form is now extant in India. True, we have our devotional songs, but formally or structurally speaking they are not really different from other forms. What distinguishes them is merely the fact that they use words which are devotional or mystic or in other ways sacerdotal. The same forms are used in an erotic thumrī as well as a kīrtan or a devotional psalm. There is no specific form, distinct in all its aspects of svara, tāla and pada which is reserved as especially sacred. Music now consists of a large body of forms basically undistinguished as either spiritual or profane; these forms are adapted to profane or devotional songs purely on the basis of individual choice. There is no doubt a small repertoire of tunes which are generally associated with devotional songs, in every area of the country. But these do not have any strictly well-defined regulated form which can set them apart from other musical forms at large. Moreover they are not in their totality—besides their pada or word-content—governed by any sacred, transcendental or ritualistic motive. They can be changed or modified at will, as they often are. None of them approaches anywhere near the gāndharva in its strict and immutable principles of formation.

It could thus be of greatest interest to the history of Indian music to study gāndahrva in detail, to know the range of its forms, its structural details and its relation to other musical forms.

Forms of music governed by a spirit similar to the one behind gāndharva have been known in other countries and cultures, too. The most significant and outstanding example seems to have been the musical forms of ancient Egypt. Plato, the ancient Greek philosopher, was a great admirer of these forms. He has made interesting remarks about them. He says that in Egypt all art-forms including music followed a set ordained inflexible pattern, in which no alteration was permitted. In other words, these forms were inviolable because (like gāndharva) they were divinely ordained. To quote Plato : "in Egypt they have a tradition that their ancient chants which have been preserved for so many ages are the compositions of the Goddess Isis."1

In China, too, there existed a ritual-music which was played in imperial courts and in Confucian temples till recent times. This music was part of an elaborate ceremonial in which "the actions of the emperor, the words of the hymns, the number of musicians, their positions and their instruments and all the minutia of ritual were fixed by ceremonial law".2 Musical movements were similarly governed and believed to have a more than human significance. Ritual music in China was said to relate to primordial universal principles.

1 The Dialogues of Plato, trans. by Jowett, Laws II, para 657.

2 Groves Dictionary of Music and Musicians, Vol. II, See under Chinese Music.

Europe, also, has an analogous form in the Gregorian chant. This is a form of liturgical church music sung during 'Roman' rites. Every ceremony has its fixed songs, sung to a music that has been maintained intact in an unchanged form since the late Middle ages.1 Alec Robertson, the celebrated authority on Gregorian chant, says that it is impossible to understand Gregorian chant without understanding the spirit of Christian rituals or liturgy. For this music was seeped with the liturgical spirit.2 European music since then has undergone fundamental changes-melody has given place to polyphony ; an entirely new way of envisioning musical forms has developed and transformed all music. Yet, Gregorian chant still maintains its original melodic nature, it is still sung to the various modes of Medieval Europe (corresponding to our thāṭ or mela-kartās) and its forms have remained unaltered. For, a sacred ritualisic importance is attached to these forms-not only to the hymns sung but also to the very musical movements with which they are sung.

We now have an idea of the spirit motivating gāndharva. We can understand it in greater depth if we follow Abhinava who highligts the peculiarities of gāndharva by contrasting it with the musical forms he terms gāna. Gāna forms were profane forms and were malleable according to the demands of the theatric context in which they were employed. Gāna formed an integral part of the ancient Indian theatre and was employed within the actual plot being enacted. It helped in the unfoldment of the human drama depicted on the stage (Abhinava characterises gāna as : naṭyasā-magrimadhyamajanijatasvara, A.B. on N.S. 33, 1). The musical structure of gāna was moulded to a comparatively lighter vein, for gāna was composed basically with an eye towards immediately pleasing effect (raktyanusāreṇa pravṛttitḥ, ibid.). It was set to appropriate metres with the purpose of emphasising and accentuating the ethos of the moment (rasabhāvocita vṛttajātinibaddham, ibid.). With this end in view the composer had a free play with the music in composing gāna. Gāna was meant for the audience, its appeal was directed basically towards the viewer ('tadavagamaṃ pradhānam sā māijikaṃ prati', ibid.). But gāndharva was meant for the performers, the adṛṣṭa arising from gāndharvāta accrued, in principle, to the musicians, since adṛṣṭa (it is said) by its nature attaches to the performer (gāndharvaysa prayoktari prādhāny-enādrṣṭaphalatvāt).3

1 Higin Angles writing on Gregorian Chant for the New Oxford History of Music (Vol. II) says : "it is a surprising fact that the spread of the neums (special musical notations for recording Gregorian Chant) over the different nations of Europe had no influence whatever on the essential nature of religious music. Variation is seen only in exterior detail, where 'orthography' and musical spelling are modified. The neums, though they can be classified as the work of one school or another, never provide us with a 'particular national verrion' ; they transmit the primitive chant with a melodic and even sometimes a rhythmic uniformity which is really astounding" (p. 105).

2 See his account of the Gregorian Chant in the Pelican History of Music, Vol. I.

3 The printed A.B. on N.S. 33,1 reads : गान्धर्वस्य प्राधान्येन दृष्टफलत्वात An akāra, it appears, has been dropped by an inadvertant scribe. A more appropriate reading would be प्राधान्येनादृष्टफलवात। In the context where the line occurs Abhinava is contrasting the different phalas resulting from gāna and gāndharva (the sentence preceding the line being (Contd. on next page)

It may be worthwhile here to take a somewhat detailed look at the role of the ancient theatric dhruvās. This will reveal their character and their function and will give us an idea of the spirit underlying gāna forms.

There were basically five dhruvās : prāveśikī, ākṣepikī, prāsādikī, antarā and niṣkrāmikī; these were also called praveśa-gāna, ākṣepa-gāna-prāsādika-gāna etc. (N.S. 6, 29-30). Within each of these were multiple sub-forms or samsthānas—“dhruvāstu pañcavijñeya nā nāsaṃsthana-saṃsrayāḥ” (N.S. 32,9). The samsthānas differed according to different syllabic arrangements, number of letters, metric length, the rasa and mood as well as the particular stage character or nātyapātra who sang the poem forming the dhruvā: “saṃsthāne tu gurulaghusanniveśo’ ksaramātreyattā rasabhāva-prakṛtibhedaśca tadāśrayeṇa prakāraṇānātvam” (A.B. on N.S. 32, 9).1

As is well known, characters were divided into types and categories; kings and other high born exalted characters were the type (prakṛti) of uttama or jyeṣṭha-characters; lowly characters like the Vidūṣaka were adhama, while characters neither high nor low were madhyama. Each character, when singing a verse, moulded it to a suitable rendering of dhruvā befitting his position.2

Different junctures and occasions on the stage had their own dhruvās: prāveṣiki—as the name suggests—was for entries; niṣkrāmiki for exits after acts; ākṣepikī was employed when there was a sudden intervention in the prevailing mood by a new element which thrust (ākṣipta) itself upon it; prāsādikī helped in building up or purifying the mood or rasa which was being presented: “prastutaṃ rasaviśesaṃ yadā prasādayati nirmalikaroti” (A.B. on N.S. 32, 314) and antarā was needed to cover up a fault, blemish or transgression during a performance.3

It appears from

(Contd. from previous page) discussed is गानेन फलवैचक्षण्यमुपाध्यायतम). He stresses the point that gāndharva has mainly (प्राधान्येन) an adṛṣṭa phala while gāna has a dṛṣṭa phala; it gives rise to immediate pleasure. Now adṛṣṭa phala mainly accrues to the performer (प्रयोक्तॄन्). Mīmāṃsakas have called it कतुसमवायि; they say : क्रतौप कर्मणःअतिशयःकर्तरिनित्यमात्मश्रितः। (Mīmāṃsa Kośa—see apūrva)

However as the line stands, it may yet be made to accord with the context by being construed as : “the fruit of gāndharva (which is transcendental merit) is seen (i.e. known) to accrue mainly to the performer (as contrasted with gāna the fruit of which, being immediate pleasure, accrues mainly to the sāṃjikās).”

1 Also in the N.S. itself :

ध्रुवास्तु पञ्चविधास्तु नानावृत्तसमुद्भवाः। यथास्थानरसोपेता ह्यव तमाधममध्यमाः॥ —N.S. 32, 35.

Uttama, adhama and madhyama refers to ‘exalted’ ‘lowly’ and ‘middling’ characters appearing on the stage.

2 एतेन च वध्या अपि विधि प्रकृतिसंभवम्। जयेﬞष्ठानां वृत्तसंयुक्त्तम् कुर्यादारी तथैव च॥ —N. S. 32, 18.

also मध्यमंसमयः कार्या चतुरﬞथवसानिकी। मध्यमाधमानां कर्त्तव्या वृथा चैवावसानिकी॥ —N. S. 32, 23.

3 नानारसार्थयुक्ता नॄणा यां गीयते प्रवेﬞशे तु नाम्नि। विनेया सा ध्रुवा तज्ज्ञैः॥ ऋगमूलत-डघ विघ्निजैः क्रियते या हृदयेन नाट्यविद्यो। आलेपिकी ध्रुवासो हृता द्वितीया स्थिता वापि विनेया॥ याच रसांतरसुपगतामार्गे प्रवﬞशात् कुलं प्रसादयति। रागप्रसादजनन्या विद्यात् प्रासादिकी तां तु॥ विपण्णे; मुहूच्हिते ग्रान्ते वस्वमरणसंयमे। दोषप्रच्‍छादना या च गीयते सान्तरा ध्रुवा॥ —N. S. 32, 311-315.

Abhinava's clarification that antarā was employed to cover up a slip on the part of an actor due to physical over-exertion or mental disturbance or confusion. The antarā dhruvā—which must have been sung in the background, as the acter himself was momentarily disabled—evidently gave a performer time to control and compose himself and adjust his costume.1

Abhinava has given us interesting examples from current plays illustrating the actual use of the āksepikī dhruvā. These will give us a somewhat concrete idea of the function of dhruvās and of their purpose.

In the play, Udāttarāghava, where Rāma is the hero, a scene occurs in which the prevailing mood is that of love (prastuta śṛṅgārakrama); suddenly Rāvaṇa, a picture of fury, enters upon the scene, totally subverting the ruling ethos of śṛṅgāra and addresses Rāma angrily in a verse:

Stop hermit ! Go not now.

You have given me extreme pain

Humiliating my sister,

The mortification suffered by my kinsmen,

Khara and others, acts like the wind

Fanning the flame of my fury,

It can only be quenched

By the blood spurting from the breast

Of your mutilated frame.

Because of the fiery nature of this new intruding sentiment, the above verse was sung to a variety of ākṣepikī dhruvā having a fast tempo (ākṣipyamāna rasasya diptatayā drutā).2

In the well-known play Veṇīsaṃhāra, we are confronted at the beginning of the third act with Aśvatthāman, full of eagerness for battle and desirous of watching his father vanquish the enemies in the raging Mahābhārata battle. Suddenly a voice speaks from the background (nepathyā): “Alas! Where now is your father!” The voice was that of Droṇa's charioteer who had come to report of Droṇa's death in battle. This is an example of the sudden intrusion of poignant pathos in a scene dominated by the heroic sentiment (vīrarasa). On this occasion—as befitting

1 अनक्तुं यन्दनयाsऽधिकतन्वियादयलुब्धद्रुतप्रयोगश्रमसंच्छदौ । धामादिदोपसमभावनां । वस्त्राभरणावकाशादित्सया या नीयते सान्तरां ध्रुवा

—A. B. on N. S. 32, 315.

2 आक्षेपिकी । तत्क्षिप्यमानरसान्तरदीप्ततया द्रुता ।

यथा उत्तररामस्य प्रसृतशृङ्गारक्रमोल्लङ्घनन

अरे तपस्विन् स्वप्तरी (राः) भव । वन्दितान् न गम्यते ।

स्वसुरतं परामवप्रेप एकतत्सथ्यं ।

(वरप्रभृति बान्धवोद्वलनवातसंधुक्षितः ।

तवेह विदलितभवतुसुमुलच्छोभितच्छल

क्षराच्छुरित वक्षसि प्रथममेतु कोपानलः ॥)

इत्यादिना (रावणावचेन) ।

—A.B. on N.S. 32, 313

(A.B. does not contain the last two lines of the verse quoted)

the sudden introduction of Karuna rasa where Vīra was ruling—the ākṣepikī was rendered with a slow tempo.1

In the Kuṭṭanīmatam of Dāmodaragupta (8th Cent. A.D.), we find an extremely interesting description of a contemporary performance of the first act of Harṣa's Ratnāvalī. An account of it would be interesting in this context. The performance began with the flute player giving the sthānaka while the prāveśikī dhruvā was rendered as a dvipada in the rāga bhinna-paṅcama.2 Then entered the sūtradhāra.3 The sūtradhāra danced around the stage to the accompaniment of a dhruvā with appropriate tāla and rhythm set to eight kalās (an ancient measure of musical time). He called the naṭī to the stage and then began the time-honoured prastāvanā or prologue with the two characters, naṭī and sūtradhāra, discussing household affairs4. Next, having indicated the entrance of a leading character—in this case Yaugandharāyaṇa—the sūtradhāra moved a few steps with the grace of a dance movement and made his exit along with his wife to the accompaniment of the 'exit song'—which must have been the niṣkrāmikī dhruvā.5 At the end of the act, the hero, king Udayana, moved around the stage with graceful steps and took his exit along with the other characters. The niṣkrāmikī dhruvā was rendered during his exit.6

The purpose of the antarā dhruvā was to distract viewers from any lapse on the part of performers. It was thus not a part of the action; its words could never be pre-composed, for who could anticipate the occasion when an unforeseen break in the play would happen? Antarā was a dramatic clutch, so to say, on which the

1 यथा ('दूता') वाक्याकर्णनेन वीररस्याख्यस्य तु रसस्य भावस्य । 'सिस्वेति' विलम्बिता । यथा आध्वस्तस्याम्नो युध्दवीरे क्रोधोल्लसनन कुडुःङ्घापि ते तातः ; हति नेपथ्यश्रवणाद तस्य करुणररसस्य

— A. B. on N. S. 32, 313.

2 No pūrvaraṅga was, evidently, perfomed in the instance noted by Dāmodaragupta. Yet a pūrvaraṅga must have preceded the performance of the play when rendered in any full-fledged performance. Dāmodaragupta describes the play as presented by a dance-teacher (nartanācārya) and his troupe before a prince in a private showing. Only a single act of the play was presented, the idea being to reveal before the prince the artistry of the nartanācārya as a trainer and a director. The Pūrvaraṅga, which was never part of the dramatic plot as such, was in this instance not shown; but in a full-scale performance of the complete play, pūrvaraṅga, too, must have formed a part of it.

वाणिकदत्तस्थानक उद्ग्र्याहिताेभिर्नटचने सम्यक् । प्रावेशिकया द्वृया द्विपदे ग्रहणान्तरेविकारसूत्री ॥

—Kuṭṭanimattam, 880.

4 अट कलापोरमाणि दृश्यो पारिक्रम्य तालयुक्ताम् । आहूय नटो कुडृया तथा समस्तवर्गः कवेलादिमन् ॥

—ibid., 882.

5 मृचितपताकागमनः वि चदगतव्या पदान्यन ललितानि । निष्चितद्राम गृहिण्या सार्ध नि:सरणगीतेन ॥

—ibid , 883.

6 ......चितैः श्चरण्यैसः परिक्रमं कृत्वा निष्क्रामिकया घ दृया विनिर्ययो नायकौङ्ङपि सह सर्वे: ॥

— ibid. 927.

We meet with certain technical terms in the above description :

(i) sthānak is a term of uncertain meaning. It probably meant an accompaniment where the flute player indicated to the singer the proper pitch or sthāna of the note to be sung.

(ii) Bhinna-paṅcama was an ancient grāma-rāga like the bhinnaṣadja, an oft-mentioned form of this variety.

(iii) Dvipada secms to be the same as a dvipadī, used also by Kālidāsa for songs in his Vikramorvaś-īyam.

continuity of sentiment—if not of action—could rest and a break in the audience's state of absorption with the plot be avoided. Antarā was rendered with meaningless words ('śuṣkākṣara': somewhat like the syllabic formations that are used in a modern tarānā) because meaning was quite unnecessary and could only be harmful since any new lyric suddenly intruding itself into the play was bound to introduce elements of alien feeling or sentiment, foreign to the plot. One of the antarā dhruvās current during Abhinaya's time was known as the latikā: "Kevalaṃ chidrācchādanamātraprayojanāyaṃ na sārthakapadakadambayojanamupagītíśuṣkākṣaraireveyaṃ laksye ca latikādināmnā prasiddhā ca gīyata iti nāṭyadharmaprāyeyam" (A.B. on N.S. 32, 315).

Bharata has given many examples of verses that were sung to dhruvās in his days (N.S. 32, 47-151; 160-177: 181-235; 255-301). But these are loose verses ('muktakas') quoted outside the dramatic context and do not reveal the intimate relation between the dhruvā and the theatre as effectively as the examples from Abhinava and Dāmodaragupta do. Their illustrations show that plays in ancient times were incomplete without dhruvā. It is not surprising, therefore, that Bharata concludes the chapter on dhruvā with the statement that "one's very first effort should be directed towards the song, for the song is the bed on which rests the theatre".1 He stresses his point with a metaphor: "Just as the night is illumined by the stars, similarly the dhruvās rendered with due rasa at appropriate occasions lighten up a dramatic performance."

tathā rasakṛtā nityaṃ dhruvāḥ prakaraṇāsritāḥ nakṣatrāṇīva gaganaṃ nāṭyamudyotayanti tāḥ

(N.S. 32, 430)

The dhruvās have a modern parallel in the nāṭyapadas of the traditional Marathi stage and also in the Hindi films (to give a more vulgar example). These provide instances of a kind of music which, like the ancient dhruvā-gāna, is governed by the poetic content and the mood befitting the scene being depicted; musical forms in them are inspired basically by the fancy of the composer seeking a popular appeal; they are often derivative of older, more classical and rule-bound forms : this last observation being more true of Marathi nāṭyapadas and only rarely so of film-songs.

1 नीते प्रयत्नः प्रथमस्तु कार्यः। शय्यामाहि नाट्यस्य वर्दन्ति गीतम् ॥

—N.S. 32, 436.

CHAPTER V

GĀNDHARVA AND GĀNA

With the observations made in the last two sections in mind, let us turn to a more detailed examination of gāndharva and its distinction with gāna. Abhinava has taken great care to differentiate gāndharva and gāna from a comprehensive structural and theoretical view-point, both as to their form and spirit.

He analyses a fundamental four-fold difference (vilakṣaṇya) between gāndharva and gāna :

1. they differ as to the nature of their formal structures (svarupāt) : they are composed of a distinct set of svara, pada and tāla.

2. they differ as to the end of purpose that they serve and consequently also in the nature of their resulting reward or fruit (phalāt) : gāndharva has an adrṣṭa-end while gāna aims at an immediately pleasing aesthetic effect —a drṣṭa-end.

3. they differ as to the occasion of their application or employ (kālāt) : gāndharva is used in the pūrvaraṅga, gāna within the plot of a play.

4. and lastly, they differ as to their distinct functions (dharmāt).1

Abhinava has discussed these points of difference at length in the beginning of his commentary on the 33rd chapter of the Nāṭyaśāstra. Moreover, throughout his Abhinava Bhāratī, wherever the context touches upon gāndharva, he comments upon some aspect of its distinct nature.

Where distinctions between two forms are concerned, elements relating to their svarūpa or structure are the most fundamental. Musical forms consist of svara, pada and tāla : gāndharva and gāna were different in all three respects. We shall take up svara first, for this, as Dattila has said, was the primary element.

Abhinava's illustrations reveal some very basic differences pertaining to svara between gāna and gāndharva :

1. First and foremost : in gāndharva the rule governing the distance in śrutis between one svara and another had to be rigidly maintained : ṛṣabha was always on the third śruti above ṣaḍja, gāndhāra on the second śruti above ṛṣabha and so on. This arrangement of śruti-distances between svaras had a slightly altered form in the madhyama-grāma, but

2. तत्रैव गान्धर्वस्य वि लक्षणमुक्तमथ चतुष्टयेऽपि मुनिना । तथाऽन्यैरुनुसन्धानवन्ध्यो ध्यं महामागं बोधयितुमनुसन्धीयते । स्वरतालपदादिविश्रयातमकं प्रवर्त्तिनिर्वृत्तप्रधानदृष्टादृष्टफल-सामवेदप्रभवमन।दिकालनियतत्वमनःश्रयोपरञ्जनं गुणताविहीनं गान्धर्वमिति स्वरूपफलात् कालाद् धर्माच्च विधमानमवश्यं गानवैलक्षण्यं भेदैकसम्प्रदानम्

—A. B. on N. S. 33, 1.

here, too, the seven notes in the octave were formed according to a strict number of intervening śrutis. True, śrutis in an octave were recognised to be twenty-two yet only seven of these became svaras. Svaras were nine, if we include the intermediate (sādhāraṇa) svaras – antarā-gāndhāra and kākali-niṣāda—but these were quite weak notes and had no important musical function : they could not, for instance, become aṃśas or the main governing notes in a jāti; their use was restricted to extremely rare occasions. There were, thus only nine musically significant intervals or svaras recognised in gāndharva (with a slight difference in each of the two grāmas) and with these alone could svara-structures be formed.

Gāndharva evinces a very rigid selectivity; since other śrutis were also felt as musically expressive and were, indeed, copiously used in gāna forms. In gāna no rigidly fixed law regarding śruti-intervals (antarāla-niyama) between notes seem to have been honoured.

Abhinava illustrates this difference in the śruti-structure between gāna and gāndharva by referring to actual renderings of musical forms. Indeed, he appeals in this matter primarily to the evidence of actual practise. He remarks : “what I say is only an elaboration of what can be directly recognised. But for those who are not aware of theoretical or conceptual aspects, even facts known from direct experience can remain (unrelated) as in the knowledge of a child (who knows facts without perceiving their significance).”1 With this exhortation Abhinava proceeds to remark: In gāna profuse movements are made over kākali and antarā śrutis (i.e. intermediate śrutis between notes) and thus a large number of śrutis are utilized.2 He quotes the example of mālava-kaiśika which apparently had a greater number of svaras with a four-śruti interval than was permitted in gāndharva and further observes : “so large is the variety of śrutis utilised in the svaras with which rāga, bhāṣā, vibhāṣā, deśī, mārga and other forms (utilised in composing gāna) are rendered that any number of examples can be adduced.”3

In gāndharva, on the contrary, it was imperative that the ordinance regarding śruti-intervals which assigned each svara its well-defined place in the octave or its sthāna be faithfully observed, because this ordinance, says Abhinava, in its essence, formed the base for determining aṃśas (the main notes in a jāti) and thus was responsible for the different duration of performance that was to be accorded to different notes.4

1 उक्तमात्रं च प्रतीतममुच्यते । प्रतीतानामप्यलक्षणज्ञानां बालविज्ञानवदवैच्यम् । — A. B. on N. S. 33, 1.

2 गाने तु काकल्यन्तरस्वरितरङ्गणं विचित्र श्रुति ग्रहणम् । — Ibid., 33, 1.

3 स्वराणां मालवकैशिके न्युतितयैकृत्य दर्शयन्नातु कियता रागमार्गा विभावा देशीमार्गादिगतास्तानां स्वराणां श्रुतिविच्छिदयं भूमः । — A. B. on N, S. 33, 1.

4 किं चान्तरालनियमोऽतः प्रमाणस्थाने स्वरकलांष बधान्न (? वादत्) सारतया गान्त्यैचर्यसंपद्यते; नतैव गाने । The manuscript reading is faulty. The printed edition suggests a correction : ‘बधान्न (शाद) सारतया’ which is not happy, as it does not convey the sense of contrariness between gāndharva and gāna which is obviously intended here.

2. In their primary forms the jātis of gāndharva were constructed with seven notes. But they could also be rendered hexatonic (ṣāḍava) or pentatonic (auḍuva) by dropping one or two notes. Dropping of notes in each jāti was governed by specific rules and injunctions. One general rule was that in jātis of the madhyama-grāma, pañcama was never dropped; in those of ṣadja-grāma, dhaivata was indispensable while madhyama could never be dropped from any jāti of either grāma. Besides, specific injunctions (on the basis of amśa) regarding each particular jāti had also been decreed.1

But in gāna any note could be dropped at will to render a form ṣāḍava or auḍuva; there was no rule governing the process. The principal aim was the creation of the desired dramatic effect, the rest was left to the choice of the musician. Madhyama could never be omitted in gāndharva; but it was omitted in forms such as the bhinnaṣadja (a grāma-rāga) used in gāna.2

The indispensability of madhyama was a feature specifically gāndharvic. It seems to have been borrowed from sāmagāna. Bharata speaks of this characteristic as being peculiar to gāndharva and sāma :

sarvasvarāṇām pravaro hya nāśi madhyamah smr̥tah gāndharvakalpe3 vihitaḥ sāmasvapi ca madhyamah (N. S. 28, 65)

"Madhyama is the chief of all notes and is known as indispensable—it has been so decreed in the laws of gāndharva and in sāma songs."

Dattila, too, says, that madhyama is indispensable in both the grāmas (Datt. 20). Other ācāryas of gāndharva must have made similar injunctions.4 We shall later enter upon some discussions as to the nature of the central position assigned to madhyama in gāndharva.

3. A jāti could never drop more than two notes because an octave with less than five notes was not permissible in gāndharva. But in gāna a four-note octave was not unknown. Bharata states : "in avakr̥ṣṭa-dhruvā a four-svara formation is also employed."5 Abhinava comments : "the formations (of only

1 See the jātis, ch. IV.

2 लोपोनापि नियतगान्ध्वं दोषितो ग्रामधरमपेक्षेन च जात्याभपेक्षेन दर्शितः । गाने तु रक्यनुसारेण प्रवर्तते (वृत्ते) रसान्वितः । तथापि गान्धर्व यथादानशिल्पमेव समर्थितं तस्यापि मध्यमस्य नित्यपड्कालिनां लोप: —A. B. on N. S. 33, 1.

also

धैवत्योमान्तरादपि मध्यमस्यापि विनाशनप्रयोजनमुक्तं भवति । —Ibid. 28, 64-65.

3 The word कल्प is here explained by Abhinava in a somewhat different way. His remarks are interesting and we shall discuss them later in the chapter.

4 Viśākhila had stated that Pañcama in the ṣadja-grāma and dhaivata in the madhyama-grāma were indispensable notes. Regarding madhyama, he must have ordained that it was indispensable in both grāmas. This seems implied in Abhinava's following statement :

अन्ये तु धैवतपञ्चमयोगोप्रामोर्विभागो यदनाशिलत्वं विशाखिलादिमिरुक्तं नानुमन्यते मुनेर्मध्यम एवाविनाशि मत इत्याहुः: —A. B. on N. S. 28, 64-65.

5 पड्स्वरस्य प्रयोऽड्‌स्ति तथा पञ्चस्वरस्य च । चतुःस्वरप्रयोगोऽपि ह्वाक्रुष्टध्रुवास्विह ।। —N. S. 28, 77.

hexatonic and pentatonic octaves) is a binding rule pertaining to gāndharva alone; but in dhruvā songs, used in a dramatic context, a four-note octave is also employed." He adds that the avakrṣṭā dhruvā (with the four-note octave) was sung during pathetic scenes ; its metric form was dominated by long syllables.1

4. Both grāmas had consonances between certain notes, known as samvāda. Notes standing in the samvāda relation with each other were called samvādīs. Samvāda was based upon a natural harmonic and accoustic2 law : all notes standing in a relation with each other as that obtaining between the tonic and the natural fourth (ṣadja-madhyama-bhāva) or the tonic and the natural fifth (ṣadja-pañcama-bhāva) were samvādīs. The ancients expressed this by saying that two notes which were nine śrutis or thirteen śrutis apart from each other were mutually consonant or samvādīs.

In the ṣadja-grāma, ṣadja had a samvāda with pañcama; rṣabha with dhaivata and gāndhāra had a samvāda with niṣāda (these three pairs being 13 śrutis apart); ṣadja had another samvāda with madhyama (the two being 9 śrutis apart). The same samvādās obtained in madhyama-grāma except that there could be no samvāda between sa and pa (they being only 12 śrutis apart); but instead a samvāda here existed between rṣabha and pancama.3

A general maxim observed in gāndharva was that when a note in a jāti was taken as the amṣa, the dominant note, then its somvādi was not omitted in making the jāti auḍuva or ṣāḍava-pentatonic or hexatonic. Abhinava observes : "samvādinah tata eva na jätucillanopo kāryah; the samvādi should never be dropped" (A.B. on N.S. 28, 23); and again : "yena prakāreṇa samvādikramo na lupayate sa yatno gāndharve kāryah: one should endeavour to observe the maxim of not dropping samvādi notes in gandharva" (A.B. ibid.).

But in musical forms like rāga, bhāṣā etc., employed in gāna, the samvādi maxim was not binding either in rendering melodies ṣāḍava or auḍuvita. Just as one could, in gāna forms, move over intermediary śrutiis at will, similarly one could drop

1 'अस्तोति'......तमावसमानो नादस्वरोपेतनियमसमाह इहोच्यते । यदिदमं नाटचोपप्रयोगे नृत्तं चतुःस्वरप्रयोजनौषदि... (कु) वे त्याह । अवकृष्टतां करुणोपयोगिनीषु गुरुतरायामु ध्रुवासु ।

—A. B. on N. S. 28, 77.

2 Abhinava's demonstration brings out this element : वीणायां च पडजादिस्थोनिर्कं स्वरतरोत्पन्नयुक्तिकर्मविभूतिमाने तत्सम्प्रायर्मिति व्यपहारः |

"on the vīṇā if a note such as the ṣadja is plucked while stopping the string with another finger, then it (the samvādi) is produced on its own. This can be observed in practice." The process can, indeed, be demonstrated. On plucking many of the sympathetic harmonics of the note to which the string is turned are naturally sounded. The harmony of the natural fifth and the natural fourth being more prominent can be easily demonstrated as Abhinava describes.

3 ययोश्च नवश्रुतीकरणमन्तरं तावयोरेव संवादिनो । तथैथा पडजपञ्चमौ, ऋषभधैवतो गान्धारनिषादवदन्तो पडजमध्यम-

विति पहजग्रामे । मध्यमग्रामेऽप्येवमेव । पडजपञ्चमवर्ज पञ्चममध्ययोश्चान्तराद: |

संवादी मध्यमग्रामे पडचमयोः स च ॥ पडिजग्रामे तु पडजषष संवाद: पडचमस्य च ।

—N. S. 28, 22-23.

See also Datt. 18.

whatever note one deemed fit for creating the desired effect. Abhinava adduces evidence from actual practice in saying that “even with ṣadja as the amśa, pañcama is, indeed, dropped to render forms (born of the ṣadja-grāma) pentatonic”.1

5. The sādhāraṇas, the two intermediary notes of the gāndharva octave, were, we have observed, rarely used. When used they followed strict ordinances. Bharata gives an ordinance regarding the occasion of their employ : “there are three jātis”, he says, “where the learned have enjoined the use of the sādhāraṇa notes; these are : madhyamā, ṣadja-madhyamā and pañcamī. The amśas (which permit the use of sādhāraṇa in these jātis) should be known as ṣadja, madhyama and pañcama”.2 Two of these three jātis mentioned were renderable with a great number of alternate notes as amśas. In the ṣadja-madhyamā all seven notes in turn could become amśas and in madhyamā, five notes, excepting ni and ga were the permissible amśa notes. Pañcamī could be rendered with two amśas : ri and pa. In the first two jātis the sādhāraṇa notes could be used only when the ruling amśas were either ṣadja or madhyama or pañcama. Abhinava says : “ṣadja-madhyamā saptāmśā tatra niṣādagāndhārāyostāvadamśayoh nāsti sādhāraṇam; samapeṣu sātiśayastatprayogah, madhyamāyā apyeṣa vidhiḥ, sā hi dviśrutivarjam pañcamāmśāṃśā ” (A.B. on N.S. 28, 44-45). This may be translated as : “the jāti ṣadjamadhyamā can have all its seven notes as amśas. When either ni or ga is the amśa, the sādhāraṇa notes are not to be used. Only with sa, ma and pa (as amśas) is their extended employ permitted. This very rule applies also to the jāti madhyamā, which has five amśas, barring the two two-śruti notes, i.e. ni and ga.”

In the jāti pañcamī, when pañcama was its amśa and gāndhāra was dropped in order to render the jāti ṣāḍava, then the antara-gāndhāra could sometimes be used as an extremely weak note.3

6. Jātis were governed by a host of rules which enjoined the initial and final notes, the strong and weak notes and other structural elements. Further, the melody in every jāti followed a specific path characteristic of that jāti. This was known as the antaramārga (internal path); it manifested the rūpa of a jāti (N.S. 28, 75). The analoguous term in current Hindustani music is ‘chalan of a rāga’. ‘Chalan’ indicates the specific movements that manifest a rāga. In gāndharva, antaramārga followed a strictly codified course laid down in authoritative manuals (antaramārge’ pi vidhāyivargā gāndharvavedagītāḥ, A.B. on N. S. 33, 1).

1 माने तु साधारणत्वात् संचरः: श्रुतितः कौशिकदेव चेत्संनिधायास्ति ततः: संवादनपञ्चमंर्जन स्वरमोक्षपलब्ध इति संवादनन्तरं लुप्यतां नाम । तथा च पड्जादिष्वपदेश:, पड्जेर्जन्ये पञ्चमलोपकृतमोड्विकं भजत एव —A.B. on N.S. 28, 23.

2 स्वरसाधारणतांवस्थितो यो यास्तु जातयः । मध्यमा पड्जस्मध्या च पञ्चमी चैव सूरिभि: ॥ आसामंशास्ति विर्जों यः पड्जमध्यमपञ्चमम् । यथास्वं दुर्ब्बलतरं व्यवासायस्वतन्तु पञ्चमो ॥ Note A B. : तन्रासां कोऽंश: साधारणविपय इत्याह आसामंशा इति ।

3 पञ्चम्यास्तु षड्जपञ्चमावंशो आहो — वीतं ये पञ्चम एकार्यां काकल्यन्तरोमप्रसंगाद् स्यादन्यदाह 'दुर्ब्बलतरस्थ्यत्न्यासी' हुर्ब्बलत्वगत् कर्तव्यम् । तेन पाडवकारिणो गान्धारस्य स्थाने दुर्ब्बलेनांतरस्वरेण यत्न भाव्यं तत्रासीदुर्ब्बलतर: कर्तव्य: । ओडुवितकारिनिरपिादर्शने दुर्ब्बलताकलीसदृशो दुर्ब्बल: काकल्यो न कर्तव्य इति । —A.B. on N.S. 28, 45.

In gāna, antaramārga was not governed by rigid laws; creating an aesthetically charming movement was the primary aim of antaramārga here (śobhānurodha eva tatra karanam, A.B. ibid).

7. Ancient music—like most modern classical music—used three octaves. known as mandra, madhya and tāra. All jāti's moved over a span of three octaves but they were not permitted to move over the entire range at will; there were rules to restrict the movement of each jāti both in the tāra and mandra (N.S. 28, 70-71).

But forms like the grāma-rāgas, utilised in gāna, had no such restriction and a performer could move over mandra and tāra at will.1

8. We have spoken in the last chapter of the ancient tānas. Seven-note mūrchanās when rendered hexatonic or pentatonic became tānas. This rule alone can give quite a number of tānas. In the ṣadja-grāma there were seven mūrchanās; by dropping each of the seven notes successively from the seven mūrchanās, one could obtain 49 possible ṣādava tānas. Similarly in the madhyama-grāma, too, there could be 49 ṣādava tānas. The pentatonic process could give many more tānas.

Every mūrchanā in a single grāma, by successively dropping different groups of two notes, could give 21 tānas; thus seven mūrchanās could give 147 tānas for each grāma. There could then be possible the number of 392 tānas, ṣādava and auḍuva, in the two grāmas. But in gandharva, only 84 of these tānas were permissible : “tatra mūrchanāśritāstānāścatitih” (N.S. 28, 33).

Only certain notes or groups of two notes could be dropped from either grāma. For making ṣādava tānas in the ṣadja-grāma only sa ri ni and pa could be dropped and in madhyama-grāma only sa ri and ga could be dropped. Thus in ṣadja-grāma there could be only 28 ṣādava tānas and in madhyama-grāma 21, giving a total of 49 ṣādava tānas for the two grāmas : “lakṣaṇam tu ṣaṭsvarāṇām saptavidham ṣadjarṣabhiṣādāpañcamahīnāścatvāraḥ ṣadjagrāme madhyamagrāme tu ṣadjarsabhagāndhārāhināstrayah evamete ṣaṭsvarāḥ sarvāsu murchanāsu kriyamānāḥ bhavantyekapañcāśattānāḥ” (N.S. ibid.).

Similarly, for making auḍuva tānas only three groups of two notes could be dropped; in the ṣadja-grāma these were : sa-pa, ri-pa, and ni-ga and in the madhyama-grāma : ni-ga and ri-dha. These gave a total of 35 auḍuva tānas in the two grāmas : “pañcasvarāṇām tu pañcavidhameva lakṣaṇam—ṣadjagrāme ṣadjapañcamahīnāḥ, ārṣabhapañcamahīnāḥ, niṣādavadgāndhārāhinā iti trayaḥ; madhhyamagrāme tu gañ-dhāraniṣadadvaitinīḥ ṛṣabhadhaivataīh iti dvau, evamete pañcasvarāḥ sarvāsu mūrchānāsu kriyamānāstānāḥ pañcatrimśadbhavanti” (N.S. ibid.).

Thus there was in gandharva a total number of 84 tānas including both ṣādava and auḍuva.2

1 Abhinava commenting on N.S. 28, 70, where a general rule regarding tāra movement in jāti's is enjoyed, comments :

इहैत जातिय, ग्रामरागादौ नायं नियम इति विभावः

2 तत् मूर्छनाश्रिता इति। मूर्छनानाश्रिता यतस्तत्सामेव तेजःस्वाभाव्योषः। तत् सरिनिप इत्येतेपामन्यतमेन हीना: पट् (स्वर) मूर्छना: सप्त चत्वारिख़िे षट्के प्रस्ताराभ्यासतस्तानाः। सर्वाङ्गलयेतिहासमेतन्मेन होना मध्य(म) मूर्छनैकाख़िसतस्ताना इत्येकान्तपक्षान्त पाङ्चात्। द्वौद्र्विकास्तु पड्जग्राम एकविशति: अन्यत्र तु चतुर्द्दशोति पञ्च त्रिशदिति। उपयतः चतुर्गीति -A.B. on N.S. 28, 33.

Forms used in gāna were bound by no such rule. In gāna any note or group of notes could be dropped at will. In fact at times more than two notes could be dropped if desired as in the avakrṣṭā-dhruva.

9. Beside tānas, there were kūṭatānas—permutations of two or more notes in all possible disarray. We have seen in the last chapter, that the seven notes of an octave could be permuted and combined in 5040 ways. Using two notes one can have only two combinations; with three notes the possible combinations are six, with four notes 24; with five notes 120 and with six notes 720. Seven notes give 5040 tānas. In gāna all possible combinations, using as few as two notes upto all the seven could be utilized: “yadaivaṃ svarasvarūpasampattim gāne tu dvisvarāt prabhrti pūrṇasvaraparyantaṃ dvidhā śoddhā caturvimśatidha tathā vimśasaptaśatatadhā catvārimśatapañcasahasradheti kathitāniyā yaḥ kūṭanabhedastatkrtaṃ svaravaicitryaṃ varṇavṛttināmeva ca hyetat” (A.B. on N.S. 33. 1). No restrictive limitation obtained (“na hi iyattā kāciditi sthitam” A.B. ibid.). To quote Abhinava again : “melodies like grāma-rāgas and bhāṣās (used in gāna) could be manifested with ease (jhaṭiti) by rendering the notes of its mūrchanā in a successive and non-successive (i.e. kūṭatāna) order. For this reason musical forms such as the grāma-rāgas utilised mūrchanās and kūṭatānas in multiple combinations.”1

One of the most prominent features of gāndharva that emerges from Abhinava's cominents is that it used only seven svaras or at the most nine. This was, evidently, not because other intervals in the octave were not cognised as musically significant; for among the 22 śrutis others besides the seven svaras of gāndharva, were perceived as musically expressive and were used in gāna. The selectivity excercised in gāndharva must have given it a limited range of musical scales : amounting only to seven (or at the most nine) produced by taking each of the seven (or nine) svaras separately as the tonic.2

What was the reason behind this selectivity? Could it be that only certain musical scales, when rendered according to certain melodic laws, were considered as suitable for expressing the sacred and the holy ?

This appears plausible. It has been recognised since ancient times that different musical scales express and arouse different basic impulses in man. Plato stresses this repeatedly. In the Laws (Laws II, para 655) he states: “the figures and melodies which are expressive of virtue of soul or body or of images of virtue are without exception good, and those which are expressive of vice are the reverse of good.” Again in the following dialogue he reasserts: “Athenian: And now do we still hold to our former assertion, that rythms and music in general are imitations of good and evil characters in men ? What say you ?

1 cf. ग्रामरागाभावादि पुण्णत्ते संज्ञाणकान्तकयुक्तमुच्छ्थना प्रयोये झटिति रामप्राप्तः, अपुण्णत्ते तु रामतान प्रयोेगवर्जनम् । प्रत्येकं ग्रामरागादि सर्वेषि मुच्छ्थनादीनां वध्दा प्रयोग इति मुच्छ्थनानानात्वं युक्तम् । नेहि इयत्ता कार्चिदिति सिथतम् । —A.B. on N.S. 28, 34.

2 Notes within each of the grāmas were fixed at precise intervals and their fixed śruti-interval with one another could not be changed in gāndharva. The only way to arrive at different scales could be by taking different notes as tonics. The process can be easily demonstrated on a harp tuned to an octave with seven svaras.

98 A Study of Dattilam

Cleinias : That is the only doctrine which we can admit." (Laws VII, para 799).

Plato and his master Socrates were advocates of the virtuous and the just life. The qualities of virtue and justice, they thought, should permeate into every human activity including the arts. Music of all the arts occupied a primary position in Plato's scheme of things. Music was 'the tamer of the soul', 'guide to virtue', the art which softened the inner harshnesses and made man prone to the just life. Education in music was, according to him, a must for every budding citizen. But not all music qualified. Plato wanted to banish music which was not expressive of the positive values in life and human character from his 'republic' or the just State. There was in his State no need of 'lamentations and strains of sorrow' and the 'harmonies expressive of sorrow'. Such where the 'harmonies' known as the mixed or tenor lydian1 and the full-toned or bass lydian, and such like. Again, since 'drunkenness and softness and indolence' were 'utterly unbecoming', the 'soft or drinking harmonies', the ionian and the lydian (described as 'relaxed') were to be shunned.

Regarding melodies which Plato did want in his republic, he has Socrates utter: "I want to have one warlike (melody) to sound the note or accent which a brave man utters in the hour of danger and stern resolve; or when his cause is failing, and he is going to wounds or death or is overtaken by some other evil, and at every such crisis meets the blows of fortune with firm step and a determination to endure; and another to be used by him in times of peace and freedom of action, when there is no pressure of necessity, and he is seeking to persuade God by prayer, or man by instruction or admonition or, on the other hand, when he is expressing his willingness to yield to persuasion or entreaty or admonition, and which represents him when by prudent conduct he has attained his end, not carried away by his success, but acting moderately and wisely under the circumstances, and acquiescing in the event. These two harmonies I ask you to leave; the strain of necessity and the strain of freedom, the strain of the unfortunate and the strain of the fortunate, the strain of courage and the strain of temperance; these I say leave."2 These two harmonies which expressed the noble and the manly impulses were the dorian and the phrygian.3

The Greek 'harmonies' were apparently akin to our own rāga-forms; for Greek music was more like the melodic music of the East than the polyphonic music of the modern West. Each of the melodies such as the lydian, ionian etc. had a certain specific scale or 'thāṭ' or'melakartā' which was elaborated upon in accordance with definitive rules of formation and structure, as with our rāgas. The basic thāṭ of the lydian corresponded with our 'bilāval', of the ionian with 'khamāj', of the dorian

1 As in India, so in Greece many melodies were known by the name of the region or country of their origin. The lydian was from Lydia. The ionian from Ionia on the coast of Asia Minor.

2 Dialogues of Plato, Jowett, Vol I, pp. 662-63. Republic III, para 399.

3 "And these, he replied, are the dorian and the phrygian harmonies of which I was just now speaking." ibid.

with 'bhairavī' and of the phrygian with 'kāfī.1 Plato approved of the last two but

thought the first two as unsuitable.

We do not know why Plato wanted to banish melodies born of 'thāṭs' like the

'bilāval' and 'khamāj' from his republic; rāgas born of these thāṭs are not considered

by us as dissipative or effeminate in contrast to those born of the 'bhairavī' and the

'kafī' thāṭs. But to know Plato's reason one should be able to refer to actual Greek

musical renderings, which is not possible.

Yet one thing is clear: Plato gives expression to a kind of musical selectivity,

which in principle, appears analogous to that of gāndharva. The motive, too, for

such selectivity seems to have been analogous. Just as only certain scales were con-

sidered as 'ethical' by Plato, similarly only certain scales were considered as suitable

vehicles for gāndharva jātis, the purpose of which was to express the sacred and the

holy.2

In gāndharva-structures with a rigid seven-svara or seven-tone base, chromatics-

m were ruled out or delegated as subsidiary; this is clear from decrees ordaining

a limited use of the sādhāraṇa or intermediary svaras or tones. The reason apparently

was the fact that an indiscriminate use of chromaticisms is belived to lead to loose-

ness of musical structures and the resulting ethos. Plato in his Republic would have

nothing to do with multiplicity of notes and the resulting complex or panharmonic

(chromatic) scales. Thus after accepting only the dorian and phrygian harmoniai

(scales) as suitable, Socrates continues :

Socrates: Then I said, if these and these only (he is referring to dorian and phry-

gian) are to be used in our songs and melodies, we shall not want multi-

plicity of notes or a panharmonic scale ?

Glaucon: I suppose not.

Socrates: Then we shall not maintain the artificers of lyres with three corners and

complex scales, or the makers of any other many-stringed curiously har-

monised (i.e., tuned to strange scales) instruments?

Glaucon: Certainly not.

(Republic III, para 399)

1 "Taking the white notes of the piano as easy equivalent, the seven 'harmoniai' in the diatonic

genus are as follows (descending) : a to A (Hypodorian, later Aeolian); g to G (Hypophrygian,

later Ionian), f to F (Hypolydian); e to E (Dorian); d to D (Phrygian); c to C (Lydian) and b to

B (Mixolydian later Hyperdorian). The names of the 'harmoniai' were drawn from the differ-

ent districts of Asia Minor where the modes are said to have been in use. ...."

-Pelican History of Ancient Music, Vol I, p. 102.

2 If we scrutinise the seven mūrchanās of the two ancient grāmas which give us all the possible

scales in gāndharva musical structures, we see that both bilāval and khamāj were obtainable in

gāndharva. Thus, although Plato banished the Greek equivalents of these scales from his repub-

lic as too soft and dissolute, they were considered as suitable vehicles for expressing the sub-

lime in gāndharva. But scales alone are only raw material for musical forms, the same scale

differently formed can express very different impulses and sentiments-bhairavī, which Plato

approved of can be very light, sorrowful or dissipated when sung as a thumrī, but bilāskhānītodī

formed from the same scale has an entirely different ethos. So, unless we know the musical

formations through which Greek music rendered these scales in Plato's time, we can not have a

true idea of the basis of his judgment,

100 A Study of Dattilam

The motives behind the avoidance of chromatic, 'curiously harmonised', scales in ancient gāndharva and the music acceptable to Socrates and Plato were certainly analogous.

Gāna, unlike gāndharva, was guided by the principle of producing a pleasing effect—it was not governed by any ethical principle like the music acceptable to Plato or by the sublime and sacred like the gāndharva forms. All possible musical scales and complex chromatic tonal formations were permitted in it; even quarter-tones or notes with a single śruti-interval were part of the normal fare. Thus Vṛddhakaśyapa (in a passage quoted by Abhinava) says that "in rāgas and bhāṣās all the kākali and antara notes (i.e., microtones and half-tones) are permissible, and all notes whether with a four-śruti interval or a three-śruti interval or a two-śruti and even a one-śruti interval, should be used: "kākalyantarayogena catustrīyektaḥ śrutīn/svarān sarvān prayuñjīta rāgabhāṣāsu sarvathā"'(A.B. on N.S., 28, 35).

Gāndharva also differed from gāna in according to the note madhyama a unique position. It was in gāndharva a note that could never be dropped. We do not know why madhyama was the only note in gāndharva which was indispensable. Gāna forms observed no such scruple regarding madhyama.

Abhinava does suggest a reason for the indispensability of madhyama in gāndharva. Madhyama, he says, is the central note in the octave: above it we have the three notes, pa dha nī and below it, ga ri sa; it thus occupies a pivotal position between the upper and lower tetrachords.1 This, on the face of it, seems to present a solution to the problem. A musical system which places a fundamental importance on dividing the musical scale into lower and upper tetrachords, madhyama, the centre or pivot, can, indeed, tend to become indispensable. In the Greek musical system where the octave was split into two complimentary tetrachords, the upper and the lower, madhyama (Greek mese, Skt. madhya) apparently occupied an 'indispensable' position, for it was the note joining the two tetrachords. Aristotle thus calls madhyama, the 'central' note of the heptachord.2 Without this middle note, the mese, the two tetrachords could not be conjoined and hence the note was pivotal.

Abhinava's conception of the centrality of madhyama in gāndharva assumes the note to have been pivotal in the same manner as in ancient Greece. But a closer look will reveal that madhyama could not have been the central note in every gān-dharva scale and hence the argument that it was indispensable because it was the pivot between two tetrachords, cannot reasonably be maintained.

We observe that in order to arrive at different scales in gāndharya (where each note was fixed at an unmovable śruti-interval from another) the only way was to take different notes as the tonic. Now, with any other note but the ṣadja as the tonic, madhyama could not remain central. In Greek music, on the other hand, the 'mese' always remained stationary ; intervals between the other notes of the two tetrachords shifted and this gave rise to different scales. And so madhyama in this

1 For particulars of Abhinava's description see ch, IV, p. 86

2 See Greek-English Lexicon: Liddell and Scott, under 'mese'. For the role of the mese in Greek music, see Greek Music in the Pelicon History of Music, Vol. I

system was truly central. But in gāndharva the principle of scale formation was, evidently, different.

We can, however, conjecture that in the music of the sāma, from which gāndharva is said to have been born, madhyama did occupy a central position as in ancient Greek music. Yet by the time gāndharva was formulated into an independent system madhyama had lost its pivotal position. It perhaps remained an indispensable note as a legacy of the past.

Two other notes besides the madhyama were indispensable in gāndharva, though in a more limited sense: pañcama was indispensable in the madhyama-grāma and dhaivata in the ṣaḍja-grāma. The reason is again obscure. Perhaps there was no reason; it just so happened that in the limited number of melodies accepted in gāndharva pañcama was never actually dropped from jātis of the madhyama-grāma and similarly dhaivata was always found to be used in jātis of the ṣaḍja-grāma. The rule that these notes should not be dropped could be a generalisation of musical practice and not a rule governing structures. This is what Utpaladeva has suggested.1

Whatever be the reason for the indispensability of certain notes in gāndharva, gāna was not bound by these rules. In this, gāna was akin to current musical forms.

In tāla, too, gāndharva had many peculiarities specific to it. Abhinava says that tāla in gāndharva was 'an immutable rule-bound entity which measured (time) through a fixed number of demarcations; it united svara and yati with the instrumental vṛtti and occupied a secondary position (aṅgāṅgib­hāva) (to svara); its solitary aim was to establish a sāṃya or equipoise, the nature of which cannot be explained in words.'2 Abhinava, here points out three things: firstly, tāla-structures in gāndharva were just as much governed by imperative decrees as svara-structures. Secondly, he suggests that tāla was subservient to svara. And thirdly, Abhinava remarks that the goal of tāla in gāndharva was to establish sāṃya. This third point is important.

Sāṃya was a concept central to gāndharva tāla. Dattila introducing the nature and function of tāla says that "tāla results in sāṃya, and through sāṃya one attains fulfilment and perfection both in this world and the next: tālāt sāṃyaṃ bhavet, sāmyādiha siddhiḥ paratra ca" (Datt. 110). According to Dattila thus, sāṃya was not merely a musical function of tāla in gāndharva; when truly attained

1 श्रीमदुत्पलदेवास्त्वाहुः । जातिलक्षणे वद्यमाने मध्यमस्य लोपो न तु वाच्य इष्यत । तन्मध्यमस्य नाशो न कतॄव्य इति कि प्रतिपेधन । प्राप्तेरभावात् । सर्वस्वराणां नाश इत्यनेनापि न चिकित्सितं । नियमस्वरलोपस्य यथास्वं जातिपु वक्ष्यमाण-स्वात् ।

-A.B. on N.S. 28, pp. 64-65.

This may be translated as:

In laying down the characteristics of jātis, nowhere has it been stated that madhyama may be dropped in any of them. To what purpose, then, an especial prohibition here ruling that madhyama should never be dropped, since the possibility of (its being dropped) does not at all obtain (prāpterab­hāvāt)? Neither is it necessary here to enumerate all indispensable notes, for this, too, serves no purpose because in describing the jātis each note which is enjoined as dispesable is in any case laid down separately for individual jā­tis,

2 तालोऽपि गान्धर्वे नियतत्वेन संरूप्यते भङ्गजैन परिच्छेदोभयां यतिस्वरेभ्यां तिमेव मेलनमात्रदयोगम क्र्रियां भावव्यावृ-ध्यमानः साम्यमात्रलभ्यमिति न शक्यं वक्तॄम् ।

-A.B. on N.S. 33, 1.

it could result in transcendental fulfilment or adrșṭa. Viśākila (as quoted by Abhinava) has made an identical statement regarding sāmyā.1

Dattila does not define or expound the function and nature of sāmyā. But Abhinava often uses the term in relation to tāla. Some inference can be drawn as to its meaning from his account.

Abhinava says that percussion playing during the staging of plays was of three types, (1) that which depended upon varṇa : varṇānusāra, (2) that which followed the svara : svarānusāra and (3) that which aimed at attaining sāmyā : sāmyātmā.2

By varṇa Abhinava here denotes the range of sounds or 'bols' that were rendered on a drum (puṣkaravādya).3 These 'bols' could be played either softly (masṛṇa) or agitatedly (dīpta) or in a neutral manner (madhya). Though these varṇas were not denotative of any meaning, they could by their different arrangements and tempos (laya), arouse or manifest different states of mind. Drum-playing which utilised this expressive aspect was varṇānusāra drum-playing. When accompanying a dramatic song different varṇas could form a fit accompaniment for the different types of verbal structures in the accompanied songs. As a result tāla in the dramatic context could be directly helpful in augmenting the sentiment expressed in a song.4 In this manner of playing the drummer had, within limits, a free hand to show his virtuosity. Alternatively, drum-playing could be more subdued and concentrate upon following the music in all its contours (svarūpamanuvartamānam), playing when necessary and remaining still when the song so demanded (tadgānamśasthitasthāyitvam). This manner was imitative of a song hence svaranusāra5 Bharata

1 विशाखिलाचार्यः 'साम्याद्विघः सिद्धिः; परलोक' इति वदन् प्रतीयते ।

-A.B. on N.S. 33, 1

2 तत् पौष्कर वार्ष प्रकटयति नाट्योपयोगिनि (गीते) । वर्णानुसारेण स्वरानुसारेण साम्यातमना ताले च ।

-A.B. on N.S. 34, 36-40.

3 What Abhinava denotes through the term varṇa (lit. 'alphabet') has been termed akṣara (lit. 'letter') by Bharata. Bharata lists 16 akṣaras:

क ख ग घ ट ठ ड ढ त थ द ध म र ल ह इति योज्याक्षराराणीह;

-N.S. 34,39.

These were 'bols' like our dhā dhin nā tun nā kat tā tiṭa dhīta etc. that could be rendered on drums by variously forming the strokes with the hands and the fingers. (N.S. 34,46). Bharata says that these 16 akṣaras were to be variously rendered on the drums, some on the right side of the instrument while others on the left etc. (N.S. 34,41-42). Abhinava calls these akṣaras: दक्षिणस्य पुष्करस्य शब्दतो वामतश्चेष्टितं मुखडयम् । तस्य दक्षिणस्य दक्षिणपुबे पद वर्णाः । वामे द्वयः ।

-A.B. on N.S. 34, 42.

The 16 akṣaras were further combined with each other and with some vowels, resulting in a large variety of 'bols' (N.S. 34, 42-45).

4 वर्णाः करणमीचिदर्पि वाद्यविशेषलया प्रभवन्त्योध्वगामात्रार्श्रुतिमादलेध्विन् नगरग्राम्योपनागरिका रमरुचिर्नयाणि ताल (लय) भेदेषु दोषमध्यमसुरूपसर्वचि, तिर्व्यक्तिक इति तावत् प्रसिद्धमेव ।

-A.B. on N.S. 34, 36-40.

5 स्वरसंयोगांंश नादोभूतवर्णंसमुद्यनभाजि सदनुबन्धरूपं तदमानांशस्थितस्याविलकं स्वरूपमनुवर्तमानं तत एवं तद्गतरचन-वृत्त्यधीनं एव मूलरागस्वरमानाध्येतोति यत्तत् उपदिष्टंमानसरागानुरागाणां इति ।

-ibid. 34, 36-40.

seems to refer to this manner of percussion playing stating that "the (percussion) instrument should follow the metric arrangements of words in the song being sung: yadvṛttam tu padaṃ gāne tādrśaṃ vādyamiṣyate" (N.S. 34, 183).

We can form a more concrete idea of the above kinds of tāla-accompani ments indicated by Abhinava with refernce to current musical experience. We are well aware of the variety of expressions a drum is capable of. Depending upon the ‘bols’ played, their combinations, the type of attack (soft, staccato, agitated, loud, etc.), the tempo, the rhythm and other such factors, a drummer can lend lustre to

song or instrumental-playing by creating rhythmic counterpoints to the melodic line. This is, evidently, what Abhinava has called drum-playing according to varṇa: varṇā-nusāreṇa. We also know the numerous ways in which a skillful drum-accompanist can echo the melody being rendered by following and imitating the rythms created by the svara-structures. This is apparently what Abhinava meant by svarānusāra playing.

The two above types of renderings were used only in gāna. The ‘sāmyāatma’ rendering of tāla was common to gāna and gāndharva. Abhinava does not give a straight-forward description of it, but we can deduce this much that in contrast to the other modes, this manner of accompaniment consisted of a ‘neutral’ association of svara and tāla : where the tāla contained no overtones of expression to augment or highlight the mood in the accompanying musical composition. In other words, sāṃya was evidently attained when tāla kept the time by strictly keeping to pre-ordained beat-formations and patterns, while melody followed its own well-defined

forms; their association—to borrow a simile from Abhinava—was like that of two independent kings who had formed a union with the aim of defeating a single common enemy, and both retained their separate identity—śatrujvalanapravṛttāmārsātta-sābhimānanarapatidvayavat.1

We have experience of sāmyāatma or netural drum-playing, too, in some contemporary forms. A talā is fundamentally rendered on the drum through a certain given pattern of ‘bols’; this pattern being called (in Hindustani Music) the basic thekā of a tāla. Often the drummer accompanies the melodic line, set to a certain tāla, by just keeping to the thekā ; he does not attempt to create variety by playing colourful patterns of his own or by attempting to echo or recreate the rhythmic structures sung or played by the musician he is accompanying. This netural accompaniment is perhaps like ancient sāṃya playing. Such sāṃya is, in fact, the basic function of tāla: to demarcate rhythms and keep time.

The nature of ancient samya might become clearer if we consider yet another difference between gāndharva and non-gāndharva tāla. The very instruments used in rendering tāla in gāndharva and gāna appear to have been distinct. Tāla in gāna was played on avanaddha instruments. These were drums capable of producing a range of expressive sounds (vide N.S. ch. 34). They could thus provide an appropriate

1 The simile occurs in the A.B. on N.S. 4, 248-259 where it is used to denote the “neutral” nature of association between the dance in pūrvraṅga and its accompanying gāndharva music. In this association, too, both music and dance kept to their own ordained forms; no attempt was made to mould one to the other by deviating from set structures.

accompaniment to gāna-forms by svarānusāra or varnānusāra playing. But the distinct instrumental basis of tāla in gāndharva was apparently the 'ghana'. Ghana instruments consisted of cymbals generally made of bronze. ('kāmsyatāladikam ghanam', Amarakośa 1, 7, 4).

The very first verse of the Nāṭyaśāstra, with which the tāladhyāya on gāndharva begins (ch. 31, where topics regarding gāndharvatāla are expounded), connects tāla to ghana instruments.1 Abhinava rightly remarks that percussion instruments such as the various drums - through the pliancy and resulting inflections of the taut skin on which they are played - produce a number of notes at different pitches as well as a wide variety of different expressive sounds. Unlike these, a ghana instrument can produce only a single monotonous sound, unchanging in character; a ghana instrument can be used only to keep the measure in a tāla by being sounded at right intervals - it cannot be used for expressive accompaniment like a percussion instrument :

"tatra hanyata iti ghanah kathinataikarūpaḥ tata iva namannonmanaśaithilyādiyogā-bhāvadakṣaravaicitryam cānurudhya, mānāmatrenopayogī kāmsatālikādirucyate" (A.B. on N.S. 31, 1). The only function of ghana could be the attainment of sāṃya, a neutral 'balance' or 'equipoise' between tāla and svara-structures. Abhinava, therefore, in another passage, says : "ghana is so termed because of the hardness of its form; as a result of this hardness ghana instruments cannot play specific 'bols' which can express different emotive situations or sentiments. The only function of tāla on a ghana is to accomplish sāṃya : mūrtikāṭhinyena ghanam tadanusāreṇa rasabhāvo-payogi viśiṣṭavaranadvāreṇa ca nopayujyata iti sāṃyāmatropayogitvāttāla ityuktam" (A.B. on N.S. 28, 2).

The svara-structures in gāndharva, we have seen, were its main musical content; they were the object to be measured through tāla which was subservient to it: "geyātmānā meyena ca gāndharvatālāsya melanā" (A.B. on N.S. 33, 1; cf. Datt. 3, 5). The process resulted in sāṃya.

Yet the accompanying tāla was no ordinary measure. Our own tālas are basically (i.e., in their thekā-form) simple structures, outlining a rhythmic-cum-metric pattern, repeated cyclically: thus tritāla is a simple pattern consisting of four sets of quadruple beats 4 4 4 4 forming 16 mātrās-the third quarter is unstressed to give the 16 beat cycle a distinct character; jhap-tāla has 10 mātrās divided as 2 3 2 3 with the third quarter unstressed. The same principle holds for other tālas, even for tālas with as many as 60 mātrās.

Tāla-structures in gāndharva (namely, the seven gītakas) were, in comparison, extremely complex. They were not simple patterns, cyclically repeated. A single gītaka was composed of a group of distinct patterns with a set organisation of sound-ed and unsounded beats as well as tempic and rhythmic formations. It contained

1 यत्र तु यदघनं प्रोक्तं कलापातलयान्वितम् । कालस्तस्य प्रमाणं हि वक्ष्यं तं तालयोगतः ॥ -N.S. 31, 1.

"The instruments which are called ghana, along with topics such as kalā, pāta and laya should be known in connection with tala- (musical) time forms their measure."

numbered repetitions of smaller or larger patterns within greater patterns; introduc-

tions of fixed rhythmic irregularities, and other such features : all pre-decreed and

formed according to gāndharva manuals (we shall see these complexities when we take

up the gītakas in detail).

Current music has no tāla-structures analogous to the gāndharva gītaka. An

analogy from geometry might make the difference clearer : tāla in current practice is a

'regular' structure-each different tāla may be'compared to structures like the square,

the round, the triangle, or even the hexagon or the pentagon. Tāla in a gāndharva

gītaka was not made up of a single regular structure but was, evidently, a conglomerate

of many different structures arranged into a single whole.

Keeping to the strict time-demarcations, tempos and yatis (tempic regulations)

were a must in gāndharva; for only thus could sāṃya and the resulting siddhi (or, in

other words, adṛṣṭa) accrue. Care was being execised for this purpose. A singer

could, in principle, himself keep time through proper attentiveness (avadhāna) and by

marking tāla with bis hands, yet in order to keep a check on the singer (who was

after all capable of human error) and also to allow him to sing at ease, it was neces-

sary to appoint someone to keep time with a cymbal.1 Gāndharva had a specific function

in the pūrvaranga of ancient dramas where it was used to accompany certain specific

dances. The aim of the entire ensemble was the attainment of adṛṣṭa. Abhinava

says that “if in this ensemble the singer alone were to keep time, a procedure which

involves unsounded and complex gestural beats like āvāpa and vikṣepa, then the ins-

trument players and the dancers will not be able to keep an eye on the singer and will

render the tāla imperfectly in a slip-shod, some-how-or-other manner (kathāñciccaranti);

thus 'sāṃya' will not be attained”.2 Adṛṣṭa, consequently, will not accrue. A singer

could make sounded beats through clapping his hands, but this could not be so clearly

audible as the sounds produced from cymbals: “saśabde' pi pāte karaśabdasya tathā

sphuṭamākrṇanāṃ na bhavati yathā ghaṇaśabdasya” (A.B. on N.S. 33, 1). Therefore

cymbal playing with its neutral yet audible sounds was employed during gāndharva

for a better attainment of sāṃya.

Tāla in a gāna, however, served various dramatic functions. Not only could

it expressively accompany songs, on its own, too, it could be quite effective. In the

1 This view regarding the purpose of cymbal playing during a performance of gāndharva has been

asciibed by Abhinava to Viśākila:

विस्फ़ारिकावाद्ययोगस्यनुपूर्वी तथा ललितं सुसंपादयितव्यं यत्नेन कृत्वाऽपि पृथग्वे तालकलानां विच्छेदे वाऽवधानवस्तामात्र-क्रियया आवापादिक्रिया च सम्पाद्या प्राणीतत्वतः गातः: प्रसिद्ध परिपालनप्रयोगः क्रियतां महताम् ।

Taking his cue from this or from a similar ancient account, Ṣārṅgadeva says:

गान्धर्वमार्गकुशलः: कास्यातालधरो नरः । यातुः सहायः व तत्स्थः: प्रमादविनिवर्त्तये ।।

• S.R. 5, 38-39.

Kallinātha explains:

अयमार्थः- गान्धर्वस्यायतनियतत्वेनाङ्गत्वफलसाधनात्। गान्धर्वप्रयोगे मुख्यो गाताऽत्र नियमादिक कुर्वन् गायेत् ।

दितीयो गाता कांस्यातालधरः स्यात् यथा प्रमादो न जायेत तथा महयेताः: साहाय्यं कुरीत।

2 यथा पूर्वं ज्ञातादृष्टसिद्धं संयतगीतकर्द्धमानादि प्रयुज्यते । यो गायको यतिनामा (वाङ्) न वेत्ति नर्त्तक-

वादकाद्यस्तद्बलोकनं विना तत्वप्रकाशसंं लय कयाचिच्चरित(त्ति) । प्रयोक्तॄ साम्यं न त्यात् ।

-A.B. on N.S. 33, 1.

theatre as envisaged by the ancients, different actors according to their various social status as [uttama, madhyama and adhama and their distinct dramatic functions, moved with different stylized gaits. Their movements were accompanied by suitable sounds on the percussion.1

Bharata has given a list of various percussion ‘bols’ which typified movements of various characters and other specific movements made on stage : a king moving with a natural, spontaneous, unagitated grace (rājñāṃ svabhāvagamane) was accompanied with guru-syllabic ‘bols’ played as ghem tāṃ kem tāṃ khem tāṃ etc. (N.S. 34, 153); whereas when a servant moved he was accompanied with a mixture of guru-laghu bols—tham ke tāṃ ke ṭa kinā etc. (N.S. 34, 162) which by their very rhythm signify a restless sprightly movement; other characters had similarcharacteristic bols (N.S. 34, 150-170).2

Tāla in gāna was a flexible thing; it could improvise patterns to create a desired effect. Contrary to this, in gāndharva no deviation from the set codes was permitted and sāṃya alone prevailed and denoted an equipoise of co-formation between svara and tāla : svara never going out of tāla and tāla setting off the rhythmic demarcations in svara, each keeping to its own enjoined forms. Abhinava has remarked : “tāla in gāndharva is formulated with pre-regulated demarcations numbered in their measure and fixed in their method of indicating time (as in āvāpa, vikṣepa etc.); in it a concurrent occurrence of set tempic regulations (yati) together with svara-structures is all that constitutes a (svara-tāla) association—for in it ātodya or drum accompaniment is not used and tāla is not a subservient element to the melody as a limb is to its principal structure; its only function is the attainment of sāṃya : tālo’pi gānd-

harve niyatatvena saṅkhyāparimāṇaṃ bhañjanaṃ paricchedopāyaṃ yatisvarairvṛt-timeva melanamātodyogamaṅgāṅgibhāvavyavaruḍhyamānaḥ sāmyamātraphalamiti” (A.B. on N.S. 33, 1).3

The effect of these pre-ordained tāla-measures with their set intricate patterns sounded on cymbals and ritually indicated through gestures, and accompanying the equally well regulated gāndharva melody must have been a solemn and sublime experience. Abhinava, who had evidently heard this music, remarks that the sāṃya resulting from the association of tāla and svara in gāndharva, was beyond description: “sāmyamātra phalaṃ na śakyaṃ vaktum” (A.B. ibid.).

Many other more specific distinctions of formal detail can also be noted between tāla in gāndharva and gāna :

1. Bharata points towards one when he lays down that the three gītis—three ways of singing words on the basis of three different patterns of laya or tempic arrangements—were applicable in gāndharva alone and not in dhruvā (N.S. 29, 49).

2. Tāla, we have seen, was marked by a set of sounded and unsounded beats. The unsounded beats evidently performed more of a ritualistic rather than a purely

1 गाने पुनरुक्तमादि भेदभिन्नप्रकृतित्वमतययोज्चितमसूणमन्थररोदितादिमेदभिन्नसर्ïत प्राणवल्लन..... -A.B on N.S. 33, 1

2 See also in this connection page 90 of the informative article by Dr. Raghavan entitled “Music in Ancient Indian Drama”, J.M.A., Vol. XXV.

3 See Appendix for a discussion of the role of drums in gāndharva.

tāla-like function of indicating time. Hence their formation was not necessary in

gāna forms.1 The unsounded and sounded beats were of equal importance in

gāndharva and were set to specific pre-ordained patterns and arrangements. Gāna

followed quite another principle in its formation of beats and rendering of tāla.

Abhinava has remarked that "in gāna as well as in gāndharva, demarcation of time is

impossible without the help of certain actions (like formation of beats). There is,

however, this distinction : although both in gāndharva and gāna, time-demarcation is

sometimes made with an action common to them (kutracit paricchedye paricchedavi-

dheyasya sāmye'pi), yet a set of distinct actions like the folding of certain fingers are a

necessary part of the conditional causes which effect damarcation in gāndharva (paric-

chedahetutvenāśrīyate), because only thus can transcendental merit (adrṣṭa) be acquired.

The function of time-demarcation itself is common to gāna and gāndharva, but

demarcation alone is in gāna the purpose of this function and any action or kriyā

whatever can become the basis of effective demarcation provided it is in concurrence

with the melody sung or played".2

3. The basic time-units which formed the foundations of time-measures in

tāla were quite distinct in gāna and gāndharva. The basic unit of time in gāndharva

was called the kalā, itself measured through the mātrā. Mātrā is a prosodic measure

and is defined as the time taken to pronounce a laghu-letter formed with a short vowel

such as 'a', 'u', 'i', 'ṛ'. In gāndharva five such mātrās constituted the basic standard

time-unit: "cītakālāsya paricchedakālāntaram pramāṇam pañca nimeṣā hrsvākṣaram

pañcakam yāvat, kakhagaghana iti" (A.B. on N.S. 31,4).3

In dhruvā-gāna such a time-unit was unsuitable because a dhruvā was gover-

ned by its lyric, and the lyric in turn was governed by its metric structure consisting

of prosodic divisions according to guru or laghu syllables. Tāla had to follow the

metric pattern, for it was subservient to the lyric. The gāndharva time-unit compris-

ing five laghus could not remain the basic time-unit for this would have been too large

a unit to be suitable. The standard time-unit was thus the smallest metric unit.

1

ध्रुवागाने तु दृष्टफलं गायनस्यैव सोप्तु व्यपारः; । अवापादीनां तत्तानुपयोगमात् ।

—A.B. on N.S. 31, 1.

2

गाने गान्धर्व वा क्रिया विना परिच्छेदो न सम्भवद्येव । एतावान्तु विशेषोऽस्ति सङ्क्षेपतः पदगान्धव कुत्रचित् परिछेदये

परिच्छेदो विधेयस्य साम्प्रत्यसूक्तया क्रियया काचिदेव क्रियया कवचित् संधीयते,व्यपदेशे वत्तमानो परिच्छेदहेतुत्वेना-

श्रिमते तद्वाभिव्यक्तः(ग) गाने तु तद्वारा(दृष्टं)() गाने तु तत्कार्यमात्रस्य साम्प्रदायात् तद्द्वारेण सम्पादनस्योक्त(तत्) रानिदानत्वं साध्यत्वाद् या

काचित् क्रिया गानवादन संवादिनी परिच्छेदहेतुत्वेन हि इति ।

—A.B. on N.S. 31, 1.

See also:

तत्र चिरशीकृतादो क्रिया यच्चापि या काचिदप्युपायस्तथा नियमादृष्ट (फल) सिद्धये विरशष्टहस्ताङ्कुलिखित्येयोपयोगिनी

—ibid. 29, 372.

and

तत्र क्रिया चात्र परिच्छेदोपयोगिनी युक्तक्रमस्माभिः। साच गान्धर्वे विशिष्टदृष्टैवेत दर्शनात् ।

—ibid. 31, 30-32.

cf. मात्रातु लौकिकी नेहू किन्तु पद्यगुणा ततः;

—Dattila. 117.

For details see 'Kalā', ch. IV.

Bharata has said that "the kalās and components contained in gītakas, should be rendered according to the metric units and divisions, when transformed into dhruvās".1

Abhinava here remarks: "Kalā and similar elements (in dhruvās) should be rendered on the basis of metre and syllables of prosody (tena kāḷādayo yāvacchandogītai-reva vṛttairviracyante).

In dhruvās, the metric-unit was one of the pivotal and vital aspects of their structure. It was one fundamental point through which one dhruvā was distinguished from another (N.S. 32, 35-45).

Different syllabic arrangements were thought to be aesthetically suited for different emotional situations (as has been competently analysed by the ālaṅkārikas).

Bharata thus says that the avakṛṣṭā dhruvā when expressing grief should be composed largely with guru-syllables.2

Conforming to the metric structure was, for this reason, essential in dhruvās. Distortion in metre could distort the purport of the poem sung and defeat the very purpose of a dhruvā.

In gāndharva, too, the hymns or psalms that were sung, had a metric base. But in singing them the metric arrangement of words or letters was not necessarily honoured.

An immediate communication of meaning was here unimportant and the words could be dragged or compressed as decreed.

4. In gāndharva, formations of tempo, yati (regulation of tempo) and similar elements had to be made as prescribed, for gāndharva was used for a ritualistic propitiation of a Deity: "gāndharve devatāviśeṣaparibhäṣāṅanusāreṇa vinoyogo" (A.B. on N.S. 33, 1).

Sudden reversals of laya and introductions of abrupt yati-formations were not unknown in gāndharva.

These had to be observed as enjoined.3 Gāna was not controlled by ritualistic ends and, therefore, in it, abrupt changes in the flow of tāla were not introduced on the basis of decrees enjoined with a view to adṛṣṭa ("na yaterākasmikatvenādṛṣṭamātraphalālatā", A.B. on N.S. 33, 1).

5. Parivarta (lit. repetition) was a term signifying repetition of a smaller or larger portion of a tāla-structure.

In gāndharva parivartas, too, followed a time-honoured (kālāyātarūpah) code.

Thus in vardhamānaka the parivarta of certain of its components followed a well regulated order.4

In dhruvā gāna, parivartas did not

1 यान्यूनानि कलाश्चैव गीतकान्तर्गतानि तु । तानि छन्दोगतैर्‌वृत्तैर्विभाव्यन्ते ध्रुवास्वयम् ॥ -N.S. 32, 14. See also N.S. 32, 3: एवंस्तद्भिः(न्यथोक्तैर्‌वृत्तानि) नानाच्छन्दःकृतानि तु । ध्रुवास्वेव यानि गच्छन्ति ।

2 Bharata says that the avakṛṣṭa when composed of guru-syllables, should be sung to triple time (ojakrti); such dhruvās express grief: गुरुप्रायावकृष्टं तु ध्रुवं त्रिगुणमस्मृतम् । वेदनोन्मादयोः शान्तं ध्रुवं यत्नेन गादि च ॥ -N.S. 32, 40-41.

3 गीतकादावपि मध्ये स्तोत्रगतलयादिरिल्यादिकं नान्यथा क्रियते । -A.B. on N.S 4, 248-259. also : नियतक्रमशास्त्रेऽपि मध्ये स्तोत्रगता यति (तिर्‌र्‌) त्यादावदृष्टोपयोग प्रधानत्वे -ibid , 33, 1

4 परिवृत्तिरपि विशिष्टा सङ्क्ता चेत्यादिवृत्तं भेदे च नियतैव । -ibid., 33, 1. For details see 'Vardhamānaka', ch. IV,

follow an obligatory or imperative rule because as Abhinava says: “gāndharva iva dhruvāgāne parivartanāmadṛṣṭārthetikartavyatarūpābhāvāt” (A.B. on N.S. 5, 86-92).

The above observations, we think, suffice to reveal that as in melodic structures, so in tāla, gāndharva had certain very distinct features, characteristically its own. Let us further discuss the more important features.

We have seen that the very acts or actions set for demarcating time in gāndharva were as important as the demarcation itself. This was, strictly speaking, a non-musical element. Certain actions were incorporated as a ritual—not surprising in a body of music which was basically sacred. The ritualistic actions constituting gāndharva tāla, appear at the first glance to be somewhat bizarre. We are, today, quite unfamiliar with such gestures being part of any existing musical form in India. Our tālas are all like gāna-tālas : rhythmic demarcations are the only aim of all the actions made by the drummers or those who keep time, and no fixed set of gestures are prescribed. If under the sway and forceful impulse of rhythmic patterns, players sometimes make curious bodily gestures, these are not part of the tāla, but involuntary reflections of the impact of rhythm; which with some players becomes a mannerism resulting in a stylised gamut of gestures. But none of the current gestures are anything like the prescribed gāndharva gestures of āvāpa, vikṣepa etc.

But once we realise that gestures in gāndharva tāla were ritualistic, it would not be difficult to discover likenesses in the great repertoire of ritualistic gestures that form part of our pūjā-viniyogas or rites of worship. In the liturgical viniyoga of any japa or homa or in the proper ritual (sopacāra) propitiation of a deity, we use many gestures, both of the hand and of the fingers, which are quite similar in nature and form to those of gāndharva. Gāndharva, in all probability, must have borrowed its set of gestures from some already existing body of rituals.

If so, the question arises, from which cult ? The answer perhaps is : from Śaivism. We see that there exists a strong Śaiva influence in gāndharva : Śiva is its chief deity, all the hymns in it are addressed to Śiva and its complimentary dance forms, the tāṇḍava and the lāsya, were created by Śiva and Devī.

Sāma singing, too, had many ritualistic gestures as part of the music. Every note was indicated through a fixed gesture on a specific finger : the note kruṣṭa (pañcama) was indicated on the tip of the thumb, the note prathama (madhyama) at the root of the thumb, gāndhāra on the forefinger, ṛṣabha on the middle finger, ṣadja on the ring-finger, dhaivata on the small finger and niṣāda at the root of the small finger.1

This ritualistic practice is still observed by traditional sāma-singers.

1 अंगुष्ठस्योपरि क्रुष्टो हुयंगष्ठे प्रथमः स्वरः । प्रदेशिन्यां तु गान्धारो मध्यमायां तु पडचमः ॥ अनामिकायां ऋषभस्तु कनिष्ठायां तु धैवतम् । तस्याधस्ताच्च योऽनुत्तमनिर्वाद तत्र निषादयेत् ॥ —Nāradi Sikṣā, 1,7, 4-5. also Māṇḍūki Sikṣā: बाह्यांगुष्ठतु क्रुष्टस्वरत अंगुष्ठे मध्यमः स्वरः । प्रदेशिन्याम्तु गान्धारो मध्यमायां तु पडचमः ॥ अनामिकायां ऋषभस्तु कनिष्ठायां तु धैवतः । तस्याधस्तात्तु योऽनुत्तमस्यानिर्वाद इति तं विदुः ॥

There is some controversy regarding the exact denomination of kruṣṭa. But scholars generaily agree in calling it the sāmic equivalent of pañcama. See Bh. Sang. Iti (Bengali), Vol. 1, p. 117. Also Dr. Paranjape's Bh, Sang. Iti (Hindi), p. 95-96, who ably discusses the problem.

Abhinava argues a link between the gestures indicating accents in Veda and those made to indicate tāla in gāndharva (A.B. on N.S. 31, 40-50; for details see ch. IV, p. 81). But though both had ritual gestures the purpose was not common; for in gāndharva these gestures were part of tāla and not svara. Yet, gāndharva had certainly borrowed many element from Vedic forms and its propensity for ritual and sacred gesticulation was certainly a legacy from the past, even if the context of the employ of rituals had changed. One may conjecture that tāla in sāma-music also had gestures and these were passed on to gāndharva. But little is known of tāla in sāma music and we do not know if this had its peculiar gāndharva-like gestures. Ancient works are silent as to the nature of tāla in sāma singing and the surviving tradition, too, offers no clue. Whatever little survives of sāma singing is always unaccompanied by tāla. And it is likely that sāma was always free of tāla accompaniment. Had tāla been an important aspect of sāma, indications regarding its form would certainly have been noted by the Prātiśākhyakāras and the Śikṣākārās.

A distinct feature of gāndharva tāla was the peculiar nature of the instrument used. Tāla-measures were demarcated chiefly by sounding bronze cymbals.

The association of cymbals and gongs with religion and worship—with the sacred—is, indeed, an ancient phenomenon. To this day in India, the jhāñjha and manjīrā (large and small bronze cymbals) are specifically connected with music in a religious or devotional setting. In the West bronze bells are a part of every church. Bell-playing has been turned into a high art. There are musical forms played with church bells such as the carillons: a set of bells placed on a church tower and played by means of a keyboard or another mechanical device. These bells, 30 to 50 in number, are tuned to the twelve-note chromatic scale like the paino or the harmonium and a number of compositions are played on them.1 Another sophisticated technique of bell-playing is called change-ringing. In change-ringing a set of bells is rung by a set of men, one for each bell. The bells are rung in a methodical order according to certain schemes of arithmetic permutation. For instance, a set of three bells 1 2 3 may be played six ways without repetition: 123, 132, 231, 213, 312 321 (this is not unlike our own khaṇḍameru). Standard combinations with the help of a number of bells are known under traditional names such as "Grandsire Triple", "Treble Bob" etc. The art is still widely practised in England.2

In India, too, jhāñjh playing has been turned into a great art demanding considerable skill and training, especially in certain parts of Gujarat. But generally, cymbal accompaniment is of a simple and uninvolved nature—it consists merely of sounding the instrument at appropriate rhythmic points to demarcate time. This is how cymbals appear to have been played in gāndharva, too; for as we can infer from Abhinava, the purpose of tāla in gāndharva was to provide just the tāla; colourful and complex rhythmic playing was the province of gāna.

As to the size, shape and nature of the cymbals played in gāndharva, Bharata affords no information, neither does Abhinava. Śārṅgadeva, however, has a section

1 See the Harvard Dictionary of Music, under ‘Carillon’.

2 See ibid., under ‘Change-ringing’.

on cymbals. He, too, was aware of the importance of cymbal-playing in gāndharva (S.R. 5, 38-39) and his statements regarding cymbals may be of interest. Cymbals, he says, were to be made from properly heated and tempered bronze (ghanavādye syātkām-syamgnau suśodhitam ; S.R. 6, 1170). They were to be circular, of equal measure, 13 angulas in diameter (other sizes were also possible) and similar in appearance and size to the leaf of a water-lily.1 At the centre of each flat cymbal disc was a regular dent, one angula in measure in the middle of which was a tiny hole smaller than a guñja berry2 (randhram ca madhyagam pādona-guñjamātram ; S.R. 6, 1171). Through these holes was passed a thread woven from a special cloth, the ‘paṭṭavas-tra’ (perhaps silk) and the two cymbals were secured together at an appropriate distance.3 The method of holding and playing the cymbal discs was as follows : one of the cymbals was held with the spread out left palm, the grasp being secured with the help of the thumb and the fore-finger ; the other cymbal was similarly held in the right hand, but further secured by twisting the attached thread around the thumb and ring-fingers. The left cymbal was held facing upward and was struck in the middle by the right cymbal to give the desired sound (S.R. 6, 1174-77).

The method of holding and playing cymbals as described by Śārṅgadeva sounds very familiar to us, for cymbals are played in an almost identical manner to this day. The tradition was presumably quite old and perhaps, during the performance of gāndharva, the method of playing was the same.

We do not know if the gestures described for rendering the unsounded beats like āvāpa and vikṣepa etc. were indicated by the cymbal-player or by the singer himself. It would seem more practicable that the singer made these gestures. The cymbal player with two large discs in his hand could not have effectively made them, as his palms and fingers were not free enough to make the appropriate actions. The likely arrangement seems to have been that a cymbal-player marked time in its prescribed demarcations while the ritualistic gestures were made by the main musician. Śārṅgadeva, too, implies, this when he says that the “cymbal-player was to act as an assistant to the gāndharva-singer so that mistakes in tāla could be avoided” (S.R. 5, 38-39).

We have seen that gāndharva had a much larger basic time-unit than gāna. The reason was that gāna was bound by the metric or prosodic unit of the lyric to which it was composed while gāndharva was not. Movement in gāndharva, thus, must have been much slower than in gāna. This befitted the sombre, solemn and sacred character of gāndharva.

1 नलिनीदलसंका्शो कांस्यतालो समाइचती । तयोदरशङ्कलो वक्त्रे कांस्यजे इयत्तन्न तो तलै ॥ —S.R. 6, 1181.

2 Guñja is a red and black tiny berry, abrus precatorius, which forms the smallest of the jeweller’s weights. Sanskrit English Dictionary, Monier Williams.

3 नेत्रवस्त्राच्छादयेत् । रज्जुकृत्य निवेशयेत् । रन्ध्रे डोराणाम निःसृत्ये प्रन्थिं च सज्जयेद दृढम् ॥ —S.R. 6, 1173.

Sudhā comments : नेत्रे वस्त्राच्छादयाणि पट्टवस्त्रविच्छेदानि, तैः रज्जुकृत्य रज्जुं विधाय रन्ध्रे निवेश्य अग्निमार्गनिगत्ये प्रन्थिं कुर्यात् ।

The large time-unit in gāndharva also shows that it was a more 'musical' form than gāna, in the sense that its rhythmic structure was not dominated by any extra-musical consideration such as the metric structure of the poem being sung or the necessity to adhere to the syllabic-form of the word content. Musical-form was, in it, the first and foremost consideration and the words sung had to conform to it.

There was also a very fundamental difference between gāndharva and gāna on the basis of pada or words. In gāna, pada dominated over svara and tāla. Dhruvā songs were an integral part of the ancient drama and helped in furthering the plot or the sentiment in a play. Abhinava says that a good gāna should merge its own distinct voice in the (orchestra) of the various elements constituting a play (gānam tu nātyasāmagrīmadhyanimajjitanijasvaram saphalāyeti, A.B. on N.S. 33, 1). In it the lyrical or the poetic rather than the musical element had the upper hand :"gītisāratvāc-cedam gānamiti" (A.B. ibid.).

In his chapter on dhruvās (ch. 32), Bharata pointedly indicates the subsidiary position of pada in gāndharva. He states: "In the gāndharva form which I have already delineated (in chapters 28-31) pada acts only as an aid for projecting svara and tāla (svaratālānubhāvakam)."1 Bharata wanted to signify that in gāndharva, pada served only as a peg on which svara and tāla were hung. Indeed, Abhinava comments thus on the above verse: "these latter two (i.e., svara and tāla in gāndharva) cannot be formed on their own and need a prop on which to stand. This is provided by pada."2

We have an analogue—though not ruled by the same spirit—in contemporary classical singing. Here, too, purely musical elements, svara and tāla, are the primary concern of the artiste; the word sung, the pada, acts as a foothold for the other two. And the artiste with his predominantly musical impulse and preoccupation with svara and tāla, often distorts the verbal forms of words sung beyond recognition; or at least he stretches, splits, repeats or accentuates syllables in such ways as to render the literary content beyond apprehension (except perhaps to the initiated). This is to a great extent inevitable if music is to predominate.

Such a tendency prevailed even during the period of the sāmagāna. Distortions of the mantras sung in order to better formulate and perform the svara aspect, were not uncommon in sāma. Six ways in which they could have been enumerated: vikāra, viśleṣaṇa, vikarsaṇa, abhyāsa, virāma and stobha.

Vikāra was distortion of a word: 'agne' was sometimes sung as 'ognayi'.

Visleṣaṇa was disjoining the syllables of a word and prolonging or stretching individual syllables during a song: 'vītaye' could be rendered as 'vo yi to yā 2 yi.

Vikarṣaṇa was similar to viśleṣaṇa and denoted 'pulling' or stretching of a syllable: 'ye' could be rendered as 'pā 2 3 yi.

1 गान्धर्व यन् मया प्रोक्तं स्वरतालपदात्मकम्। पदं तस्य भवेद वस्तु स्वरतालानुभावकम् ॥ —N.S. 32, 27.

2 तत्र हि स्वर तालौ प्रधानम् । तौ च नाधारो न भवयो प्रयोक्तुमिल्याधारतया यदुपयोपि तदाहु । स्वरतालानुभावकम्पति । —A.B. on N.S. 32, 27.

Abhyāsa, as the word indicates, was the repetition of one or more padas: ‘to yā yi’ to yā yi’; ‘śrddhayā śraddham’ ‘śraddhayā śraddham’ ‘srddhayā śraddham’ etc.

Virāma stood for a process in singing where a break was made within the body of a pada in a way quite unwarranted from the semantic point of view: the two padas ‘guṇāno havyadātaye’, could be rendered as ‘guṇānoha vyadātaye’, quite distorting each.

Stobha was introduction of one or more meaningless syllables quite foreign to the mantra: these were, ‘au ho, vā’ ‘hā u’ ‘ehā u’ ‘ho yi’ ‘ho yi i’ ‘auho-i, ‘ohā-i’ etc.

Many of these processes of distortion were carried over into gāndharva. Abhinava says that in rendering words during gāndharva-singing the ‘meaning of words could be quite overlooked’ (arthānusāramanavekṣyaiva):1 that is to say, it was not necessary for a gāndharva singer to render words in such a way as to make its literary meaning clear to the listeners.

It would be interesting to compare the nature of verbal or syllabic distortion (vikāra)in gāndharva with that of sāma from which gāndharva has, indeed, been said to have emerged.Not all the details are available, but much interesting material may be gleaned and inferred from Bharata, Abhinava and others.

Bharat (as we had occasion to remark in the last chapter) has forbidden certain alaṅkāras in dhruvā-singing because in rendering them a single syllable within a word was likely to get unduly stretched (varṇa-prakarṣa). The relevant passage in the Nāṭyaśāstra reads :2

“The alaṅkāras described here pertain to saptarūpa (i.e., gāndharva); not all of these are desirable in the dhruvā because they stretch the syllables beyond the comprehension of listeners; a dhruvā cannot be properly rendered when syllables are so stretched. Alaṅkāras like śyena, bindu and others cause such a stretch. These alaṅkāras should not, therefore, be rendered in dhruvās in their true measure (svapramāṇataḥ). And for this very reason only ascending notes (i.e., alaṅkāras of the ārohī varṇa) should be utilised in dhruvās.3 A dhruvā should suit the meaning expressed ; therefore, it should be so composed that it acts as a beacon to the literary content (‘kāryā arthadarśikā’). Syllables in the words of a dhruvā should be sparingly used.”

The sparing use of syllables (kṛṣatvaṃ padasaṃśryaṃ) according to Abhinava denoted the following: “in dhruvā, just as the lyric was to be sung in such a way that its literary meaning as a poem was not lost sight of, similarly, on the same principle,

गान्धर्व तन्न वार्तानुसारमनवेदक्ष्यैव जाल्यन्तरद्विच्छितवर्णा॰न्नाधारमसं तालप्रपितसर्वरघानतया गुरुत्वेन प्रतिपाद्यते ।

—A.B. on N.S. 33, 1.

सङ्गतरागता जेया बल॑कारा वर्णविस्तृतमे । नैते सर्वे ध्रुवावस्थिता: (श्रुति) श्रुतवर्णप्रकरान्वितात् ॥ न हि वर्णप्रकर्षन्तु ध्रुवां वाचां सिधारयेत् । श्येनो शाल्यदयथा विन्दुर याव ति (तु) प्रकर्षिन: ॥

ते ध्रुवां वाचं प्रयोक्त॑ न काव्य: स्वप्रकाशनात् । तदवस्थाहुरह ति हि ध्रुवा कार्योऽर्थदर्शिका । वाणीनां तु पुन: काय कुत्राल्पं पदसंश्रयम् ॥

—N.S. 29, 26-29.

Abhinava on this point says that the phrase ‘ascending notes’ (which according to him refers to ārohī alaṅkāras), in conjunction with the ‘api’ shows that the avarohī alaṅkāras are also implied:

आरोहिणो वर्णस्य सम्बन्ध्रत; । अविशदादावरोहिणोऽपि ये स्वरा' इत्यदूरा: । ते प्रयोक्त॑व्या: ।

the letters constituting the words of a dhruvā were to be such as would be easy to pronounce and into one so that they could be easily carried across to the audience."1

Bharata has singled out bindu and śyena as alaṅkāras prone to stretch a syllable. Bharata has described the bindu (N.S. 29, 38) but not śyena; the lakṣaṇa of śyena has been quoted by Abhinava from another authoritative text which he does not name (A.B. on N.S. 29, 27). Abhinava remarks that some thought śyena was another name for an alaṅkāra called veṇī which Bharata includes among avarohī alaṅkāras (N.S. 29, 25).2 Although bindu and śyena have been singled out by Bharata as instances of alaṅkāras that could lead to excessive stretching of a syllable (varṇaprakaṛṣa), the rule really was that all alaṅkāras with a similar tendency were to be avoided.3

Bindu and śyena were permissible in gāndharva. We do not know the exact nature of the varṇa-prakaṛṣa they effected. It is likely that beyond a simple stretching of a syllable they also distorted it as in the sāma vikāras called viśleṣaṇa and vikaraṣaṇa.

In sāma, the distortion called virāma occurred when two words sung one after another were syllabically split in such a way as to render both unintelligible. This is quite common in classical singing even today; a sentence like ‘mere ghar āye (‘he came to my house’) is quite liable to get distorted into ‘meregha rāye’ and so forth. This happens when a single syllable such as ‘gha’ is unduly prolonged. In gāndharva, single syllables were similarly prolonged and virāma must have occurred frequently.

There was another factor which could lead to an occurrence similar to virāma in gāndharva. This was the use of the gītis. The gītis in gāndharva, namely, māgadhī, ardha-māgadhī, sambhāvitā and prthulā, depended upon certain patterns of syallabic formations. Two of the gītis consisted of repeating syllables or words. In singing the ardha-māgadhī gīti, a line such as ‘devaṃ śarvaṃ vande śarvam’ was rendered in the following manner:

‘devaṃ, vaṃ śarvaṃ, vaṃ vande, de śarvaṃ’4

1 एतदुक्तं भवति । यथा हि गीतिर्यपंणमनवधीर्यन्ती कर्तॄणां तद्वैब पदगतानां यक्षराण्यपि दुःषणसय्योगादधी नाझ्ञान्यधिकानि कर्तॄव्यनि ।

—A.B. on N.S. 29, 29

2 श्येन इत्यनेन तन्वातरश्रितान् व्यपदिशारोणालङ्काराणां सङ्क्रयानियमं प्रत्यनादरं दर्शयति । तन्न श्येनलक्षणम्:

श्येनश्रकातरो जात: कलामात्रान्तरे स्थित:

तस्माच्छव स्वरे बृंहित गृह्णात्येव विलक्षणम्

अन्यो वेष्टि: श्येन इत्याहः ।

—A.B. on N.S. 29, 27.

3 ननु विन्दुश्येन व निषिद्धौ । नेत्याह । ये चान्ये तु प्रकारणाः । तेऽपि न ध्रुवागां प्रयोक्तव्या: ।

—ibid., 29, 27.

4 पदस्य समन्ध्यावेपदाक्षरस्तदावृत्तौर्धमागधीति

—A.B on N.S. 29, 47.

Ṣārṅgadeva notes as an actual example of ardha-māgadhī. The sentence he takes up for illustration is similar to that of Abhinava, which we have noted above :

पूर्‍वंयो: पदमोक्षे चरमे दिर्घदीर्घते । तथाार्धमागधी प्राप्तः:

—S.R. 1, 8, 18.

यथा—म रा ग सा । सा सा ध नी। प ध प म

व र द्र वं दे वं

Such a rendering dissects words in a manner quite inimical to their sense. In gāndharva it was permitted as in sāma. Abhinava speaking of the ardha-māgadhī significantly says : “in Sāmaveda where the song (that is the melody) dominates, the sense of a word is not heeded to in making repetitions (as in ardha-māgadhī): yataḥ sāmavede gītapradhāne āvrttisvartho1 nādriyate” (A.B. on N.S. 29, 47). Abhinava also gives an example from sāma-gāna recalling a practice similar to the formation of ardha-māgadhī. In singing the sāma mantra ‘udutyam jāta-vedasam’ (Sāmaveda 1, 1, 3, 11), syllables upto ‘udutyam jā’ were repeated in succession as in ardha-māgadhī and then finally ‘tavedasam’ was sung; this process completely disfigured the normal pada-formation—“evamādi tu jāta-vedasamiti hi atra jāśabdaparyantam-āvrttiparyantaparam-payayā gītvā ‘tavedasam’ iti gīyate” (A.B. ibid.). Kallinātha, in the same context, also quotes this example, but he says that the syllabic repetition was carried on upto the word ‘veda’ in ‘jāta-vedasam’ after which the entire pada ‘jāta-vedasam’ was sung without being split.2

Abhyāsa or repetition of a phrase or a word was common in gāndharva as in sāma singing. In the gīti māgadhī an abhyāsa occurred which was quite complex both musically and syllabically. In singnig the line ‘devam śarvam vande śarvam’ in accordance with this gīti, the first pada ‘devam’ was sung to a slow tempo in the initial kalā (a time-unit), in the next kalā the two padas ‘devam śarvam’ were sung in medium tempo (double the slow tempo) and in the third kalā two more words ‘vande śarvam’ were added to ‘devam śarvam’ and all the four words were sung together to a fast tempo (double the medium). The same process was repeated with a gradual addition of more words in further kalās.3

Bharata forbids the rendering of these gītis in dhruvā; they were to be used only in gāndharva: “etāstu gītayo jñeyā dhruvāyogam vinaiva hi/gāndharva eva

1 In A.B., the reading is :आवृत्तिस्वर्धों नाद्रियते. This makes no sense. This line was probably borrowed from the Vrtti on Mataṅga’s Br. Kallinātha quotes the line in the context of the ardha-māgadhī. His reading is :

यथा—सामवेदे गीतप्रधान आृवत्तिस्वधा वण्ण पादबुध्या पुनरुक्तिदोषे पदादौ सांगमात्रं कतश्च वाड᳀श्रूय मत᳁क्नेन परिहृतं यथा—सामवेदे गीतप्रधान आृवृत्तिस्वधा नाद्रियते इति।

—Kalā on S.R. 1, 8, 18.

2 तथा वेदनॊवोद्धृतं च—उदुत्यं जातवेदसम् इत्यलं वेदशब्दपयंन्तमावृत्तिपरंपरया गीत्या जातवेदसमिति गीयते; पदखण्ड-नादर्थमज्ञॊ न भवत्यनाप्रतीत । अतः सामवेदप्रकृतिके गीते गानवशात्तु स्वच्‍छन्दतानां पुनरुक्तिर्योऽनिष्टसच्‍च न दोषायेतिमन्तव्यम् ।

—Kalā, ibid, in continuation of fn. above.

3 देवमिति पदं गीत्या प्रथमकलां निवीर्ंशा विलम्बितेन लयेन यदा द्वितीयां कलां मध्यमलयेन देवमित्यतेन पदेन शार्वंमिति शब्दसहितेन निर्वहयति ततः देव इत्यं शाब्दः कलाव्यव्यानात् विनिवर्त्तिनिर्वह॑ एव॑ शार्वमित्यादयोड्यप्यस्तर-कलाव्यपिनस्तदा मागधीति गीति।

—A.B. on N.S. 29, 47.

For Śārṅgadeva’s description see S.R. 1, 8, 16-19. He notates the gīti as :

म ग म घ धनि धनि सनि घ रिग रिग मग रिस

दे व᳀ं दे व᳀ं श रु᳀ं दे व᳀ं रु᳁द्ध्र व᳀ं दे

yojyāstu nityam gānaprayoktrbhih" (N.S. 29, 49). Looking at the above examples

one can understand the reason behind this prohibition.

Pada in gāndharva occupied the lowest rung in the bierarchy of importance.

Svara and tāla were the dominant elements. Among these, too, svara was the leader

("tena svarāḥ pradhānāṃ tālonāmātmā tatsāmyenopakārakah, A.B. on N.S. 28, 11-

12). But, though of subsidiary importance, pada, too, was formed and regulated

according to vidhis. Ostensibly, just any song could become the basis of gāndharva

because unlike svara and tāla, the form of the pada in gāndharva has not been speci-

fied by Bharata or Dallita; yet only certain hymns, handed down in tradiṭion, seem to

have served as the word-base for music in gāndharva. Bharata quotes three hymns,

all Śiva stotras.1 Abhinava gives part of another Śiva hymn: "devaṃ śarvaṃ vande

śarvam."

Pada in gāndharva consisted also of a string of meaningless syllables. These

were technically known as śuṣka or bahirgīta and also nirgīta. Śuṣka syllables, too,

were preordained and Bharata ascribes them to Brahmā himself (N.S. 31, 104).

Abhinava remarks that just any string of syllables cannot serve an adṛṣṭa purpose, and

only specifically ordained syllables, current since time immemorial, gave rise to spiri-

tual benefit (abhyudaya) because they were at par with Vedic mantras: "nanvevaṃ

bhinnākārādirapi kriyatāmittyaśaṅkya nādiśāstraṃ bandhaprasiddhyavasiṣṭābhyudaye

hetvarthaṃ keśāncidevetyaāha—"yāni purā brahmaṇā gītāni"; brahmarūpatvena ca

vedamantravādigītīni" (A.B. on N.S. 31, 104).

Sometimes meaningless syllables were combined with an epithet of Siva :

jhaṇṭum jagatiya valitaka jambuka jhaṇṭum titi ca

laghu ca jhaṇṭum (titi cā)

diṅgle ganapatipaśupatijambuka diṅgle varabhujia

digi-nigi cā

titi ca digi nigi cā paśupati niticā

(N.S. 29,88)

1 (a) देवा देवा: ससृजत्नमितम्। देवैर्यज्ञहेतुभि: पितृभि: प्राणमितचरणम्।

दैलोक्यहेतुभिर् ऋषिभि: रहु शरणमहंमगत: !

—N.S. 29, 107.

(b) भूताधिपति भवनेभ्यो देवेन्द्रवै सुरसभमहोदयम्।

रोदं भयदं गजचमूपट शम्भुं द्रयक्षं ज्वलननिभजटम्।

भुजगपरिकरं त्रिदशणवनंर्द देवैर्-निधि-यै: परिपालितचरितम्।

उमार्तिनिर्मथितमथसुरवृंद शरणं सुरनुतमहिमि समुपगत:॥

—ikid., 29, 110.

(c) अमरं प्रवदं मदनाज्झरं भुवननिकायनटमभयपद्म।

नितपुरनाशकर देवं तमहं प्रणमत (पात) सुरश्रृङ्गणतचरणम्।

पृथिवी सलिल ज्वलन: पवन: सूर्येश्चन्द्रौ यजमानो व्योमाधय: काव्य:।

मुनिभिर्-यंस्य प्रोक्र्तस्य लोकगुर्हि तमचिन्त्यमर्-ज (तमजं) विधिनिलय भैरव रुपम्।

खट्वाङ्कुपालगह्रे स्पितयूपपतिप्रलयनिमित्तम्।

चन्द्रार्धधरे तिलकाक्षधरं मुण्डाधरं रशनाधरगरुम्।

बहुभिर्-भिवृक्ती: विविधविकटवेषैस् तृविद्येर्विमुखैर् अपि तै:।

प्रमद: परिकतमहमीश सुखदं सततं प्रणत:।

—N.S. 31, 113.

Other such examples are also to be found in the Nāṭyaśastra (N.S. 29, 90, 29, 101, 29, 106 etc.). Most śuṣka constructions, however, contained, no recognisable meaningful word :

jhaṇṭum jagatiya digi nigi jhaṇṭum

titijhala kucajhala etc.

(N.S. 31, 104)

The pada in gāndharva, whether meaningful or meaningless, was dedicated to the gods and hence to adṛṣṭa.1 But in gāna, pada was the dominant element and it had fundamentally a dṛṣṭa or immediate result to aid nāṭya in creating the aesthetic delight (rasa) towards which it was directed.2 Abhinava, therefore, once again stresses : svara in gāndharva plays the dominant role and on it depends the pada, which is therefore secondary ; contrary to this in gāna communication of meanings is of the primary importance and thus pada has the dominant role. Svara is subsidiary in gāna; it aids pada by imparting to it the charm and colour of music (uparañjana).3

In defining dhruvā, Bharata states :

dhruvā varṇāstvalañkāra yatayaḥ pāṇayo layāḥ,

dhruvamanyonyasambaddhā yasmād tasmād dhruvāḥ smṛtāḥ (N.S. 32, 8)

“The beats beginning with dhruva (‘dhruvā iti pātādyāl, A.B.), the varṇas, the melodic figures, the tempic regulations (yati), the syncopations in tāla (pāṇi) and the different tempos (laya)—these are all connected in a ‘dhruva’ (steadfast) manner with each other within a dhruvā; this is why the name dhruvā.”

Abhinava commenting on the above verse remarks : “the one constant, unchanging element (avyabhicāri) which causes the ‘dhruva’-ness (dhruvarūpe nimittam) in dhruvā, is pada. It is the fundamental factor on the basis of which the fitness or suitability of all other factors is judged; pada (in dhruvā) is the pillar on which varṇa and other musical elements stand : eteṣāṃ dhruvādīnāṃ anyonyasambandhe’ vyabhicāritaya dhruvarūpe nimittam padam ādhāratvādaucityayojanācca; na hi nirādhārā varṇādaya iti darśitam caitat” (A.B. on N.S. 32, 8).4

Words in a dhruvā governed its musical form : whether a certain alaṅkāra was justified or not, whether a certain tempo was indeed the suitable one, whether a rhythmic syncopation should or should not be formed—such questions in a dhruvā

1 पदमपि विनियुक्तं स्फुटकुलकादि भेदवयस्य नवधा भवेद् देवलस्तुतो । गानधर्वं तद्धि वाथानुसारमन्वेष्येद् ।

—A.B. on N.S. 33, 1.

2 गानस्य तु पादवर्गेसाम्यसम्पत्ति: साम्यिकजनने चापरञ्जनपुंसां प्रागुरसाध्यायोजितसाधारणसनिध्यरुप्यात्मकरसभाववृत्तानु-स्त्वति हेतुसुचितसादृश्ययोगप्रक्षित दृष्टं मुख्यं फलम् ।

—ibid., 33. 1.

3 एतद् गानधर्वं स्वर: प्रधानं तदाधारतैव गुणभूतं पदमिति, यत् तद्विपरीतं गानधर्वं *(? गाने)ऽर्थसंवेदनयोगेन पदस्य

प्राधान्यमात् तदुपरञ्जनताच्च गुणावात् स्वरसयेतदपि श्यकार्येत भवति ।

—A.B. on N.S. 33, 1.

• The A.B. reads गानधर्वेऽसंबदन...etc., Obviously a wrong reading for the whole context discusses differences of gāndharva and gāna ; so when attributes contrary to ‘gāndharva’ are delineated, ‘gāna’ must be the word intended ; a repitition of the word ‘gāndharva’ is here self-defeating.

4 Abhinava condudes :

ध्रुवं पदं तस्मादेतो: प्रावेशिकयायो ध्रुवा इति ।

could be decided only with reference to the nature of the poem being sung and the situation or rasa being depicted through its 'pada'. Some predecessors of Abhinava had introduced the simile of a fresco or wall-painting (kuḍyacitra) to express the relation between words and music in dhruvā. In wall-paintings, a wall is the ground on which a painting comes into being. Similarly in a dhruvā, pada acted as the ground on which other elements of a song were manifested.1 Even when (as it sometimes happened) the music in a gāna-form was not actually set to words, it yet remained subservient to overall poetic contention or context. Instrumental Music in plays often acted as a 'filler' to cover gaps in the mood left unuttered by a song. Gāna-music was in such cases set to appropriate grāma-rāgāṅgas, bhāṣāṅgas or similar forms. Such music could be delicate or charming (surocitam), or if the occasion demanded, full of musical graces, adorned with brilliant alaṅkāras (as Abhinava says, in bhiṇḍikā and dvipadi-gāna forms). On other occasions it could be full of an agitated fury of swift tāla patterns (uttālīkṛta tālavibhramam) which emphasised the restless movements of a character in situations of valour (vīra) or fierce action (raudra). It could also underline the plaintive mood of sorrow in separation through befitting compositions (vipralambhaśṛṅgārocitabandhatva). In such cases gāna-forms helped to strengthen the mood in a dramatic scene during those periods when nothing was being said or sung on the stage. But such 'background' music, too, was subservient to the literary dramatic content.

Like gāndharva, gāna-songs, too, it appears, sometimes utilised śuṣka or non-sense syllables (like humkāras) and were on occasion rendered without any words whatsoever (svarālāpaprabhṛti gānamanakṣarameva dṛṣṭam). But such śuṣka songs were subservient to the moods expressed by the dhruvā songs proper. They acted as prologues to dhruvās and introduced the sentiments in the imminent dhruvā with appropriate instumental accompaniments.3

There was thus quite a distinction between gāndharva and gāna in all their three constituent factors of svara, tāla and pada : "evam tāvat svarapadatālātmakas-varūpavaicitryavailakṣaṇyam" (A.B. on N.S. 33, 1). These formal differences were, in spirit, caused by the fact that gāndharva and gāna were motivated by entirely different objectives.

1 अन्येऽपि कुड्यचित्त (तु) दृष्टान्तेन तेन ध्रुवानामाधारः पदमिति पर्यवात् तद्विधते यस्यां वृत्तजातो सा ध्रुवेति ।

अत एवं ललख्य गीयमान रुपकस्मैव ध्रुवेत्याहुः।

—A.B. on N.S. 32, 8.

2 गाने पुनःश्रध्याकारे मुख एवं साम्भर्या रसभावोचित वृत्तजाति निवद्धम् । यत्तु काव्ये तन्नोक्तमित्युक्तं रसपुऱणोपयोगिनाम् प्राधान्यमुपसृत्यन । ततश्चित्तवृत्त्यनुसारेण गानस्य गीतवृत्तस्य च प्रयोगे समारम्भे वरवचनं सुरोचितम् ।

भिण्डीका दिवीगानादौ स्वरालापकूजितगोष्ठा इत्येष वर्चितु वीरोचितपरिक्रमप्रादर्शनानुदानिकृतततल-विग्रहं सम्पादयत कविचित्रतगानं च विप्रलम्भग्ज्ञारोचितवृत्त्यन्वितं विचित्रादिमान इव नाट्यायितावकाशमादधात् स्वं वाच्यं प्रथमांजयदवभाविती स्वरतास्विभावाहालम्बमानभावाग्रास्त इति ।

—A.B. on N.S. 33, 1.

3 ननु नाट्येयु स्वरालापप्रभृतिप्रयोगानन्तरमेव दृष्टमिल्याशङ्कयाह । यत्तु वाक्यार्थनोपेतमिति । वाक्यार्थन वाकुक्रियामात्रस्वरा-लापे हुम्कार इति वदुक्तप्रस्तुत येऽनुसन्धाने यावद्यति यावकृत्यैव गानं द्वय रसापेक्षाकरणाल्लूयतेनियमे।भावेन यथाकविस्वातन्त्र्यं तदेवविगमातलं व्यवस्थामादर्शयिष्यामिहीनं ध्रुवागानात्मकमेतच्च भवतीति यावत् ।

ननु च तथा विधस्य प्रयोगनेत्याह—“स्वतोऽर्थे चिति” । सर्वतोऽर्थानां तत्सुपिरवादीनामनुरञ्जकं स्वाध्याश्रय-भावि ध्रुवारूजनोपगिनस्तद्वीयांशाविधानं च पुरक्रमिति यावत् ।

—A.B. on N.S. 32, 30.

different purposes and served quite distinct ends. Therefore Abhinava adds : “The

above differences as to svarūpa or formal structure are also indicative of the distinct

‘phala’ resulting from the two forms—“anena phalavailakṣaṇyamapi vyākhyātam”

(A.B. ibid.).

The phala or end-result of an object—the purpose it serves, the fruit it bears

—is connected with its function, which Abhinava calls its dharma. The function of

gāna, we have seen, was intimately connected with the drama. Gāna could be ‘sa-

phala’ only by being successfully incorporated in an ancient play as one of its many

constituent elements. Its entire form (svarūpa) was moulded and patterned to this

end. So subordinate was gāna to nāṭya that Abhinava seems to have felt that, out-

side the play, gāna could be quite feeble in its effect : “na hi nāṭyādbhīrayabhaṅgyāpi

dhr uvāgānaṃ gīyamānaṃ sukhapādāmuttpādayati” (A.B. on N.S. 33, 1).1 The phala

or the goal of gāndharva was adṛṣṭa. Abhinava often reminds his readers that forma-

tion in gāndharva should be made exactly as directed; for only thus could adṛṣṭa be

achieved. We have already noted some of Abhinava’s comments. We here note a few

more pertinent points to bring the matter home :

1. The adṛṣṭa-motive behind the seven tāla-gītakas and the vardhamānaka

has been indicated by Bharata and also by Yājñavalkya (in his Smṛti). Bharata says :

gītakeṣu pravuktasteṣu dēvāstusvanti nityaśaḥ

vardhamāne prayukte tu rudrastuṣyati sānuṣaḥ

(N.S. 5, 47, 48)

“The rendering of gītakas pleases the gods and vardhamāna pleases Rudra

and his retinue.”

The Yājñavalkya Smṛti forbids practitioners of profane music from taking part

in holy rituals (3, 167); such musicians (kusīlavas) are also not permitted to take part

in śrāddha ceremonies (3, 155). But regarding sāmagāna and also gāndharva the

Smṛti says :

yathāvidhānena pāṭhansa māgāyanavieyutam

sāvadhānastadabhyāsātparaṃ brahmādhigacchati

aparāntakamullopyaṃ madrakaṃ prakarīṣyathā

auveṇakaṃ sarobindumuttarāraṃ gītakāni ca

rggāthā pāṇikā dakṣāvihitā brahmagītikā

geyametattadabhyāsakaraṇānmoḵaśasamjñitam

viṇāvādanatatvajñaḥ śrutijātiviśāradah

tālajñaścāprayāsena mokṣamārgam nigacchati

gītajño yadi yogena nāpnoti paramaṃ padaṃ

rudrasyānucaro bhūtvā tenaiva saha modate(Yājñavalkya Smṛti 3, 4, 112-116)

1. The printed A.B. text reads gīyamānamukhapādamuttpādayati, a reading which makes no sense in the

context. We suggest that the reading should be gīyamānantu sukhapādamuttpādayati. This agrees in spirit

with the sentence that follows. Moreover, orthographically, ‘न’ being close to ‘स’, a scribal mis-

take is not unlikely.

“If sāma-gāna is intoned properly and without errors, according to the due procedure and the enjoined rules, one can attain the ultimate Brahma through this discipline. (Similarly), the gītakas : aparāntaka, ullopya. madraka, prakarī, ovenaka, along with robindu (i.e., rovindaka) and uttara as well as rk, gāthā, pānikā and the brahmagītikā ordained by Dakṣa, are songs, the disciplined singing of which leads one to mokṣa.1 The musician who knows the mysteries of vīṇā-playing, who is an expert in śruti and jāti2 and knows the ways of tāla, can without effort travel the path to moksa. If (perchance, such a) musician cannot attain the ultimate liberation through this yoga, he becomes a companion of Śiva and delights forever in His company.”

We notice here that the gītakas are said to lead to mokṣa, the ultimate goal of adṛṣṭa. All the seven gītakas of gāndharva are mentioned by name—aparāntaka, ullopya (i.e., ullopyaka), madraka, prakarī, auveṇaka, obindu (‘rovindaka’ in N.S. 31, 200–201 and Datt.), and uttara (cf. N.S. 31, 200–201 and Datt. 161–221).

2. The jātiṣ of gāndharva had adṛṣṭa as their aim: “jātiyadrṣṭam” (A.B. on N.S. 28, 36). Rāgas (forms used in gāna) on the contrary had a dṛṣṭa phala for they were primarilydesigned to have an immediately pleasing effect: “rāgamiti raktirdṛṣṭaphalam” (A.B. ibid.). Abhinava defining jāti says : svaras when arranged in specific assemblages are called jāti; jātiṣ are pleasing and they result in the propitiation of adṛṣṭa: “tatra keyaṃ jātirnāma ucyate svarā eva viśiṣṭasanniveśabhājo raktimadrṣṭā-bhuyudayaṃ ca janayanto jātirityuktāḥ” (A.B. on N.S. 28, 66).”

Abhinava here quotes an ancient authoritative (āpta) verse : “the (jāti) nandayanti if properly formed according to the enjoined rules (yathāvidhi), can expiate the sin of having done a Brāhmaṇa to death” (sakṛtprayuktāpi hi nandayantī yathāvidhi brahmahanaṇam punāti, A.B. on N.S. 28, 9–10). This verse is repeatedly quoted by Abhinava to point out the adṛṣṭa aspect of gāndharva (A.B. on N.S. 28, 95–97 and 31, 54).

3. Not only the singing of gāndharva but also its playing in accordance with the right techniques—dhātu—resulted in adṛṣṭa (besides giving rise to aesthetic pleasure): “dadhātyadrṣṭaṃ viśiṣṭaṃ dṛṣṭaṃ raktiṃ ca taranti dhārayanti ca vīṇāvādyas-varūpamiti dhātavah” (A.B. on N.S. 29, 50). The dhātus resulted in adṛṣṭa only when rendered as decreed in the śāstra : “śāstroktayādrṣṭopayogino’pi” (A.B. on N.S. 29, 70).

According to Abhinava, another difference between gāna and gāndharva was that they were employed on quite different occasions (‘kāla-vailakṣaṇya’). The difference in occasion is to be understood in relation to theatre.

Gāndharva, unlike gāna, could not be an essential part of the plot itself because its forms did not permit alterations to suit the mood of a passing scene.

1 Vijñāneśvara, in his famous commentary Mītākṣarā, adds that the vardhamānaka and its āsā-ritas were also implied here along with the seven gītakās known collectively as prakaraṇa-अपरान्तकोल्लास्यमद्रकप्रकरोवीणकानि सरोभिन्दुसहित चोत्तरमित्येतानि प्रकरणाख्यैः सप्त गीतकानि । च शब्दादास-रितवद्भिमानकार्दिमहागीतानि गृह्यन्ते ।

2 The term ‘jāti’ in this context stands for the 18 jātiṣ of gāndharva as the Mītākṣarā clarifies : जातयस्तु पञ्चदशः सप्त शुद्धाः संकृतजात्याद्वैश्वादेशमष्टादशविधास्तु विद्यासारस्तु विचारः: प्रवीणः ।

Gāndharva could not, to use Abhinava's expression, drown its voice within the symphony of art-forms emyloyed in ancient plays.

Yet it played an important role in the staging of ancient plays. It was employed during the pūrvaraṇga as an act of worship through which divine powers were propitiated ("pūrvaraṇge gāndharvameva pradhānam", A.B on N.S. 33, 1).

Pūrvaraṇga in the ancient theatre was a prologue to the play proper. But it was a prologue in a special sense, for it did not introduce the play nor did it essentially form any part of the scheme as envisioned by the play-wright. Abhinava, regarding pūrvaraṇga, says : "nahyam bhāvānukīrtanātmakaṅyākhyānānukārasvabhāvah (A.B. on N.S. 5, 5-7) : its form is not imitative; it does not depict a narrative or plot where feelings and actions are shown." Though distinct from the play proper it was yet enjoined as a necessary beginning to all plays (N.S. ch. 5). It was an opening ritual,1 an act of veneration. But it was no ordinary ritual. It was worship in the form of dances accompanied with music. The ancient play was a great aesthetic spectacle; it was appropriate that the propitiatory act performed at its beginning should itself be an artistic paean or pageant in honour of the gods. The aim of pūrvaraṇga was, naturally, adṛṣṭa.2

The ancient pūrvaraṇga was a complete entity in itself, a well-defined whole divided into definite sections or components.3 Its ritual propitiation consisted of dances in the tāṇḍava or lāsya forms. The music to which they were set was gāndharva. Any of the gītakas could be used though the tāṇḍava was chiefly set to vardhamānaka :

gītānāṁ madrakādīnāṁ yojayamekāṁ tu gītakaṁ vardhamānamathāpiha tāṇḍavaṁ yatra yujyate

(N.S. 5, 13)

Dance in the pūrvaraṇga (as we have said earlier) was performed in conformance with well-ordained forms and also gave rise to adṛṣṭa. The following pronouncements of Abhinava will bring out this point :

(i) "śivalokagamanantaphalam nṛttasya yasmāddevasya bhagavato mahādewasya stutimittam paritoṣanaprayojanametannṛttamathetorniyamā-

1 Abhinava indeed explains the word pūrvaraṇga as :

अपि तु देवताप्रितोषणफलं पूर्वविलपनादिदानवत् सत्यमेवेति पूर्वं रङ्क इरयेष समासः ।

—A B. on N.S. 5, 5-7.

Dr. Raghavan speaking of the ancient theatre says : "Lot of ceremony and propitiatory acts marked its beginning 'and, all this preliminiary was referred to as Pūrvaraṇga, literally the pre-performance acts." 'Music in Ancient Indian Drama', J.M.A., Vol. XXV, p. 82.

2 cf. नाट्यस्य रचनाप्रधांयैर्दि चत्रैण योजनीयमदृष्टसम्पत्तये चेतिं मङ्गलध्यम् ।

—ibid., 5, 27-29.

3 यस्माद् रङ्क प्रयोक्तव्यः पूर्वमेव प्रयुज्यते ।

तस्मादयं पूर्वरङ्को विज्ञेयो डित्सतामिह ।

अस्याङ्गानि तु काराणि यथावदनुपूर्वशः ।

तन्वीभावनसमयोगे: वाण्ययोगकृतस्थया ।

प्रत्याहारौड्वतरणं तथा हारम्म एवं च ।

आश्रावणा वषट्कारानिसृथया च परिषडटना ।

सह्घोटना ततः कार्यो मार्गांसारितमेव च........

—N.S. 5, 7-10.

122 A Study of Dattilam

dr̥ṣṭasampattyai" : nr̥tta (dance without abhinaya) is performed as an act of praise in honour of Lord śiva, the highest of gods; its fruit is the eternal Śivaloka—the dwelling of Śiva himself; and since the aim of nr̥tta is to please Śiva, it results in adr̥ṣṭa when performed as duly ordained" (A.B. on N.S. 4, 302).

(ii) "ihādr̥ṣṭaviṣeṣasampattihututvaṃ vardhamānādiprayogasya sopakaraṇr̥ttaprayogasya ca : here (in the pūrvaraṅga) vardhamāna etc. are employed beecause they result in adr̥ṣṭa; nr̥tta, with all its aparthernalia, is employed for the same purpose" (A.B. on N.S. 4, 246-259).

(iii) "tena prayoktuḥ prekṣāpravartayiturarthapateḥ sāṃājikavargasya ca pāpaviśuddhīḥ śivalokagatamanantaphalam nr̥ttasyeti nānyādr̥ṣṭaikaprayojanādbhinnaprayojanatvamuktam : the aim of nr̥tta is to expiate the sins of the performers (of a play), of those who arrange for the performance and of the spectators; the ultimate end of nr̥tta is eternal Śivaloka itself; apart from such an adr̥ṣṭa end there is no other end which nr̥tta fulfils" (A.B. on N.S. 4, 319).

Gāndharva acted as the musical compliment to the propitiatory dance in the pūrvaraṅga. It had no other use. The occasion of the use of gāndharva was in this way quite distinct from that of gāna.

Including this difference on the basis of kāla (kālāt), gāndharva and gāna had quite a few distinctive characteristics. Thus Abhinava in true śāstric fashion concludes: "when things differ as regards to their nature, function and effects (svarūpeṇa, dharmeṇa, kāryeṇa), they are quite distinct, as one substance from another: "yadyat svarūpeṇa kāryeṇa dharmeṇa ca bhidyate tattu tena vilakṣaṇaṃ bhavo bhāvāntarādiva" (A.B. on N.S. 33,1).1

Besides a propitiatory function through its employ in the pūrvaraṅga, gāndharva had yet another importance for ancient theatre. It was the original embryonic form from which the forms employed in gāna were derived. And because the roots of gāna lay in gāndharva, a knowledge of gāndharva was important for a true understanding of gāna forms. This was certainly a major consideration in Bharata's mind for having included a long section on gāndharva in the Nāṭyaśāstra.

Dattila states that all existing songs could be traced to the jāti form of gāndharva (Datt. 97). This seems to have been the universal opinion among ancient ācāryas. They spoke of musical forms like bhāṣās, rāgas, grāma-rāgas etc. as derivative from the jātis.2 These derivative forms were born through mixing and

1 He continues :

भियते च स्वरूपादिना गानम् । गानधर्वाद् वेलक्षणयस्य हि तन्मात्रानिमित्तकं व्‍यापारान्तरमिति व्यापकानुपलक्‍ष्यापि पृथक्‍सिद्धये वेलक्षणयेन व्याप्त इति । तत्त च गान-स्ववेदोचितविधिप्रबन्ध प्रधालवादिदं गानधर्वम् । गीतसारवाच्चेदं गानमिति ।

Thus clinching the argument in the manner of a logician.

2 Abhinava quotes a predecessor, a 'commentator'—ṭīkāāra—on the Nāṭyaśāstra who had preceded him as remarking :

जातेरसंप्रति श्राम्य इति वा सकलस्य रामादेर्जन्महेतुजारिति टीकाकारः ।

—A.B. on N.S. 28, 46.

combining elements taken from two or more jātis in various ways and were hence sometimes called hybrid or saṅkara forms. These forms when moulded to the theatrical context were called gāna.

Saṅkara forms were created through combining the aṃśa (main note) of one jāti with the nyāsa (the final note) of another and the apanyāsa (a subsidiary nyāsa) of a third and so forth. Endless such permutations and combinations were possible and these gave rise to a host of forms : “geyaṃ grāmādilakṣaṇaṃ tadyadi kasyāpi jātyaṃśakasyā sambandhyinaśs’onyasya nyāso’nyādio’ panyāsa ityādi bahutaravaicitryam” (A.B. on N.S. 28, 141).

These new forms, however, were not heterogeneous mixtures but homogeneous wholes. To illustrate the fact that although born through a mixture of many jātis, the saṅkara forms were, in truth, independent entities, Abhinava gives the example of ‘pānaka’ a favourite drink of the period. Pānaka was produced through mixing various ingredients like jaggery, pepper etc., yet its taste had a flavour all its own : “guḍamaricādirasāyojanāmaye’pi pānaka iva rasāntaratvam” (A.B. on N.S. 32, 332). Grāma-rāgas and other derivative forms were similarly new and independent wholes (yathāvihitaṃ pānakamidamiti na tu kramamiśritam”, A.B. on N.S. 28, 141).

A saṅkara form could be traced back to its principal source jātis through analysis. The jāti or jātis, the structure of which predominated in them was said to constitute their root. Dattila thus states: “in hybrids, (the parent) jāti can be known through the form predominating: “saṅkare rūpabāhulyājjāātinirdeśa īṣyate” (Datt. 96).

Bharata has no such general dictum. But during Abhinava’s period certain recensions of the Nāṭyaśāstra did contain a verse at the end of chapter 28 to the same general effect. The verse states : “the source of a mixed song (miśrageya, i.e., grāma-rāgas etc.) should be sought on the basis of the predominating (form). In these hybrid forms much intermingling occurs between a variety of ornamentations for the sake of creating a pleasing effect.” Abhinava introduces this verse by saying that “it speaks of the process of creating grāma-rāgas and similar forms from the jātis”.1 Abhinava doubted the authenticity of this verse for it was not to be found in the manuscripts he had consulted : “asmad pustakeṣu na tathā dṛśyate” (A.B. on N.S. 28, 141). But he agreed with the contention.

He argues that its acceptance as an authentic passage from the Nāṭyaśāstra was not really essential to the point at issue, since Bharata had implied the matter expressed in it through other statements2 where he enumerated the aṃśas of jātis and suggested the great wealth of derivative forms.

1 अथ जात्युद्भवग्रामरागादिसम्मावनारचनार्धमायां केचित् पठन्ति: मिथो गेयं यदिदं तदचिन्त्येयमधिकबाहुल्यम् । सङ्कीर्यमाण सन्ततिविच्छेदात् याथातथ्येन रञ्जनातिशयात् ॥ —A.B. on N.S. 28, 141.

2 न च तया विना न सङ्गरगोतीत एतद्व्रतम, जात्यंशके हि सर्व स्वीकृतम् । तद्वापि चा (मा) ज्जनोपयोगिनो वैचित्र्यस्य- भ्यानुज्ञां मुनिनेव कृतम् । चतुःस्वरप्रयोगेऽपि 'हावकृष्टध्रुवास्विह' —(भ० ता० 28-79)

इति वदतोत्तरत्र चालङ्कारवैचित्र्यं विमृश्यतेति —A.B. on N.S. 28, 141.

124 A Study of Dattilam

The amśa or amśas of the jātiś, indeed, seem to have been the basic factor in deriving hybrid forms. Matańga says: “jātiś (√jan—to be born) are called so because they give rise to grāma-rāgaś through their amśas” (Br. 240).1

Many ancient authors on music had written about these derivative hybrid forms and had named the root jātiś from which some major forms were born. Abhinava names Matańga, Nandi, Kaśyapa and Yāṣṭika among these authors: “tadevamśabhūtam vanmiśrageyam tadasya jāterarthitamiti vannivamyate matańganandikaśyapayāṣṭikādibhistadhikenarūpabāhulyena” (A.B. on N.S. 28, 141). He quotes a passage of 21 verses from the work of one of these authors (probably Kaśyapa) where many well-known contemporary derivative-forms are traced to their original jātiś (A.B. on N.S. 28, 141).

This passage concludes with an interesting statement: that seven gītiś were utilised in dhruvāś (dhruvāganopayogāya tadetadgītisaptakam). The term gīti here is pāribhāṣic with a meaning different from what it has in gāndharva. The grāma-rāgaś derivative of the jātiś were rendered in various modes or styles—these were the gītiś. Each gīti was ruled by a characteristic musical idiom or manner of expression.

In the gīti called śuddhā the musical style was gentle or soft; the melodic line followed a straight, simple, unbending path : “śuddhā syādavakrailalitaịḥ svaraiḥ” (S.R. 2, 1, 3). The bhinnā gīti or style, on the other hand, was full of subtle but fast notes; the melodic line was studded with soft and gamakas followed a curved path : “bhinnā vakkraịḥ syaraịḥ śukṣmair madhurairgamakairyutā” (S.R. ibid.). The gīti called vesarā was rendered in a brisk and fast but charming style, with a kind of virtuoso quickness pervading all its movements: “vegavadbhiḥ svarairvarṇacatuṣke’pyatīrak-titah/ vegasvarā rāgagītirvesarā cocyate budhaiḥ” (S.R. 2,1,6).

Śārńgadeva following Durgasakti divides gītiś into five classes : śuddhā, bhinnā, gaudī̄, vesarā and sādhāraṇīti” : “gītayaḥ pañca śuddhā ca bhinnā gaudī ca vesarā sādhāraṇīti” (S.R. 2, 1, 2). Matańga, however, has seven gītiś. He describes their characteristics in the Brhaddeśī (Br. 291-308).

The gītiś appear to correspond to the notion of the various vāniś in dhrupad such as the ḍāgara-vāṇī, khaṇḍara-vāṇī etc. The ḍāgara-vāṇī has a slow and solemn style of singing with the ālāpa dominating; the khaṇḍara-vāṇī, is quick, fast, full of swift gamakas and set to running rhythms.

The various gītiś must have been applied to augment the rasa of the moment just as the rītiś of poetry were used for a similar effect.

We must take into account one possible confusion before we move on. In the first few kārikāś of the twenty-ninth chapter of the Nāṭyaśāstra, Bharata

1.cf. जायते शुद्धिमनगमो रागसाधनरभाषाविभावात्मक रो (?गी) तिसप्तक दृश्टसिद्धिमांरसोपर रञ्जनं व यतस्ततो जातय :1 —ibid., 28, 37-44.

Grāma-rāgaś were divided into seven classes called the seven gītiś. This was the classification according to Matańga. Yāṣṭika believed in three gītiś, Durga-śakti in five and Sārdūla in only one.

—See Sudhā. on S.R. 2, 1, 2-5.

—Kalā on S.R. 2, 1, 6-7.

sets forth the application of the jātis (on the basis of their dominant notes) in the different rasas : the different moods depicted on the stage (N.S. 29, 1-13). This, on the face of it, would seem to show-contrary to what we have been s aying-that the jātis (and hence gāndharva to which they belonged) were also used during dramatic action.

We must, however, be cautious in understanding Bharata here. Abhinava in this context remarks : “muniṇa jātyaṃśakāvinyoga ādhānābhiprayoktṛḥ (A.B. on N.S. 29, 8) - Bharata has enjoined the jātis not in themselves but as forming the base for (other forms) ” According to Abhinava's teacher, too, (upādhyāyāstvāhuḥ) the intention of Bharata in relating jātis and svaras to rasas was to enjoin the use not of the jātis themselves but of the jāti-aṃśas ; this meant the use of derivative saṅkara forms, of which these aṃśas were the base. Hence, he says, it was the rāgas, bhāṣās etc. that were entailed in Bharata's injunction : “tasmādāyamayabhiprāyo muṇeh iha jātyaṃśakā eva paramārthataḥ samastagītāsarvasvaṃ tatra hi rāgabhāṣādayopyantar-bhuktāḥ” (A.B. on N.S. 29, 8).

Indeed, at the end of the section enjoying jātis in rasas, Abhinava quotes a long passage of 75 verses from Kaśyapa where detailed prescriptions of the actual grāmarāgas, bhāṣās etc. to be used in particular dramatic situations have been noted.

Just as the jātis gave rise to the various rāgas, similarly the gītakas, too, appear to have been the source of tāla-structures used in gāna. We have spoken of Bharata's dictum that the components of the saptarūpa or gītakas, suitably transformed, resulted in the dhruvās (N.S. 32, 3). These gītaka components when transformed into dhruvās became independent forms with a character all their own just as the rāgas born of the jātis. Abhinava explains the point through an interesting analogy. He says: “an army arrayed eminently into a composite group (saṅghāta) with all its parts including the infantry (puruṣabhaṭa), the elephant battalions (gaja) etc. can be regrouped to form a different and totally new array which can transform it into an entirely distinct congregation (saṅghātantara). Yet there remains a cause and effect relation between the old and the new formations. Similarly, the components of the gītakas which had a specific arrangement ordained for the achievement of adṛṣṭa, could be regrouped with a different end in view : they could be suitably arrayed to express rasas and moods. They were, to this end, rearranged according to the metrical structure of the song being sung. In this new regrouping, a causal relation with the original compostion (i.e., the gītakas), though present, was not immediately felt. The new group, in fact, could become the basis for further formations.”1

Bharata gives a list of gītaka components that were thus transformed into dhruvās (N.S. 32, 3-6), and states that these formed the limbs of all five classes of

1 एतदुक्तं भवति । यथा शास्त्रेऽनु(हो)त्पत्तिं सङ्घातेन रस(च)नायामङ्गैर्गजादिभि(टे)गजादीनि केनचिद्वर्करणं स्थितितान्नीत्या कृत्स्ना सङ्घातान्तरे(ण) सम्पाद्यमानौवैचित्र्येण कारणकायद्भावोऽप्यभिनि(न)यत्, तथा गीतकाद्दू(भू)दन्यदृष्टप्रधाने गीतकाद्दूवैपसङ्गतोत्रित्वात् । कृ(कृ)ताद्भुतोचित्य समर्पणार्थमकरुणा(णां)योग्यतयागदुक्तर्या(र्या)दुक्तर्या(दृ)चित्रनीयमानवत्तव-शोभार्थ(य)येण सङ्घातातमकरुणं प्रतिपन्नकारणकं प्रागू(गू)दश्वाया अनुगम्य कार्यमो(मौ)त्तरकलिक् भजते कारसाङ्गद्रव्यपदेश्यै(श्या)नि च भवन्ति ।

-A.B. on N.S. 32, 3.

126 A Study of Dattilam

dhruvās: “dhruvāṇāmaṅgasamjñāni pañcānāmapī nityaśah” (N.S. 32, 6.)1 Abhinava gives details of the use to which these limbs could be put in dhruvās (A.B. on N.S. 32, 4-6): The gītaka component called mukha when transformed into a dhruvā was to be employed in dramas when the main core, or the principal story of the plot, was to be indicated (pradhānārthasūcanā mukhāt). The episode which followed the indication of the principal story (the pradhānārtha) was ushered in with a dhruvā movement based on the gāndharva component, pratimukha: (tadanu vastusūcanam pratimukhāt). A song when repeated was accompanied with the gāndharva component called upavartana (punarāvṛttyā gītamupavartanāt). The indication or the introduction of a new element in the plot was made with antāharaṇa (bhāvyārthāntarasamākṣepo’ antāharaṇāt). When the main plot was being enriched with an addition of new elements, the dhruvās used were derivative of the components venī and praveṇī (ekasyaivārthasya vaicitrayā-yogo venīpraveṇībhyām). Similarly, other components of the gāndharva gītakas transformed into dhruvās were used at other dramatic moments on the stage.2

We can deduce two structural principles which appear to have been followed in transforming a gāndharva gītaka into a dhruvā : (1) the gāndharva time-unit (five times a laghu syllable) was shortened to conform to the metric unit of Sanskrit poetry, and (2) the complex structure of the gītaka (which was musically quite independent of the metric structure of the poem being sung and was quite free from any subservience to the poetic content) was in dhruvā made to defer to the metre in the song.

Besides being, structurally, the parent body of subsequent forms, gāndharva was also the basic theortical frame-work in terms of which other forms were understood. The physical and metaphysical processes which were believed to give rise to musical sounds; the concept of śruti, gramā, mūrchanā, amśa (dominant note), nyāsa (final note), antarāmārga (typical movements characteristic of a melodic form) etc. were elements analysed by the theoreticians and codifiers of gāndharva; the same conceptual and formal notions were then also applied to its derivate hybrid forms. Abhinava, therefore, says that an understanding of gāndharva was necessary for knowing the very manner in which nāda (musical sounds) arose: “nādotpattau tadupayogi kalpitam” (A.B. on N.S. 33, 1).

1 The whole passage reads :

सप्तस्वरप्रमाṇं हि तत्तु ध्रुवेयभिनिर्जितम् । एतद् स्वरगानव्यतिक्रमं तु नानात्वेन च कृतानि तु ।

ध्रुवं वाद्यं यानि मृच्छति तानि वक्रमार्गं पुनः । मुखं प्रतिमुखं चैव वृत्तं बाह्यमपिसर्पति तु ।

स्वरतप्रवृत्तं वचः सच संहारणं तथा । प्रस्तारो मायपक्षातः स्वातुपवर्तनमेव च ।

उपपातः प्रवृत्ती च सुवर्णं संश्रयणकत । संश्रयेष्टमनुसंधानं महाज्ञानिकमेव हि ।

द्विजातिमार्गसंवादं च पदैकतालनिदर्शनम् ।

-N.S. 32, 2-6.

2 तथा हि प्रधानार्थसूचना मुखात् तदनु वस्तुसूचनं प्रतिमुखात् । वस्तुनः्यलङकारयोगेन वैहाससात् । यथास्वितप्रस्थितानुबन्धः स्थितिान् निवृत्तिरां संधाने प्रवर्तते । अर्थोपादानं वच्यामनो ह्यनुसंधानं सदृशः । वाक्यार्थसमाधिः संहारणात् ।

वर्णानुस्करः प्रस्तुतारत । संपिण्डितततालयोगेनो मायपक्षातात् । पुनरावृत्त्या गीतमुपवर्तनात् । प्रह्वानार्थान्वयायातिः क्षेपकात् । छन्दः कमायात-ग्रहाद्वादिसमुपर्णनैच गतिर्वैचित्र्य यां संश्लिष्टकात् ।

भाष्यार्थान्तरसमाक्षेपोताहरणात्...मिति पूर्वोक्तं माहाज्ञानिकम् एवमनेयं वोढुमनुसंधेयम् ।

-A.B. on N S. 32, 4-6.

Some idea of these components may be had from Part III, ‘prakaraṇa’.

Gāndharva and Gāna

127

Gāndharva was governed by a great liturgical spirit but it was no mere ritual. It was also aesthetically significant.

The 33rd chapter of the Nāṭyaśāstra recounts the qualities expected of a good musician and the defects to be avoided (guṇadoṣa-vicāra). Some theorists were of the opinion that this guṇa-doṣa section applied only to dhruvās because these could create the desirable effect only if rendered with proper skill and aesthetic impeccability. Gāndharva being basically a ritual for the attainment of adṛṣṭa, its aesthetic quality was really immaterial. Abhinava notes this opinion and makes a fitting reply to it. He remarks : "a vidhi (enjoined action) results in its fruits only when performed according to its true nature without mutilation. Gāndharva, though a vidhi, was meant to be sung and rendered on the vīṇā; this being the case, how could one, in the absence of a suitable voice and the required skill, acquire the right knowledge of 'laya' and other musical elements without which gāndharva could not be truly formed? How, in other words, could a mutilated formation of the right time-measures lead to adṛṣṭa? Therefore, has it been said (by Yājñavalkya in his Smṛti, 3, 115): 'a person with a true knowledge of vīṇā-playing, an expert in śrutis and jātis, and one who is well versed in tāla, can attain mokṣa without effort.' Thus in gāndharva, too, a discrimination of guṇa and doṣa is essential. A pleasantly sweet and effective voice is the soul of svara, and skill in handling is the soul of tāla and of singing; these constitute musical effectiveness (guṇa) and on it depends the attainment of adṛṣṭa."1

Gāndharva, Abhinava thus asserts, was musically significant, too, apart from having an adṛṣṭa fruit. This is why he defines it as having a dṛṣṭādṛṣṭaphala (on N.S. 33, 1) : both an immediately pleasing aesthetic effect as well as an adṛṣṭa fruit. Elsewhere-commenting on Bharata's statement that gāndharva was dear and pleasing to the gods (N.S. 28, 9)—Abhinava, making a similar remark, says: "it is pleasant to the ears and is enjoyable, thus it has a dṛṣṭa fruit; it is anādi (without a beginning) and thus also has an adṛṣṭa fruit: śrūyamānam prītivardhanamityevanāditvāddṛṣṭādṛṣṭāphalatvāt......"

Throughout his commentary Abhinava has taken care to distinguish between gāndharva and gāna. His detailed analysis of the matter occurs fairly late in the Abhinava Bhāratī text, at the beginning of chapter 33, but even in his introductory remarks at the beginning of the geyādhikāra (chapter 28) he indicates a clear distinc-

1 अथ केचित् मन्यन्ते । उपरञ्जनात्मनि गान एव सङ्तरां गृणपग्रहणं दोषावबन्धनं चादाय उपरञ्जनप्राणं हि गानम् । उपरञ्जन्ति ऋ गुनदोषाविचारतद्विवेकादिभिः । अत्र एवं तद्विद्वेकस्य गानं प्रति प्राधान्यं प्रयोजयितॄ प्रपञ्चयित्वा प्रतिपाद्यमारम्भः । गानध्वं तु रञ्जना न तथा प्राधान्यं तु च्छन्दप्रयोग्त् रसादिक फल तदिति । नैतत् । तथा हि यथास्वरूपं (म) विकलं प्रवृद्धो विधिः फल प्राप्नुते । शारीरद्वारवद्वारकैषा सम्पत्तिरेव च रककफादिष्वदितशास्त्वादिना विद्वान् कर्थ लयादितत्त्वज्ञानम् । यदत्र विकृतं कालसम्पत्तिरतत्स्वज्ञानस्य क सङ्गति रदृष्टफलसङ्गतिः । उक्तं हि "वीणावादन तत्वज्ञः श्रुतिजातिविशारदः" इति । "ताल" इति च । तस्माद गानध्वं सुत्रार्ं गुनदोषाविवेकौद्योग्वान् । स्वरस्य हि तावद्गुणुरगुरञ्जनात्मकमेव लचः (घु) तालस्यापि गानं प्राणास्तद्गुणमपि गुनायत्तं तदायत्तौ वादुष्टसिद्धिः

-A, B. on N. S. 33, 1.

tion between the two forms, remarking : “the distinct nature of gāndharva will be explained by us in the dhruvā section (dhruvādhyāye); still, some indications are being given here at the beginning, lest lack of discrimination be a cause for confusion.”1 Abhinava then touches upon some of the basic characteristics of gāndharva as distinct from gāna.2 He takes up the definition of gāndharva given by Bharata (viz. “gāndharva-miti tajñeyam svaratālapadātmakam”, N.S. 28, 8) and remarks that this definition could cause a certain confusion for it does not clearly indicate any distinction between gāna and gāndharva and could equally apply to both for both consisted of svara, tāla and pada. He adds that one could go further, and say that this definition applied to gāna alone (viparyayo'pi kasmānnā bhavati). One could even infer from it that gāna and gāndharva are identical as both are equally comprehended by the same definition.

He then states that in order to allay such doubts regarding the nature of gān-dharva and to pin-point its distinct nature Bharata has further specified the nature of gāndharva in the very next verse as :

atyarthamiṣṭaṃ devānāṃ tathā prītikaraṃ punah

gandharvāṇāṃ ca yasmāddhi tasmādigāndharvamucyate

(N.S. 28, 9)

“it is very dear to the gods and gives great pleasure to the gandharvas—for this reason it is called gāndharva.”

These remarks, according to Abhinava, distinguish gāndharva from gāna. They indicate that unlike gāna, gāndharva was an eternal, timeless form, for it has been called 'dear to the gods'—and naturally, the gods never part with what they cherish and so have possessed it since the beginning of time. Moreover, Bharata's words also indicate that gāndharva has an adṛṣṭa phala, for it 'pleases the gods' ; hence leads to adṛṣṭa : “tulye svarādyātmakatve gānaṃ gāndharve'ntarbhūtaṃ iti kā bhāṣā ? viparyayo'pi

1 संचायं गान्धर्वविद्वेको घृ ध्रुवाध्याये नृत्योऽस्माभिरभिधास्यते । इह इव तु दṛṣṭिको वाच्योयुक्त.यत्नाभिज्ञानमधुनैव हि संमोहो हेतु: स्यादिति ।

—A. B. on N. S. 28, 9-10.

The purport of mentioning the nṛtta section in this passage is not quite clear. It is true that in describing the distinct nature of nṛtta (ch. 4-5), Abhinava often remarks about its close connexion with the gāndharva and the adṛṣṭa significance attached to both. But as this section precedes ch. 28, गर्भाधिकार्य्य: is not the right word to use and, moreover, the nṛtta section does not contain a detailed gāndharva-viveka as such.

2 Such as the fact that the transcendental merit arising from gāndharva accrued to the performer, while the pleasing effect aimed at through gāna was directed basically towards the listener; and that the fabric of gāndharva was the fundamental implement, a upakarana with which the structure of gāna was fashioned :

मान्धर्वोडपि (हि) प्रयोक्त: फलमिति तदनुसारेण शारीरप्राघान्य फलं दṛṣṭatam | गाने तु श्रोतॄफल रक्तिप्रधानत्वेति गानस्य (गान्धर्वं) उपकरोति मनीषा ।

—A. B. on N. S. 28, 2.

Abhinava also tells us that the function of gāndharva tāla was to achieve sāṃya alone and not to echo or suggest different sentiments :

मृदतिकाठिन्यनिचयनं तदनुसारेग रसाभोगोपयोगि विशिष्य्टवर्गोद्दारेण च नोपयुज्यत इति साम्यमात्रोपयोगित्वात्ताल इत्युक्तम् ।

—A. B., ibid.

kasmānna bhavati, tādātmyameva vā katham na syādityāśaṅkāṃ śamayitumāha 'atyarthamiṣṭaṃ devānāmiti' anenānāditvaṃ sūcitam, devā hi kathamīṣṭam vijahyur 'tatheti' tena devatāpariṣadvāreṇa prītiṃ dadātītyadrṣṭaphalatvaṃ darśitam” (A.B. on N.S. 28, 9-10).

In this context Abhinava also makes some very interesting remarks regarding the aesthetic effects of gāndharva music. He says that gāndharva was capable of raising one’s consciousness to a transcendental level of ānanda or bliss, such as is associated with liberation (apavarga). Hence, he comments that Bharata's statements (in the above verse) indicate that gāndharva could fulfil the supreme puruṣārtha.1 Abhinava then quotes the oft-repeated line from an ancient Smṛti which says that “the jāti nandayanti rendered even once in accordance to the injunctions governing its forms, can expiate the supreme sin of having murdered a brāhmaṇa”. He goes on to state that the function and purpose of gāna was mainly to please the listeners while gāndharva pleased the gods too, thus serving both a dṛṣṭa and an adṛṣṭa end. Two bodies of music with such different results could not be identical.2

Thus the reader was forewarned right at the beginning of the description of gāndharva that the exposition to follow was that of a distinct body of music.

We have argued that, according to Abhinava, the first 13 verses of Chapter 29 (where Bharata has described the employ of the jātiS in various rāsas) did not strictly pertain to gāndharva. The description was important from the nātya point of view but was indicative of jāti-born forms and not of the jātiS. This rasa aspect of jātiS has been described in a new chapter and not in chapter 28 where jātiS are expounded and where this matter too logically belongs. Bharata's aim, comments Abhinava, in initiating a new chapter was to distinguish extraneous matter from gāndharva proper.3

परसंवित्सहृदयतिलकमालक्षणमुपलक्षणं, तथा तेन प्रकारेण प्रतीतेरपवर्गोचितानन्द-स्वभावाविभावनीयत्वात्तन्मयत्वपरिग्रहवैशिष्ट्यं द्रष्टव्यम् । तथाप्रकृतान्न गानादितिव्यपदेशात् 'माध्यन्दिणामिति' प्रयोत्कृप्तलक्षणं तेन ह्यन्त संविदप्रवेशलाभे गानु: फलयोग:..........—A. B. on N. S. 28, 9-10

“सकृत्प्रयुक्तापि हि नगदयन्ती यथाभ्रहत्यांऽपनाति” इति प्रयुक्तिगतमतमस्मै फलम् । ननु गान इव मुख्यतया श्रोतुनिष्ठं गन्धर्वाणां प्रीतिवर्धनमिति सम्भवः; इत्यापि च श्रुतिमार्गं प्रीतिवर्धनमित्येवंनानादिलक्षदृष्टफलत्वाच्च प्रधानं गानधर्वं प्रयोजयति” इति व्युत्पत्या, एवं देवानामिप्टं प्रीतिकारणत्वं गानधर्वाणां भवतीति सम्भः । तस्माच्च प्रीतिमेव विशेषण वर्धयति यत्तद्गानमिति॥व्यााहारः । न गानं हि केवल प्रीतिकायै वतते । तेन तादात्म्यं तावदयुक्तं ।

—A. B. on N. S. 28, 9-10.

गानधर्वस्यैव चारुत्वस्य मा सम्भाव्येतदर्थं मध्यायारचछेदः । यथा च वक्ष्यति 'ऐवमेता बृंहणीयो जातयो नाट्यसंख्या:' (म.ना. 29.13).

—ibid., 29, 1

CHAPTER VI

GĀNDHARVA AND BHARATA

Abhinava, in drawing his conclusions, refers constantly to the Nāṭyaśāstra text, its tradition and practice. His analysis which reveals gāndharva as a distinct musical entity is based on an analytical enlargement and detailed clarification of matter found in the Nāṭyaśāstra itself—where it is present in a seed-form.

Bharata has described gāndharva in a section devoted separately to it ; in delineating its distinctive features he has pointedly made injunctions showing certain features which were exclusive to gāndharva. He also compares gāndharva with dhruvā on many points, implying clearly the distinctiveness of both. Abhinava thought that these pointers in Bharata conveyed suggestions enough for the knowing to conclude that gāndharva had a distinct and independent identity as a specific body of music. At the outset of his long discussion on the differences between gāndharva and gāna (A.B. on N.S. 33, 1), he remarks : “the characteristics of gāndharva have been sufficiently described in four chapters by Bharata himself. Still, as the matter is difficult of access (anusandhānavandhya), we are here proceeding to inquire into it so that such illustrious personages (sic) (as are incapable of investigating on their own) may come to percieve it : “nanvevaṃ gāndharvasya kiṃ lakṣaṇmuktamadhyāyācatusṭayeṣu muninā ; tathāpyanusandhānavandhyo, mahābhāgaṃ bodhayitumanusandhī-yate” (A.B. on N.S. 33, 1).

During Abhinava’s times it was clear to the initiated, the ‘sampradāyavid’ that a definite section of the Nāṭyaśāstra was devoted exclusively to gāndharva as a distinct musical form. Such knowledge could not naturally be popularly current knowledge because it needed a detailed doctrinal and technical knowledge of the entire Nāṭyaśāstra where many arts and sciences have been juxtaposed together. Bharata has himself stated : “It is difficult to reach a final understanding of this nāṭya. Why ? Because it involves a knowledge of many branches of learning and of innumerable arts and their techniques. Even a single branch of learning can be like a sea, impossible to cross, what to speak of a deep and true understanding of many arts and sciences :”

na śakyamasya nāṭyasya gantumantam kathaṃcana

kasmādbahutvājjñānānāṃ śilpānāṃ vāpyanantatāḥ

ekasyāpi na vai śakyasvanto jñānārṇavasya hi

gantum kiṃ punaranyeṣāṃ jñānānāmarthatattvataḥ (N.S. 6, 6-7)

For those who wished to handle a variety of art-foms specialization was difficult and it was easy to confuse distinctions within a single art. The whole of the

geyādhikāra deals with the same art-form, namely, music; one could quite understandably loose sight of distinctions between musical forms themselves, even though the distinction between gāndharva and gāna-music was major and crucial one. This is apparently why Abhinava took pains to show in detail the distinctiveness of gāndharva.

In ancient India, a certain authoritative text or body of texts often formed the basic core on which was founded the study of a specific śāstra or vidyā. Still, very many ideas were part of the lore or teachings of sampradāyas, which were only implicit or not very explicit in the text constituting the backbone of study. These ideas were basic to a full understanding of the text and were said to have originated from the author of the text himself or from his immediate, intimate circle.

Pāṇini's grammar provides a good illustration. To arrive at a true understanding of Pāṇini's system as outlined in the Aṣṭādhyāyī, a number of key notions and explanations have to be taken into consideration, which form an essential part of the Pāṇini sampradāya, though they are only hinted at in his text. The Nāṭyaśāstra, too, contained many such hints or clue-ideas which needed elaboration and clarification.

One such idea was the clear distinction of gāndharva as a distinct musical form. Bharata's delineation of music as of the other arts constitutes basically a practical manual depicting actual forms and techniques to serve as a guide for formers of these arts, especially in relation to staging plays. His work is, to this end, a storehouse of injunctions, ordinances and decrees. Discursive, analytical problems are not disregarded, but they are secondary and they, too, consist of short dictums : as when Bharata says that certain alaṅkāras should only be formed in gāndharva and not in dhruvās and that the gītis should be formed in gāndharva alone. Bharata has, in view of the scope of his work, naturally taken for granted much of the conceptual and discursive framework underlying the art-forms he describes.

Yet though Bharata does not enter into detailed comparisons, he does give many pointers regarding the distinctiveness of gāndharva. We have already discussed the most significant hints in Bharata concerning gāndharva as an exclusive entity in dealing with Abhinava's analysis of the matter. It would not be out of place here to recapitulate some major pointers and also take up some others :

(1) The gāndharva uddeśa—the comprehensive and definitive enumeration of topics defining the limits and scope of this body of music—acts as a sure guide for judging what is and what is not gāndharva. It reveals that the matter relating to gāndharva ends with the 31st chapter.

(2) There are some verses that indicate the sacredness attached to gāndharva and show the adrṣṭa-motive behind it. One such verse we have discussed, namely, verse 9 in chapter 28. Another verse—though occurring somewhat out of context—in the 33rd chapter recounts gāndharva as a very ancient form, its creator being Prapitāmaha, Brahmā himself :

gāndharvametat kathitaṃ mayā tat (vaḥ) pūrvaṃ yaduktam prapitāmahena (N.S. 33, 23).

"Gandharva, which I have described to you, was propounded in ancient times by Prapitāmaha (Brahmā)."

A variant reading for ‘prapitāmahena’ in this verse is ‘ihanāradena’ (see footnote, N.S., Vol. IV, p. 402). Nārada was traditionally credited as being one of the earliest perpetrators of gāndharva. Dattila calls Brahmā the creator of gāndharva and Nārada the sage who taught the art to the mortals (Datt. 2).

3. We have observed that the chief use and function of gāndharva in ancient theatre was in the pūrvarañga where it performed a propitiatory role. Bharata gives a mythic story relating the reason and the circumstances which led to different aspects of gāndharva being accorded slightly different propitiatory roles in the purvarañga.

We have seen that the pada or the word-content in gāndharva could either be a hymn to Śiva, recognised as part of the gāndharva repertoire, or a collection of nonsense syllables strung together like a ‘tarānā’ and called nirgitā, bahirgītā or śuṣka. It was thought in Bharata’s time that the meaningful hymns were pleasing to the gods, while the nirgitā served as propitiatory songs for mollifying the demons (daityas). The story, as told by Bharata, goes that saptarūpa or gāndharva was once being performed by Nārada and the gandharvas in an heavenly assembly where both gods and demons were present. Both nirgitā and hymns to the gods were sung. The demons were greatly enraged at hearing the praises of the gods. They then conferred together and announced that the nirgitā portion of gāndharva would belong to them. Gāndharva thus acquired the power to propitiate the gods and demons alike :

citradakṣiṇavṛttau tu saptarūpe pravartite sopana sairgite devastutyabhinandite nāradāyāstu gandharvaih sabhāyām devadānavāḥ nirgitām śrāvitaḥ samyaglayatalasamanvitam tacchrutvā tu sukhāṁ gānam devastutyabhinanditam abhavankṣubhitāḥ sarve mātsaryāddaitaryārakṣasāḥ sampardhāryā ca te’nyonyamityavocannavasthitāḥ nirgitām tu sāvaditramidam gṛhṇīmahe vayam saptarūpeṇa santuṣṭā devāḥ karmānukīrtanāt vayam gṛhṇīma nirgitām tuṣyāmo’ traivasarvadā

(N.S. 5, 31-35)

A few verses later, Bharata again reiterates :

gītakeṣu prayukteṣu devāstuṣyanti nityaśaḥ vardhamāne prayukte tu rudrastuṣyati sānuṣaḥ

(N.S. 5, 47-48)

In the 31st chapter, Bharata has a verse extolling vardhamānaka which indicates its adṛṣṭa-end :

laksyalakṣaṇasampatyā mārgayuktividhikramaiḥ vardhamānāprayoktāro yāsyanti śivagocaram

(N.S. 31, 73)

“Those who perform vardhamāna, in accordance with its (śāstric) exposition and render it in the various mārgas (i.e., citra, dakṣiṇa and vṛtta) according to

the proper decrees following the true order of performance, can attain the nearness of

Śiva Himself."1

(4) In his chapters on music, Bharata has left many indications revealing the

distinctiveness of gāndharva in comparison with dhruvā, such as in describing alañ-

kāras in the two foms, gitīs, the nature of padas. There is another hint which, though

a comparatively minor one, does add to the body of evidence. In delineating gāndharva

Bharata addresses himself to the ‘gāndharvavedins’. Thus in specifiying a certain point

in gāndharva formations he states : “the jāti ṣadjodīcyavā should be rendered ṣāḍa-

vita (hexatonic) by the gāndharvedins through dropping the note ṛṣabha (ṣaṭsvary-

amṛṣabhāpetam kāryaṃ gāndharvedibhiḥ N.S. 28, 108)”". Gāndharvedin, as is

clear from the context, evidently refers to those who knew gāndharva as a particular

body of music and not generally to ‘knowers of music’. In the section on dhruvā,

Bharata never speaks of gāndharvavedins but invokes the experts who know the rules

of dhruvā’ (dhruvāvidhānajñāḥ).2 He also uses neutral addresses such as ‘those who

know the subject’ (tajjña) or those who are “experts in performing” (prayokṛ) or those

who know (budhah) or “those who know singing”,3 but never gāndharvavid or

gāndharvavedin.

(5) Gāndharva was as an art-form independent of nāṭya. Though it played

a functional role in the pūrvaraṅga of the ancient theatre, yet it was not moulded to

the use of dramatic plots for creating programmatic effects. Bharata, therefore,

mentions gāndharva and nāṭya separately in some places showing the independence

of gāndharva from naṭya :

(a) gāndharvaṃ caiva nāṭyaṃ ca yaḥ samyaganupaśyati (N.S. 36, 78)4

(b) gāndharvaṃ caiva nāṭyaṃ ca dṛṣṭvā cintāmupāgamat (N.S.37, 2)

1 Abhinava comments :

लध्य लक्षणसम्पद्यते । मार्गाश्चित्तनिदशः; युक्तितर्योगनम् परितर्त्तिमको विधिः । यथासदृशसंख्यातद्विसंख्यातरूप-क्रमपरिवर्त्तनेनैवारितोल्लक्षणं च सम्पदा श्टितत्वीरुपलक्षण द्वयात्मकं ये प्रयुञ्जते ते शिवगोचरं

हृद्याहृष्टपर: श्रेय: प्राप्तिलक्षणं यास्ति। भविष्यत्प्रयोगणं विविध फलानां धिय: फलमूलनतत्वाकरणमित्याह । बहुवचन-प्रयोगेण गान्तुवादकनटंकानां तुल्यफलत्वमाहू।

—A.B. on N.S. 31, 73.

2 एप नयकरपिण्डो विज्ञेयो व ध्रुवाविधानज्ञै:

—N.S. 32, 240.

3 ध्रुवा प्रासादिको कार्यो तर्ज़ीमंङयलयान्वय:

—N.S. 32, 325.

सानुबन्धा बुद्ध: कार्यो गतिभैरस्तरा ध्रुवा

—ibid., 32, 326.

उद्दता तूद्यता तस्मात् न या ध्रुवा बु्रधी:

—ibid., 32, 330.

एवं भावान् विदित्वा तु श्वस्ता कार्यो प्रयोक्तृभि:

—ibid., 32, 344. etc.

4 We have quoted this verse from Kane’s History of Sanskrit Poetics, p. 19. In the G.O.S.

edition of the N S. the verse occurs in the 37th chapter towards the very end of the text, as

No. 30. The line has a slightly variant reading :

गान्धर्वं चेह नाट्यं च य: सम्यक् परि (स्यमन्त) पालयेत् ।

134 A Study of Dattilam

Professor Kane interprets the term gāndharva in such usages in a non-pāribhāṣic sense and understands them as meaning music in general : and as such- since music formed an integral part of ancient drama-he is not in favour of according an independent stature to gāndharva as separate from nāṭya. He says: “The Nāṭyaśāstra declares that sṛṅgāra is excited by flowers gāndharva and the reading of kāvya etc. (N.S. 6, 47). Too much emphasis should not, therefore, be laid (as is done in I.H.Q., Vol VI., pp. 72-80) upon the fact that gāndharva and nāṭya are separately mentioned in some places by Nāṭyaśāstra (as in N.S. 36. 49 and 78). This separate mention is due to the sāmānyaviśeṣanyāya or brāhmaṇaparivrājakanyāya” (History of Sanskrit Poetics, p. 19). Prof. Kane’s remark would be quite true if gāndharva meant music in general. But in view of the convincing evidence which reveals gāndharva as a specific body of music, which did not form an integral part of nāṭya, it would not be unjustified to interpret Bharata’s statements as indeed implying a separateness between gāndharva and nāṭya.

A statement of a somewhat similar intent by Bharata occurs in the 34th chapter of the Nāṭyaśāstra :

gāndharvaṃ caiva vādyạṃ ca svātinā nāradena ca

vistāraguṇasampannamuktạṃ lakṣaṇakarmataḥ.

The verse does not occur in the reading accepted by Abhinava,1 but is verse 3 in a somewhat different recension of the chapter (see N.S. Vol.IV, p. 464) which describes the various drums and modes of drum-playing used as accompaniment of ancient dramas. The above verse means : “Svāti and Nārada have described gāndharva and vādya in detail along with their specific qualities and have expounded them in accordance with their functions.” The chapter where the verse occurs deals with vādya as related to gāna in Abhinava’s sense of the term and hence gāndharva, in this context, must be understood as denoting a form quite separate from it. The association of Nārada with gāndharva is also significant, for Nārada has been reputed to be one of the earliest perpetrators of gāndharva. Narada’s connection with vādya in general - not restricted to gāndharva alone-as indicated in this verse is also not surprising because apart from gāndharva Nārada’s name is associated also with dance, drama and music of all variety. Svāti was another ācārya of almost mythic repute like Nārada. His name occurs also among the ancient sage-like authorities on music whom Rāṇā Kumbha, for one, had studied: “vāyusvātimahendrakaśyapamamarutsūnu…” (S. Raj. 1, 1, 1, 40).

1 Where the corresponding verse is:

यथोक्तं मुनिभिः पूर्वं स्वातिनारदपुष्टकरः।।

सर्वलक्षणसंयुक्तः, सर्वां (तथा) गीतविधूतिम्‌ ॥

—N.S. 34, 2.

CHAPTER VII

SOME MINOR GĀNDHARVA FORMS

We see in the Yājñavalkya Smṛti that besides the seven gītakas of the gāndharva (which have been collectively termed ‘prakaraṇa’), other forms are also named in association with these and are also said to result in mokṣa or the attainment of Śiva’s supreme company. The Smṛti names these forms with the words :

ṛggāthā pānikā dakṣavihitā brahmagītikā geyametattadabhyāsakaranān mokṣasajñītam.

(Yājñavalkya Smṛti 3, 114)

We find that four forms are named, ṛk, gāthā, pānikā and brahmagītika,1 the third viz. pānikā is desbribed as ordained by Prajāpati Dakṣa’ (dakṣavihitā), who according to Purāṇic accounts was a mind-born son of Brahmā.

An intriguing question here arises: What were these forms and what was their relation with gāndharva? The gāndharva topic ‘prakaraṇa’ encompasses only the seven gītakas: madraka, ovenaka, prakarī, utara, ullopyaka, aparāntaka and rovindaka (these were signīcantly, also named the saptarūpa). How were these other forms related to the seven gītakas ?

A passage from an unnamed ancient text quoted by Abhinava (perhaps from a Purāṇic chapter now lost) recounts some more forms of the same category as the four we have already mentioned :

ṛco vai brahmagītaśca brahmaṇābhitāḥ kila dakṣeṇa pānikāścapī gāthā vai kaśyapena tu mātrbhiśca kapālānī sāmānyuktāni nandinā gītakādi tu saptaiva nāradenoditāni vai

(A.B. on N.S. 31, 54)

“the ṛks and the brahmagītās have been laid down by Brahmā. Dakṣa has ordained the pānikās and Kaśyapa, the gāthās. The kapālas have been propounded by the Mātṛkas and the sāmas by Nandi. The seven gītakas have been propagated by Nārada”.

Here, we are told of two other forms, kapāla and sāma, besides the four: ṛk, brah-magīta, pānikā and gāthā, mentioned in the Smṛti list. As in the Smṛti account, these forms are mentioned in conjunction with the seven gītakas constituting the topic prakaraṇa of gāndharva. Like the gītakas, these other forms, too, are accorded divine authorship.

1 Vijñāneśvara in his Mitākṣarā, commenting on the verse, says :

ऋग्गाथाद्याच्चतस्त्रोगीतिकाः इत्येतद्परान्तकादिगीत जातमध्येारो पिततामभार्ं मोक्षसाधनत्वान्मोक्षसंज्ञितं मन्तव्यम् ।

136 A Study of Dattilam

Śārṅgadeva also names these above mentioned forms along with the seven gītakas and the vardhamāna. His enumeration and accompanying remarks are interesting:

etaih prakaraṇākhyāni tālairyāni jagurbudhāḥ tani gītāni vaksyāmastesāmadyam tu madrakam aparāntakamullopyam prakaryoveṇakam tataḥ rovindakottare sāpta gītakanityāvidhisuḥ chandakāsārite vardhamānakaṃ pāṇikam tāthā ṛco gāthā ca sāmani gītānīti caturdaśa śivastutau prayojyāni mokṣāya vidadhe vidhiḥ

(S.R. 5, 53-56)

"The gītas termed prakaraṇa which the wise have described as composed of these (aforementioned) tālas, we shall now describe : the first of these is the madraka, followed by aparāntaka, ullopya (ka), prakarī, oveṇaka, rovindaka and uttara—these have been termed the seven gītakas. Chandaka, āsārita, vardhamānaka, pāṇikā, ṛk, gāthā and sāma; (counting these) the gītas are fourteen in number. They are to be employed in singing praise to Śiva; Brahmā had performed them for attaining moṣka."

We observe that Śārṅgadeva intimately associates forms such as chandaka, āsārita, vardhamānaka, pāṇikā etc. with the saptarūpa and yet accords to them a separate category : after listing the gītakas of the saptarūpa (madraka etc.), he calls them ‘the seven gītakas’, then after recounting other sc ven forms he brackets them together with the gītakas. The fourteen forms are together called the ‘gītas’ that have a sacred significance. Of the non-saptarūpa forms that he enumerates, we have already become somewhat familiar with vardhamāna. The āsārita (named in Śārṅgadeva’s list) was a limb or component of vardhamāna and perhaps also, at times, had an independent entity. Chandaka seems misplaced in this company and has indeed not been associated with pāṇikā and others by the older passages from the Smṛti and the ancient text quoted by Abhinava.1 We notice that the gītas, kapāla and brahmagīta, are missing from Śārṅgadeva’s list.

Kapāla, however, is mentioned in another context. Śārṅgadeva associates kapāla with the jātiṣ though not with the gītakas. He says that the kapāla gītas are of seven kinds and each is born of a śuddha jāti (these being : śāḍjī, ārṣabhī, gāndhāhrī,

1. Chandaka was a component rendered at the end of the gītakas. Its formation seems to have been optional. In any case it seems to have had no independent entity. Śārṅgadeva himself qualifies chandaka with the words :

अन्ते गेयं गीतकान्तं छन्दक

—S.R. 5, 180.

Kallinātha makes the purport explicit :

इदं छन्दकं मृदङ्गादिषु, एकं गीत्वा तदन्ते गेयमिति । अस्य स्वातन्त्र्येण प्रयोगो नास्तीति प्रतीयते ।

—Kalā on ibid.

madhyamā, pañcamī, dhaivatī and naiṣādī). He adds that the kapālas resembled the rāgas born of these jātiṣ (S.R. 1, 8, 1).

Although kapālas were born of jātiṣ and were in this respect similar to rāgas, yet they formed a class quite apart from the rāgas. They were classed with the jātiṣ both structurally and in spirit.

Of the 18 jātiṣ, 7 were śuddha and the other 11 were mixed (samsargaja); they were born of the śuddha ones, so were the kapālas. Kallinātha commenting on Śārṅgadeva says that the kapālas were akin to the samsargaja jātiṣ because they shared a common source with them. He clearly implies that kapālas should be bracketed with the jātiṣ rather than the rāgas.1 His words are worth quoting: “After the jātiṣ have been described, now (beginning with S.R. 1, 8, 1) the kapālas, born of the śuddha-jātiṣ, ṣāḍjī etc. —this proclaims that the kapālas are not born of mixed (samsargaja) jātiṣ. A question here arises : both śuddha and samsargaja jātiṣ have common jāti-hood (jātiṭva); why then may the kapālas not be derivative of the latter? The answer is : the samsargaja jātiṣ, like the kapālas, are also born of śuddha jātiṣ; both thus share the same causal source from which they have directly and immediately descended; sharing this trait in common with the kapālas, the samsargaja jātiṣ cannot be said to have caused the kapālas.” Kallinātha clearly equates the kapālas with the samsargaja jāti and implies that they are as much a part of gāndharva as the 11 mixed jātiṣ. He continues: “The purpose of knowing the kapāla forms is stated by (Śārṅgadeva) with the words ‘rāga’ etc. (see second line, S.R. 1, 8, 1). By the word ‘janakajāti­nām’ in this context are denoted the ‘śuddha’ jātiṣ. Rāgas born of them are the janya-rāgas and we shall later (in the rāgādhyāya) come across dictums—such as the one describing śrīrāga—which state ‘śrīrāga is based on ṣaḍja and is born of the ṣāḍjī jāti’. Rāgas of this type are akin to the kapāla born of the some parent jāti, as for example the ṣāḍjī. These rāgas exhibit a structure similar (to the corresponding kapālas). The purpose underlying the description of kapālas is to help in understanding the raga-forms. Just as on seeing a kapāla (the bowl portion of a pitcher) which is, verily, a part of the pitcher-structure the pitcher itself can be recognised, similarly a rāga can be recognised in a kāpāla-gīta, which, has a structure akin to a portion of a rāga-structure. This is also the reason why kapālas have been so named.” Thus kapālas had, like mixed jātiṣ, some traits in common with rāga forms, and yet they were different from rāgas for they bore a closer kinship with the parent jāti.

1 अधुना जातिस्वरूपनिरूपणानन्तरं शुद्धजातिसमुद्भूतकपालानि शुद्धजातिम्यः; षाड्ज्यादिम्यः सतस्तभ्यः समद्भूतानि कपालानि एतेषां संसर्गजातित्वम्। कपालोत्पत्तिनियम्युक्तं भवति । ननु जातित्वाद्विशिष्टताभ्योऽपि कपालोत्पत्तिः कुतो नेति चेत; उच्चते । कपालान्यपि तास्मै एव उभयेषामप्येक कारणाद्वैधावैगोल्यनतस्याविशिष्टत्वाद्वांशसंर्जनां कपालजातीनां प्राय कपालन्वयमेवादिति; कपालपरिज्ञानस्य दृष्टप्रयोजनं तादृशोऽयं—“राग” इति । जनकजातीनाम्; ग्रह शुद्धजातयो जनकत्वेन विवक्षिता:; ता:षां राग: तत्त्कालीनो रागा:; तत्सां पाड्ज्यादीनां कपालेपु समिता: सद्दशाकारा:; प्रत्येकदेशे:वेन कपालानि तेषां संज्ञाज्ञानवत्त्वया । कपालानीति तेषां संज्ञाज्ञानवत्त्वया ।

-Kalā on S.R. 1,8, 1.

The kapālas were not only structurally akin to the jāti, they were also governed by the same sacred, propitiatory spirit. Śārṅgadeva says that their musical structure as well as the hymns sung to them were composed by Brahmā. Also that kapālas were dear to Śiva and a person who sang them according to their ordained forms achieved spiritual welfare:

iti sapta kapālāni gāyanbrahmoditaiḥ padaiḥ savaraiśca pārvatīkāntastutatu kalyāṇabhagbhavet (S.R. 1,8, 10).

"Such are the seven kapālas, which when sung to the words and music composed by Brahmā in praise of Śiva, result in spiritual well being (kalyāṇa)."

Kallinātha relates the mythic story of how Brahmā composed the seven kapālas. Once Śiva was out collecting alms. He was singing songs set to jātis such as the ṣādjī. So moving were Śiva's songs that on hearing them the Moon who dwells on Śiva's crest was moved to melt. Its amṛta (life giving water) began to drip on the skulls of Brahmās (dead Pitāmahas who had ruled over bygone 'kalpas') that Śiva was wearing as a garland. These skulls—kapālas—of bygone Brahmās immediately came to life and began singing songs in imitation of jātis being sung by Śiva. These songs sung by the kapālas of bygone Brahmās came to be known as kapāla-gāna.1

Rānā Kumbha describes the kapālas in a like context—after having described jāti. In addition, he records hymns to Śiva supposedly sung by the kapālas. He also gives notations (svaralipi) for each of the kapāla songs. The source of the notations is unknown. Their authenticity appears dubious although Kumbha traces them to no less an authority than Mataṅga.2 Unlike his predecessor, Śārṅgadeva, Kumbha notes three sets of kapāla-gāṇas, the first two sets contain 8 kapālas, the third contains 7. The first set is ascribed to Brahmā and termed brahma-kapāla-gāna; each kapāla of this set is named after a synonym of Brahmā. The second set is called the nānā-deva-kapāla-gāna: the eight kapālas in this set are ascribed each to a different god such as Āditya, Indra, Kāma, etc. The third set corresponds to Śārṅgadeva's description of the kapālas: each of the seven kapālas here are said to have been born of the different śuddha jātis.

Regarding the mythic origin of kapāla-gāna, Kumbha relates a story similar in outline to the one related by Kallinātha, whom Kumbha in this instance, evidently, follows. In Kumbha's slightly altered version, the melting of the Moon takes place

1 पुरा मिक्षाटनसमये शङ्कना पाइज्यादिपु नीयमानास निरविष्टयरसामिध्यवत्या रसादिमकार्या तन्मीलितमतचन्द्रकाल्यो च स्वन्त्यां तदमृतरसिक्ततनि तदभूषणबहुकपालानि तदा सजीवन्ति तद्गानमनुकृत्यागायन् किल । अतः कपालगीत-स्वाल्कपालसङ्जानीतिसिद्धम् । पाइजोकपालादिलक्षणानि स्पष्टार्थानि । एवंपर्या जातिभिः स्वरूपमेदस्तु न्यायादिभेदाद उच्चतमनियमेन कपालानां गातुर्दृष्टफलयोग्यं चाह—‘इति सप्त कपालनीति’

—Kālon S.R. 1, 8, 1-10.

2 कपालगानं तदिंद प्रसिद्ध वेदसमितमतम् । विधृणोति महोपाली मतङ्गस्य मतादिह ।

—S.R, 2, 1, 4, 424-425.

The extant Br. names and describes the jātis along the sanctioned lines of gāndharva, but mentions no kapāla-gāna.

during Śiva's dance; the kapālas that consequently come to life belong not only to the Brahmās of the past but also to the other gods.1

Like Śārṅgadeva, Kumbha has in many passages noted the sacerdotal character of kapālas; he connects them like jātiś, to sāma-gāna.2

Ancient texts, we have seen, speak of a brahma-gīta, ascribing its composition to Brahmā. Neither Kumbha nor Śārṅgadeva mention it. Perhaps it was this brahma-gīta, which was later embroidered with a mythic legend of origin, and was then termed kapāla-gāna. This explanation is attractive, but has one major snag: the brahma-gīta has been associated with the gītakas, not with the jātiś and was probably a tāla-structure basically, but the kapāla as described above is clearly a svara-structure associated with jātiś.3

Another gāna similar to the kapāla, namely, the kambala-gāna, is also described by Śārṅgadeva and Rāṇā Kumbha. Śārṅgadeva indicates its basic musical structure and possible variations on it, in two kārikās (S.R. 1, 8, 11-12). He ascribes its creation to the famed sage-musician of antiquity, Kambala. He remarks :

prītaḥ kambalagānena kambalāya varaṃ dadau purā purāriradyāpi prīyate tairaṭaḥ śivaḥ

(S.R. 1, 8, 13)

"In ancient times Śiva, being pleased with kambala-gāna, granted Kambala his desired boon; even to this day these songs give Śiva joy."

Kallinātha says that the Kambala who composed the kambala gāna was a famed Nāga of this name.4

Kumbha also describes kambala-gāna after kapāla. His description is similar to that of Śārṅgadeva whom he has obiously followed in this context. Like his predecessor, he observes that kambala-gāna was sung by Kambala-Nāga in order to please Śiva: "yatkambalākhyanagena prītaye pārvatīpateḥ gītam" (S. Raj 2, 1, 4, 471).

1 Kumbha deals with the kapālas at a considerable length.

–See S. Raj 2,1,4, 416-69.

2 न तथा प्रोक्तं देवो नानाभानोः । यथा कपालगानेन संगानसमेन च ॥

–S. Raj 2,1,4, 425-426.

कपालगानं वेदैर्नामनस्मितवर्ण्यते । तेनेष्टं सकलेश्वर्यैर्गीतं सामिरनवहम् ।।

एतत् स्वरूपद्वय यो गायति शिवानुग्रहात् । स सङ्गसङ्घर्षं प्राप्नुहि महतामपिसंश्रयम् ।

–ibid. 2,1,4, 436-44. etc.

3 We must also not forget that the passage quoted by Abhinava ascribes different authorships to the brahma-gīta and the kapāla. Abhinava, moreover, quotes another passage in the same context, a line from which reads :

ऋषिगणैः ब्रह्मगीतस्य कपालानि दशोदयः ।

– A.B. on N.S. 31, 54.

Here, too, brahma-gīta is clearly distinct from kapāla—it cannot be said that the pada ‘brahma-gitāśca’ qualifies ‘kapālāniti for they bear different genders. It is not, however, unlikely that some text did speak of kapālas as sung by Brahmā, and this is what gave the later authors their cue.

4 कम्बलनाम्ना एवण्डपरशण्कुण्टलेन कुण्डलीन्द्रेण गीतवादस्य कम्बलसंज्ञा ।

–Kalā on S.R. 1,8, 11-13.

140 A Study of Dattilam

The Mārkandeyapurāṇa (ch. 21) relates the story of two Nāga brothers who were granted supreme knowledge in music by Sarasvatī herself. Kambala was one of the brothers. The Purāṇa relates that these brothers worshipped Śiva through the seven gītakas—evidently the saptarūpa of gāndharva. There is no mention, however, of Kambala having composed new songs in praise of Śiva. It is the achievement of Kambala's brother Aśvatara which in the Mārkaṇdeya account comes out as the more creditable one : it was Aśvatara who worshipped Sarasvatī and received the boon of music, both for himself and for Kambala, from the Goddess (see also section VII). It is not unlikely, however, that other versions of this Purāṇic story described Kambala as the author of kambala-gāna.

Though kapāla is mentioned in connection with the gītakas (in the ancient passages that have come down to us), the kambala-gāna is not mentioned either in connection with jāti or with gītaka ; nor is its name to be found in the surviving ancient texts on music like the Nātyaśāstra and the Dattilam.1

Bharata mentions by name only ṛk, gāthā and pāṇikā, and does so in the context of the gītakas :

"ṛggāthāpāṇikādīnāṃ saptarūpaṃ prakīrtitam" (N.S. 31, 368).

Here Bharata has used an 'etc.' after ṛk, gāthā and pāṇikā. This according to Abhinava was a pointer towards other similar forms such as sāma and kapāla: "ādiryeṣā-mityupalakṣaṇadviniyuktetena sāmakapālādīnā ṛggāthāpāṇikādi hi svarūpeṇaiva lakṣitamiti" (A.B. on N.S. 31, 368). Abhinava, too, we notice, uses an 'etc.' after sāma and kapāla. Probably he intended to include the brahma-gīta too within this class. We do not know what other forms besides these were also included in this class, for no definitive ancient list is available.

Other comments of Abhinava on the above verse from Bharata are worthy of attention. He says, firstly, that ṛk, gāthā etc. were tāla-forms. Secondly, he explains Bharata's statement as denoting the fact that the structure of these forms conformed to the gītaka.2 Another line from the Nātyaśāstra reiterates: "yā recaḥ pāṇikā gāthāḥ saptarūpāṅga eva ca" (N.S. 32, 2)—ṛks, pāṇikās, gāthās are, verily, composed with the components of the saptarūpa (also translatable as 'they form part of the sapta-rūpa').

1. The extant Br. also makes no mention of kambala-gāna. It is significant that while Sārṅgadeva in his very uddeśa names kapāla and kambala-gāna along with the jātis :

जात्यादिषु महागीतं कपालानि च कम्बलम्

-S.R. I, 1, 36,

in Br. and its Vrtti—though long sections on the jāti are found—nowhere are these two forms mentioned. The Br. section on the gītakas, if such a section was there, is not available.

2 अथान कम्प्यधिकामुगादितालतस्त्ररूपमस्ति । ततः कसमांनोऽपि मित्याशङ्क्याह । 'ऋगादीनां यतस्त्ररूपं गीतकरूपान्तरित्यमेवेति यावत् ।

'ऋगाथापाणिकादीनां सप्तरूपमिति' ।

-A.B. on N.S. 31, 368.

elsewhere Abhinava says :

नन्वारितबद्धेमानुगोतकपान्तरपानिकादिच्छन्दकप्रभृतीनां तावत् तालात्मकत्वम् ।

-ibid. 31, 26.

We find a similar observation in Dattila. After describing gītakas, Dattila describes the vardhamānaka and then has 4 kārikās on the pāṇikā (where the structure of this form is described in his compressed sūtra-like style, Datt. 232-236). He then remarks :

athehānuktatālānāṁ gītāvayavasambhavah

tālaḥ sa + + + + + + + + bhyāṁ samsādhyo

yaktitaḥ kvacit

(Datt. 236-237)

"Now regarding tālas not here described ; their tāla (structure) is derivative of gīta(ka) components and can be fitly reconstructed with the help of (the two beats śamyā and sannipāta ?)."

These undescribed tālas (spoken of in context of the pāṇikā) obviously point to ṛk, gāthā, sāma, kapāla, etc. which were of the same class as the pāṇikā. Dattila says that the structures of these tālas were built of the same components as the gītakas—a statement parallel to that of Bharata. Dattila, indeed, thought that his description of the gītakas, along with vardhamāna and pāṇikā was enough for an expert to be able to formulate the other tāla-structures akin to these. Like Dattila, Bharata, too, has described the pāṇikā, and the pāṇikā alone, among tāla forms of this class. Uncharacteristically, he has only a single verse on pāṇikā (viz. N.S. 31, 325). Even Dattila, who was being deliberately brief, gives a fuller description.1

We know of another gītaka, which though not a part of the saptarūpa form. This was the vardhamānaka. Both Bharata and Dattila have devoted a large section to the vardhamānaka. Vardhamānaka was formed with the same structural elements and on the same principles as the gītaka. Pāṇikā and other such forms emerge as kindred of the vardhamānaka ; they, too, were shaped like the gītakas and were governed by the same transcendental, propitiatory adṛṣṭa spirit.

The saptarūpa, we have argued, resulted in the dhruvā through structural transformations. Ṛk, gāthā and pāṇikā were akin to the saptarūpa in this respect also (N.S. 32, 2). Abhinava names them along with the gītakas as the formal basis for the five dhruvās: “among the limbs or components of the gītakas and of ṛk, gāthā and pāṇikā those components which shine out as the most excellent should, on the basis of the fact that they share some general characteristics (with dhruvā-forms), be turned into the five assemblages (saṅghātās) such as the prāveśika and other (dhruvās)—

gītakānāṁrggāthāpāṇikānāmavagato yat kiñcitutkṛṣṭamaṅgarupan lakṣyate tat sāmā-nyānuguṇaiḥ prāveśikyādibhih pañcabhih saṅghātaiḥ sthāpitavyam” (A.B. on N.S. 32,

1. This is obviously the reason why later authors like Sārṅgadeva and Kumbha in describing the pāṇikā have borrowed substantially from Dattila, though without acknowledging the debt. Bharata's descriptions on this point were too short. See ‘pāṇikā’ ch. IV.

4-6).1 The gāthā structure, Abhinava giving an example remarks, was the source for the use of sweet and charming rhythms in dhruvās : “gāthābhyo hṛdayālhādakavṛtto-payogah” (A.B. ibid.)

Though considered as part of gāndharva, pāṇikā and others of its class are treated in a somewhat off-hand manner by Dattila and Bharata and are clearly not accorded the same importance and status as the saptarūpa and vardhamānaka. The reason is difficult to envision. There is, however, a short laconic statement by Abhinava which might have some bearing on this matter. Regarding the pāṇikā, he says that “the tāla-arrangement in pāṇikā though also constituted of śamyā etc. (like saptarūpa) has the measure or form of an inferior or destitute saptarūpa : pāṇikāyāstālavibhāgaḥ śamyādikamapi hīnasaptarūpaprapramāṇam” (A.B. on N.S. 32, 4-6).

In contrast to the gītakas and the vardhamānaka, pāṇikā and others of this class were, apparently, softer, gentler, more tender forms, perhaps also lighter in their ethos and expression. Abhinava quotes the opinion of one Bhaṭṭa (“bhaṭṭamatenāpi”) who remarked “gāthānāṃ pāṇikānāṃ ca gānaṃ masṛṇasampadā : the songs of gāthās and pāṇikās are pervaded with softness.” He adds that such qualities passed from these forms to the dhruvās and were the source for similar attributes (in them) : ityādi lhruvākārye kāraṇavad dharmānugamāt.” (A.B. on N.S. 32, 4-6).

The ‘feminine’ character of pāṇikā and others is also testified by the fact that these forms were used as accompaniments for the lāsya in the pūrvaraṅga. Lāsya, we have seen, was the dance suited to the sukumāra variety of plays where the dominant sentiments were śṛṅgāra and hāsya. Abhinava relates pāṇikā to the lāsya, and introduces Bharata’s verse on the pāṇikā with the words : “atha sukumaraprāyāllasyad-vāreṇa pūrvaraṅgoyoginīṃ pāṇikāṃ lakṣayati : now the pāṇikā is being expounded which being largely delicate in character is useful in the pūrvaraṅga through the lāsya” (A.B. on N.S. 31, 325). Later, commenting on the same verse, Abhinava again says : “(pāṇikā) being itself delicate- sukumāra-is predominantly used in the pūrvaraṅga of a sukumāra play when the (śāstra) enjoined dance movement of such a pūrvaraṅga are being rendered.” He adds : “It (the pāṇikā) belongs to the same category as the gītaka and thus its exposition in this context is quite appropriate.”2

It seems likely that forms like the pāṇikā stood in the same relation with the gītakas of gāndharva as does the ṭhumrī with the Hindustani classical forms of today, the khayāl and the dhrupad. The ṭhumrī though quite as classical as the more austere khayāl and dhrupad is a ‘light’ classical form—it is classical and yet not quite classical—being more pliant and feminine in structure and spirit.

And yet pāṇikā and others of this class seem to have been quite unlike the ṭhumrī of our analogy in one principal aspect : they do not appear to be, chronologically,

1 See also :

नारददैवैःगीताङ्गानि वर्णांङ्गानि सप्तरूपमिमानि मुख्यप्रतिमुखानि सप्तरूपप्रमाणं तु यथेच्छतस्तथादि याश्चर्य ऋग्मायाः; पाणिका इह निघृतानि तदैवं “ध्रुव इति संहितम् । नात्रतं किश्चित्वदिति नपु सक्केन गीताङ्गसप्तरूपप्रप्रमाणानां सम्बन्धः ।

या इत्यनेन ऋग्माषापाणिकानाम

—A B. on N.S. 32, 1-2.

2

सुकुमारप्रयोगे हि पूर्वरङ्गे तदङ्गे वा चारीप्राये नियमे च प्रधानं सुकुमारत्वात् । एवं च गीतकतुल्यकथयतयैवेयं लक्ष्यायते

युक्ता ।

products of a later development like the ṭhumrī which bears a similar relation to older

forms like khayāl as do Baroque and Rocco to the historically earlier Renaissance

forms in European architecture. If names are at all to be taken as guides, pāṇikā etc.

would, rather, seem to be older than the gītakas. Some of these forms have names such

as ṛk, sāma and gāthā1—implying a relation with the Vedas. The relation is not clear but

some connection is borne out by the description of these forms as found in Sārṅgadva,

who evidently based his descriptions on the authority of ancient sages.2 In the ṛk, he

says, the pada or word content could constitute of either secular or Vedic passages:

"laukikairvedikairvāpi gātavyamṛcamūcire" (S.R. 5, 224). Further, such time units—

kalās—as were devoid of words in ṛk could be filled in or 'completed' (kalānāṁ pūraṇam)

with the help of 'words from mantras' or, alternatively, with a series of prescribed

nonsense syllables uttered by Brahmā (S.R. 5, 225).

The gītaka called sāma was rendered the stobha (nonsense) akṣaras of the

Vedic sāmagāna : "stobhaṅgīṁ vijānīyatsāmno vaidikasāmavat" (S.R. 5, 230). The

Vedic sāmagāna was sung in (S.R. 5, 231-232) seven parts known as prastāva, udgī-

thaka, pratihāra, upadrava, nidhana, hiṅkāra and oṅkāra. The gāndharva sāma had a

corresponding arrangement : its first five parts, parallel to those of sāma, were named

udgrāha, anudgrāha, sambodha, dhruvaka and ābhoga; the last two were hiṅkāra and

oṅkāra which were used as pūrakaṣ or fillers of unworded, 'empty rhythmic time units'.3

Gāthā, too, had structural elements akin to the gāndharva sāma, for besides other

standard gītaka features it contained 'many limbs of sāma' : "sāmāṅgāni ca bhūyāṁṣi"

(S.R. 5, 230).

1 ṛk is the mantra in Ṛgveda; sāma means the musical composition to which a mantra was set

and gāthā in Vedic times denoted 'song'.

2 Describing ṛk he says

विकलः पङ्कलो वाऽन स्तोभः स्यात्पूर्ण संमतः ।

-S.R. 5, 226

and describing gāthā he says :

कला मुनिजनरुचि गाथायाश्चतुरङ्गरा।

-S.R. 5, 228

We do not know which sages Sārṅgadeva had here in mind. Dattila and Bharata have not

described ṛk, sāma and gāthā. Viśākha, however, might have described them. It seems from

a remark of Abhinava's that Viśākhila had devoted a greater space to the lāsya with which these

forms were intimately connected than had Bharata. Abhinava, commenting on Bharata's short

description of the catuspadi, which seems to have been an appendix to forms such as the pāṇikā

and belonged to a genre of songs known as the lāsya gāna, says :

अनया च विशाखिलादिलक्षितं सङ्क्षेमव लास्यगानं स्वीकृतमुपलक्षितं च ।

-A.B. on N.S. 31, 328-329.

3 बहरणा च पुरा गीत प्रस्तावोद्दिश्यते तदा । प्रतिहारोपद्रवो च निधनं पञ्चमं मतम् ॥

ततो हिङ्कार शौकारः सप्ताङ्गीते तनु तु । उद्ग्राहः स्वादनुद्ग्राहः संबोधो ध्रुवकस्तथा ॥

आभोगश्चेति पञ्चात्मालास्यगीतस्योच्यते । कामात् । हिङ्कारौकारौस्तन कलापूरकता मता ॥

-S.R. 5, 231-233.

Kalā comments :

बहरणा च पुरा गीतमित्यत्र सामेऽप्यहर्तव्यम् । प्रस्ताव इत्यादि । वैदिके सामनि प्रस्तावादिष्यास्य अङ्गुस्य गीते सामन्युग्राह

इति संज्ञा । उद्गीथस्य प्रतनुद्ग्राह इति संज्ञा । प्रतिहारस्य संभोग इति संज्ञा । उपद्रवस्य ध्रुवक इति संज्ञा । निधनस्य

आभोग इति संज्ञा । एवं कमो द्रुयात् । स्वत्व गीते हिङ्कारौकारौः कलापूरकता दृष्ट्या ॥

The question then is : Were these forms—ṛk, sāma—etc. then really chronologically older than the saptarūpa as the description given by Śārṅgadeva (also Kumbha and others) suggests ? Were they, in other words, closer to the ancient sāma-gāna than the saptarūpa? And if so, how was it that they were accorded an inferior position in the gāndharva scheme? These are questions which cannot have satisfactory answers till more ancient works come to light. Perhaps we might still venture to make a conjecture—forms like pāṇikā etc. did really have a closer affinity with the Vedic sāma music, but they were at the same time (perhaps as a corollary to their being old) comparatively simpler structures and, consequently, when the more sophisticated and intricate saptarūpa evolved, these older forms were relegated to a lesser position of honour.

CHAPTER VIII

THE GENESIS OF GĀNDHARVA

Gāndharva was sacred music—it must have grown and developed as a mucical expression of the ancient seer’s sensitivity towards the divine and his experience of it. Over a period its forms became fixed and were treated as a reserved territory, so that they may not be contaminated by more vulgar, less divinely inspired forms. These stratified forms were dedicated as ritual and were preserved unchanged. This had happened also in Egypt.

Plato was, in principle, against innovations in music made simply for the sake of newness out of flippancy or a desire to seek sensuous titillations. He approved only of certain old and established musical forms which were known to have a sound moral influence on the character of citizens. He was in favour of dedicating the approved melodies to the gods, like the Egyptians did, because thus consecrated, the melodies could be perpetuated for generations to come; their purity could be guarded due to the divinity attached to them. (See Laws VII, paras 799-800).

Gāndharva, too, was perpetuated intact because of its supposed divine origin and its ritual dedication. Historically, however, the period when it became codified into a strict ceremonial form, looms as a large and tricky question. As in other fields of ancient arts and sciences not much material is available to help our speculations. Some conjectures, however, may be made.

Abhinava often calls gāndharva ‘anādi’ : existing since the beginning of time (like the Vedas). Yet its beginninglessness was apparently not its own but was dependent upon the Sāmaveda, which was its source : “gāndharvavedavat svayamanādirvā sāmavedaprabhavate’ pi vā” (A.B. on N.S. 33 1).

There was a general saying among ancients that all songs are born of sāma : ‘sāmabhyo gītam’. Abhinava quotes this saying in the beginning of his comments on N.S. 28, 9-10 and adds that sāma was the penultimate cause of all musical forms; gāndharva, he says, originated in sāma and the gāna forms in turn originated in gāndharva : “sāmabhyo gītamiti kathitam, sāmāni cātra kāraṇakāraṇāni, gāndharvaṃ hi sāmabhyastamādbhavanaṃ gānam.” Again, in his lengthy analysis of gāndharva with which he opens the 33rd chapter, he defines gāndharva as : “dṛṣṭādṛṣṭaphalasā-mavedaprabhāvānādikālanirṛttam . . . (gāndharva) has both a dṛṣṭa and an adrṣṭa phala, it is born of Sāmaveda and is without a beginning. . . .

Regarding the gītakas of gāndharva (the saptarūpa), Dattila himself says: “ityevaṃ ṛṣibhirgītam sāma-vedasamudbhavam saptarūpam . . . : such is saptarūpa, sung by the ṛṣis and born of the Sāmaveda” (Datt. 222). Bharata has a short cryptic

statement regarding the origin (yoni) of gāndharva: “its origin”, he states, “is song, vīṇā and flute’ (asya yonirbhavedgānaṃ vīṇā vaṃśastathaiva ca, N.S. 28, 10).

From this statement alone it is not quite clear whether Bharata is referring to a historical source of gāndharva or to the musical factors with which gāndharva is built-up (which in other words originate it) because the word ‘yoni’ can be construed in both the senses. Abhinava interprets the line as signifying a historical origin. The word gāna he says, (in this context) denotes the mantras sung as sāma song: “gānamiti sāmayoni gītiṣu samākhyā.” ‘Vīṇā’, according to him, denoted the vīṇā used during the mahāvrata ceremony: “viṇeti mahāvratopayoginī auḍumbarī”. ‘Vamśa’ was indicative of the tradition of flute-playing initiated by Nārada: “vamśa iti nāradādigurusantānānuyāyīti”. These were the sources of gāndharva: “gāndharvasya prabhavah.”

The Sāmaveda had numerous ṣākhās, many more so than any other Veda. The mantras sung were taken from the Ṛgveda. Ṣākhā differences thus did not pertain to literary material and must have included differences as to musical renditions of ṛk mantras.

Sāma-gānas were classified into grāma-geya, āraṇyaka, ūha and ūhya songs. The first two were the basic compositions (yonigāna) of which the latter two were variations (vikṛtigāna). These vikṛtigāna forms may also have contained elements of improvisation since ‘rūpāntara,’ or formal changes during its learning and study, played an important role.1 Grāma-geya (lit. ‘to be sung in habitations’) sāma-songs were compositions suitable for performance in the presence of the populace or the community. Āraṇyakas (lit. ‘to be sung in forests’) had a mystic significance attached to them and were also called rahasya-gāna; they appear tho have originally developed in the secrecy and solitude of forest retreats. It is not unlikely that the nature of music in grāma-geya and araṇya-geya songs was also different and that grama-geya songs leaned toward folk forms while araṇya songs were composed with more intricate art forms.

Evidence reveals sāma as a rich heritage of music comprising a wide range and variety of musical forms. Many characteristics certainly passed into gāndharva as a legacy and left their indelible stamp upon the younger form and, through it, upon all subsequent music. Yet, between sāma and gāndharva there seems to have occurred a great transition. This cannot be traced in detail because information regarding early musical forms is scanty. However, some symptoms of change are evident. The very nomenclature of the notes, we discover, underwent a basic change. In gāndharva the seven notes of the octave were named ṣaḍja, ṛṣabha, gāndhāra, madhyama, pañcama, dhaivata and niṣāda. These names have persisted in the musical tradition of India down to the present times—a period of over 2500 years, since these names already appear in the early epics and the Śikṣās as quite well-established. The sāmic

1 The spirit of a free musical - interpretation of the ṛks to be sung (at least during the early formative period of sāma music) was, perhaps, also the reason why many mantras were composed to over 25 tunes ; cne mantra (Sāmaveda, 9, 107, 4) in particuar having as many as 61. Vide Veda Mīmāṃsā by Sri Anirvaṇa (Calcutta Sanskrit College Research Series, Vol. I. p. 59.

names were very different kṛuṣṭa, prathama, dvitīya, tṛtīya, caturtha, mandra and atisvārya. These according to the śikṣākāra Nārada1—by whose time the modern names had already become more current—corresponded to pañcama, madhyama, gāndhāra, ṛṣabha, ṣadja, dhaivata and niṣāda.

The notes of sāma, we notice, are counted in the avarohī order and not in the ārohī as in gāndharva and ever since. We also notice an odd point: the sāma octave does not follow a straight course in its succession of notes—dhaivata precedes niṣāda instead of following it. The reason for this irregularity is quite unknown; the irregularity had straightened out by the time gāndharva evolved.

When the Śikṣās were composed, there existed only two grāmas—the ṣadja and the madhyama—as in gāndharva. The musical structure of these grāmas must have been a legacy to gāndharva from the older form. The Śikṣās, however, speak of a third grāma—the gāndhāra, ‘no longer known in the world of men’. But evidently, it was a grāma current during an early period of sāma singing, though it went out of use later by the time gāndharva began taking shape.

Apart from grāma, the sāmic frame-work seems to have contributed tāna and mūrchanā to gāndharva.

Every mūrchanā and tāna had a distinct name.2 They were used in specific Vedic rituals. Dattila, too, mentions the sacred character of tānas: “devārādhanayogena tatpunyotpādakā” (Datt. 31). Abhinava quotes injunctions from authoritative texts indicating the ritualistic functions of both mūrchanās and tānas.

In sāma-music every mūrchanā and tāna had evidently a specific independent ritualistic role. In gāndharva, mūrchanās and tānas were incorporated within the total musical structure and had no independent application. Abhinava thus, regarding mūrchanās, remarks: “mūrchanās (in themselves) have no application in this śāstra (i.e., gāndharva), but in sāmic rituals one can clearly observe their application. Thus it has been said (as part of a certain sāmic ceremony) ‘three gāthās should be sung to the mūrchanā uttaramaṇdrā’. The same is true of tānas which have, in the Vāyupurāṇa etc., been thus described: ‘the first tāna should be the one called agniṣṭomīya, the second vājapeyika etc.’ their names themselves indicate their connection with the sacrifices corresponding to these names: “etaduktaṃ bhavati—mūrchanānāṃ yadyapīhāgame naṣṭyupayogaḥ, tathāpi dṛṣṭasāmakriyayāṃ va sphuṭa evopayogaḥ tathā hi darśitaṃ ‘uttarandraȳa tisro gāthāḥ’ iti; tānānāṃ agniṣṭomīyamādyaṃ syād dvitīyaṃ vājapeyikaṃ” ityādināmanirūpanadvāreṇaiva vāyupurāṇādu yajñopayogi sāmar.ggāthopayogaḥ pradarśitaḥ” (A.B. on N.S. 28, 34).

1 His Nāradi Śikṣā is an aṅga or branch of the Sāmaveda and was composed in order to help Sāma-vedins to know sāma music. The work is not very early and has been assigned to a period between the 1st and 2nd centuries A. D. The other Śikṣās like the Yājñavalkya Śikṣā, the Māṇḍūki Śikṣā are equally late.

2 The Nāradi Śikṣā lists 21 mūrchanās, the basis of which is not quite clear as it recognises only two grāmas as current. Two sets of mūrchanās have names entirely different from those in Bharata and Dattila; while the third set contains names in common with the ṣadja-grāma and madhyama-grāma mūrchanās as given by Bharata and Dattila. The names of tānas are not found in the extant Nāradi Śikṣā. But Dattila (cf. Datt. 31) speak of Nārada in connection with them. To tāna names are found in Vāyupurāṇa., (86, 20-27). Br. (106-17), S. Raj (2, 1, 1, 476-93) etc,

Abhinava also speaks of a sāma-song called the agniṣṭomika which was evidently composed to the tāna of the same name. A person who praised Śiva with this sāma acquired great adṛṣṭa as a result. Also—as a verse quoted by Abhinava testifies—‘the members partaking in a yajña ceremony (sadasya) could all be expiated of their sins on hearing the agniṣṭoma sāma and could conquer the transcendental worlds’: “agniṣṭomikasāmena śivam stutvā tatphalamitit ca prayokturadṛṣṭaṃ śrūyate tathā ‘sadasyamagnisṭomasāmā sṛnvataḥ pātakairmucyate, lokān jayati” (A.B. ibid.).

Although in gāndharva, mūrchanās and tānas had no independent status, yet the fact that they did form part of gāndharva and were preserved as limbs of gāndharva in their ancient unaltered form, shows the indebtedness of gāndharva to sāma.

A basic musical factor which gāndharva owed to sāma was the indispensability of the note madhyama (“madhyamavyatiriktāḥ sarve svarā nāśinaha madhyamāsya tu nakvacidvināśaḥ” A.B. on N.S. 28, 61-65). Bharata himself states that madhyama occupied an indispensable position both in sāma and in gāndharva, thus implicitly acknowledging the influence of sāma. Abhinava in this context, quotes a line from the Nāradi Śikṣā and remarks that madhyama in sāma was the initial (prathama) note (the tonic?) and hence it was always used: “sāmasvapīti,” tatra hyasau prāthamyani yataprayogah. yadāha nāradamuniḥ, ‘yah sāmagānāṃ prathamaha sa viśenmadhyamasvarah’, ‘venmodharma’ itayanye paṭhanti” (A.B. on N.S. 28, 64-65).

Our knowledge of sāma forms is so meagre that we can form no idea of the nature of the indispensability of madhyama in it—but the unique position this note occupied in gāndharva was certainly due to the influence of sāma.

We have already noted that the attitude to pada or the word-content of songs bears a great similarity in sāma and gāndharva. The nature of syllabic-distortions (vikāra) in gāndharva show a family resemblance to those in sāma singing, and this was certainly not accidental or due to a coincidence. Sāma must have exerted an influence on gāndharva in this region too.

Sāma must have also contributed many other factors of a more subtle nature to gāndharva, like graces and alaṅkāras; also it must have influenced the younger form in the very manner of throwing one’s voice, the inflexion of notes in singing or playing and a score of such seemingly minor points that loom so large in giving a distinct individuality to any musical culture. It also must have contributed melodies or melodic lines and phrases which later became stratified into gāndarva as part of jātiṣ.

Though influenced by sāma, gāndharva was distinct from sāma. Gāndharva, historically, fell between the sāmic forms out of which it was born and later grāma-rāga, bhāṣā, rāga and other ‘gāna’ forms which were born of it: “gāndharvaṃ hi sāṃabhyastasmādbhavaṃ gānam” (A.B. on N.S. 28,9). There must have been non-sāmic influence on gāndharva.

Abhinava has enumerated two other factors besides sāma-songs as sources of gāndharva : the audumbarī vīṇā and the flute as played in the tradition of Nārada.

The Vedic people took a keen delight in music and a variety of instruments were in vogue and professional musicians were quite common.¹ Many vīṇās were

1 History and Culture of the Indian People ; The Vedic Age, p. 456,

known of which audumbarī was one of the most popular, Shri Ramakṛṣṇakavi has opined that "a careful examination of the Vedic rites and Śikṣās thereupon drives one to the irresistible conclusion that the origin of Indian music lay in certain rites where the priest and the performer chant some gāthās alternately while the wife (yajamānī) plays on vīṇā and the closing of the sacrifice was enjoined with a peculiar dance. The kind of vīṇā mentioned for the above purpose is called the picchola and in another place it is called audumbarī, that is, made of audumbara wood".1

Abhinava mentions the audumbarī in connection with the rites of mahāvrata. This was a festival in which girls and matrons whose husbands were still living danced around a fire to the music of lutes and flutes; 'dance music and song filled the whole day of moving'.2 The lute played was apparently the audumbarī. It was evidently a kind of harp vīṇā rather like a cembalo and contained a hundred strings made of muñja grass. It was played by striking the strings with sticks made of the vetasa plant. The vīṇā was called vāṇa. The epithet audumbarī was attached to it because it was made of audumbara wood.3 It appears to have been similar to the modern santūra popular in Iran and Kashmir and recently reintroduced to Hindustani classical music. It evidently was the precursor of later vīṇās like the mattakokilā - with 21 strings that could be tuned to the three octaves4-and also vipañcī and others that were used in rendering gāndharva and later musical forms.

Abhinava mentions the vamśa (flute), too, in connection with gāndharva. Flutes must have been known in a great variety. Though the term vamśa indicates a bamboo flute but Abhinava in his commentary on the flute section of the Nāṭyaśāstra, quotes an ancient kārikā which mentions flutes made of many materials like silver (raupyah) bronze (kāmsyaja), gold (sakañcanah) and wood (khadira).5 These must have descended from ancient models. The nature of flute-playing as developed by Nārada and his tradition is not known. Presumably-apart from technical peculiarities-the nature of the music played must have been the same as in sāma songs.

Yet it is very probable that the tradition of the flute and audumbarī-vīṇā-playing incorporated new musical elements or carried old forms to further elaborations or intricacies not to be found in sāma songs. A demand for strict regulation attached itself to the musical movements with which mantras were sung in sāma especially during the later Vedic period when the sāma songs lost their creative impetus and

1 Quoted in Bh. Sans. Iti (B) by Prajñānanand, Vol. I, p. 199.

2 The Sanskrit Drama, Keith, p. 29.

3 औदुम्बरस्य वीणादण्डस्य दशदशाधिकैकस्मिन्तिमभिते मोक्ष्यास्तनूःप्रवयन्ति स वाणः शततनुः । श्रृण्वन्ति केयां भुवस्वरिसंवृततन्तव इति वयस्ति प्रतिनोतिः । भुवस्वरिस्थित शतन्तव इति व यासित्र शत होता । सुवस्वरिस्थित-शतन्तव इति वयस्ति भातत्सुमुद्गात गृह्यपतिमुत्सृजम् ।

-Kātyāyana's Srautasūtra, 21, 17.

4 तत्र मत्तकोकिला प्रधानभूता । एक विंशतितन्त्रीकलवेनानूगताधिक निस्थानगस्वरसारणा जातीयति वीणाशरीरमुख्यते ।

-A.B. on N.S. 29, 12.

5 वदिरादिविनिर्मितौष्यसो भवत्येव । तथा चोक्तं वंशे सुष्टु यदा पूर्वं वंशसंज्ञा तु वेणवी । वंशास्तु खदिरा रौप्या: कांस्यजा वा सकाञ्चनाः ॥

-A.B. on N.S. 30, 1.

became fixed and static. But instrumental music by its very nature cannot be so mantra-bound and perhaps evolved new forms which it handed to gāndharva.

Gāndharva must also have absorbed forms borrowed from the everchanging popular traditions. The gāndharva form is intimately connected with Śiva hymns.

Tāṇḍava and lāsya were dances complementary to gāndharva which formed their musical accompaniment. These dances, too, were associated with Śiva. This fact signifies a very definite Śaivaite inspiration behind gāndharva. Śaiva sampradāyas

have always been liberal in adopting non-Vedic forms and ideas. It would, therefore, be quite in character for them to have fused non-Vedic popular musical forms with sāmic traditions in gāndharva.

Some influence of regional or folk factors in the jātis is, in fact, suggested by their names. Thus the musical idiom current in the North of India seems to have contributed in giving the jāti ṣadjodīcyavā its form ('udīcya'– the country to the north of the river Sarasvatī, i.e., the northern region)1. We, too, have our 'regional' rāgas,

incorporating stylised folk-idioms such as sorāṭha, bangāl-bhāirava, mānd etc.

Two other jātis evince a northern influence–the madhyamodīcyavā and the gāndhārodīcyavā. Another jāti, the āndhrī (also called andhri) may have been influenced by local forms popular in the Āndhra region.

Some accounts of sāma-melody are available, but of sāma-tāla we know almost nothing. In gāndharva, tāla-structures were built of very complicated patterns.

They may have borrowed some elements from sāma; but in tāla, more than in svara, non-sāmic influence seems to have acted as a great formative force. Many terms and topics relating to the svara aspect of gāndharva can clearly be traced to sāma-music,

but no aspect of gāndharva tāla is similarly traceable. Indeed, accounts of sāma singing evince no trace of tāla playing.

It is likely that tāla, rendered as a rich structural pattern with its own independent forms, was absent from sāmamusic (as is also testified by the fact that none of the existing sāma-śākhās have anything like a tāla-accompaniment with the sāma chant) but evolved in the popular percussion music of Vedic times. This, then, was passed on as a legacy to gāndharva and through it to subsequent art-forms of Indian music.

The earliest mention of terms used in gāndharva in connection with tāla, is found in the epics and the Smṛtis. The Sabhāparva of the Mahābhārata contains a description of gandharvas and kinnaras who performed in Yudhisthira's newly built Sabhā or council hall. These musicians are called experts in śamyā and tāla and in laya (śamyatālaviśāradāḥ pramāṇe ca laye sthāne kinnarāśca kṛtaśramāḥ, MBh. Sabha. 4, 46). Śamyā and tāla were names of two of the sounded beats in gāndharva.

1 The geographical location of Udīcya, that we have given is from the Mbh. as quoted by Monier-Williams in his dictionary under 'udicya'. Regarding ṣadjodīcyavā Abhinava remarks :

उदीच्या: प्रयोक्तारो वाहुल्येन सत्यस्यामिभवदीच्याः । वान्ति मच्छल्नीतिम् । उदीचे्येडपि व दृश्यते । गीतं च देशनामध्येर्व-हुल्यादिमिराश्च तालयेक्षा व्यपदिश्यतेव टबकरागो मालवपञ्चमो गौडी मालवी काम्भोजीयादि ।

–A,B. on N.S. 28, 106-109.

The Anuśāsanaparva, mentions two other terms-pāni and sāma-along with śamyā and tāla ('pānitālsatālaiśca śamyātalaiḥ samaistatathā', MBh. Anu. 25, 19). Śamyā and tāla, indeed, are mentioned repeatedly (Sabhā 4, 44; Droṇa 69, 11 etc.).

The Uttarakāṇḍa of the Rāmāyaṇa mentions kalā and mātrā which were technical terms denoting time-measures in gāndharva tāla. The mention is made in connection with the singing of the Rāmāyaṇa by Lava and Kuśa ('kālāmātraviśeṣajñau', Ram. Uttara, 94, 6).

Evidently, some factors and aspects of gāndharva tāla were an established part of musical practice during the epic period. However, historically speaking, it is almost impossible to draw any conclusions regarding chronological priorities in this matter. Whether gāndharva was influenced by the tāla-forms referred to in the Rāmāyaṇa and Mahābhārata or whether in actuality it was gāndharva which arose first and in truth influenced the aforementioned forms, will remain a question which can be answered only in the realm of pure conjecture. Perhaps there were common currents and cross-currents of influences which entered into the formation of both the gāndharva and the more popular epic forms.

The language used in the Śiva hymns sung in gāndharva, shows a Sanskrit that has left the Vedic stage quite behind it. The language bears an affinity to that of the epics. Gāndharva padas must have been composed sometime during the long period when the Sanskrit speech of the epics was taking shape. By the time of Bharata and Dattila, these hymns, as also the nonsense syllables sung as nirgīta or śuṣka along with gāndharva music, had already become stratified-unalterable to its last syllable as a Vedic mantra. A long time must have elapsed between the formulation of the gāndharva hymns and their attaining the sanctified status accorded to them by authorities like Bharata and Dattila.1 The extent of this intervening period, however, remains a matter of pure conjecture.

It must, indeed, have taken a long time for gāndharva to reach the final immutable forms that it ultimately acquired in its total aspect of svara, tāla and pada. These were then put into codes. A period of more or less loose and flexible formulations must have preceded. No accounts of this stage, when gāndharva was in a state of flux, are available. The differences between the various descriptions of gāndharva found in early codifiers like Bharata, Dattila, Viśākila (of whom we know only from quotations) and others, centre on relatively minor matters. They are not unlike slight differences in rituals found between different Vedic śākhās in their observances of ordained sacrifices. Yet it is not unlikely that these residual differences are traces of a period when a greater freedom in rendering gāndharva forms was not unknown.

1 The gāndharva hymns have elements of pre-Pāṇini grammatical features. Commenting on a line from a hymn which reads : पृथिवी(म्) सजिल ज्वलन(नः) सूर्यश्चन्द्रो यजमानो व्योमाऽभू:(म) कार्यः

Abhinava points out an irregularity in grammar and justifies it by saying that such usage is permissible, for the hymns are 'sung by Brahmā' or in other words they are 'ārṣa' : पृथिवीमित्यादि वाङ्मात्रं मित्य नते हनद् उपिज्ज्वलन इत्यादावलमु ब्रह्मगीतत्वात् ।

—A.B. on N.S. 31, 114.

CHAPTER IX

GĀNDHARVA AND POST-DATTILA WRITERS

By the age of the Nāṭyaśāstra and the Dattilam, gāndharva and its codifiers already formed a long and established tradition. The tradition must have continued but works are not extant.

After Dattilam there is a long break in the available saṅgīta-literature, the next extant work being the Brhaddesī of Mataṅga. This work avowedly deals with the popular, ‘profane’ desī forms ; and, moreover, only a portion of the work is available. Brhaddesī is generally believed to have been composed in the seventh century.

After Mataṅga there is again a long period of over three or four centuries from which no works survive. Yet this period appears to have been quite fertile in the field of music. We do not hear of many independent works, but commentaries on Nāṭyaśāstra were continuously being written by great scholars of repute, versed in all the arts used in theatre including music, its various forms and traditions. Abhinava, who was born in the middle of the 10th century1 and lived a long life till well into the 11th century2 speaks of many preceding commentaries on the Nāṭyaśāstra and often refers to commentators of repute.

Abhinava recounts the names of Lollaṭa, Śaṅkuka, Bhaṭṭa Nāyaka among other scholars who had commented upon the Nāṭyaśāstra. Their commentaries, evidently, covered the entire span of Bharata’s work including the section on music as did Abhinava’s since Śārṅgadeva in his Saṅgītaratnākara, (a treatise devoted almost exclusively to music and dance), speaks of these authors as ‘vyākhyātaro bhāratīye’. He apparently had in mind those aspects of their commentary which touched upon music and dance.3 Śārṅgadeva also names a Kīrtidhara, who is quoted by Abhinava too.

Other names mentioned by Abhinava are those of Bhaṭṭa Yantra and Śrī Harṣa. Besides these, we hear of Mātṛgupta, Bhaṭṭa Gopāla, Ghaṇṭaka and others. These scholars flourished at different dates from the 5th century (probable date of Mātṛgupta) to Abhinava’s own times.4

Many of these commentators – besides other authors who wrote independent works on music—had apparently discussed the distinctness of gāndharva and its specific

1 Abhinava Gupta by K.C. Pandey, p. 9

2 Ācārya Viśveśvar, in the introduction to his translation of 3 chapters of Abhinava Bhāratī into Hindi called Abhinava Bhāratī ke Tīn Adhyāya, says that Abhinava flourished from 950-1125 A.D., preface, p. 22.

3 व्याख्यातारो भारतीये लोल्लटोद्भटनायकाः । भट्टोत्पलभट्टलोल्लटभट्टश्रीमुक्तिगंधरः परः ॥ – S. R. 1, 1, 19.

4 For details on their dates, see Abhinava Gupta by K.C. Pandey, pp. 187-198. also : Abhinava Bhāratī ke Tīn Adhyāya, preface, pp. 12-15.

characteristics. Abhinava, we have seen, often refers to views of his predecessors in discussing certain aspects of gāndharva.

Although it is true that no saṅgīta-texts from the time of Bharata and Dattila to the 7th century (when the Brhaddesī was composed) survive, yet we do have the Purāṇas, some of the earliest of which, contain sections on music. The Purāṇic tracts deal with the subject only incidentally and at second hand : lifting verses from standard works and heterogenously juxtaposing and confusing them with semidigested matter borrowed from several sources. Still the Purāṇic records regarding music are valuable material. They give us an idea of the nature of music which the Purāṇas thought as worthy of description. Significantly for us, their account throws light on gāndharva and reveals the esteem in which it was held.

One of the earliest Purāṇas is the Vāyupurāṇa. The main body of its text is said to belong to the 3rd century A.D. Another early text is the extent Brahmāṇda-purāṇa (sometimes called the Vāyavīya Brahmāṇda). This is said to have once formed part of the original Vāyupurāṇa. Dr. R.C. Hazra is of the opinion that the separation of the two took place around 400 A.D.1 Both these Purāṇas contain a description of music they term gāndharva. The context of the descriptions and the descriptions themselves are almost identical in both texts, with minor deviations which could be due to insignificant pāṭhabhedas.2 It is evident that the two extant Purāṇas, in their description of gāndharva, relied on slightly different recensions of the same original undivided Purāṇic text. Hence the description occurring in them is earlier than 400 A.D.

The delineation of gāndharva in these two texts and its context is worthy of attention. The description occurs within an account of the genealogy of Manu (manuvaṃśakīrtana), where the story of King Raivata, son of Revata, is related. Raivata was the father of Revatī who became the wife of Balarāma, elder brother of Kṛṣṇa.

The Purāṇa recounts the episode of Raivata's visit to Brahmā with his daughter in order to seek His advice regarding a suitable bridegroom for her. The story goes that Raivata was distracted by the 'gāndharva' being performed before Brahmā and tarried in Brahmaloka for a few moments. Meanwhile many mortal yugas passed away, for the time-scale in Brahmaloka is very different from that on earth : a single

1 Purāṇic Records, p. 18.

2 Vāyu ch. 86 is almost the same as Brahmaṇda ch. 61, 18-53. Some verses occuring as Vāyu 86, 43-52 are, however. missing from .he Brahmāṇda account. Curiously, some matter which should have been there is absent from both accounts : Vāyu verse 19, in connection with mūrchanas, says : गान्धारग्रामिकाख्यानि केतवःमानं निबोधत but the names of the mūrchanās are missing; the next verse enumerates names of tānas.

The same lacuna occurs in Brahmāṇda, after 61, 34. This would suggest that the matter in the recensions from which the accounts in the two Purāṇas are derived had an identical gap. Chapter 62 of Brahmāṇda is the same as chapter 87 of Vāyu, but about 3 verses occuring in Vāyu are missing in the Brahmāṇda account after verse 28.

The readings in both the Furāṇic texts concerning music are thorougly corrupt. However, a critically edited and admirably restored text of the Vāyu chapters on music may be found in Textes des Purāṇa sur le Theorie Musicale. Vol. I by Alain Danielou and N.R. Bhatt, pp. 20-100. Unfortunately, a comparison with the Brahmāṇda text does not seem to have been made by the authors.

154 A Study of Dattilam

day of Brahmā is equal to a mortal caturyuga—a complete cycle of existence over millions of years after which pralaya or deluge takes place.1

Hearing this episode the ṛṣi to whom the matter was being related became curious about gāndharva;2 which is then described in two chapters.

Gāndharva, we have seen, was in general literature a term referring to music : any music whatever, and not a special form of the art. Yet interestingly the description of gāndharva found in the Vāyu and Brahmāṇḍa Purāṇas reveals that what the Purāṇakāras here had in mind was basically the special body of music called gāndharva by Bharata and Dattila.

The Purāṇic account is by no means a complete account of its subject-matter. But though fragmentary, it is yet an account of what constituted the distinct gāndharva corpus of music. It contains an exposition of the following topics: svara, mūrchanā, tāna, varṇa, alaṅkara (among svara topics) and vṛtti, pādabhāga and the gītakas (among tāla topics). The exposition is partial, mainly containing classificatory lists.

The treatment is not as clear and methodical as one would expect of a truly śāstric account. But the account is clearly that of gāndharva in its restricted sense. Indeed, in one or two ways it contains more details than the expositions of Bharata and Dattila. Neither Bharata nor Dattila have given us the names of the tānas. Dattila has a single indicative verse : “they (the tānas) are known by such names as agniṣṭoma etc. as stated by Nārada and others.” (Datt. 31). The Vāyu and Brahmāṇḍa Purāṇas enumerate the names of all the tānas (Vāyu 86, 21-30).3 Adhinava was aware of the Purāṇic text on this point for he says that the names of the tānas are to be found in the Vāyupurāṇa; he quotes a line from this text (which is 86, 21A of the Vāyupurāṇa).4

One aspect of the purāṇic text tellingly reveals that the form being described was really gāndharva in Dattila's sense of the term. This is the inclusion of gītakas. Other topics discussed in the Purāṇa could also occur in a Śikṣā work dealing with sāma-music but gītakas were peculiar to gāndharva. The names of all the seven gītakas occur (Vāyu, 87, 34; 35; 40; 43; 44) and though their description is rudimentary and couched in terms which hardly make much sense in terms of music, yet their very inclusion is signicfiant. The term saptarūpa, as meaning the seven gītakas, also occurs (Vāyu, 87,38). At one place the Purāṇa hints vaguely at the nature of the melody to be associated

1 रेतस्य रेतः पुत्रः प्रकृत्या नाम धामिकः । ज्येष्ठो भ्रातृषुतस्यासीत राजा प्राप्य कुशस्थलीम् ॥

कन्या सह श्रुता च गाथर्वं बहुपुत्रौप्तिके । मुहुर्त देवदेवस्य मात्यं बहूयंश्च विमोः ॥

—Vāyu 86, 3-4.

See also Brahmāṇḍa 61, 20-21.

2 एतच्चतुष्प्रमाणान्तो गान्धर्व बद चर्वाहि ।

गान्धर्वं प्रति च्छायां पृथकस्तु मुनिसत्तमा: ततोध्वं सप्तपक्ष्यामि याथातथ्येन सुव्रता: ॥

—Brahmāṇḍa 61, 26-28.

See also Vāyu 86, 13 which is the same as Brahmāṇḍa 61, 28.

3 The two Purāṇas list the names of tānas and of the mūrchanās of gāndhāra grāma as well as the other two grāmas.

4 We have quoted the A.B. passage in another context: but we repeat :

तानानां, “अग्निष्टोमादयः स्वाद् द्वितीयं वाजपेयिक” इत्यादीनां निष्पत्तिहेतुरेणीव वायुपुराणादौ यज्ञोपयोगी

सामग्रामाथर्वणोक्तः: प्रदर्शितः

—A.B. on N.S. 28, 34.

with the gītakas (the relation is of a kind which neither Bharata nor Dattila speak of). The account, however, is couched in extremely cryptic terms and is so indistinct and indiscernible, besides being hopelessly brief, that hardly much of value can be made of it (Vāyu 87, 34.38). The Purāṇic text also speaks of the three vṛttis: citrā, vṛtti and dakṣiṇā,1 which were modes of instrumental accompaniment to song. It also mentions certain tāla terms specifically involved in the technical descriptions of gītakas, such as pādahbāga, mātrā, kalā, upohana (Vāyu 87, 40; 42; 44; 47), but their use is such that beyond revealing the fact that the account shows an awareness of the relation of these terms with the gāndharva-gītakas, nothing more is manifest.

Howsoever vague and confused the Purāṇic record may be as a śāstric account of musical forms, this much is clear that its exposition is not of gāndharva in a general sense but of the specific body of music called gāndharva, for only specifically gāndharva forms and topics have been discussed—and gāndharva is, indeed, the name under which the Purāṇa expounds the subject. The association of gāndharva with Brahmā (in whose presence Revata heard gāndharva being performed) further strengthens our point, because both Bharata and Dattila speak of Brahmā as the original perpetrator of gāndharva (Datt. 1, 2; N. S. 28, 39; 33, 23).2

Another Purāṇa, the Mārkaṇḍeya, also contains a passage on music, though a comparatively small one (consisting of a few verses only). Mārkaṇḍeya is another early Purāṇa which is believed to have reached its present form before the 5th or 6th century, when its latest part, the Devīmāhātmyā, is thought to have been inserted in it. Other parts are earlier and the section on music, which forms part of chapter 21, was probably composed between the 3rd and 6th centuries.3

The Purāṇa relates the story of Nāga Aśvatara, (brother of Kambala), who performs tapas in order to please the Goddess Sarasvatī and sings a beautiful hymn in praise of her (Mārkaṇḍeya 21, 32-48). Sarasvatī is pleased and appears before him She tells him to ask for a boon. Aśvatara asks her to reveal to him and to his brother Kambala “everything relating to svara or music”: (samastasvarasambaddhamubhayoh samprayaccha ca, Mārkaṇḍeya 21, 51). Sarasvatī promises Aśvatara that both he and Kambala will become experts in music. She enumerates various

1 तिसृणां चैव वृत्तीनां वृत्तिस्वरूपं च दक्षिणा । —Vāyu 87, 47.

See ‘Vṛtti’ in Part III for its definition in Bharata, Dattila etc.

2 गान्धर्व नारदादिषु: प्रमाणानि स्वयम्भुवां । —Datt. 2.

जातयोष्ठादशेत्येवं ब्रह्मणा|मिहितं पुरा । —N. S. 28, 39. etc.

3 R.C. Hazra, discussing his views regarding the chronology of Mārkaṇḍeya chapters, concludes : “The above conclusion about the date of the chapters under discussion agrees remarkably with the view of Pargiter who says :

‘The Devīmāhātmyā, the latest part, was certainly complete in the 9th century and very probably in the 5th or 6th century A.D. The third and fifth parts (i.e., chapters 45-81 and 93-136 respectively). which constituted the original Purāṇa, were probably in existence in the third century and perhaps even earlier ; and the first and second parts (i.e., chapters 1-9 and 10-44 respectively) were composed between these two periods’.”

elements of music in which they will become proficient (Mārkandeya 21, 52-56). The elements include mūrchanā, 49 tānas, the three grāmas, the three layas, the three yatis and also the seven gītakas. The two brothers then come to acquire an exceedingly specialised kowledge (agryam vijñānam) of pada, tāla and svara : “vijñānamubhayoragryam padatālasvarādikam” (Mārkandeya 21, 60). Aśvatara and Kambala later set out to worship Lord Śiva. The worship consists of propitiating Him through the gītakas. Śiva is finally pleased by their gītakas and appears before them.1

This Mārkandeya passage is interesting from our point of view. Its enumeration of the elements of music is, evidently, an off-hand list of some aspects and terms connected with the subject which general audiences (who formed the readers and listeners of the Purāna) must have come to acquire as part of their common knowledge and must have associated spontaneously with the art. Thus a Purāna-like text of a later period, when speaking of music, would enumerate terms like rāga, rāginī etc., because these are terms which for the past few centuries are most generally associated with music.2 The enumeration in Mārkandeya is suggestive ; for the topics listed are —with a single exception of the grāma-rāga—all part of gāndharva ; indeed, mention is made also of the seven gītakas.

Further, the gītakas are said to have been used by the two Nāga brothers in worshipping Śiva. This episode significantly suggests that the music taught by Sarasvatī to Aśvatara and Kambala was none other than the gāndharva. This form, we have seen, was said to be exceedingly dear to Śiva and was used in his worship.

We also note that the brothers are said to have acquired a knowledge of pada tāla and svara; this is reminiscent of the accepted śāstric definition of gāndharva as : “gāndharvamiti tajñeyam svaratālapadātmakam” (N. S. 28, 8).

The very names of the two brothers are again suggestive. Both Kambala and Aśvatara were renowned ancient authorities on music. Śārngadeva mentions them along with Viśākhila, Dattila, Bharata, Kaśyapa and other ancient authorities.3 Evidence shows that they, too, had written authoritative works on gāndharva : Śārngadeva brackets the names of Aśvatara and Kambala together as holders of a specific opinion— differing from Bharata—regarding the use of the sādhārana notes in certain gāndharva jāti-forms 4. Rānā Kumbha claims to have studied their works.5 He too mentions the opinion of Kambala and Aśvatrara regarding the prayoga of sādhārana notes in

1 तत्रः कालेन महता स्तूयमानो वृपःकृतः । ततोऽष्ट गीतकस्तो च प्राह संग्रीतकार्ता वरः ॥ —Mārkandeya 21, 63.

2 See, for example, the Brhaddharma Purāna, Pūrvakhanda, chapter 44.

3 विशाखिलो दत्तिलश्च कमबलोऽश्वतरस्तथा । —S. R. 1, 1. 16.

4 पञ्चभिर्मध्यमाद्जन्य मध्यमाद्वयसनिःसृताः जातियः । स्वरसाधारणं प्रोक्तं मुनीनिमित्तं रतादिमः ॥ अनेध्य सम्रदेवद्वया स्वनियामद्भवेत । एतदल्पनिगद्याहुः कमबलाश्वतरादयः ॥ —S. R. 1, 7. 21-22.

5 वामस्वतीमहेंद्रकव्यकपमरसूदृशो नार्थः । कृतान् रस्मातुम्बुरू कमबलाश्वतररक्षोराजसन्विभतान् । श्रीसोमेश्वर भोजराजरचितान् ग्रन्थान् विलोक्य त्वम् तत्सारेण समुचितेन कुशले श्री कालसेनो नृपः ॥ —S. Raj 1, 1, 1, 40

jātis (S. Raj 2, 1, 4, 50-55). In addition, he speaks of a certain view held by Kambala regarding the beat-formation in one of the gītakas (the ullopyaka);1 this evinces that Kumbha had either read a work on gāndharva ascribed to Kambala or had come across a quotation or an opinion ascribed to Kambala in an early work by another established author.

Another important work of Purāṇic literature is the Viṣṇudharmottara, classified as an Upapurāṇa. It is a large encyclopaedic compendium dealing not only with the usual Purāṇic legends, stories and myths but also with a host of varied subjects such as law, politics, science of war, anatomy, medicine, cookery, manufacture of perfumes, horticulture, lexicography, rhetorics, dramaturgy, painting, dancing, etc. Music also forms one of the subjects dealt with. The Purāṇa has been dated to the 5th century A.D.2

The Viṣṇudharmottara has two short chapters on music (khaṇḍa 3, chs. 18 and 19). The subject is treated under the name of gīta-śāstra. The delineation is sketchy, replete with gaps and incomplete on many points. Yet, in keeping with the śāstric character of this encyclopaedic Purāṇa, its account of music is the most systematic and methodical of the Purāṇic treatments.

Gīta-śāstra is divided into two main sections : song (gīta) or vocal music and ātodya or instrumental music. Chapter 18 (khaṇḍa 3) is devoted to gīta, the next chapter to ātodya. The topics covered under gīta are sthāna, svara, grāma, mūrchanā, tāna, coksā-prayoga (?), jāti, alaṅkāra and gītaka. Three grāmas are mentioned and murchanās are named as 21 in number—7 in each grāma. Tānas, too, are similarly classified on the basis of three grāmas ; they have also been named.

The jātis are not named but three lines summing up the salient jāti-charactesristics (jāti-lakṣaṇas) have been quoted with the words: “bhavanti cātra ślokāḥ” (‘following are the ślokas on the topic’). However, inspite of the plural ‘ślokāḥ’, the number of ślokas quoted amount to only one and a half, hardly justifying the use of the plural.3

The first two lines on the quotation are identical with the verse enumerating jātilakṣaṇas in the Nāṭyaśāstra, that is, N.S. 28, 66. The last line corresponds with N.S. 28, 141A. In the Nāṭyaśāstra, more than 70 verses from 28, 66 to 141A expound jāti-lakṣaṇas in detail – verse 66 initiates the topic while verse 141A concludes it. In the Purāṇic account, only the lines initiating and concluding the exposition of the subject are

1प्रपञ्च कम्बलं प्राह पूर्वमेव कलविण्मिथः ।

—S. Raj. 2, 4, 1, 187.

2R.C. Hazra after presenting his arguments in favour of this date concludes :

“From all the evidences adduced above it is clear that the Viṣṇudharmottara cannot be dated earlier than 400 A.D. and later than 500 A.D. Bühler also is of opinion that the date of its composition cannot be placed later than about 500 A.D.”

—Studies in the Upapurāṇas, Vol. 1, p. 212.

3भवन्ति चात्र श्लोका:— ग्राह्यास्तरम्यस्ते च न्यासोजप्न्यास च ।

अल्पत्वं च बहुत्वं च पादवोधकत्वे तथा ॥

एवं च बृद्धौ जातयो दशलक्षणा ॥

—Viṣṇudharmottara, khaṇḍa 3, Ch. 18.

to be found. It is likely that the Purāṇa once did contain the intermediary ślokas also as is testified by the use of the plural ‘slokāḥ’ with which the quotation is introduced.1

After its all too brief treatment of jāti, the Purāṇa has a single verse on alañkāra, which is almost the same as N.S. 29, 20. Then come the gītakas.2 They are named and characterized in lines identical with the Yājñavalkya Smṛti 3, 113-114 where gītakas have been enumerated and are said to lead to mokṣa. The Purāṇa, too like the Smṛti says that ‘repeated performance of the gītakas leads to mokṣa’ and adds: “I have recounted the gītakas, (though briefly, because they propitiate the gods—

“saṅkṣepena mayā proktam surādhanakaranārārat.” The next verse is again borrowed from the Smṛti (where it occurs as 3, 116) :

gītājño yadi gītena nāpnoti paramam padam

śivasyānucaro bhūtvā tenaiva saha modate.

It is evident, from the account we have analysed above, that the Purāṇa in discussing music had gāndharva in mind. Though the description is short, it certainly pertains to gāndharva for it incorporates many of its topics and delas with jātis as well as gītakas calling them propitiatory forms.

There is, however, one puzzling note. The Purāṇa speaks of a kokṣā-prayogah in the context of ‘prakaraṇa’ (i.e., the generic term for the seven gītakas) : “cokṣā-prayogah svarorādhane rūpakeṣu prakaraṇānugatath manyuvyapadeśena te bhavanti.”

The passage is extremely mystifying. Also, it obviously contains scribal errors. The phrase “manyuvyapadeśena” is obviously meaningless. ‘Svarorādhane’3 should be either ‘svarārādhane’3 or ‘surārādhane’ in which form the phrase occurs in the same context a few lines later. In the light of these obvious errors, it is not unlikely that ‘cokṣāprayogah’ is also a faulty reading.4

1 It is, however, not unlikely that the Purāṇic lines have been borrowed not from the Nāṭyaśāstra, but from another text; a text such as the Dattilam where the three successive lines corresponding to those in the Purāṇa occur in a similar unbroken order as :

ग्र’हांशो तु रसस्यै च पाडवोऽधिकृत्यते क्रमात् । अल्पत्वं च बहुत्वं च न्यायोऽन्यास एव च ॥

स्वरमेतद् यथाजाति ददर्श जातिलक्षणम्‌

— Datt, 55-56.

True, such an hypothesis would stili fail to account for the plural in the introductory words with which the three lines are presented in the Purāṇa, but such incongruities - we must not forget - are not uncommon in Purāṇic texts.

2 अपरान्तकुल्लूभट्टः मद्रकः; प्रकृतिः तथा। उवेचकं सरोविदन्मुख्त्तम् गीतकानि तु ॥

ऋग्माथा पौराणि दशविधिहिता ब्रह्मगीतिका। गीतमेतत्तदभ्यासः कारणामोक्षसंज्ञः ॥

The original reads : कारणाचमोक्षसंज्ञः apparently a faulty reading which we have corrected on the basis of the Yājñavalkya Smṛti 3, 114.

3 As amended by the authors of ‘Textes des Purāṇa sur la Theorie Musicale’, p. 134. A variant reading for prakaraṇānugatath: मन्तव्यथ्यपदेशैन्, is noted as प्रकरणानुगतमतनयव्यपदेशैन् (see ibid. fn. 7) which though better yet makes no sense.

4 The Br. speaks of a kokṣā-gīti :

स्वरैश्च तु तिमिः: पूर्णा चौक्षा गीतस्वदाहता, Br. 292

But this has nothing to do with the matter dealt with in the Purāṇa, where the word kokṣā is totally out of place.

Another feature of the Purāṇic description which is discordant with gāndharva, is its mention of three grāmas and of mūrchanās and tānas pertaining to all three of them. This, however, seems to be a common Purāṇic trait—the two other Purāṇas we have just discussed also share it—and the explanation might be that the Purāṇas were in this matter leaning upon early Śikṣā works. The Nāradī Śikṣā, for example, while stating that only two grāmas exist in the world of men and that the gāndhāra-grāma is now to be found only with the gods, yet classifies the mūrchanās and tānas according to three grāmas, listing 21 murchanās by name. This perhaps was a tradition of enumeration going back to an early date when all three grāmas existed in musical practice.

Chapter 19 (khaṇḍa 3) of the Viṣṇudharmottara, which follows the chapter on gīta, treats of ātodya. Here, too, occur some topics clearly belonging to gāndharva-tāla. They are mātrā, kalā, vṛtti and prakaraṇa. Their exposition, though brief, has been patently based on manuals of gāndharva : (1) the measure of a mātrā is described as consisting of five laghu akṣaras,1 a characteristic peculiar to gāndharva; (2) the three gāndharva vṛttis—citrā, vṛtti and dakṣiṇā—are described in lines which closely echo Dattila and correspond to Bharata;2 (3) the seven sounded and unsounded beats (the number as enumerated by Dattila) are listed and termed kalā (again as in Dattila);3 (4) the basic yathākṣara-tālas of gāndharva are listed4 and (5) the two ways in gāndharva of rendering the ‘prakaraṇa’ (i.e., the seven gītakas)—with a relatively shorter or longer frame—are mentioned : “prakaraṇaṃ dvividhaṃ kulakaṃ cchedyakaṃ ca” Viṣṇudharmottara, khaṇḍa 3, ch. 19).

What follows this brief section on tāla in the Viṣṇudharmottara is a small passage on avanaddha, percussion instruments, the sixteen basic syllabic ‘bols’ of drum playing are listed and the method of playing them is indicated. The association of different ‘bols’ with different rasas is enumerated and then follows the kutapa-vinyāsa—seating arrangement of a dramatic orchestra—with which the chapter ends. The avanaddha passage is clearly inspired by the Nāṭyaśāstra,—chapter 34 (or a similar account from another text of the same nature); the kutapa-vinyāsa echoes lines of Bharāta (N.S. 34, 215) almost exactly.

1 पञ्चलघ्वक्षरोच्चारणकलामाता ।

—Viṣṇudharmottara, khaṇḍa 3, ch. 19.

2 तिस्रो वृत्तीः । जताः पूरितद्रवाणा च । द्वयोरक्षरकालविश्रान्तिः चतुरक्षरा वृत्ता अष्टमात्रादाक्षिणा ।

—ibid.

compare द्विमात्रा स्यात् कला चित्ते चत्वुमात्रा तु वार्त्तिके । अष्टमात्रा तु विधिदृश्विदक्षिा समुदाहृता ।

—Datt. 16-17.

and चित्ते द्विमात्रा कर्तव्या वृत्तो सा दृगुप्ता स्मृता । चत्वुर्य्य’ना दक्षिणा स्यादित्येवं कलासमूho ।

—N.S. 31, 3-4.

3 ग्रहया (आवाप) निष्क्रमो (निष्काम) विकल्पः: प्रवेशः शम्यातालः सन्निपातभेदः कलासमूho ।

—Viṣṇudharmottara, khaṇḍa 3, ch. 19.

cf. Datt. 113-114.

4 युक्त चचवुटः अयुक्त चाचपुटः लः प्लुतान्तः ।

—ibid.

We have quoted the text as restored in Textes des Purāṇa sure la Theorie Musicale. p. 150.

The chapter on ātodya in the Viṣnudharmottara also evinces that the music intended to be expounded was gāndharva; but here some matter from the non-gāndharvic ātodya section of the Nāṭyaśāstra (ch. 34) has also crept into the account.

The early purāṇic accounts of music thus show an awareness of gāndharva—of its distinct forms and its characteristic adṛṣṭa or transcendental propitiatory function.

Gāndharva was the most prestigious form of ancient music (apart from sāma); it was the form-supreme, the source of all other forms and the Purāṇas, therefore, when describing music, describe the gāndharva, taking it as the eminent representative of all musical forms. Gāndharva was, moreover, a sacred 'revealed' form and must consequently have been the most noble, the most towering form for compilers and listeners of Purāṇas. The purāṇic audiences, indeed, must have heard gāndharva being performed as liturgical music in the worship and propitiation of the gods and must have associated this form with the idea of the highest in music, To them it must have been the essence of all music.

CHAPTER X

GĀNDHARVA AND POST-ABHINAVA THEORISTS

Let us now consider some later, post-Abhinava theorists : their opinions regarding gāndharva and the notions they nurtured regarding its nature, purpose and extent. The most influential, comprehensive and voluminously versatile author on music in the period following Abhinava was, certainly, Śārṅgadeva. His Saṅgītaratnā-kara, which treats with dance as well as music, remains one of the most authoritative works on ancient and medieval music. Śārṅgadeva composed his treatise in the early years of the 13th Century1, about two hundred years after the Abhinava Bhāratī.

A period of two centuries is not a long one in view of the centuries of unbro-ken traditions of art and learning which India had known; but the period separating Śārṅgadeva from Abhinava was one of the most turbulent that India has passed through. Among other things, many traditions of śilpa—art and craft forms—and śāstra—the related doctrinal and technical knowledge—were broken or at least seriously impaired. Gāndharva, too, seems to have suffered.

We do not know if Śārṅgadeva, like Abhinava, had a first hand knowledge of gāndharva as a living art. It seems that his knowledge of the subject was mainly doctrinal; still, there are some hints in his treatment of the subject which tend to point towards a close second-hand if not a direct personal knowledge of some forms as actually rendered. His doctrinal knowledge of gāndharva, however, was quite thorough.

Śārṅgadeva has expounded all the topics and elements of gāndharva (including its musical structures, the jāti and the prakaraṇa or saptarūpa) with a wealth of detail. His exposition is mainly based upon that of Bharata, yet he had also studied other authorities like Dattila and Viśākhila and he often gives corroborative evidence from these authors or refers to their views as differing on points of detail from that of Bharata.

Śārṅgadeva clearly points at the sacred character and propitiatory-cum-expia-tory function of the jātis. He remarks that their forms have to be maintained strictly as decreed and that no tampering is to be permitted in this respect; for, the jātis were as inviolable and sacrosanct as Vedic mantras. He also points out that the jātis have given rise to the rāga-forms and that the knowledgeable can detect the subsequent derivative rāgas in the pristine structures of the parent jātis :

drśyante janyarāgāṃśastajjñairjanakajātiṣu

brahmaproktapadaiḥ samyakprāguktāḥ śaṅkarastutau

1 Under the patronage of King Siṅghana of the Yādava dynasty who ruled in Deogiri, modern Daulatabad, from 1210 to 1247 A.D.

api brahmahanaṃ pāpājjātayaḥ prapunantyamūḥ

reo yajūṃṣi sāmāni kriyante nānyathā yathā

tathā sāmasamudbhūtā jātayo vedasammitāḥ

(S. R. 1,7, 113-115)

"The experts (tajjña) can detect portions or limbs of rāgas in their parent jātis. (These jātis) sung to the words uttered by Brahmā—words we have already quoted—have their proper employ in praising Lord Śiva. The jātis can expiate even the sin of one who has killed a Brāhmana. Just as the ṛk, yajus and sāma mantras cannot be altered in their form, similarly the jātis which are born of sāma and which are comparable to the Vedas, are also inviolable."

We observe in this characterization of the jātis, all the salient features we have noted as peculiar to gāndharva. Though Śārṅgadeva has not actually used the word adṛṣṭa (as Abhinava does) in describing the function of the jātis, he has ascribed to jātis the same adṛṣṭa attributes as Abhinava. And Kallinātha, in commenting on Śārṅga-deva, has here used the term adṛṣṭa—"atha yathāvidhi jātigānairadṛṣṭaphalātiśayam darśayati—'brahmaprokteti'". (Kalā on S. R. 1,7, 113).

Śārṅgadeva states that an expert could detect the outlines or partial structures of a rāga-form in its parent jāti; and, in fact, while describing some of the jatis, Śārṅgadeva points out the names of the rāga-forms one could 'see' in them : in jāti śāḍjī, he says, the rāga varāṭī was visible;1 in āṛṣabhī, two rāgas could be detected, namely, deśī and madhukarī;2 in madhyamā, three rāgas, cokṣasādava, deśī and andhālī, could be felt as partially present;3 in śaḍjakaiśikī one could similarly 'see' the rāgas gāndhārarapancama, hindolaka, deśī and velāvalī;4 and so on.

Śārṅgadeva has also recorded the verses which Brahmā had presumably composed in praise of Śiva to be sung in the jātis; in addition, he gives the actual musical notation to which these verses were to be sung. For example, the verse tr which śāḍjī was sung was as follows :

tam bhavalalāṭanayanāṃbujādhikam

nagasūṇupranāyakelisamudbhavam

sarasakrtatilakapāṅkānulepanam

pranaṃāmikāmadehendhanānālam

(S.R. 1,7,64)

1 अस्यां वड्जादयो न्यासः । गान्धाररुष्टचुम्बमपन्यासो । वराटी दृश्यते ।

—S. R. 1, 7, 64.

Kallinātha qualifies the statement वराटी दृश्यते with the observation :

'वराटी ग्राह्यते' इति । विकल्पवशात्संप्रदायात्प्रविचारलोपाद्वा वराटी प्रतীয়ते । अस्माद्वराट्येकदेशमात्रं तु रागलक्षणम् ।

2 अस्यामौष्ठ्यमुपभोः त्यासः । अंशा एवापन्यासा देशीमधुकुर्यो दृश्यते ।

—S. R. 1, 7, 66.

Kalā : 'देशीमधुकुर्यो दृश्यते' इति देशी मध्युकुर्यो सदृशौमार्गत्वादुपलक्षयेत् ।

भावः ।

3 चोक्षषाडवदेश्याम्धाल्यो दृश्यन्ते ।

—S. R. 1, 7, 72.

4 प्रामुख्य गान्धारपञ्चमः; द्वौलकदेशीवेलावल्यो दृश्यन्ते ।

—S. R. 1, 7, 80.

Gāndharva and Post-Abhinava Theorists

163

We do not know the source of these verses. Neither Bharata nor Dattila have them. The source of the notations seems to be the extant Vṛtti on the Bṛhaddeśī—or a similar work. The Vṛtti gives a series of notations for some jātis calling them ‘prastāraś’ (cf. mōdern Hindustani ‘vistāra’) of the jātis; that is, musical movements typical of the jāti. The Vṛtti does not note any verses to be sung to these prastāras and the notations are not arranged word-wise as in Śārṅgadeva.1 Abhinava has given no verses or prastāra-notations of the jātis, though a few stray notational phrases do occur as illustrative of structural details.

The verses noted in connection with the jātis by Śārṅgadeva bear the stamp of a literary style belonging to a period certainly much later than the age of gāndharva: they are much more sophisticated and are dominated by a much later manner of Kāvyic diction than the few simple hymns quoted in connection with gāndharva by Bharata.

Śārṅgadeva speaks of a chronological development of musical forms which agrees with that given by Abhinava. Gāndharva, he says, was born of sāma, and then (in detecting living rāgas within the ancient jāti structures) he demonstratively reveals how gāndharva in turn gave rise to the rāgas. Regarding jāti-born forms, Kallinātha adds that some were direct derivatives of the jātis while others were indirectly derived through other intervening forms: “jātīnām sākṣātparamparayā vā janyarāgā eva” (Kalā on S.R. 4, 7, 113).

Gītakas, the tāla-counterparts of jātiś in gāndharva, have also been described by Śārṅgadeva, and he does so with an eye to every available śāstric detail. He illustrates the forms of gītakas through charts giving exact beat formations and other structural details. Here again Bharata is his main source, but his indebtedness to Dattila—and also others—regarding tāla is quite obvious.

Śārṅgadeva has not described all gāndharva forms in a separate independent section. His fundamental framework, like Dattila’s, is to divide music into two broad categories: svara and tāla. He deals first with svara, of which the first chapter, the svarādhyāya, includes the jātiś and an exposition of almost all the svara-topics listed by Bharata and Dattila under gāndharva. The second chapter deals with rāgaś, the third and fourth with composers of music and rules governing the popular compositions called prabandha. Then follows the section on tāla. This begins with the gītakas, which are dealt with in a chapter covering all the topics concerned with gāndharva tāla. Next follow what have been termed deśī tālas.

1 The first line from the above verse was, for example, sung as :

सा सा सा सा प नीध प धनी

तं भ व ल ला ट

रि गम गा सा रिग घसा घ

न य नों कु जा धि

रिग सा री गा सा सा सा

क

—S. R. 1, 7, 64.

164 A Study of Dattilam

Gītakas, too, like the jātiṣ, are described by Śārṅgadeva as having a sacerdotal function :

śivas tu tau prayojyāni mokṣāya vidadhe vidhiḥ (S.R. 5, 56).

"(These seven gītakas like madraka etc. and seven gītas like pāṇikā etc.) are to be employed in hymns addressed to Śiva ; they were performed by Brahmā in order to attain mokṣa."

Śārṅgadeva uses the term gāndharva in order to distinguish gītakas (and their component tāla-limbs) from deśī tāla-forms. He uses the term gāndharva also in context of the demarcation of tāla in the basic tāla-structures with which the complex gītakas were formed. With a view to the fact that it was necessary in gāndharva to render a tāla exactly as enjoined, he states :

gāndharvamārgakuśalaḥ kāmsyatāladharo'paraḥ gātulṅ sahāyaḥ kartavyaḥ pramādavinivṛttaye (S.R. 5, 38-39)1

In using the term gāndharva here, Śārṅgadeva obviously had the special sense of the form in mind. Kallinātha, in fact, makes this explicit in commenting: "gān-dharvasyātyantiyatvena drṣṭaphalasādhanatvāt... : gāndharva is totally bound by rules, for only thus can it result in adṛṣṭa..."

Śārṅgadeva's indebtedness to Abhinava, whose influence (especially regarding the doctrines of the Nāṭyaśāstra) was far and wide, is evident, and, indeed, he mentions Abhinava as one of the vyākhyāṭṛs (commentators) on Bharata whom he had studied (S.R. 1, 1, 19-21). Like Abhinava, Śārṅgadeva has contrasted gāndharva and gāna forms, using the terms in the same sense as his illustrious predecessor. Introducing his fourth chapter which deals with prabandha, Śārṅgadeva says :

rañjakaḥ svarasandarbho gītamityabhidhīyate gāndharvaṃ gānamityasya bhedadvayamuditam anādisampradāyaṃ yadgandharvaiḥ samprayujyate niyatam śreyaso hetustadgandharvam jagarbudbāḥ yattu vāggeyakāreṇa racitaṃ lakṣaṇānvitam deśirāgādiṣu proktaṃ tadgāṇaṃ janarājanam tatragandharvamuktam prāgadhunā gāṇamucyate (S.R. 4, 1-3)

"A pleasing conglomeration of notes is called a gīta. It is of two kinds : gāndharva and gāna. That which has been handed down in an unbroken tradition since the beginning of time and which is sung by the gandharvas is called gāndharva by the experts ; it is gandharvamuktam prāgadhunā gāṇamucyate"

1 We have already discussed and translated the śloka in connection with the use of cymbals in gāndharva ; ch. V.

undoubtedly a source of spiritual well-being. That gīta, which is created by musician-composers (vāggeya-kāra) in deśī rāgas and is endowed with qualities appropriate to it, is known as gāna ; it has a great popular appeal. Gāndharva has been described earlier, now gāna shall be described."

Indeed, Śārṅgadeva uses the terms gāna and gāndharva as denoting two distinct musical forms even in his first chapter, the svarādhāya, where the jāti s and related topics are delineated :

gāndharve mūrchanāstānāḥ śreyase śruticoditāḥ gāne sthānāsya lābhene te kūtatāncopayoginaḥ (S.R. 1, 4, 91)

"Mūrchanās and tānas have been ordained in gāndharva by Vedic injunction, but in gāna they are of practical use (only) along with kūtatānas, and help in attaining the location of notes (sthāna)."1

Kallinātha clearly connects gāndharva—which he calls a particular species of song—with an adṛṣṭa goal and quotes some lines from an unnamed Smṛti which ordains the use of particular tānas and mūrchanās for propitiatory functions and promises great transcendental merit thereby. He adds that kūtatānas when associated with gāna have only a dṛṣṭa or non-spiritual goal.2

One clearly notices the influence of Abhinava in Śārṅgadeva's statements. But Śārṅgadeva, we observe, uses the term gāna in a wider sense than is evinced by the use of the term in the Abhinava Bhāratī. Abhinava uses the term to mean profane and popular gāndharva-born forms adapted and moulded for the purpose of theatre ; Śārṅgadeva, however, uses gāna as denoting all profane and popular forms.

Though Śārṅgadeva uses gāndharva to denote a specific body of music, gāndharva according to Śārṅgadeva's notion included many more melodic forms than are warranted by Bharata's and Dattila's definitive description and by Abhinava's analysis. In S.R. 4, 1-3, (quoted above) Śārṅgadeva implies the extent of forms included in gāndharva; these include not only the jāti s but also non-gāndharva forms like grāma-rāgas, bhāṣās, antarā-bhāṣās etc. Śārṅgadeva says : 'gāndharva has been described before, now gāna shall be described.' This statement comes as an introduction to the fourth chapter, from which the realm of gāna avowedly begins. The inference, clearly, is that forms from jāti s to the antarabhāṣās, all of

1 Kallinātha commenting on sthāna here says : स्थानस्य तत्त्वमूर्छनावशेन सारिरतानां स्वराणामाधारस्वरतेनोभयेन परिज्ञाने..... -Kalā on S. R. 1, 4, 91.

2 'तौधैव' बद्यमाणलक्ष्यो गीतविनियोगस्थल तत्साहचर्येणोच्यते । अपि शुद्ध एवं विवक्षितः । श्रेयःसेष्टसुफलाय......कूटतानातां तु केवळं दृष्टफलत्वेनोपयोगमात्रं = 'गान' इति । गानं च बद्यमाणलक्षणं तावत् । -ibid.

which have been expounded in preceding chapters are to be included under gāndharva. Indeed, Kallinātha on this passage comments : “gāndharva denotes forms beginning with jāti upto antara-bhāṣās described in the first and second chapters” (svaragatarā-gavivekayorjātyantarābhyāsāntam yaduktam tadgāndharvamityarthah; Kalā on S.R. 4, 1-3). Again in the chapter on the gītakas, Kallinātha in commenting on the term gāndharva (used by Śārṅgadeva in S.R. 5, 38-39) names the same extent of forms; however, in this case, he also names the gītakas as included in gāndharva: “the fourteen gītas, such as madraka, the jātis such as ṣaḍja, the six variety of rāgas such as the grāma-rāgas : these are the forms denoted by the word gāndharva.”1 The six varieties of rāgas alluded to were: grāmarāgas, upa-rāgas, rāgas, bhāṣās, vibhāṣās and antara-bhāṣās listed and named by Śārṅgadeva in the second chapter of his work (S R. 2, 8-47). These were the so called mārga-rāgas.

Mārga is, in fact, another term which Śārṅgadeva uses as synonymous with gāndharva and its use is more copious. Right at the beginning of his work Śārṅgadeva introduces a two-fold classification of musical forms, mārga and deśī, distinguishing them on the same lines as gāndharva and gāna :

mārgo deśīti tad dvedhā tatra mārgaḥ sa ucyate yo mārgitō viriṇcyādyaiḥ prayukto bharatādibhih devasya purataḥ śambhorinitābhuyudapadaḥ deśe deśe janānām yadrucyā hṛdayarañjakam gītam ca vādanam nṛttam taddeśītvabhidhiyate

(S.R. 1, 1, 22-24)

“That (i.e. saṅgīta) is two-fold : mārga and deśī. Mārga is that which has been discovered (mārgita)2 by Brahmā etc , and was performed by Bharata and others before Lord Śiva. Mārga certainly bestows spiritual well-being. But that music—instrumental and vocal—and that dance, which delights people of different regions and is born of different aesthetic tastes, is called deśī.”

We observe only one small difference here from his statements distinguishing gāndharva and gāna : the terms gāndharv1 3 and gāna in Śārṅgadeva's usage concern music alone while mārga and deśī embrace dance as well.

1 ‘gāndharv’ iti prāgvat. vāsavyamaṇapi madrakādinī čaturdaśa gītaṅ gītāni prāguktaḥ। pāṭhyādyavijñāyamo grāmarāgādayasča paṭṭabhidhā rāgā gāndharvavedanotsannnte । —Kalā on S.R. 5, 38-39.

2 mārgita iti ‘mārga anveṣaṇaṁ’ ityevaśabdāt: karmaṇi niṣṭhāyāṁ hṛvam । Kalā. ‘yo mārgita iti । mārgitoñveṣitō dṛṣṭaḥ । I Sudhā. The phrase may also be translated as ‘which has been revealed to Brahmā etc.

3 Śārṅgadeva has also used the word ‘gāndharva’ in classifying vāggeyakāras, i.e., musician-composers. Gāndharva was the name for the most superior vāggeyakāra : he was proficient in handling both mārga and deśī forms :

mārge deśī vā yo deśi sa gāndharvamidhyate —S.R. 3, 12.

This use of the word is clearly quite distinct from its denotation as an ancient body of music.

Gāndharva and Post-Abhinava Theorists

167

In his chapter on tāla, too, we notice that Śārṅgadeva uses the word mārga as synonymous with what he understood as gāndharva. In his preamble on tāla he remarks: “the nature of tāla is known as two-fold by the wise : one pertains to mārga, the other to deśī. Mārga is denoted by two kinds of actions, sounded and unsounded . . . . .”1 This section on mārga tālas ends with gītakas and pānikā etc.

Then deśī tālas are taken up and Śārṅgadeva introduces this new section with an appropriate maṅgala-śloka (a customary propitiatory verse). Bowing to Śiva, Śārṅgadeva says:

śrīkīrtivijayānandaṃ pārvatīlocanotsvaṃ rājacūḍāmaṇiṃ natvā tālaṃ deśīgataṃ bruve

(S.R. 5, 236)

There is a double-entendre (śleṣa) here. Though the epithets used apply obviously to Śiva, they also denote names of prominent deśī tālas.2

The section on mārga-tāla treats with the same topics as are included under gāndharva tāla by Bharata and Dattila. In fact, Śārṅgadeva has also used the word gāndharva to denote this mārga body of tāla-forms (S.R. 5, 38-39),3 thereby clearly indicating that he thought gāndharva as co-extensive with mārga. Two lines further Śārṅgadeva again uses the word gāndharva as denoting what he had called the mārga-body of tāla forms. He uses the word deśī in this context as denoting a set of forms quite distinct from gāndharva.4

1 स च द्वेधा बुद्धः: स्मृतः। मार्गदेशीगतं तल्लाक्ष्यस्य क्रिया हि द्विधा। निःशब्दाशब्दयुक्ता च......

—ibid. 5, 3-4.

2 ‘श्रीकान्तविजयानन्दमिति’। अत्र ताळेश्वरयोरविशेषणानि श्लिष्टानि दृष्टव्यानि।

—Kalā on S.R. 5, 236.

3 We have already had occasion to discuss the verse in question. It says that in gāndharva a person playing cymbals should assist the main singer so that errors can be avoided. In this verse Śārṅgadeva uses the word mārga in conjunction with gāndharva गान्धर्वमार्गकुशल; the terms according to Kallinātha were used as synonyms : तस्मिन् गान्धर्वे मार्गे कुशलः.

4 In classifying yugma and ayugma tāla-structures in mārga-gāndharva forms, he says that these structures gave rise to many mixed (saṅkirṇa) forms; but adds that according to some authorities only 5 saṅkirṇa structures were permitted in gāndharva. Then he points out that tāla-structures had many sub classes consisting of irregural number of time-units ; these he says will be described in the section on deśī tālas, implying that structures not permitted in gāndharva were quite valid in the deśī forms:

युग्मस्य ये तयो भेदा: पादवायुगमस्य हि क्रियाः। तेऽप्यमुन्योन्यसंकर्यात् संकीर्णा: बहवो मताः।। अन्ये पृथक्चव संकीर्णान् गान्धर्वेऽपि विद्धि बुधाः। पञ्चसप्तनवापिस्येदंर्‌केश तत्रलक्षणम्।। तालाक्षरस्वरार्द्रे तित्थं चतुरङ्गकालादिका:। चञ्चत्पुटादिमेदास्तु सन्ति खण्डाश्रयादयः परे। देशीतालप्रपञ्चेन तानपि व्याहरामहे।।

—S.R. 5, 39-42.

(see part III section II for technical discussion on the point).

Again, in defining the specific features of the mārga dances, Śārṅgadeva clearly reveals this coextention of mārga and gāndharva forms. The mārga dance, he says, is of two varieties, tāṇḍava and lāsya; their musical accompaniment consists of vardhamāna, āsārita etc.1 As in Bharata, gāndharva has here been related to its sacredotal dance counterpart, again showing that in Śārṅgadeva’s scheme of classification the term mārga was interchangeable with gāndharva.

The word mārga has not been used as a broad category to classify musical forms by ancient authorities like Bharata or Dattila.2 Mataṅga uses the term, though, evidently, not quite in the sense as found in Śārṅgadeva. The latter has defined mārga as a corpus of music both ancient and immutable, quite distinct from and contrary to deśī forms which were more or less contemporary and were still in a variable and fluctuating state. Mataṅga on the other hand uses mārga in a context where the term evidently denotes a sub-category of deśī forms and not a corpus of music distinct from the deśī. Defining deśī-music he says :

abalābālogapālaiḥ kṣitipālaiḥ nijecchayā

gīyate sānnurāgeṇa svadeśe deśīrucayate

nibaddhāścanibaddhaśca mārgo’yaṃ dvididho mataḥ

aplapadi? (ālāpādi) nibaddho yaḥ sa ca mārgaḥ prakṛrtitaḥ

evamrakaro deśī yā jñatavyā gītakovidaiḥ

evametanmayā prokatam deśyā utpattilakṣaṇam

(Br. 13-15)

1 आबालकामनयेऽपि भावानां ध्यानोदित यत् । तन्नामतो भोगशब्देन प्रसिद्धं नृत्यवेदिनाम् ॥ गानविच्छेदप्रमाणं तु सर्वाङ्गनयरञ्जितम् । आङ्गिककोक्तप्रकारेण नृत्तं नृत्यविदो विदुः ॥ ताण्डवं लास्यमित्येतद् द्वयं हि नृत्तं निगद्यते । वर्धमानादिसारितादिगीतस्तत्तद्गुणायुतम् ॥ —S.R. 7, 26-28.

2 Mārga is a topic in gāndharva tāla where it denotes the various ‘ways’—fast, slow etc.—of tempic usage. This meaning, clearly, is entirely different from the sense in which Śārṅgadeva uses the term as a general category defining a body of musical forms.

Mataṅga's definition of deśī is the same as that of Śārṅgadeva (S.R. 1, 1, 23-24), but his notion of mārga clearly does not comply with what was considered as mārga by the latter. Mārga in the above passage is a name given to those deśī forms which were comparatively more regulated or rule-bound than the others.

Abhinava, too, has used the word mārga. He, too, couples the term deśī with mārga; but it is quite clear from his usage that mārga according to him was quite distinct from gāndharva. It will be worthwhile to consider his use of the term :

1. There were 18 jātis accepted in gāndharva. 7 were śuddha and another 11 were born of two or more śuddhas combined in certain ways. Beyond these 11 mixed or saṃsargaja svara-forms, no more mixtures were permitted in gāndharva. But other forms, born of the jātis through a further series of combinations and permutations did exist; these were forms such as rāgas etc. derived mainly on the basis of jāti-aṃśas. A possibility of their being mistaken for a jāti was to be guarded against. Thus Abhinava, commenting on N.S. 28, 46 where the 7 śuddha and 11 saṃsargaja jātis have been listed, remarks : “anyāstu bhavantu jāṭaya iti aṃśarāgavibhāṣādeśī-sthādaśamārgyāsāvityarthah.-other forms could (imaginably) become jāṭis; for the aṃśa (born) rāga, vibhāṣā. deśī, mārga etc. bear the appearance of (jāṭis) since they copy them in their major features such as antarāmārga (specific movements) and nyāsa (the note of rest) etc.; (to avoid this confusion) the (antara) mārga and the nyāsa of the 18 jāṭis has been definitively given.” Here mārga, bracketed with the deśī, has been clearly listed as a non-gāndharva form.

2. In his long analysis of gāndharva and its distinctiveness at the beginning of N.S. ch. 33, Abhinava lists both mārga and deśī as belonging to the gāna form of music and associates both with non-gāndharva forms like rāga, bhāṣā and vibhāṣā : “kiyadvā rāgabhāṣāvibhāṣādeśīmārgadigatānāṃ svarāṇāṃ śrutivaicitryaṃ brūmaḥ.”1

3. In another passage he links both mārga and deśī with the great variety of gāna forms such as grāma-rāgas and bhāṣās that had arisen out of various jāṭi-aṃśas (the dominant notes in a jāṭi) : “vistaravivakṣāyaṃ tu jātyaṃśakagrāmarāgagabhāṣādeśī-mārgavaicitryeṇānanantā eva svarāḥ” (A.B. on N.S. 29, 8).

4. Yet again he brackets mārga and deśī with forms like bhāṣā where unlike gāndharva, madhyama was a dispensable note and could be dropped at will : “anena bhāṣādeśīmārgadiṣu madhyamasyāpi viṅśitvamabhyupagatam bhavati” (A.B. on N.S. 28, 64-65)

Abhinava gives no specification or definition of mārga as a form. All we can gather from him is that he includes it among forms which he classes collectively as gāna and which arose from gāndharva.

1 This occurs in a passage where Abhinava describes the great variety of śrutis that can be employed in gāna in contrast with gāndharva.

गाने तु काकल्यनतरस्वरितिप्रभृतौ विचित्रस्वरुप्रबन्धम् । स्वराणां मालवकेशिके चतुःश्रुतिकाक्रूस्य दर्शनातः कियद्वा रागभाषाविभावादेशीमार्गदिगतानां स्वराणां श्रुतिवैचित्र्यं ब्रूमः ।

• A.B. on N.S. 33, 1. see also ch. V.

170

A Study of Dattilam

Kallinātha has the most explicit definition of mārga vis-a-vis deśī. “Those forms”, he says, “that are created in various areas and regions (of the country) for the plearuse of the populace are in the deśī state; with this end in view their composition follows a free and spontaneous course.”1 He quotes an authority named Añjaneya who had defined deśī rāgas as :

yeṣāṃ śrutisvaragrāmajātyādinīyamo na hi nānādeśagatitchāyā deśīrāgāstu te smṛtāḥ

(Kalā on S.R. 2,2, 159-160)

“Those (rāgas) which are not governed by rules regarding śruti, svara, grāma, jati, etc. and contain echoes of musical movements popular in regional (or folk) music of various areas, are known as deśī rāgas.”

Kallinātha adds that like song-forms, even instrumental music and dance, are known as deśī when they are composed on the basis of pure whim, without imperative rules to control them: “evaṃ vādyanṛttayorapi kāmacārapravartitayordeśītvamavagantavyam” (Kalā. ibid.). Mārga, he says, is opposed to deśi and hence from the above definition of deśī it follows that musical compositions and forms that were governed by rules were called mārga : “niyame tu sati teṣāṃ gītādināṃ mārgatvameva” (Kalā. ibid.).

Śārṅgadeva, too, evidently had a similar notion regarding mārga in his mind. And mārga understood in the above way would naturally include the jātiś within its fold, for jātiś were forms in which rules were imperative—but mārga could also include much else besides. And it did.

Kallinātha says that svara forms beginning with jāti and extending upto antara-bhāṣa are mārga (or gāndharva, for he took the terms as synonyms).2 Thus mārga according to Kallinātha includes grāma-rāgas, upa-rāgas and rāgas born of grāma-rāgas, bhāṣās, vibhāṣās, and finally the antara-bhāṣās. In other words, in addition to jāti, all those forms which Śārṅgadeva has listed in the first prakaraṇa (section) of his second chapter (called rāgavivekādhyāyah) are called mārga forms.3 Let us explore the possible reason for this classification.

Grāmā-rāgas were very ancient forms and were very intimately connected with the jātiś. They were the first-born of the jātiś. The Vṛtti on Bṛhaddeśī quotes a dictum, ascribing it to Bharata, that grāma-rāgas are direct off-springs of jātiś : “tathā cāha

1 deśītvam ca tattadanujamanorujjanekfalavetan kāmacārapravartitastatraḥ. —Kalā. on S.R. 2, 2, 159-160.

2 gānadhvamārgaḥ.1 gānantu vāgrameyakāradīni pratamatsvāt pṛthayitamevaḥ. tatra gāndharvamuktaṃ prāpnoti. svaragatarāgavibhāgedyojyāyābhantar bhāpāntyaduktam gāndharvamityamīyateḥ:

—Kalā on S. R. 4, 1-4.

3 The colopnph of S.R. 2, 1, thus defines the scope of the first prakaraṇa, chapter 2 as : इति द्वितीये रागविवेकाध्याये ग्रामरागोपरागरागभाषाविभाषावनतरभावविवेकाध्यायं प्रवदं प्रकरणम्

bharatmunịḥ-jātisambhūtatvādgrāmarāgāṇāmiti" (Bṛ. Vṛtti on 321). Bhāṣā, vibhāṣā etc. were born of the grāma-rāgas and not directly of the jātiṣ. The Bṛhaddeśī notes a chronological genealogy of these forms as follows :

grāmarāgodbhavā bhāṣā bhāṣābhyasca vibhāṣikāḥ vibhāṣābhyaśca sañjñātastathā āntarabhāṣikā

(Bṛ. p. 105. This verse occurs in the the chapter incorporated into the Bṛ. from Sarvāgamasamhītā)

"Bhāṣās are born of grāma-rāgas, vibhāṣikās are born of bhāṣās and from the vibhāṣās arise the antara-bhāṣikās."

Even a late author like Kallinātha, explaining the significance of the name grāma-rāga and its relation to 'grāma', connects this form directly with the jātiṣ : "though the grāma-rāgas are not directly born of the grāmas but only by way of the jātiṣ, yet they are less removed from the grāmas than such forms as bhāṣās, rāgas etc.-hence are they called grāma-rāgas."1

Realising the closeness of the grāma-rāgas to the jāti-form, Abhinava, indeed, calls them gāndharva-kalpa : 'almost gāndharva'. The remark occurs in connection with the compulsory use of madhyama in gāndharva. Bharata-we have seen-has a dictum that the note madhyama was never to be dropped in sāma gāna and in forms that were gāndharva-kalpa (N.S. 28, 65). Abhinava interprets the term gāndharva-kalpa as embracing not only gāndharva, but also the grāma-rāgas which were gāna forms closest to the jātiṣ; they were so to speak just on the other side of the line demarcating gāndharva. "By the phrase 'gāndharva kalpa,'" Abhinava says, "is meant the gāndharva śāstra, but also forms where the extent of gāndharva almost does not end, namely, the grāma-rāgas; this decree, however, clearly indicates that in bhāṣā, deśī, mārga etc. madhyama is a dispensable note."2

Abhinava's observation reveals the great proximity of the grāma-rāgas to the jātiṣ. Already in the bhāṣā, a direct offshoot of the grāma-rāga in the lineage of forms descended from the jatiṣ, the note madhyama had lost its honoured position. We also notice significantly that Abhinava here relegates mārga-forms to a position not only removed from the jatiṣ but also away from jāti-born grāma-rāgaṣ.

1 ग्रामरागा इति । ग्रामयोर्जातिनिर्यद्भावेनोल्पन्नानामपि भाषारागाधपेक्षया व्यवधानाल्पत्वादित्यां ग्रामरागत्वव्यपदेशः ।

• Kalā on S.R. 2, 1-8-14. Explaining the connection of grāma-rāga with grāmas Kumbha also says : रागस्य ग्रामस्य सम्बन्धे या जातिरिह वर्तते । तयो यो जन्यते रागः स तदग्राम इरीरितः ॥
• S. Raj, 2, 2, 1, 141.

2 गान्धर्ववंशान् ईपदसमाप्ते गान्धर्वे ग्रामरागरागरूप इति वानen भाषावेदिभिर्मागंदिवु मध्यमस्यापि विनाशित्वमम्युपगतं भवति ।

• A.B. on N.S. 28, 65. However regarding bhinnamaṣadja, a grāma-rāga, he says that 'ma' could be omitted in it (see ch. V, p. 93).

The grāma-rāgas must have shared other rules with the jātis besides the indispensability of madhyama. Certain grāma-rāgas—those classed as the śuddha ones —were so close to the jātis that each of them was born of a single unmixed jāti.1 Others were born of a mixture of jāatis through a process which they shared with the 11 samsargaja or mixed jāatis. Their structural resemblance to the jātis must have been very great.

Yet, howsoever ancient and howsoever close the grāma-rāgas may have been to the jāatis, they were in essence malleable forms and usable as gāna. They were, indeed, the prototypes of all gāna-oriented forms. Abhinava, to this effect, quotes a predecessor of his—referring to him as the Ṭikākāra—who had aptly shown the basic aesthetic and musical difference between jāatis and grāma-rāgas in saying : “the jāatis serve as the basic svara-structure from which gāna forms have originated and are thus useful for gāna. But jāatis are not the source of the peculiar charm (rakti) of gāna forms; this quality is imbued in them through grāma-rāgas.”2 This suggests that though jāatis and grāma-rāgas had a close structural kinship, the musical approach in rendering the two forms was entirely different in spirit : jāatis were sacred liturgic, hieratic melodies, absolutely unbending and uncompromising in their form and spirit; grāma-rāgas, on the contrary, seem to have been early embodiments of a freer, more pliable and hence a basically secular and profane attitude to music (many rules—we have seen in ch. V—binding in the jāātis, in fact, did not apply to them in a strict manner). They must have had a more ‘popular’ appeal and were thus called ‘rāga’ : Mataṅga defines rāgas as those forms which had the power to please the populace (rañjako janacittānām sa ca rāga udāhṛtaḥ; Br. 281) and he includes grāma-rāgas with forms belonging to the rāga-mārga3, the corpus rāgas as distinct from the jāatis.

Like other gāna forms, grāma-rāgas, too, played an inherent role in the ancient dramatic structure : the plays employed them as essential nāṭya-sāmagri (as important to the unfoldment of the plot and revealing the moods of characters). The Vṛtti on the Brhaddesī has an interesting passage where it distinguishes what it calls a gīta from a rāga : rāgas, it says, are songs such as the grāma-rāgas, which conform to the ten characteristics that govern musical structures;4 these rāgas become gītas when

1 Regarding these, the ‘śuddha’ grāma-rāgas, Mataṅga says : निरूपणान्यजातीयो स्वजातितिमनुबन्धका: । स्वजात्यनियमापाश्चैव त एव ग्रामा: परिकीर्तिता: ॥ —Br. 317-318.

2 टीकाकारस्त्वाह: । आद्ययसवरस्वरुपेण जातीनां गान उपयोगो न तु रक्तिसम्पादकत्वेन तत् ग्रामरागाणामेव । —A.B. on N.S. 29, 14.

3 Grāma-rāgas are the very first of the rāga-forms which Mataṅga describes, obviously, because of their being closest of all ‘rāgas’ to the jāatis.

4 See the ten jāti-lakṣaṇas, part III section I. Like the jāatis, grāma-rāgas were governed by the same lakṣaṇas or formal characteristics such as aṁśa (dominant note), graba (initial note), nyāsa (note of rest) and other modal features; though, of course, a greater freedom was permitted them.

they are moulded to the five kinds of dhruvās used in dramas.1 In support of his view

the Vṛttikara quotes the authoritative statement (āptavacana) of the ancient ācārya

Kaśyapa who had spoken of the intimate connection between grāma-ragas and

dhruvās :

kvacidamśaḥ kvacinn yāsaḥ sāḍabauḍubite kvacit

alpatvam ca bahutvam ca grahāpanyāsasamyutam

grāmarāgā prayoktavyā vidhivad daśarūpake

praveśākṣepaniṣkrāmaprāsādikamathāntaram

gānaṃ pañcavidhaṃ yat tad rāgairebhiḥ prayojayet

pūrvaraṅge tu śuddha syāt bhinnā prastāvanāśrayā

vesarā mukhayoḥ kāryā garbhe gauḍi vidhīyate

sādhāritāvamarśe syāt sandhau nirvahaṇe tathā.

(Vṛtti on Bṛ. 364A)

"The wise should know the aṃśa (dominant note), nyāsa (note of rest), apan-

yāsa (subsidiary note of rest), graha (initial note), the weak notes (alpatva),

the notes used copiously (bahutva) the structural range in the lower and upper

registers (mandratārau) and (rules regarding) sāḍava and auḍuva (governing

grāma-rāgas) and formulate them accordingly. These grāma-rāgas should be

appropriately employed in dramas: the five kind of gānas—praveśa, ākṣepa,

niṣkrāma, prāsādika and antarā—should be sung to these rāgas. The śuddha

(variety of) grāma-rāgās should be employed in the purva-rāṅga, the bhinna

during prastāvanā, the vesara during the two mukhas, the gauḍa during

garbha and the sādhārita during avamarṣa, sandhi and nirvahaṇa."2

We must make a clarification before proceeding. We have said earlier that

śuddha, bhinna, vesara and others were names of gītis and that different gītis were

different styles of rendering rāgas (see p. 121). In the above passage from Kaśyapa,

śuddha, bhinna, vesara etc. are used in the sense of different classes of rāgas, not just

different styles of rendering the same rāga. This apparently introduces an incongruity.

How can a gīti be both a style of rendering and a classificatory category of forms?

In current musical practice the same rāga sung in a different style does not become a

different rāga unless significant structural differences are also introduced. Different

gītis evidently entailed such structural differences besides a difference in style. After

describing the stylistic characteristics of different gītis (Bṛ291-308), Mataṅga proceeds

to enumerate rāgas and to place them in different classes, again on the basis of the same

gītis. This was not just as an odd way of classifying the same forms on the basis dif-

ferent styles. For, the names of rāgas enumerated under different gītis are not identi-

cal. And when the same rāga does occur under diffferent gīti-classes, a difference of

1 ननु गीतरागयोः को भेदः । उच्यते । दशलक्षणलक्षितं गीतं रागस्वराभिधेयम् । गीतं चतुरक्षरोपेतं ध्रुवायोगात् पश्चाद्-

विधम् ।

— Bṛ Vṛtti. on 364A.

The term, we notice, closely approximates the sense of the term gāna as used by Abhinava.

An ancient dramatic plot was divided into, various parts; prastāvanā mukha, garbha, avam-

arṣa, sandhi and nitvahaṇa were technical names for these parts.

more than style or manner of rendering is clearly implied. An example discussed in

the Vṛtti will bear this out. The rāga kaiśika is listed both as śuddha form (śuddha-

kaiśika) and a bhinna form (bhinna-kaiśika). Both śuddha and bhinna kaiśika were

derived from the same parent jātiṣ (kaiśika and kārmāravī). The Vṛtti on Bṛhaddeśī

in describing bhinna-kaiśika raises the question : is this rāga identical with śuddha

kaiśika? The reply made is that it is not. The Vṛtti then points out important

structural differences and concludes: bhinna kaiśika is different for it has a different

form : rūpānyatvenāmbhidhīyate" (Vṛtti on Br. 330).1

With this in mind we can return back to the matter at hand.

At the end of his commentary on the 28th chapter of the Nāṭyaśāstra, Abhi-

nava quotes a long ancient passage in which specific grāma-rāgas are traced to their

jāti-origins. The passage is presumably from Kaśyapa, but whether or not it is the

same Kaśyapa who is quoted in the Vṛtti on Bṛhaddeśī we do not know.2 Here, too,

the grāma-rāgas are described as being employed in dhruvāgāna :

eṣā vibhāṣā gīti tu grāmarāgāśrite mate

dhruvāgānopayogāya tadetadgītisaptakam

Bharata also (in the 32nd chapter, the dhruvādhyāya) has a passage defining

the connection of grāma-rāgas with the dramatic plot. Like Kaśyapa he lays down the

types of grāma-rāgas appropriate to the various junctures of a play (N.S. 32, 426-429).

Though Bharata does not use the word grāma-rāga as such, it is clear from the epi-

thets he uses that the reference is to grāma-rāgas.3

1 We quote the passage at length. The purport is clear inspite of a faulty reading of some crucial words:

भिन्नकैषिको मध्यकैषिकग्रामसंभवः | कैषिकीकामरविचार्जितायुष्पन्नवतु । गृह्योऽपि स पञ्चमो न्यासः । निषादौ चैव धैवतौ ।।

काकली । पूर्बंषडजं ।।

(cf. śuddha-Kaiśika which has been identically described in Vṛtti on Br. 321)

शुद्धकैषिकं उ दय(?) । यदापि रागं तथा अपि भेदोऽस्ति । मदनबहुलोद्गार । ये तु संस्थाने (न) शुद्धकैषिके स्वरलापः । क्रियते

तत्संस्थानं विह(ङ्गं) । तारेक स्वरैरालापः । कर्तव्यः । विविधकैषिके हि तारस्वैरालापः । कर्तव्यः इति । रूपान्यत्वेनायं

भिद्यते ।"

To us it appears that the above difference between śuddha and bhinna kaiśikī was somewhat

akin to the difference between such Hindusthani rāgas as sohni and pūriyā.

2 It appears not, because Kaśyapa in Abhinava's passage has seven gītis whereas Kaśyapa in the

Br. passage has only five .

3 पूर्वरङ्गविधाने तु कतर्वयो रागजो (पाठभेदः 'पाडवो') विधिः । देवतानांधिकारस्तु तत्र सम्प्रकीर्तितः ।।

ततश्चकार्यवशेनु नानाभावसमाश्रयम् । ग्रामध्रेषु तु कर्तव्यं यथासादृश्यनाश्रयम् ।।

मुखे तु मध्यमः ग्रामः । पञ्चः । प्रतिमुखे भवेत् । साधारितस्थया गर्भे विमर्षं चैव पञ्चमम् ।।

कैषिकं च तथा कार्य गान निर्वहणे बुधि । सन्निवृत्ताश्रयं चैव रसभावसमन्वितम् ।।

—N.S. 32, 426-429.

Bharata is here referring to the grāma-rāgas named ṣadja-grāma, madhyama-grāma, śuddha,

pañcama, kaiśikamadhyā, sādhāritā and kaiśika (see Kaśyapa quoted at the end of A.B. on

N.S. 28, 141. also S. R. 2, 189.) Note the comments of A.B. :

काश्यपमुनिप्रणीतेऽपिनिह्नुप्तरागपि ग्रामरागनगरागन् स्वीकरोति । मुखे मध्यमग्रामः । पञ्चः प्रतिमुखे गर्भं साधारितः ।

पञ्चमौडुम्बरीणां ।

—A.B. on N.S 32, 428.

Gāndharva and Post-Abhinava Theorists

175

The Vṛtti on Bṛhaddeśī, in fact, quotes this passage from Bharata (with a reading somewhat variant from the Nāṭyaśāstra) at the end of the section on the śuddha grāma-rāgas: the purpose being to authoritatively establish the connection of these śuddha grāma-rāgas with the theatre.1 Kallinātha has also quoted this passage with yet another variant reading. He quotes it as beings quoted in the Bṛhaddeśī; his version of the passage establishes the connection of the entire gamut of grāma-rāgas (and not only the śuddha ones) with the theatrical plot and its various divisions.2

We observe that the grāma-rāgas were not only the closest to gāndharva of all later forms, some of them like gāndharva were also accorded a role in the propitiatory pūrvaraṅga. The grāma-rāgas which were thought exalted enough for this function were the śuddha grāma-rāgas; perhaps, as the name implies, the nearest in structure to the jātis.

Indeed, Mataṅga has said that in the pūrvaraṅga even the jātis employed should be the śuddha ones : “pūrvaraṅgakṛte śuddhā kartavyāṃ gītayoktr̥bhiḥ” (Br̥. 241). The śuddha grāma-rāgas were each born of a single jāti and hence they, too, were comparatively śuddha or ‘pure’, unadulterated forms.

Of the seven śuddha grāma-rāgas, the rāga called ṣādaba has been named most often in connection with the pūrvaraṅga by the authorities. The Vṛtti on Bṛhaddeśī describes it as “ṣaṭsu rāgeṣu mukhya iti ṣādabaḥ—it is supreme among the (other) six rāgas, hence called ṣādaba” (Br̥. Vṛtti on 318) and adds “pūrvaraṅge pracuraprayogatvādāsya śuddhaṣādabasyādau nirdeśaḥ : śuddhaṣādaba is the first (rāga) to be described because of its profuse employ in the pūrvaraṅga” (Br̥. Vṛtti, ibid.).

The ancient pūrvaraṅga was predominated by the gāndharva forms like gītakas, vardhamānaka etc.3 which were propitiatory forms per se, but some dhruvā forms, too, seem to have had a place in the total pūrvaraṅga scheme. Bharata has, in fact, named the avakr̥ṣṭā dhruvā in connection with the pūrvaraṅga.4 Its propitia-

1 ननू ग्रामि द्विविधयोगवि शोध: कथमालमभ्यते। वचनादेव लभ्यते। तथा चाह भरत:— mध्ये तु मध्यमग्राम: षड्ज: प्रतिमुखे भवेत्। गर्हौ सारङ्किरातचैववंश्यं तु पञ्चमं ॥ संहारे कैशिक: प्रोक्त: पूर्वरङ्गे तु पादकः। चित्तस्याप्यादशा द्रष्टुं त्वत्ते कैशिकमध्यमं ॥ षड्जानां विनियोगोऽयं ब्रह्मणा समुदाहृतः। इदानीं भित्त्रानां लक्षणमाह........ —Br̥. Vṛtti on 322. We can clearly see that this passage names the very seven grāma-rāgas enumerated in the foregoing footnote.

2 Kallinātha quotes the same passages as quoted in the footnote above. However, he adds three extra lines at the beginning of the passage which are not to be found in the extant Br̥ : पूर्वरङ्गे तु शुद्धा स्यादिमन्ना प्रस्तुतावस्थया। वेशरा मुख्यो; कार्यो गम्भीर गौडी विधीयते ॥ —Kalā on S.R. 2, 30-32.

3 Note Abhinava's dictum: पूर्वरङ्गे गान्धर्ववेद प्रधामं —A.B. on N.S. 33,1. This implies a dominance of gāndharva but not the absence of other forms.

4 ततश्चोद्यापनं कार्यं परिवर्त्तनमेव च। नान्दी श्लोकावकुष्टा च रङ्ङद्वार तथैव च ॥ —N.S. 5, 14.

tory function was to please lesser deities like the pitṛs and the nāgas.1 The avakṛṣṭā when employed in the pūrvaraṅga may have been moulded on austere lines, more in accordance with the gītaka form. One of the grāma-rāgas, perhaps the ṣāḍaba, may have formed its melodic content.

One gathers from Abhinava that other dhruvā-forms were also apparently employed as part of the pūrvaraṅga, and these were perhaps composed to grāma-rāgas when so employed. But, though the grāma-rāgas could have an additional function in the pūrvaraṅga, their main function was to accompany the unfoldment of the dramatic plot proper, as the verses we have discussed from Kaśyapa and Bharata indicate.

The grāma-rāgas—in contrast to the jātis which were essentially fixed and closed forms—must have been open forms, capable of assimilating new elements and of change. Through a process of change, grāma-rāgas must have given rise to bhāṣās and these in turn to vibhāsās etc. Over a period as yet newer forms were evolved, the grāma-rāgas themselves became old or ‘classical’ and also, apparently static. They became in other words, ‘ribaddha’ forms and, evidently, for this reason, came to be included in the category of mārga. In our own period, we may take the example of dhrupad and compare it with the later forms, the khyāl and the ṭhumrī. The former is now a more or less closed form and does not permit of new innovations or creations, while the other two are still open forms and capable of assimilating new elements both from folk music and parallel musical art-forms such as those of Karnatic music. Yet khyāl and ṭhumrī, too, have the seeds of rigidity in them and may some day become closed forms like the dhrupad.

A statement by Śārṅgadeva suggests that grāma-rāgas did have the kind of history as suggested above and were not always ‘mārga’ or rule-bound forms, which they definitely had become by the period of Śārṅgadeva. While Śārṅgadeva unhesitatingly classifies grāma-rāgas etc. as mārga or nibaddha forms, he yet refers to theorists who thought that these forms were deśī: “prasiddhā grāmarāgādyāḥ keciddeśītyapi-rāḥ: some call the well established grāma-rāgas etc., too, as deśī.” (S.R. 2, 2, 3). Śārṅgadeva is here certainly referring to the works of those ancient authorities in whose times grāma-rāga etc. must have been deśī forms—still fluid to some degree and capable of ‘kāmacārapravartitva’, or ‘being moulded according to urge’, the basic principle underlying deśī forms. Mataṅga, too, includes them in the broad category of deśī, though he certainly must have considered them relatively more ‘nibaddha’ well-regulated (or in his use of the term) a ‘mārga’ variety of deśī; but still it is in his classification a ‘rāga’ form and undoubtedly deśī.

1 युक्तायामवकृष्टाया प्रीतानां भवति हि । तथा शुष्कावकृष्टायां प्रीतः पितृगणो भवेत् ॥

—N.S. 5, 50.

Śuṣkāvakṛṣṭā was avakṛṣṭā sung to meaningless syllables (śuṣka) : अनृतमुष्काकरेरेव ह्यवकृष्टा ध्रुवा यतः । तस्माच्छुष्कावकृष्टेति ॥

—N.S. 5, 25.

Note A.B:

अवकृष्टा अस्मिन्‌तल्लक्षणोपेता केवलं शुष्करकरैरेवैकः । षण्णुमार्चिः ।

Śārṅgadeva himself includes only much younger forms within the deśī class, namely, rāgaṅga, bhāṣāṅga, kriyāṅga and upāṅga.1 These younger forms were ultimately linked with forms such as grāma-rāgas and bhāṣās, which have been classed as mārga by Śārṅgadeva, but the link was a remote one. A passage quoted by Kallinātha from Mataṅga (it is not in the extant Brhaddesī, which, as we have argued, is full of lacunae) says that rāgaṅgas were mere shadows of grāma-rāgas and that bhāṣāṅgas were constructed with elements that were shadows of the bhāṣā proper.2

Interestingly, deśī forms, too, are divided into two categories by Śārṅgadeva : (1) pūrvaprasiddha or those which were already well established as old and traditional rāgas when Śārṅgadeva wrote of them and (2) adhunāprasiddha rāgas or those rāgas which were comparatively recent formations and were in a state of flux. These adhunāprasiddha rāgas were still undergoing structural changes, some of them of great import, when Kallinātha wrote his commentary on the Saṅgītaratnākara (15th century A.D.). He notes that great incompatibilities existed between the adhunāprasiddha rāgas as described and as actually rendered and enumerates some important changes.3 Intriguingly enough, change had also affected some grāma-rāgas which were still in use, revealing the fact that they were after all deśī forms and were thus not immune to transformations. Thus hiṇḍola, a grāma-rāga of the vesara variety, was originally sung as pentatonic by omitting ri and dha but during Kalli-nātha's age, the actual usage was to drop ri and pa.4

Mārga, then, emerges to have been initially a term for the more regulated, relatively older, 'classical' forms within the deśī fold itself. These were forms the structure of which had become fixed due to traditional usage over a long period.

Mārga, in fact, in medieval musical parlance was often applied to anything old. Śārṅgadeva, we observe, classifies even a drum called paṭaha into the two types of mārga and deśī (S.R. 6, 805); mārga paṭaha has, evidently, nothing to do with gāndharva—it was the old 'classical' form of the instrument while deśī paṭaha was a latterly evolved form. Śāradātanaya (placed towards the end of the 12th

1 अथ रागाङ्गभाषाङ्गक्रियोपाङ्गानिर्णयम् । केयांचिन्मतमार्गित्वं कुर्वते सोड़लातमजः; ।। रञ्जनाद्रागता भापारागादेशरदीप्यते । देशीरागतया प्रोक्ते रागाङ्गादिवचच्यु्ट्यम् ।। —S.R. 3, 2, 1-2.

2 रागाङ्गादिषु भेदानां निषादादिमतङ्गकृत् । यथा — ग्रामोक्तानां तु रागाणां छायामात्रं भवेदिति । गीतज्ञः कथिता: सर्वे रागाङ्गास्तेन हेतुना ।। भाषाङ्गास्तेन कथ्यन्ते गायकैः स्तोतिकाविदिमः ।। —Kalā on S.R. 2, 2, 1-2.

3 इदानोमधुनाप्रसिद्धरागाङ्गादीनां लब्धये प्रतीतानां लक्षणविरोधानां परिहारार्थमुदयमः कियते.... —ibid. 2, 2, 159-160.

4 ग्रामरागेषु हि‌ण्डोलस्य लक्षणं रिध्यात्सकत्वेनोक्त्वा रिष्यत्यागेन प्रयोगः । —Kalā, ibid.

Even to this day the Hindustani hiṇḍola is sung as a pentatonic rāga, minus ri and pa; but what resemblence it has to the ancient form of this name cannot at all be ascertained. It is not unlikely that Kallinātha, too, is here confusing a rāga current as hiṇḍola in his days with the ancient grāma-rāga of the same name, while, in fact, the latter was entirely lost.

178 A Study of Dattilam

century A.D.) in a similar spirit, classifies some melodic alaṅkāras and movements as mārga-gamakas to indicates their oldnes. These mārga-gamakas, 22 in number, bear names such as bhrāmita, dīrghalalita, urastāra, siroguru etc. and are certainly post-gāndharva (they are not named under gāndharva anywhere) but they were already ‘classical’ and old melodic graces by the 13th Century, hence the appellation mārga was used to distinguish them.1

Mārga was, thus, quite distinct from gāndharva which was a class apart. But gāndharva-born forms themselves formed a very large and ever growing group. Within this group, some forms were very old and well-established; these were the mārga forms.

Abhinava clearly denotes through mārga a sub-class of what he terms gāna forms and keeps the term quite distinct from gāndharva. He, too, evidently had in mind a similar notion of mārga as that suggested by Mataṅga, though he makes no explicit statement on the point.

Śārṅgadeva, unlike Abhinava, has not been careful to distinguish mārga from gāndharva; on the contrary he confuses the range of forms denoted by each and makes them so overlap each other as to make the terms apparently synonymous. Maybe by his time gāndharva was on the way of becoming more a memory than an actually living form and for him it was just one of the old ‘mārga’ forms. He therefore groups together all old well-regulated forms—including the gāndharva—as mārga.

Yet he does not totally confound mārga with gāndharya. He does keep the gāndharva forms quite apart from other mārga forms by giving them a separate niche in his work. The adṛṣṭa-motive he so pronouncedly ascribes to the jātiṣ and the gītakas is not ascribed to grāma-rāgas etc. And though he has used gāndharva and gāna as almost synonymous with mārga and desī as connotative of the same distinct groups of musical forms, still there are instances of his use of these terms which reveal that a perspicuous impression, a gradually weakening yet definite residual imprint of the earlier distinction between mārga and gāndharva, had remained in his mind. Let us study instances of note :

In his discussion of vāgeyakāras (literally, ‘composers of words and music),2 the medieval name for expert musician-composers, Śārṅgadeva remarks that these

1 निस्वानितं व सुकृतिं विततं विद्युत्तथा । भामिनीदृश्वलितमुरस्स्तानं शिरोगुरु ॥ उल्लोलितालापकै व लीलोल्सारितकुष्टिते । प्रतिभा.तुमुरो दृश्यं कण्ठेक्षिप्तकमेव च ॥ नादप्रबन्धान् प्रविष्टं प्रवृद्धवपुरनुवर्तते । समासक्त कोमलं व मुग्धोल्लसितं विशेषकम् । उदयाद्रिप्रसादं चेतानि मार्गकमितिफ़धीरोरितः ॥

–Bhāvaprakāśana, ch. 7, lines 10–16 on p. 192 of the G.O.S. ed.

The scholarly editors, Yatiraja Svami and Ramasvami Sastri, of the G.O.S. edition of the Bhāvaprakāśana have adduced many erguments to conclude that the work was composed in “a period ranging between 1175–1250 A D.” see Introduction p. 76.

Śārṅgadeva defines vāggeyakāra as an expert who could both compose a new poem as well as set it to music.

वाड.मातुरुच्यते यो धातुरभिधीयते । वाचं गेयं च कुर्वते यः स वाग्मोयकारकः ॥ –S.R. 3, 2.

composers were free to handle and mould according to their desire—of course within the given rules–both deśī and mārga forms.

The vāggeyakāras are described in the third chapter. This chapter (on prakīrṇaka) deals not only with vāggeyakāras but also with such related matter as various types of vocalists, their qualifications and drawbacks, the musical vocal-tone in singing, its nature and qualities and its various flourishes, the various types of musical movements and their relation to the rāgas and also compositional aspects, such as the rules of ālāpa in rendering a rāga. The entire exposition contained in the chapter, according to Śārṅgadeva, applied to both deśī and mārga forms :

atha prakīrṇakaṃ karṇarasāyanamanākulam deśīmārgāśryaṃ vaktī śārṅgadevo vidāṃ varaḥ

Kallinātha's comments are elucidative. He remarks : “the actual forms of grāma-rāgas etc. described in the second chapter can only be experienced directly through the efforts of singers and composers (nirmātr̥); the chapter on prakīrṇaka has been written in order to lay down the characteristics of these latter two (i.e., singers and composers) and aspects related to their skill (tadādilakṣaṇa-param).1 Expert vāggeyakāras were thus free to make new creations not only in deśī rāgas but also in mārga,2 which included in this context grāma-rāgas as well as subsequent ‘classical’ or old forms. The best vāggeyakāra, according to Śārṅgadeva, was he who could compose with equal ease and skill in both mārga and deśī forms. Vāggeyakāras of this type were known as gāndharva (not to be confused with the ancient term). Those vāggeyakāras who had proficiency only in mārga forms were known as svarādi :

mārgaṃ deśīṃ ca yo vetti sa gāndharvo’bhidhīyate yo vetti kevalaṃ mārgaṃ svarādī sa nigadyate

One of the qualities of a vāggeyakāra of either category was that he could compose a song at the spur of a moment (drutagītavinirmāṇam, S.R. 3, 8). The best vāggeyakāras were capable of new compositions in which both the poem and the music to which it was set were new and fresh creations; these creations were, moreover not mere copies of older models: “anucchiṣṭoktinirbandho nūtanādhātuvirmitiḥ” (S.R. 3, 7).

1 दितीयेऽधबहितलक्षणानां ग्रामरागादीनां स्वरुपप्रकाशकारस्य गाननिर्मान्तरुपरतन्त्वेन तत्स्वरुपज्ञासायं तदादिलक्षण-परं प्रकरणकं वर्णयिते— कृत् प्रकरणकमिद्यादिना । —Kalā on S.R. 3, 1.

2 Kallinātha in the same passage commenting on the second line of S.R. 3,1 adds: ‘देशीमागोऽध्ययनमिति’ विशेषणमत् तल्लक्षणविशितानां वाग्गेयकारादीनां देशीमागोऽध्यासाधारणत्वात् तल्लक्षणपरस्य ग्रन्थस्यापि देशीमागोऽध्यासयत्वेन प्रकरणकत्वोक्तिनार्थः।

180 A Study of Dattilam

It is clear from the above that the mārga forms which the vāggeyakāras were free to handle could not have included gāndharva. Kallinātha, indeed, points at grāma-rāgas as their upper limit.1 This is quite within reason, since Śārṅgadeva has himself pointed out that jāti forms were as good as 'revealed' forms similar in this respect to the Vedas; they were only to be sung to the words composed for them by Brahmā himself and nothing in their decreed form was permitted to be altered (S.R. 1,7, 113-15). New creations in such forms were quite out of the question. Thus by mārga, in the context of the vāggeyakāras, Śārṅgadeva evidently means non-gāndharva forms-though, indeed, forms of an older, more 'classical' variety, such as grāma rāgas, bhāṣās etc.

Śārṅgadeva has also used Abhinava's term gāna and we have seen that he equates the term with deśī. Yet again the equation is not quite total. He has stated that gāna forms are those which are composed by vāggeyakāras in 'deśī-rāgas etc.' (S.R. 4,3) implying, evidently-through the 'etc'—the inclusion within this fold, of mārga-forms like grāma-rāgas which were according to him within the sphere of the vāggeyakāras' activity.

In spite of a covert distinction, Śārṅgadeva does confuse between gāndharva and mārga. This, we suggest, was perhaps due to the fact that in the iconoclastic period intervening between Śārṅgadeva and Abhinava, gāndharva became extinct as a living form. It was a form which demanded special protection and care and needed to be sustained and maintained intact with great diligence and perseverance. Even during Abhinava's period and for quite some time earlier, it must have remained in the hands of very special groups or sampradāyas which must have protected and propagated gāndharva as a sacred art. These groups must have been one of the first casualties of the great historical upheaval after Abhinava. Hence Śārṅgadeva probably knew of gāndharva only indirectly.

But he certainly seems to have had a first-hand knowledge of the grāma-rāgas. These were old but profane forms. They survived perhaps because of their profane character and a relatively freer form. Śārṅgadeva knew through available texts—and certainly also through musical tradition which was still close enough to living gāndharva to retain a memory of many of its distinctive attributes—that the grāma-rāgas were close relatives of the gāndharva forms. He could, consequently, imaginatively reconstruct gāndharva and record its structures through his immense textual and theoretical knowledge and the known resemblances of jātis to grāma-rāgas and other older jāti-born forms.

After Śārṅgadeva the next great comprehensive work on music (apart from the commentaries on the Saṅgītaratnākara), is that of Rāgā Kumbha who composed his Saṅgītarāja during his reign over the kingdom of Mewar in the 15th century A.D. Kumbha was a great scholar of music and had studied many ancient and modern texts on the subject. He was also a versatile musician. He claims not only to have

1 He clearly says that only 'the second chapter onward and forms beginning with grāma-rāgas etc.' described therein constituted the field of the vāggeyakāra :

द्वितीयोऽध्यायादि तत्र ग्राह्यमारागादीनाम् ......

• Kata on S.R. 3, 1.

Gāndharva and Post-Abhinava Theorists

181

studied and mastered ancient texts but also to have actually realised in practice the

forms expounded.1

It is not possible to fully assess how close his reconstructions were to ancient

forms. Yet, however much we may applaud his attempt, it is evident that much in

Kumbha's restoration was of dubious authenticity. We can, indeed, infer from his des-

criptions that contemporary forms had crept into the recaptured old ones. Take for ins

tance the gītakas of gāndharva: Kumbha has drawn up charts of the actual beat-cum-

tāla structures of gītaka-formations on the basis of descriptions in the ancient texts.

Śārṅgadeva, too, has given similar charts which, indeed, must have inspired Kumbha.

But Kumbha goes much beyond Śārṅgadeva; he gives not only the tāla-charts of the

gītakas but also adds an accompanying svara and pada structure to which the gītakas

are supposedly to be sung. The pada-structures of some of the gītakas consist of long

verses and are clearly Kumbha's own creations as they bear his name.2 The accom-

panying svara-structures are not jāti structures but compositions in rāgas, some of

them very late deśī forms like malhāra-rāga (S. Raj 2, 4, 1, 135), karnāṭa-baṅgāla-

rāga (S. Raj 2, 4, 1, 219) and turuṣka-gauḍa-rāga (S. Raj 2, 4, 1, 219). These rāgas

are listed as adhunāprasiddhā or currently popular deśī-rāgas by Śārṅgadeva (S.R.

2, 2, 9-10); they were evolved very much after gāndharva. Other rāgas like dhāraṇī

which Kumbha prescribes as melodic accompaniment for gītakas (S. Raj. 2, 4, 1, 219)

appear to be even later forms, since they do not appear on Śarṅgadeva's lists. Ob-

viously, gāndharva as a sacred, unalterable form was, by the age of Kumbha, a dead

form, living only in musical texts, and Kumbha in trying to revivify it was, to a very

great extent, making it sing to his own tune.

Kumbha, like Śārṅgadeva, uses the two pairs of terms, gāndharva: gāna and

mārga: deśī, much in the same way as his predecessor. Gāndharva forms he says are

śuddha or pure forms; they extend in his work (he says) from the section on śruti upto

antarabhāṣā and comprise also the gītakas like madraka etc. In other words, gāndh-

arva in his view, as in that of Śārṅgadeva, included not only the jātiṣ and gītakas but

1 प्रण्यतां सम्प्रघनीयं बुद्धिविषयं नीत्यानुरूपायार्थतः । कृत्स्ना दर्शितमोवरानभिनयैर्यान्नाभिनेयाश्रिते ॥

तान् विष्णुप्रतिष्ठाप्रतिष्ठापनविधिं नेदुं यथास्थितान् श्रीमद्॥ काल्जिमूतपतिः प्रयतते स्वोपकसन्दर्भम्॥

– S. Raj. 1, 1, 1, 37.

Although Kumbha here explicitly claims to have actually reconstructed only dramatic performances, the claim certainly also applied to music which formed the main subject-matter

of his treatise. And, indeed, a total reconstruction of an ancient drama must inevitably

have entailed a reconstruction of its musical aspect, too.

2 eg. the verse to be sung to madraka includes the line:

प्रणतातिहरं मायाह्रूपं, प्रणमति शिरसा, नवरतसिंहु तं कृष्णनरेशः;

another verse contains:

प्रणमामि सदां कृष्णनूपतिजयदं हरमजेयमहम;

yet another reads:

जयपद्मजपूरन्दरश्रीकालहसिनजयदनौच्यकराभयकर शिव नमस्ते

and so on.

–see S. Raj 2, 4, 1.

182 A Study of Dattilam

also grāma-rāgas, bhāṣās and their progeny upto the antarabhāṣās.1 Gāna he defines as comprising those forms which were handled by and moulded into a variety of shapes by the vāggeyakāras.2

Gāndharva, Kumbha says, was a sacred form with an adṛṣṭa end proclaimed by Brahmā himself. It was sung by expert musicians who did not deviate in the slightest from its set traditions and was for this reason called gāndharva ("gandharvairiha sampradāyaviḥitāṃ hitvā yato gīyate tad gāndharvamidam vidah praṇigadanti"; (S. Raj 2, 4, 1, 4). Regarding jātiś, Kumbha, again like Śārṅgadeva, says that jātiś were immutable like the Vedas and if sung in praise of Śiva in accordance to their decreed forms they could expiate the a.wful sin of brahmahatya. He quotes Purāṇas to conclude that a musician who sings or plays the jātiś, as well as those who listen to them, attain the transcendental abode of Śiva. He adds that jātiś should not be mutilated to the slightest degree for dire consequences might follow. Rāgas, he says, are born of the jātiś and can be identified as such (S. Raj 2, 4, 1, 408-414).

The ‘śuddha’ forms beginning with śruti and extending upto antarabhāṣās which Kumbha defines as gāndharva are again in the same context also identified as mārga.3 The definition of mārga in Kumbha is, in spirit, the same as in Śārṅgadeva: mārga was a body of forms created by Brāhmā for spiritually and transcendentally uplifting ends (mahodayānimittā). He ascribes the same characteristics to gāndharva. Deśī, too, is defined on the same lines as in Saṅgītaratnākara which views it as co-extensive with gāna.4

In Kumbha's usage of the terms gāndharva; gāna and mārga; deśī, the same ambiguity and equivocality are to be found as we have witnessed in Śārṅgadeva. Gāndharva is described as consisting of forms ranging from jātiś to the antarabhāṣās; mārga is implied as being coextensive with these and yet jātiś have been apportioned a section quite to themselves. Further, we notice that not jātiś but only the jāti-born forms ranging from grāma-rāgas onwards are divided into the mārga and deśī classes (S Raj 2, 2, 1, 23-24). A vāggeyakāra, as in Śārṅgadeva, is said to handle both mārga and deśī forms (S. Raj 2,3,1, 9-10). in which he is expert at creating ever-new musical compositions (navanavaracanādhātuninmānāvedī; (S. Raj 2, 3, 1, 3). Mārga thus (in Kumbha's scheme, as in that of Śārṅgadeva), excludes the jātiś in which, according to

1 नानाघवं च तदेव गानमिति तत्त्र मृगादिमिः । श्रुयणान्तरनिर्यिकातमुदितस्तैर्मृगकाच्चैः तम् ॥ गीतस्वरादेशाभिप्रयुक्तनृत्यप्रबन्धरत्नसूत्रो यो गीयतेः । सर्व चातिद्वैधाद्यमुत्कर्ष बहादिभिः कौतितमय ॥ —S. Raj, 2 4. 1, 3.

Sārṅgadeva has nowhere specifically and in so many words, recorded that gāndharva is ‘antar-abhāṣikānugam’: He has only implied it. Kallinātha, his commentator, however, makes the point explicit (Kalā on S.R. 4, 1-4). Kumbha has evidently followed Kallinātha.

2 अन्यत् पुनरीयते । यद् वाग्मोयकृतांवरेण विलसद्ध विलासविन्यासवत् तत्त्सकषणलक्षित पुनरसौ है ध्रुवा निवद्ध तथा अन्यः स्वादानिबद्धमिति । —S. Raj 2, 4, 1, 4-5.

3 मार्गश्रितस्त्व श्रुयादिलक्षित शुद्धरसङितम् । तत् प्रसरगात् प्रबन्धयत्न गीतकानि सविस्तरम् ॥ —ibid. 2, 4, 1, 6.

4 गीतं वाथ तथा नृत्यं यत् संगीतमुच्यते । तद्द्वेधा भवते मार्गदेशी मेदेन तत्सतः ॥ मार्गितत्वादिरिच्चेन प्रयत्नादर्शविचक्षणैः । महोदयनिमित्तवाढू विधातो मार्ग उच्यते ॥ तदेव रस्चेदेशीयात् चित्ररागजनक्जनैः । प्रयुक्तं स्वस्वदेशी यत्ततोदेष्टि कीर्तितमय ॥ —S. Raj 1, 2, 3, 2-4. cf. S.R. 1, 1, 21.24.

to Kumbha, not a single mātrā could be changed without fear of divine retribution just as in words constituting sāma-mantras: “tataḥ sāmākṣarāṇīva mātrāmātramapi sphuṭam/etaṣām nānyathā kāryaṃ pratyavāyabhiṣā kvacit” (S. Raj 2, 1, 4, 414).

After Kumbha, the terms gāndharva and gāna fell almost out of use. Mārga and deśī survived. The definition of these two terms as given by Śārṅgadeva was religiously copied and paraphrased by one author after another.1 Mārga became the general term for all old forms described by Śārṅgadeva and his predecessors. These forms had apparently become extinct by the 16th-17th centuries: Veṅkaṭamakhin, writing in the 1620s says, in his Caturdandīprakāśikā, that grāma-rāgas, uparāgas, rāgas, bhāṣās, vibhāṣikās and antara-bhāṣās--i.e., the six forms described as mārga by Śārṅgadeva--are sung only by the gāndharva deities and not by men.2 Indeed, not only mārga-rāgas but also the 264 deśī rāgas enumerated by Śārṅgadeva (S.R. 2, 2, 19) were no longer extant by the time of Veṅkaṭamakhin.3 Already in the 14th century we notice that many of the forms--both old and new--current during Śārṅgadeva's age were on their way to extinction. Prabandhas--one of the major forms of compositions to which rāgas, both mārga and deśī, were set--seem to have travelled a long path towards oblivion by the period of Sudhākalaśa, the Jain author of Saṅgītopanisat-sāroddhāra (1350 A.D.).4 He remarks that composers and singers of prabandhas were becoming rare in his days and thus excuses himself from describing these in detail:

prabandhabandhakartāro viralā bhūtaledhunā tadgāyanāśca na prāyo'to noktāste savistarāḥ

(Saṅgītopanisatsāroddhāra 1,37)

By the 18th century the range of forms described by Śārṅgadeva were definitely extinct. King Tulajā who ruled from 1729 to 1735, observes (like Veṅkaṭamakhin) that mārga rāgas were performed only in the abode of gandharvas5 and that many of the other rāgas too, current in an earlier period, had gone out of use.6

1 See for example the Ra. Kau of Śrīkaṇṭha (16th century A.D.), 1, 14–16; Rāgavibodha of Soma-nātha (early 17th century) 1, 6–7; Saṅgītadarpaṇa of Dāmodara Paṇḍita (circa 1625 A.D.) 1, 3–5; Saṅgītapārijāta of Ahobala (latter part of the 17th century) ch. 1; etc.

2 अवेतल्लक्षणगानप्रबन्धादिकं काव्यरुचिभि: । रागरागाङ्गचोपाङ्गैरागा भागैर्विभाविकाः । तयैवांतरभाषाद्यै रागाद्रुच्यैर्विभासित: परं ।। भावा: तद्भण्णि च ऋम्यादृण्णि । हा पाक्षादन्विति च ऋमस्मु । दक्षस्मेवेते रागेषु ग्रामरागेषु पुनः ।। रागास्त्वरतरमाषान्त मर्गरागा बहुधा पृथक् । ततो गन्धर्वलोकेऽपि प्रयोक्तुं व्यवस्थिताः ।। - Caturdandīprakāśikā 5, 17-20.

3 तस्माद्रागा भारक्रियाकृतोऽपार्क संज्ञिताः । रागाश्चत्वार एवैते देसीरागा: प्रकल्पिताः ।। तत्त् रत्नाकरग्रन्थे शार्ङ्गदेवेन धीमता । चतुःषष्ट्यधिकं रागषट्कतयमुदीरितम् ।। लघ्यन्ते न कुतापि लघ्यवर्मन् संप्रति । - ibid. 5, 21-23.

4 See Introduction pp. 5-6 in the G.O.S. edition of the work for date of the author.

5 लेभेनेकविधैर्देशी ग्रामरागादय: पुनः । रागा ग्रन्थरभाषान्त मर्गरागा: प्रकल्पिताः । वत्तो ग्रन्धर्वलोकेऽपि प्रयोक्तुं व्यवस्थिताः ।। - Saṅgīta Sārāmṛta 9, 8-9.

6 तत्त् प्रसिद्धिविपर्यस्तस्यक्वा रागान्स्तु काश्चन । सर्वन्त लघ्यमर्गेण्ड संप्रति प्रचरन्ति ये । त एव संप्रहीत्यादि राग इति विनिश्र्चय: । - ibid. 9, 14-15.

Thus, mārga in the 16th century, and later, became a general term for all old forms—forms by now totally lost in oblivion and described in texts such as the Sangī-taratnākara. Deśī was a more resilient term, for by definition it could apply to any form that happened to be current and thus the term was retained as applicable to all extant forms.

One or two authors did use the terms gāndharva and gāna. They used them as synonymous with mārga and deśī. The definition of the terms as found in these later authors is simply a close paraphrase of Śārṅgadeva's definition (in S.R. 4, 1-5).1

Experts during Śārṅgadeva's period seem to have had an inkling of the actual outlines of gāndharva forms through their direct knowledge of grāma-rāgas and other ancient forms which were close relatives of gāndharva. Moreover, musical forms in that period seem to have shared a doctrinal as well as schematic organisation with the ancients; music still, evidently, referred to a largely common system of interpreting and viewing musical forms, even if many of the old forms themselves had been lost. Experts like Śārṅgadeva could still imaginatively visualise the jātiṣ and discover likenesses with existing rāgas in their structure.

But gradually the musical system itself underwent some very crucial changes. Kallinātha writing in the 15th century gives some revealing clues regarding these changes.

In the ancient system the dual grāma-division of the octave was a very basic structural factor in understanding and rendering musical forms whether they were jātiṣ or grāma-rāgas etc. By Kallinātha's age the grāma-system had distintegrated and all forms were referred to a single common octave-structure. Thus Kallinātha observes that many rāgas which, according to the śāstra, should have been rendered in the madhyamagráma, with the note madhyama of the middle octave as the initial note, were actually sung with the ṣadja of the middle octave as the first note. This points towards a great transformation.

Each rāga in the ancient system was allotted a mūrcchanā within a grāma and this presumably gave the rāga its particular scale or a fixed arrangement of tones (which we now call a thāṭ or mela). Any of the seven notes of an ancient grama could become the tonic, the ādhāra-śruti. A given mūrcchanā could only become a

1 Śrīkanṭha describes gāndharva and gāna as:

गीतं स्यात् स्वरसनोद्भवः श्रोतॄणां चित्तरञ्जकः । सूरिभिरिसतदर्थदीप्तो गान्धर्वगानभेदतः; ।। गान्धर्वेऽपि गीयते वर्गस्तद्गुणैरमेवभूयात् । यतः वागर्थौकारणं कृत्वा तद्गानमुख्यते ।। पृथिव्यामप्ययोग्यस्य गाधावद्गानमेव निगद्यते । निवद्धे चानिबद्धे च तद्गानं द्विविधं भवेत् ।।

—Ra. Kau. 3, 6-8.

also see Govinda's remarks.

संगीतं द्विविधं प्रोक्तं गान्धर्व गानमेव च । देवगन्धर्वलोकेषु प्रसिद्धं समप्रायतः ।। गान्धर्वं सकलाभोगदायकं देवपुरुजितम् । गान्धर्वकलसंज्ञीयं क्त प्रेतिदं शुभदायकम् ।। गानं तु भरताचार्यैर्देवादिलक्षितम् । मनुष्याणां श्रुतिमनोरञ्जनं चेष्टसिद्धिदम् ।। लक्ष्यलक्षणसंयुक्त नित्यं श्राव्यं मनीषिभिः ।

—Sangraha Cūḍamani 2, 11-14.

distinct scale if the note on which it began was the tonic. But in Kallinātha's age, ṣadja seems to have become the common tonic for all forms.1 The grāma-mūrchanā system was giving place to the modern mela-kartā or thāṭ system with the tonic always being called ṣadja (or sa).

We notice in the ancient system that ṣadja was no more important than any other note. In many jātis, in fact, it could be dispensed with, and the same was true of it (with, indeed, a greater degree of freedom) regarding grāma-rāgas and other ancient forms.2

Pañcama in the two grāmas was fixed at different pitches. But in Kallinātha's period all notes were referred to a single ādhāra-śruti which was invariably the ṣadja, hence pañcama, which stands in a natural haromonical relation with this fixed ṣadja, too, had to become constant. The dual pañcamas of the ancient grāma-system had to go, and pañcama became, as in modern music, fixed on a single immovable pitch.3

Later authors clearly testify that only a single-octave system, which they termed the ṣadja-grāma (because of ṣadja having become the invariable tonic), had survived in practice. To have called this octave-fabric a 'grāma' was logically a misnomer, for it had a single fixed tonic, other notes having variable positions to give new scales; but old terms die hard.

Veṅkaṭamakhin in his Caturdaṇḍiprakāśikā remarks that madhyama grāma had become 'non-existent' : "asmābhirmadhyamagrāmo'pi asatprāya itiiryate" (Caturdaṇḍiprakāśikā 3, 69). No reasonable man, he adds, would accept this grāma, because it was not to be found in practice: "ayukto madhyamagrāmo lakṣyamārga-virodhatāḥ" (ibid. 3, 71). There was, consequently, only one grāma—the ṣadja-grāma: "eka eva tataḥ ṣadjagrāma ityavadhyāryate" (ibid. 3, 72). Veṅkaṭamakhin was one of the main theorists of the melakartā system, a system which was a living testimony to the fact that the grāma-mūrchanā system had been now replaced.

Somanātha in his Rāgavibodha (1609 A.D.) also implies the existence of a single grāma. The reason he gives (like Kallinātha) is that pañcama was no longer established on two different śrutis as in the ancient a moveable note : it was no longer established on two different śrutis as in the ancient

ग्रामयोःरागजात्यादिविप्रकर्षयोल्पन्नानामित्थं रागाणां मध्यमस्यपकर्षजन्मत्वात्तन्मूर्छनापकर्षस्थायिणः शास्त्रविधिते संभवस्यपि मध्यमग्रामोत्पन्नानां च मध्यममदितोदिप्रभृतिनाम् च मध्यमध्यमारम्भं विद्वान् मध्यवृन्दजत्वं एवारम्भो लक्ष्यते ।

—Kalā on S.R. 2,2, 159-160.

Madhyama in gāndharva was said to be indispensable. But this could not have made it the constant tonic like the modern ṣadja. For, with madhyama as the only tonic gāndharva would have had only a single scale, namely, the scale of the mūrchanā beginning with madhyama. This seems to be an extremely unlikely proposition. Moreover, madhyama in gāndharva could be rendered an extremely weak note in some jātis—a position quite unthinkable of the constant modern tonic ṣadja which is stressed with a continuous drone. Utpaladeva has suggested that madhyama was called indispensable in jātis because it just happened to be the only note which was never actually dropped. There seems to be quite some truth in this observation, because in many jāti-born forms which shared a common musical scheme with the jātis, madhyama could be dropped.

विच्चतुःश्रुतिकल्वेन ग्रामद्यमपेदकस्य पञ्चमस्यालोपपर्यवेन प्रयुज्यमानस्यापि सर्वरागोपवेक्ष्यता ।

—Kalā on S.R. 2,2, 159-160.

collated by kallinātha on sl

186

A Study of Dattilam

two-grāma system but had become fixed and immoveable.1 Somanātha, in fact, quotes

Kallinātha in support of his views.2

Other important post-Kallinātha theorists such as Śrīnivāsa (in his Rāgatattva-

vibodha3), Puṇdarīkaviṭṭhala (in his Rāgamāla4) and Ahobala (in his Sangītapārijāta) all testify to the existence of only a single grāma.5

Pandit Bhatkhande had as early as 1916, come to the conclusion that by the

time of the Rāgatarangiṇī of Locana-Kavi (written some time after Vidyāpati, late 14th

century A.D., whom Locana quotes) sadjagrāma was the only grāma in existence and

the system deeming mūrchanās and jātis as parents of rāgas had become obsolete; the

melakartā or thāt system had come into full usage.6 A careful perusal of Kallinātha

and the major post-Kallinātha authors cannot but lead to these conclusions.

The chief reason for this major upheaval in the system of music was per-

haps the emergence and the gradual dominance of the drone. Today, Indian music

without the drone is unimaginable. And in current practice, the tonic,—the ādhāraśruti

of the drone (usually the tānapūra except with instruments like the shahnāī) is always

called the ṣadja; in fact, the terms ṣadja and ādhāraśruti or tonic have now become

synonyms. The emergence of ṣadja as the basic note, with which as the constantly

droned tonic, all other notes are realised, was the new emergent system to which latter-

day theorists pointed in stating that ‘ṣadja is the only existent grāma’.

The pre-drone system had, in all probability, the harp as its central instru-

ment, as many eminent scholars have concluded.7 In this system a harp-like viṇā

was tuned to a certain grāma : the seven (or nine) svaras were placed at fixed śruti-

distances from each other.8 Any note could then become the tonic and the mūrchanā

based on that note gave a distict scale.

1 स्वान्त्यश्रुताबान्तरयुतो च सति पञ्चमे ऋमासु स्यात्। किं तु विकारो देश्या न पञ्चमे तद्विह स्वीकृतः॥

-Rāgavibodha 1, 40.

2 किं च रागविवेकवाक्यैःव्याख्याने कलिनाथोऽपि लक्ष्यलक्षणविरोधपरिहारप्रसङ्गादुक्तवान्। विच्चतुःश्रुतिकत्वेन ग्रामद्वय-

वेदकस्य पञ्चचत्वारिंशोध्यायेऽवलेपेन मृजादिमार्गस्त्वयापि सर्वरागग्राहकत्वेन इत्यादिनि।

-Authors own comm. following Rāgavibodha. 1, 40.

3 गान्धारमध्यमग्रामौ न तो लक्षणगोचरौ।

-Rāgatattvavibodha, 39.

तदेव मध्यमग्रामः स(तु) च रागे न दृश्यते। पड्जग्रामाश्रितानुरागानस्स्व गायन्ति गायका:॥

तस्तान्मुख्यतमः पद्जग्राम एव च मध्यमः।

-Rāgamala 1, 27-28.

We have consulted IICMS copy of manuscript No. 1985 of the Oriental Institute of Baroda.

5 Pandit Bhatkhande in his Srīmallaksyasangītam 1, 25-41, written under the pseudonym of

Catura, gives a fairly long list of quotations from post-Kallinātha authors testifying to the

existence of only the so called ṣadjagrāma.

6 See his A short Historical Survey of the Music of Upper India.

7 See for example “Emergence of the Drone in Indian Music” by Dr. B.C. Deva, J.M.A. 1952;

also B.B. pp. 167-71 (part of the editor Chaitanya P. Desai’s epilogue) etc.

For further discussion see ch. IV.

See Bharata kā Sangīta Śāstra (Hindi) by Ācārya Bṛhaspati, pp. 5-13 for a well-reasoned des-

cription of the tuning process.

All ancient musical forms, whether gāndharva, mārga or deśī (of the ancient variety) were based on the grāma-mūrchanā system. True, mārga and deśī incorporated and accepted many more śrutis and intervals as musically valid tones than were permitted in gāndharva, but the system remained common. It must have been a major transformation in musical practice that caused a change from the ancient grāma-mūrchanā system to the current drone system. During the course of this transformation the mārga-rāgas were gradually lost. Or perhaps they were moulded to suit the new framework and were in the process so transformed that latter-day theorists could not see them as ancient mārga-rāgas.

The drone probably originated in a class of deśī (or may be purely folk) forms which during medieval times used a tānpūrā-like instrument to sound the tonic throughout musical performances. This system then gradually became dominant to the exclusion of the older one. The historical upheavals during which the old order speedily disintegrated must have surely helped this drone-system to establish itself. Whatever the historical process responsible for this change, the change from the old to the new system was certainly a marked one as Pandit Bhatkhande1 and many noted subsequent authors have observed.

In the light of Śārṅgadeva, his commentators, and subsequent authors whose works are most accessible to us, it is not surprising that most modern scholars have bracketed gāndharva together with mārga. Even those perceptive scholars who have recognised that gāndharva, indeed,was a definite body of music in itself, yet reveal a vague and sometimes conflicting notion regarding its scope, extent and nature. The reason, we believe, is that Abhinava's exposition and masterly analysis has not received the attention it deserves. Śārṅgadeva has far too excessively overshadowed our vision of ancient forms. We do not wish to decry Śārṅgadeva, who was certainly a very painstaking and thorough scholar with a vast, almost encyclopaedic, knowledge of musical practice and theory. He stood at the threshhold of change and opens for us an illuminating window through which we can look back at ancient forms. But Abhinava is certainly a more reliable guide in many crucial matters of ancient music. Post-Abhinava authorities were no longer in the happy position of having a direct knowledge of ancient musical forms.

Following Abhinava, we have, in our analysis of gāndharva, made an attempt at sifting and distinguishing it from other forms and of establishing its separate identity. We have (to use an ancient philosophical term) discussed the ‘tattastha’ attributes of gāndharva: we have, that is to say, considered it mainly in relation to other forms. Let us now turn to the svarūpa-lakṣaṇa of gāndharva and try to understand its forms

See his monograph entitled A Comparative Study of Some of the Leading Music Systems of the 15th, 16th, 17th and 18th Centuries, besides the work cited earlier.

in themselves. For this purpose we can have no better guide than the Dattilam. It is the one available text which is devoted completely and exclusively to gāndharva. It has, moreover, the advantage of being short and pithy: it deals with the subject in a very precise, methodical way. It follows a logically organised manner of procedure in moving neatly from one topic to the next.

Part III

TEXT WITH EXPOSITION

INTRODUCTORY REMARKS

The Dattilam may be analysed structurally into three distinct sections:

1. the preamble, consisting of five verses;

2. the section on svara comprising 12 topics, (from verse 6 to 108).

3. the section on tāla, beginning with a short introductory remark, from verse 109 to 242. (Then the end of the treatise is announced in verse 243.)

We have also divided the exposition that follows into these three divisions.

The Dattilam is short and very often cryptically brief. It can, on most points of detail, only form the embryonic centre of our study. We shall take recourse, for elaboration, to other texts. The most ancient of these is the Nātyaśāstra. But the Nātyaśāstra, too, needs clarifications. Abhinava provides these on many points.

Other important texts like the Bṛhaddeśī, the Bharata Bhāṣya of Nānyadeva, the Saṅgīta-cintāmaṇi of Vemabhūpāla, the Saṅgītaratnākara and the Saṅgītarāja— to name the more prominent ones—also provide many useful clues and necessary details.

As is only to be expected regarding a body of music as ancient as the gāndharva, the details available in later works are sometimes conflicting, confusing and ambiguous. Our attempt shall be to sift the details as critically as possible and after examining evidence, to present gāndharva in a logically schematic manner with Dattilam providing the ground. This will give us an imaginative idea of the forms of gāndharva (as much as is feasible with a dead and bygone form) and also show us the relative viewpoint of Dattila as a gāndharva theorist in comparison with others.

The Dattilam was printed from a single manuscript. No truly critical edition is possible till more manuscripts may be discovered. Yet we have seen that quotations from the Dattilam abound in saṅgīta literature. These provide us with some important pāṭhabhedas. They also help fill in many gaps in the printed edition. Some lacunae still remain. We have attempted to bridge a few important ones through considerations of parallel matter elsewhere, and also, at a place or two, through deductive considerations.

192

A Study of Dattilam

DATTILAM

THE PREAMBLE

1A. (pranamya parameśānam) brahmādyāmśca gurūmstathā

B. gāndharvaśāstrasaṅkṣepaḥ sāratoyaṃ mayocyate

2A. gāndharvam nārdādibhyach prattamādau svayambhuā

B. vidhivannāradenātha prthivyāmavtāritam

3A. padasthaḥ svarasaṅghātastālenā sumitastathā

B. prayuktasćāvadhānena gāndharvamabhidhīyate1

4A. lokād vidyātpadanīḥa2 śabdaśāstrādyanugrahāt

B. prasiddhamavadhānaṃ tu samyagbuddhyādiyojanam

5A. dvayamanyadato vācyamihi saṅkṣepamicchatā

B. tatra svaragatam pūrvam meyatvādupadiśyate

Having bowed to the Supreme Lord and to the Teachers, beginning with Brahmā, I [shall now] in a concise form enunciate the essence of the śāstra pertaining to gāndharva. The Self-Born One conferred gāndharva first on Nārada and others. Nārada, then, duly brought it down to the world. A group of notes well measured through rhythmic beats (tāla) and set to words (padasthaḥ) when rendered with [due] intentness (avadhāna) is termed gāndharva. One should in this context, know the [nature of] words (padas) through common usage (lokāt) and with the help of grammar and lexicons etc. [The term] intentness (avadhāna) is well understood to mean the right direction (yojanam) of the intellect (buddhi) and other [faculties]. It is necessary for me, with my intention here to be brief, to expound only the two elements other than the [two just explained]. Therein svara is being elucidated first since it is that which is to be measured.

NOTE :

Dattila begins with the customary maṅgala-śloka after which he states the subject-matter of his work.

'Gāndharva', we have shown, stands for a specific body of music. Śāstra technically means the authoritative delineation of a subject in a methodical, schematic way.

Any god considered as the supreme deity can be referred to as 'svayambhū' (self-born). The epithet is thus commonly applied to Brahmā and Viṣṇu as well as Śiva. However, the two gods who are most intimately associated with gāndharva—indeed with music in general—are Śiva and Brahmā. Here the reference seems to be to Brahmā, who is stated in the Nātyaśāstra to be the prime (ādi) teacher of gāndharva (N.S. 33, 23; 28, 39, 31, 104).

1 T. ed. reads. padasvarasaṅghātastālena. Datt. 3 has been often quoted : A.B. quotes 3A with 'तदुक्तं दत्तिलाचार्येण'. The line reads पदस्: a happier reading, we think. Sudhā on S.R. 1, 4, 15-16 quotes this verse with यदाह दत्तिलः. . Here too the reading is पदस्: notation, but Dattila is not explicitly named as the source. The reading is same as in A.B. and Sudhā.

2 T. ed. reads वदा (होय् ? नीह).

Though Brahmā was the first teacher who revealed gāndharva to Nārada, the

god who was chiefly propitiated through gāndharva was Śiva.

Dattila says that Nārada had brought gāndharva to the world of men accord-

ing to vidhi (vidhivat). The phrase is suggestive of the strict rule-bound character

of gāndharva in which, as we have seen, a faulty formation meant the loss of adṛṣṭa.

After stating his subject-matter and its ultimate divine origin, Dattila defines

it. Gāndharva, he says, consists of four elements : pada, svara, tāla and avadhāna.

Bharata and Viśākila have defined gāndharva in similar terms. Abhinava has

discussed the relative merits of the definition as given by Dattila, Viśākila and

Bharata (See ch. I).

Dattila has included avadhāna as an essential element of gāndharva, but

neither Bharata nor Viśākhila incorporate it in their definition. The only other

ācārya who had included it was Dakṣaprajāpati. A definition of gāndharva ascribed

to him is quoted by Simhabhūpāla, who also quotes Dattila's definition in the same

passage. Dakṣaprajāpati states that avadhāna is the chief element in gāndharva

(Sudhā on S.R. 1, 4, 15-46).

Though Dattila includes avadhāna as an element of gāndharva, he does not

seem to give it much importance. He cursorily explains it in a single line. And his

explanation shows avadhāna as a quality which is necessary in the pursuit of any

vidyā; for a proper concentration and the right direction of the necessary faculties is

essential for any serious pursuit or study and is not specific to music. This is the

reason why, as Abhinava points out, Bharata and Viśākila had not included ava-

dhāna in defining gāndharva (“avadhānādvinā kutrāpi siddherayogāt”, A.B. on N.S.

28, 11-12).

The inclusion of avadhāna as an element of gāndharva was perhaps due to

the adṛṣṭa nature of this corpus of music. A line from Viśākhila quoted by Abhi-

nava reads : 'if a musician intently fixes his mind, (samavadhāna) as in japa, on the

self-revealed final śruti (where svara is manifest)1 then a special adṛṣṭa accrues to

him” ('tatra viśākhilācāryaprabhṛtibhiḥ svasamvedye śrutidhāmni samavadhānam

japavaddīyamānam viśiṣṭādrṣṭāya prayokturbhavati iti darśitam' A.B. on N.S. 28, 23).

Having defined gāndharva as a conjunction of four elements : svara, pada,

tāla and avadhāna, Dattila takes up two of these—pada and avadhāna—and indi-

cates their nature in a single verse. Pada literally means a word, in fact, a grammati-

cally inflected word, which can be used meaningfully in a sentence. A knowledge of

the nature of pada—the grammatical and semantic rules which govern it are matters

which lie outside the field of music. Thus Dattila rightly refers his readers to gram-

mar, general usage and dictionaries regarding matters pertaining to pada. Bharata,

too, in summing up the topics constituting the pada element of gāndharva lists a

string of technical terms associated with grammar and also with alaṅkāra-śāstra,

(N.S. 28, 16-17).

1 Such as the fourth śruti of ṣadja, the third of ṛṣabha, the second of gāndhārva etc.

However, there were certain peculiar traits pertaining to pada and its musical treatment in gāndharva which neither Dattila nor Bharata have listed as specific pada topics. But they have dealt with the matter piecemeal as it comes up in other topics.

Coming to verse 5 we notice that it is in direct continuation with the previous verse. The main link-word is ‘atah’. Dattila’s language in the first line of verse 5 is somewhat cryptic, but his purport is clear. In verse 4 he had said that the nature of pada could be learnt from non-musical fields of study like grammar etc. and that avadhāna was a commonly understood state of mind. Therefore, he now says, it is not necessary for him to expound pada and avadhāna (especially since he wants to be brief) but only to describe the remaining two elements besides these (atah anyad-dvayam)—meaning svara and tāla. Of these two, he says, he will take up svara first.

He gives a reason for his choice. Bharata also describes svara before tāla, but he does not state the reason for this preference. The reason Dattila gives for describing svara before tāla is that the svara part of gāndharva constituted the ‘meya’—the object to be measured.1 Tāla, by implication, formed the measure. In such a relationship, the meya is unquestionably the primary object ; the cloth, in other words, is more fundamental than the yardstick.

We must note that svara as an aspect of gāndharva signified not ‘notes’ alone but the entire melodic part of gāndharva constituting specific structures created through patterns of notes. In defining gāndharva Dattila calls it a svarasanghāta, a compact collection of notes or, in other words, a well-composed melody. All the doctrines, laws, rules and regulations which formulated, in a śāstric manner, the nature of this svara-sanghāta are grouped together under svara.

1 The text says ‘meyatvāt’; literally, ‘because of the quality of being that which is measured’.

SECTION I : ON SVARA

SVARODDEŚA

6A. śrutayo'tha svarā grāmau mūrchanāstānāsamyyutāḥ

B. sthānāni vṛttayaścaiva śuṣkaṃ sādhāraṇe tathā

7A. jātayaścaiva varṇāśca nānālankārasamyutāḥ

B. eṣa svaragatoddeśaḥ,1 saṅkṣepeṇāth nirṇayaḥ

The śrutis, the svaras, the two grāmas, the mūrchanās with tānas, the sthānas, the vṛttis, the śuṣka and the two sādhāraṇas-this is the uddeśa (the list of topics) linked with svara. These [topics] will now be considered2 and decreed in brief.

NOTE :

By the age when Dattilam was composed, the study and elucidation of gāndharva had become a well-organised śāstra. As with other śāstras, the subject had been logically analysed into proper divisions and subdivisions and arranged neatly into topics. Such a list of topics was known technically as uddeśa. Bharata calls it 'saṅgraha' which Abhinava explains as uddeśa.3

Dattila's uddeśa of svaras has 12 topics while Bharata has 13. Dattila has left out the topic dhātu which concerned instrumental technique, for, as he states in verse 96, he refers his readers regarding vīṇā and other instrumental playing to other works, in his zeal for brevity.

We have already discussed the comparatively different order in which Bharata lists4 and expounds these very topics and the doctrinal implications of this difference

Datt. T. ed. reads स्वरगतो (दे ? हे) मः। Nijenhuis has translated uddeśa as 'a mere discription of things relating to the notes' (Dattilam : A compendium of Ancient Indian Music, p. 17). This is misleading, for uddeśa as we have discussed in ch. I was a technical term in śāstras which indicated a definitive list of all topics pertaining to the matter under discussion.

2 Dattila says सङ्क्षेपेण निर्णयः which in this context can bear more than one rendering. It may, for example, be translated as : 'now in brief will follow an authoritative verdict on these'. निर्णय means 'to investigate', 'to ascertain', 'application of a conclusive argument', 'discussion', 'consideration', 'verdict' etc. (See Sanskrit-English Dictionary by Monier Williams). We have incorporated the two major aspects of the notion of 'nirṇaya'-to weigh or consider and to pass a final verdict-in our rendering.

3 गान्धर्वसंग्रहो ह्येष इत्यन्नेनोक्तं मया मेकविषमीकृतौति ।

A.B. : गान्धर्वसंग्रहो ह्येष इत्यन्नेनोहेत्यन्नेनैकोद्देश्यमेकविषयमीकृतौति ।

4 स्वरा ग्रामो मुर्छनातश्च तानाः स्वस्थाननि वृत्तयः । शुष्कं साधारणे वणा हालङ्काराश्च घातवः । श्रुतयो यतश्चैव नितयं स्वरगातमकाः ॥ -N.S. 28, 13-14.

(see ch. 1). All works on gāndharva, apparently, had the same standard list of topics but the order of arrangement and the implied hierarchical importance given to one topic over another evidently differed from ācārya to ācārya. From Abhinava we have some idea of Viśākhila's uddeśa. Like Bharata, Viśākhila, too, had listed and discussed dhātu, for Abhinava refers to him five times in connection with details of instrumental technique and in fact speaks of his difference with Bharata regarding sub-divisions within dhātu.1

Among topics concerning svara we notice that svara itself is a topic. Svarain a general sense implied the entire melodic (as opposed to the tāla) aspect of gāndharva. Svaras as a specific topic meant the musical tones used in gāndharva.

After listing the topics concerning svara, Dattila says that he shall now give, in brief, a 'nirṇaya' regarding these topics. 'Nirṇaya' is listed along with 'niścaya' by the Amarakośa as standing for ascertained knowledge and is opposed to such words like samśaya and vicikitsā etc. which stand for uncertain or doubtful knowledge.2 Thus Dattila, in effect, tells us that the exposition he is going to present stems from his authentic and ascertained cognizance of the subject-matter. Here, certainly, speaks an authoritative ācārya who has a confident command over his subject.

1 Bharata introduces the topic dhātu with गीतयो गदिता: सम्यग् धातू ष्चैव निवोधत: । —N.S. 29, 50.

Abhinava comments : धातुं ष्चैव निवोधतेति । चकारेणालङ्कारादीनामप्येषां समुच्चिनोति । एककारेण चतुष्टयाहरणं त्रिप्रहरणमथूलीनाम विभागो धातुं ष्चैव निवोधत: । एवं वृत्ती समवलेखा चित्नलेख्यादिकं विशाखिलाचार्यप्रोक्तं सर्वंथैव ध्रुवाराज्ञानवैकल्योपयोगान् मया नोक्कमिति सूच्यते।

2 अध्याहारस्तर्क उहो विचिकित्सा तु संशय: । संदेहादपरो चायं समी निप्रायनिर्णयो ॥ —Amarakośa, 1, 5, 3.

TOPIC I

ŚRUTI

8A. nrṇāmurasi mandrastu dvāviṃśatividho dhvanih

B. sa eva kanthe madhyah1 syāt, tārah śirasi gīyate

9A. uttarottaratarastu vīṇāyām tvadharottarah2

B. iti dhvaniviśeṣāste śravaṇācchruti samjñitāḥ

In the human chest reside sounds of twenty-two kinds [termed] mandra ; the same [range of sounds when sounded] in the throat is called madhya [and] in the head, tāra. Thus we obtain a sucessively higher pitch as we move up (uttarottara-tārastu). Conversely, on the vīṇā a sucessively higher pitch [is obtained on] successively lower [strings]. These specific sounds are termed śrutis for they are audible.

NOTE ;

These two kārikās contain in gist a great deal of matter. Dattila begins by describing the anatomical basis for the production of musical sounds in singing. In ancient singing, as in most contemporary vocalising, the range was three octaves. Each of these octaves was thought to be produced

1 T. ed. reads : कण्ठमध्ये स्यात्. However, the editor refers to another reading : कण्ठे मध्यः; स्यात् इति धीरस्वामिना नाटयवर्गः स्मृतः पाठः- Nānyadeva has quoted Datt. 8 but the manuscript reading is faulty : दत्तिलाचार्येण तु वीतमत्स्य (?प्ता) मिहितं यथा-

ṇṛṇāmurasi mandrastu dvāviṃśatividho dhvanih1 स पवक् चैव मध्यः स्यात्तार; शिरसि गीयते ॥

—B.B. (1) 6, 33.

2 Datt. T. ed. reads वीणायामधरोत्तर.: We have adopted Abhinava's reading who quotes the phrase as : यथाह दत्तिल-—वीणायां त्वघरोत्तर इति

—A.B. on S.N. 29, 54.

Verse 34 in the Nāṭyacūḍāmani of Somanārya (Government Oriental Manuscripts Library Madras, Manuscripts No. 12998) is, obviously, borrowed from Dattilam with a minor change. Datt. reads as : श्रोणीवक्षःसु विरंजिता:

Another echo of Dattila's description is to be found in the S.S.S., a medieval work :

वक्षःस्थाने तु हृकण्ठविभागीति समासतः; एकैकर्मपि तेभ्यः स्यात् द्वाविंशतिविधायुतम् ॥

द्वाविंशतिविधो मन्द्रो ध्वनिः; संजायते हृदि । यथोत्तरमसो नादो वीणायामधररोत्तरम् ॥

स एव हृद्गुणो मध्यः कण्ठस्थाने यथाक्रमम् । स चैव मस्तके तारः स्याद्मध्याद् द्विगुणः क्रमात् ॥

इति स्वरगता जेया श्रुतिः श्रुतिवेदिभिः ।

—S.S.S. as quoted by Sudbā on S.R. 1, 3, 7.

from a different section of the human frame. The lower octave was said to reside in the chest, the middle octave in the throat and the high octave in the head.1

The Bṛhaddeśī relates in greater detail the anatomical process through which sound or nāda was said to be produced in the human physique. In the centre of the body, the text says, is the brahmagrānthi,2 the abode of Brahmā. Within brahmagrānthi resides the vital breath, prāṇa. Prāṇa gives rise to a vital fire, vahni, and the conjunction of this fire and breath produces nāda or sound.

yaduktam brahmaṇaḥ sthānam brahmagrānthiśca yaḥ smṛtaḥ tanmadhyai samsthitaḥ prān̄aḥ prān̄ād vahnisamudgamaḥ vahnimairutasaṃyogānnādaḥ sampujāyate

(Bṛ. 18-19)

This anatomy and the process of sound production was clearly Tantra-inspired and, indeed, Śārṅgadeva, who speaks of the process at much greater length, gives an account, in this connection, of the human frame in terms of the Tantric anatomy of cakras (S.R. 1, 2; the section on piṇḍotpatti). Such an account of the production of nāda features in numerous works on saṅgīta.3

In Dattila’s account of the physical location within the human frame of three octaves, we notice that the higher the octave, the higher its position in the body: mandra resides in the chest, madhya in the throat and tāra still higher in the head. The musical sounds rise gradually in pitch as their location climbs higher, reaching up from the torso to the head. The Vṛtti on the Bṛhaddeśī explains the concept through the image of the stair-case. The nāda born of the conjunction of the vital fire and breath, it explains, rises from the navel through human effort. Climbing steps by step like smoke propelled by a wind on a stair-case, it assumes the form of the different śrutis.4

Other theorists explained the gradually rising sequence of musical sounds and their location at successively higher positions on the frame, through yet another metaphysical-cum-anatomical concept. Śārṅgadeva has put the concept succinctly: Tāntri c anatomy recognises many subtle nāḍīs; two main nāḍīs or arteries, which scale the frame from the base of the spinal column to the tip of the skull, are called iḍā and

1 व्यवहारेऽपि त्वचि ह्येव मन्त्रोद्भवभङ्गीयते । कण्ठे मध्यो मूर्छनीयस्तारो हि द्रुतगुणोत्तरः ॥ —S.R. 1, 3, 7-8.

also S. Raj 2, 1, 1, 24-27 and other works.

2 The Bṛ. does not give the location of the brahmagrānthi. Śārṅgadeva gives it (S.R. 1, 2, 145-48). The brahmagrānthi is said to be located near the navel and in the middle of the brahma-

grānthi is the nābhicakra of Tantric anatomy : देहस्य मारीरस्स्थ कन्दो विस्तारकरणम् । उत्पेघो दोर्घश्छ्रायः । नाभिस्थिततथ्चुरज्ं लुप्रमाणकः । तस्य नाम ब्रह्मग्रन्थिरिति । तनमध्ये इति । तस्य ब्रह्मग्रन्थेमध्ये नाडीमिश्रक्रकचक्र द्वादशदलमस्ति । —Sudhā on S.R. 1. 2, 145-50.

3 eg. S. Raj 2, 1, 1, 1-27 and Aumāpatam 1, 1-19 among others.

4 ततोदो तारवहानि(?)पवनसंयोगात् पुरःप्रयणान्तरे(?)पितोदर्श नामभेदव्यवकार(?)न देशमार्गमाददु घुमवत् सोपानपदक्रमेण पवनेच्छया आहुरोहनतभू(?आरोहणनतभूतं) पूरणप्राति निपायपातितया (?) श्रुत्यादिभेदभिन्नः प्रतिबिम्बः । —Bṛ Vṛtti on 27A..

pingalā. They are positioned on either side of the human frame (S.R. 1, 2, 150-154).

On the chest, across iḍā and piṅgalā are twenty two nāḍīs stretched horizontally one

above the other in a step-wise manner. These produce the twenty two graded sounds

of the octave. The lowest nāḍī produces the lowest pitch and the pitch gradually

rises as the higher nāḍīs are struck by the vital breath (mārutahāti). A similar set

of twenty-two nāḍīs resides in the throat and another in the head (S.R. 1, 3, 8-10 and

Kalā). This concept, too, was evidently quite ancient and Kohala, as quoted in the

Vṛtti on Bṛhaddeśī, refers to this notion (in connection with svaras3).

The image suggested by the concept is that of a triangular harp situated

within the human body with a range of three octaves, having its broad base at the

bottom, with successively shortened strings rising to a pyramidical tip; the higher the

string, the higher the pitch of its sound. Dattila clearly hints at this very image

when he says that in the human frame the sound rises in pitch, the higher its location

(uttarottaratārah) while in the viṇā, (viṇāyāṃ tu) the process is reversed; the lower

the string, the higher its pitch (adharottarah).

Abhinava expresses the notion with another analogy : the location of the

mandra and tāra on the viṇā (and the flute) was, he says, quite opposite to that on

the human frame ; one could liken the position to a mirror image where the left be-

comes the right and the right becomes the left.2

Though Dattilla does not state it in so many words, the viṇā he is referring to

was certainly a harp-type of viṇā, shaped like a bow and placed with its tips away

from the ground. Such a viṇā is found sculpted on a Sanchi gateway (2nd century

B.C.). A relief of musicians from Bharhut carved almost in the same period,

shows a man playing another such viṇā. The lady sitting on a stool who is shown

playing the viṇā in the Sanchi relief is holding it with its narrow end towards the

ground. In this viṇā, a higher note would obviously have been struck on a lower

string. King Samudragupta (ca. 4th century A.D.) in a famous coin minted by him,

is shown playing a similar viṇā. An Amaravati sculpture of approximately the same

period, shows a girl playing an identical type of viṇā.3 Examples can be multiplied:

an ancient figurine dug at Rupar shows a woman playing such a harp (see Roopalekha,

a journal concerned with the arts published by The All India Fine Arts and Crafts Society,

Vol. XXV, No. 2, 1954, p. 27) a small plaque from Pawaya, Gwalior; shows a dance

scene where an accompanying musician plays another such harp.4 Representations

तथा चाह कोहल:—

आत्मेच्छया महितलाद् बायु:रचनानिधायंते । नाडीभित्तौ तयाकाशे घ्वानिरक्त: घ्वसस्वर: स्मृत:॥

ऊर्ध्वनाडी प्रयत्नेन सर्वंभ्र्त्तिननिष्ठितानतु । मूर्छ्छनतो घ्वनिरामूर्छन: स्वरोत्सौ क्यापक: पर:॥

— Br. Vṛtti on 63.

2 प्राणाभिननेनैव हि तीब्रतीब्रेण शरीर एव वंणोऽपि स्वरनिष्ठित:। वीणायां तु आदर्शो वामदक्षिणानुबिपर्यांसवत्तारमन्द्र-

विपर्यय:; ।

— A. B. on N. S. 28, 11-12.

3 See Bh. Sang. Iti p. 258 and 482. Sketches of these sculptures may be seen in the appendix of

the same work; see sketches Nos. 7, 8, 62, 63.

4 For a sketch see Bh. Sang. Iti (B), Vol. II, Appendix p. 38.

200

A Study of Dattilam

are to be seen even in the Indian-inspired art of South-East Asia at such sites as

Borobudur and Angkor Thom.1 A harp of this variety survives to this day in Burma

(see Die Musikinstrumente Indiens und Indonesiens by Curt Sachs, p. 141). In India

we have the ‘sakoda yazh’ from Tamilnad which is a direct descendant of the ancient

harp (for an illustration see the article entitled ‘Music Instruments of India’ by

Smt. N.C. Indira Devi in Roopalekha, Vol. XXVI, No. 1 facing p. 31).

Bharata’s directions for tuning the vīṇā also reveal that he had the harp in

mind. In rendering a tāna on the vīṇā, when certain notes had to be dropped, the

technique was twofold : the process could be accomplished either through praveśa

or through nigraha. Nigraha meant to avoid touching the notes to be omitted

(nigrahastvasamśparśah, N.S. 28, 34). In praveśa the note to be dropped was

tuned either to the pitch of the note lower to it or the one higher to it (tatra prave-

śanaṃ madhurasvaraviprakarṣāduttaramārdavādvā, N.S. ibid). This, as Abhinava

says, was done either by tightening the string to which the note to be dropped was

tuned or loosening it. On rendering a tāna on the vīṇā, Abhinava says, if one is not

skillful enough or cannot avoid striking the note to be dropped, then the string to

which this note is tuned should itself be ‘omitted’ (lopyasvaratantrītvaṃ kriyate) by

loosening it so as to make it become one with another note or, alternately, tighten-

ing it to the pitch of the note bigher to it.2

Bharata himself has described the processes of lowering and raising the śrutis

on a vīṇā through the words mārdava and āyatatva (mārdavād āyatatvādvā, N.S. 28,

27). Abhinava explains these terms as: “mārdava is the loosening of a string while

āyatatva is the converse process.”3 Clearly the vīṇā Bharata had in mind was the

harp.

The process of the production of musical sounds through the human frame

was conceived on the analogy of this harp-vīṇā. As in the vīṇā, so in the human frame,

there were string-like nāḍīs (arteries, in this case, of the subtle body) stretched horizon-

tally which were responsible for the graded musical pitches produced in singing. It

is almost as though the ancients thought that there existed a concealed harp within the

human frame (as it does in the piano) except that this human-harp was constructed upside

down in contrast to the ancient wooden vīṇā or harp.

It was perhaps the physical, wooden harp itself which gave rise through

metaphor to the postulation of a harp-like process within the human frame. Influen-

ced further by Tāṇtric thought and its system of subtle anatomical nāḍīs, theorists

built up a picture of the existence of a nāḍī-built harp concealed within the body.

The concept of the body as a harp was an ancient one. It can be traced back

to the Vedic Age. In the Aitareya Āraṇyaka, there is a long passage where a man-

1

ibid., p. 18.

2

तन्त्र्यां यदा तात्क्रियते तदा लोध्यस्वरतोऽन्यो न्रियते, यदि तावत्कोशलं न, स्मृष्यते यदि वा, किश्चित्करणे (घन)

समस्वरासाम्यं पीडनेन चोत्तरस्वरसायंनीयते ।

• A.B on N.S. 28, 34.

3

मार्दवं तन्त्र्याः शिथिलीकर णं विपरोतत्वमायातत्वं क्रियत इत्याह ।

-ibid., 28, 27-28.

made vīṇā is compared to the god-created human body : “just as the human body has śira (head) similarly does the vīṇā have one; just as the body possesses a belly, similarly does the vīṇā have an ambhana (the sound box); the strings of the vīṇā correspond to the fingers; the skin covering of the vīṇā corresponds to the human skin, the tongue to the plectrum”1

In Śikṣā works this idea becomes more crystallised and the human frame is conceived of as a musical instrument and called the gātra (or śārīra) vīṇā. The harp, in contrast, was known as the dāravī vīṇā or the wooden vīṇā. Just as different notes were produced on the vīṇā through different fingers, analogously different notes were indicated on different fingers in singing sāma.2 The practice continues and is traditionally known as the hasta-vīṇā.3

The Nāradi Śikṣā says: “there are two vīṇās used for rendering various types of songs. The sāmic vīṇā is the gātra-vīṇā. Its characteristics are as follows: that vīṇā is known as the gātra-vīṇā, on which sāma-singers sing; it consists of vowels and consonants and makes use of the fingers and the thumb”.4 The term gātravīṇā here clearly refers to the human frame conceived as a vīṇā with words echoing the Aitareya Āraṇyaka. Bharata, too, conceives the body as a musical instrument that produces song and calls it the śārīra-vīṇā (a synonym of gātra-vīṇā). He calls the wooden vīṇā, dāravī-vīṇā (N.S. 28, 14-15),

There is some misconception regarding the term gātra (or śārīra) vīṇā. Some have thought it to have been a kind of ancient lute or lyre.5 But an examination of the comments of Abhinava and Nānyadeva reveals beyond doubt that gātra vīṇā denoted the human frame. Commenting on the word śārīrā vīṇā, Abhinava says: “when Sarasvatī in the form Speech arises out of the human frame this is known as vīṇā; śārīre vīṇā, vāgrūpā hi sarasvatī vīṇāśabdenocyate”. (A.B. on N.S. 28, 13-15). Dāravi-vīṇā was, on the other hand, the vīṇā with a wooden frame.6

Elsewhere, Abhinava states that in relation to the śārīra vīṇā one always speake of ‘singing’ ard the effort needed is of an internal nature: “śārīre tu gāyetye vet vyavahārah, āntarasyaitra prayatnasyā bhāvāt” (A.B. on N.S. 30, 3). But

1 अथ किलीयं देवीं वीणा भवति तदनुकृतिरसौ मानुषी वीणा भवति यथास्या: शिर: एवममुख्यो: यथास्या: उदर-मेवममुख्यो: यथास्या: जिह्वादण्डं वादकं यथास्या: स्वरा एवममुख्याः; यथास्या: सक्थि एवमुख्यो: यथा हि ह्स्तौ श्च तन्त्रीकरवती यथा हि ह्स्तौ श्च लोमतन्त्र्यौपहिता भवत्येवमेवाद्यवेतसो: श्च लोमतन्त्र्यौपहिता । लोमतन्त्री हि रस्म व च वर्मणा । पुरा वीणा ग्रपिदधति ।

—Aitareya Araṇyaka 3, 25.

The passage is quoted and explained in Bh. Sang. Iii, p. 24-25.

2 अङ्क्क ठस्योत्तमे ऋष्टोधरि। अङ्गे प्रथमं; स्वर: । प्रादेशिन्यां तु सन्धार ऋषभस्तदनन्तरम् ॥ अनामिकायां पदजस्तु कनिष्ठिकां च गृह्वेत; तस्याघस्तालाच्च योजात्तु निपादं तद्विन्यसेत् ॥

Nāradi Śikṣā as quoted in Bh. Sang. Iti (B), Vol. I, p. 318.

3 See the article entiiled ‘Sāma-Sangīta’ by V.S. Agrawal, in Nāda Nināda-Śruti, p. 33.

4 दारवी गानवीणा च हे वीणे! गान जातितः । सामिकी गानवीणा तु तस्या: श्रुति लक्षणम् ॥ गात्रवीणा तु सा प्रोक्ता यस्यों गायन्ति सामगा: । स्वरवत्यज्जसंयुक्ता अङ्गुलि लय, धरत्स्थित ।

—Nāradi Śikṣā quoted in Bh. Sang. Iti (B), Vol. I, p. 299.

5 See for example Bh. Sang Iit (B). Vol. I, p. 299.

6 दारुणोप्तं वीणा वामूला भगवतीति दारवी । तेन तन न......प्रताय: । “दारुणोह्यसो जातो बाक् एत.....स्म । गानं प्रनष्टा “नित्यादिना विशिष्टलाऽऽलार्ं; प्रादीयत इव तदाघारत्काच्च काष्टमयोदपि सतिनवेशो वीणेर्युच्यते ।

—A.B. on N.S. 28, 13-15.

regarding the instrumental vīṇā one speaks of 'playing', because the activity is external and involves the striking of strings: "vīṇāyāṃ tu vādayatyeveti; bāhyasyaiva tantrihananakāriṇaḥ prayatnāsya bhāvāt" (A.B. ibid.).

In parenthesis, we may here note a curious ancient controversy. Some predecessors of Abhinava believed that the svaras produced on the instrumental vīṇā were a 'reflection' of the śārīra vīṇā.1 Abhinava has strongly argued against this view (in the context of discussing the two-fold division of svara topics on the basis of dāravī and śārīra vīṇās by Bharata). The mere fact, he says, that the same organ of perception, namely, the sense of hearing, was the cognitive tool for experiencing the tones of both the śārīra and the instrumental vīṇā, was no reason to think that the latter was a reflection of the former. Does the human frame as reflected in a mirror, he asks, bear any resemblance to the vīṇā?: "vīṇā hi dehasya pratibimbamiti keyaṃ bhāṣā? na hi darpaṇasaṃkrāntapratibimbaśarīrapraticchandakasadrśī vīṇā vīnetyupālambhaḥ" (A.B. on N.S. 28, 21). Neither, he says, does the body itself form a sort of mirror in which the vīṇā is reflected: "na ca mukuravaddehasya pratibimbanasthānam" (A.B. ibid.). Nor, he continues, were the notes on the vīṇā, a reflection of the śārīra notes as was clear from direct observance and also from the fact that one obtained on hearing a sung and a played note was not a pair of notes-one the bimba (i.e. the śārīra note) and the other the pratibimba (the reflected vīṇā-note)-but the same identical note: "na ca vīṇāsvare śārīrasvaram pratibimbarpayatīti dṛṣṭam, dvayoḥ svaryoranupalambhāt" (A.B. ibid). The vīṇā-note he further asserts is, indeed, capable of being produced without the help of the śārīra note: how can it, then be a reflection? Moreover, quite a different kind of effort is needed to produce the śārīra and the vīṇā notes: "śārīrasvarābhāve'pi vaiṇasvarapravṛtteḥ śārīrasvaraḥ vaiṇasvarapratibimbamityasadeva; prthakprayatnanīṣpādyam" (A.B. ibid.). The reflection theory, in Abhinava's view, thus could not hold ground. Abhinava's remarks show that through the term śārīra, the ancients referred to the human frame and not to a kind of vīṇā.

Nānyadeva, too, (in his Bharata Bhāṣya) discusses the terms dāravī vīṇā and śārīra vīṇā (the context is the same where Abhinava made the comments noted above). It is obvious from Nānyadeva's remarks that the śārīra vīṇā was the human body conceived as an instrument. "The dārvī", Nānyadeva says, "is the vīṇā on which strings are attached: dārvī yā tu tatrāpi nāma tantrī samanvitā" (B.B. 1,84). It is, he adds, of various types known as vipañci, vallakī, mattakokilā etc. (B.B. 1, 84-88): these were popular varieties of ancient harps. "The śārīra, on the other hand, consists of the major and minor limbs (of the human body)-the heart, the neck, the palate, the teeth, the tongue, the nose and the chest. In it the prāṇavāyu-the breath-is responsible for the production of melodic movements (the varṇas, like sthāyī, sañcārī etc.) by creating a friction over various regions (like chest, neck and head). The

1 यत् केचित् प्रतिसमादधते—वीणाया: प्रतिबिम्बत्वात्तन समानेऽद्रियाग्राह्यो य: शारीरस्वरसमूहस्तमेव प्रतिमिम्बमपद्यतु-महंतीति तदस्यन्तमसंगतम् ।

-ibid., 28, 21.

dāravī does not consist of chest, neck, head and other limbs but still it produces similar melodic movements."1

Theorists, including Dattila, speak of twenty-two graded tones in an octave. These were the śrutis. Śruti has been a very central concept in Indian music. The ancients have discussed the concept in various ways both from a metaphysical and physical viewpoint. Śruti was inherently related to svara and we will take up the various views regarding their inter-relatedness in discussing the topic svara. An introductory clarification of the ancient concept of śruti will, however, be helpful here.

Dattila has a short definition of śruti, which hinges upon an etymological explanation of the term itself. Śrutis, he says, are specific or distinct sounds (dhvanivi-śeṣāḥ); these are termed śruti because they can be heard (śravaṇāt), i.e., heard as specific sounds, each distinct from the other. Their distinctness, Dattila clearly implies, lay in their forming a distinct interval in terms of pitch ; for the śrutis, he says, gradually rise in pitch. The śrutis according to ancients were thus the number of intervals which it was believed the unaided (but, of course, trained) ear could cognise within an octave. Hence Dattila defines śruti as 'audible', i.e., distinguishable through the ear; so have others.2

Abhinava, too, remarks that śruti is that distinction between one sound and another which is cognisable by the ear: "śrutiśca nāma śrotagamyam vailakṣaṇyam yāvatā śabdenotpādyate" (A.B. on N.S. 28, 24). Theoretically, the octave could be divided into not only twenty-two, but an infinite number of intervals. And Abhinava remarks that pitch can in the ultimate analysis be raised or lowered over infinitesimal 'atomic' intervals, but such extremely subtle distinctions in pitch cannot be grasped: "yadyapi paramāṇuto'pyutkarṣāpakarṣo vā bhaved dhvaniviśeṣasastathāpi nāsau grhītam pāryate" (A.B. on N.S. 28, 27-28).

An audible distinctness in sound (dhvani) thus was the ground for a difference in śruti. But a distinctness in sound is also audible when timbres differs even though the pitch remains the same. Some ancients, indeed, theorised that śrutis were different kinds of musical tones produced in singing, through the effect of different bodily humours (recognised in Āyurveda). Different voices have different timbres and the

1 ननु दार्वीद्रव्यकृतं किं नु खलु वक्ष्यमाणलक्षणं स्यात् । जितद्रुमादिस्वरेण तत्र स्पष्टश्रुतीनि संचरन्ति ॥ —B,B. 1, 88-89.

2 Mataṅga analyses the term śruti as: श्रू श्रवणे चास्य धातोः । वित्तप्रत्ययसमुद्भवः । श्रुतिशब्दः प्रसाद्योऽयं शब्दज्ञः । भावसाधनः । — Br. 26-27.

The Vṛtti says: श्रूयते इति श्रुतिः : —cf. Datt. 9B.

The Vṛtti quotes Viśvāvasu on śruti whose definition again parellels Dattila's: श्रवणेद्रियप्राहकवाद् ध्वनितरे श्रुतिभवेतू । —Br. Vṛtti on 27A.

distinctions are, indeed, easily cognisable. The Vṛtti on the Bṛhaddeśī thus quotes the view that śrutis were of four kinds—produced by vāta, pitta, kapha or a combination of these three ‘humours’: “apare tu vātapittakaphasannipātabhedabhinnāṃ caturvidhāṃ śrutim pratipedire” (Bṛ. Vṛtti on 27A). The Vṛtti quotes an ancient passage, ascribing it to Catura (?)1, which says that vāta produces a high-pitched shrill tone, pitta produces a deep resonant tone and kapha produces a delicate smooth and sweet tone and the tone produced from a combination of all the three humours has the tone-quality combining each (Vṛtti on Bṛ. 27A).

Abhinava, too, speaks of differences in timbre that a tone can have. One easily recognises, he says, the difference in a note when sounded on a mallaka (an ancient instrument) and on a vīṇā: “lakṣyate vīṇāsvaro’yaṃ mallakasvaro’yaṃiti” (A.B. on N.S. 28, 21). But this difference in timbre, he argues, was not the fundamental and foremost aspect of a musical sound, for timbre could not alter the note itself: ṣadja was recognised as ṣadja whether played on the mallaka or the vīṇā (ṣadjatvena tvabhedaḥ, A.B. ibid.). Abhinava explains the point through a conventional philosophical simile: fire can be produced in various ways, such as through the friction of iron rods, or the friction of marble pieces or through the effect of lightening and so forth but the smoke arising from fire produced through different causes, remains the same.2 In other words, timbre differences do not alter the musical pitch of a sound. Yet they do impart to it a different ‘colour’ and ‘feel’. Recognising this Abhinava says that just as subtle differences can exist between one smoke and another though both remain smoke, similarly the same note produced from different sources can have subtle differences of timbre which do dot take away from the fact that the note produced is identical, whatever its origin.3

Basically, śruti was thus believed to be a measured interval in pitch, Bharata consequently uses the term ‘pramāṇa’ (measure) in conjunction with the word śruti while giving the tuning process for arriving at the differently positioned pañcamas of the ṣadja and madhyama grāmas. Abhinava, in commenting on the word ‘pramāṇa’ in this context takes it as the one word which points at the most fundamental aspect of the nature of śruti. The association of the term ‘pramāṇa’ with śruti, he says, signifies that śruti is essentially a division or interval in terms of high or low positions in regard to pitch that is occupied by a sound.4 It cannot, he adds, be defined in terms

1 The same passage has been ascribed to Tuṃburu by Kallinātha in Kalā on S.R. 1,3, 10-16.

2 Abhinava gives the simile not only of agni but also of banana (kadali) which can be produced through seed as well as through a process of root-transplant; and of vṛścika—scorpion—who, it was believed, could be born from cowdung besides the usual reproductive process. These were conventional similes to establish the fact that different causes could produce the same effect.

3 ननु कारणभेदाद्‌जाते श्रव्देन कथम् तज्जातीयोऽनुरणनशब्दः; त्रियते। नैवं । कारणे नियामकत्वाद्‌; अन्यःस्फुटिकारणनिवृत्ता-दिकारणमेदसंजातीयो हि भिन्नधर्मकजातीयो जनयति । बीजाद्‌दोत्पत्तिकदली द्वयादो वा कारणमेदेऽप्यकायंता दृष्टा । न तावद्‌र्यत्रिन सङ्को मो भेदः। घूमाद् तु, तद्वेदायास्तत्र । लड्‌गहे च वीणारस्रो य मल्लकस्वरोऽयमिति, पृथक्-त्वेन स्वमेदः।

—A.B. on N.S. 28, 21.

4 cf. also Bharata’s statement: एवं श्रुत्यन्तरपदर्थां यदनन्तरं मादवायातस्वराऽ तत्रमाणं श्रुतिः।

—N.S. 28, 27.

of a unit of time (kālakalā) or as merely a particle of sound (nādāmśa) or as āyu or sthāna or karaṇa.1 “pramāṇeti vadanna kālakalā śrutirnāpi nādāmśo na cāyurna stnānam na karaṇamiti darśayati tena vaksyamaṇoccāvi bhāgo dhvanirekā śrutirititātparyam” (A.B. on N.S. 28, 27-28).

Abhinava has, in fact, defined śruti in terms of raising and lowering of pitch in very clear terms. Answering the question : ‘atha keyaṃ śrutirnāma’ (what, then, is śruti?)—Abhinava says that śruti is basically the raising or lowering of sound to the minimum audible extent. This raising and lowering creates an interval determined by a certain measure—a measure which can be specifically cognised and which thus forms the śruti; that is to say that in lowering or raising a sound the point where the lowered or raised pitch becomes clearly graspable as a distinct sound, we have śruti.2

Thus we see that an audibly distinct gradation in pitch was the basis of śruti. How is it, then, that śrutis were recognised as only twenty-two ? Certainly the human ear can distinctly perceive many more graded intervals. People with an ear for tone do so today and must have done so in ancient times, too. Abhinava has spoken of the possibility of atomic intervals—and he adds that yogīs with their extremely hightened powers of perception can perceive these intervals in their samādhi : “paramāṇumātre’pi sthānāntare yogiprāya (samādhi)samādhigamyadhvanivailakṣaṇaṃ bhavati” (A.B. on N.S. 28, 27-28). Certainly, not only yogīs, but even the trained musical ear can perceive many more than twenty-two intervals. Evidently, a well-established convention and not the ear alone, was responsible for the fact that śrutis were fixed at twenty-two.

In arriving at them, the ear certainly must have played a part but so must have other factors. When lowering and raising a note in tuning a vīṇā an expert musician would know when he reached a true ‘recognised’ śruti-interval, even though within that śruti interval itself he must have been able to perceive smaller gradations of pitch.

Some ancient ācāryas, indeed, did not subscribe to the twenty-two śruti notion. The Vṛtti on Bṛhaddeśī quotes Kohala who speaks of ācāryas ‘versed in the knowledge of śruti’ (śrutijñānavicāradakṣāḥ), some of whom recognised sixty-six śrutis and

1 Abhinava is here referring to various views posited by different theorists regarding the nature of śruti—views he held to be misconceived or not bearing on the essential aspect of the matter. The exact nature of these varied opinions and their differences is not clear. However, in one passage Abhinava has discussed the view which regarded kālāmśa as śruti and refuted it. This view apparently held that śruti was a pitch or note held for just a momentary interval of time. Such a view, according to Abhinava, was erroneous because (he seems to suggest) any sound could, besides being momentary, be also long-drawn as in the sounding of a bell. For this reason śruti was not to be confused with a temporal particle of sound (śabdāvayava):

ननु कालांशः श्रुतिः चत्वःश्रुतिकस्यापि विन्द्राघलकादारादि मितकालावस्थानाद् द्विश्रुतिकस्यापि स्थायि वर्णं प्रयोयोगेनाल्पकालत्वादेशः स्यात्, घण्टाशब्दह्रस्वदीर्घत्वेन चिरकालत्वाद् । एतन्न, शब्दावयवो न श्रुतिरियुक्तमेव ।

-A.B. on N.S. 28, 24.

2 अथ केयं श्रुतिरित्याशङ्क्य—‘एवं श्रुत्यन्तरपरिचयाच्चेतति’ । मादृशं तन्नयः; श्रुत्यधिकरणं विपरीतत्वमाश्रितं क्रियत इत्याह । श्रुति: शब्दस्य श्रवणस्योत्तरवर्तितया अपकर्षो मन्तव्य:, तदेतुल्यनादेर्वायते तु द्वे अपि तथोक्ते । एवं तीव्र-मन्द्रहेतुभ्यां यदन्तरं यो विशेषावबोधः प्रमाणं निश्चायकं यस्मात् सा श्रुति: । प्राप्तकनस्य ध्वनेर्निर्वलक्षणं यावता होनाधिकेन या तीव्रमन्द्रतमा रूपेण सा श्रुतिरिति यावत् । यद्यपि परमाणुतोड्युतश्रुत्यन्तरपकयो’ वा भवेद् ध्वनिविशेषस्थापि नासो गृहीतुं पायते ।

-ibid. 28, 27-28.

others who were not willing to point at their number and believed śrutis to be numberless.1

Later on, we shall discuss the ancient process for arriving at the measure of a śruti.

The Vṛtti on Bṛhaddeśī records different metaphysical views concerning śruti and propounds a monistic view regarding the nature of śruti which is, apparently, Vedanta-inspired: śruti, it says, is in reality one, its plurality is only a pratibhāsā—an illusory phenomenon—connected with the fact that a single indivisible ‘nāda’ arising from the navel ascends up the human frame in distinct specific steps ; at each of these steps the nāda appears as a different śruti but in reality it is one.2 In Viśvāvasu’s view (as quoted by the Vṛtti) śruti was ultimately dual in nature, because it could be both svara and śruti.3

1 तन्न केचिन्मोमांसामनस्लितधियो धीरा द्विविधान्ति श्रुतीर्मन्यन्ते । केचिदेवं (श्रुति) पक्षेऽभेदभिन्ना: श्रुतय इति मन्यन्ते ।

अन्ये पुनरानित्ये वदन्ति श्रुतिसंख्याम् । यथा चोक्तं हि— “द्विविधात् कौशिकाहरान्त (श्रुति:) श्रुतिसंख्यानवचारदक्षा: ।

पटुप्रष्टिभिन्ना: खलु केचिदश्रान्तमानसत्वमेव प्रतिपादयन्ति ।

— Br. Vṛtti. on 28.

श्र यन्ते इति श्रुत्तथा: । सा चैकानेका वा । तल्लकै श्रुतिरिरिति । तदथा—तद्वादे तावद्विधानुपवचनसंयोगात् पुरुषप्रयत्नप्रेरि-

तोदृशं नामधेयद्वारेणामाक्राम ध्वनवत् सोपानपदक्रमेण पवनेच्छया ग्राहाहरह्न्तभूत (?आरोहन्तभूत) त् पुणेरणप्रति-

निपातोऽपि चायता (?) प्रतिभास इति मामकीय मतम् ।

— Br. Vṛtti on 27A.

अन्ये पुनद्विधकारा: श्रुत्तिमन्यन्ते । कथम् । स्वरान्तरविभागात् । तथाह च विश्वावसु:—सा चैकापि द्विधा जेया स्वरान्तर-

विभागतः ।

— Br. Vṛtti on 27A.

TOPIC 2

SVARA

10A. tābhyah kāṃścidupādāya gīyante sarvagītisu

B. adriyante ca ye teṣu svaratvamupalabhya te

11A. svarāḥ ṣadjādayah sapta

Of these [śrutis] some are selected and sung in all the gītis. Among these [śrutis] those that attain the state of a svara (svaratva) are esteemed as the seven svaras beginning with ṣadja.

NOTE :

In one and a quarter verse Dattila describes svaras and their relation with the śrutis.

We have seen (in ch. 1) that quite a controversy existed among early theorists regarding the relative importance of svara and śruti. Dattila lists śruti as the first topic and he says that svaras arise from the śrutis : among available śrutis some attain the state of svara. Bharata has taken up svara as the first topic and speaks of śruti as dependent upon svara (svaramaṇḍalāsādhitaḥ).

In fact, Bharata expounds śruti as the third topic (after grāma), and implies that the notion of śrutis is subsidiary not only to svara but also to grāma : “atha dvau grāmaṯu ṣadjagrāmo madhyamagrāmaśreti atrāśritā dvāviṃśatirutayah svaramaṇḍa-lasādhitāḥ” (N.S. 28, 24). Viśākhila, too, had placed śruti after grāma and his reasons for doing so (as Abhinava points out) were similar to those of Bharata (see A.B. on N.S. 28, 21).

But Mataṅga like Dattila, evidently, believed that śrutis give rise to svaras and not vice versa. He describes śruti before taking up svara and, speaking of their relation, observes : “through the process of deduction (arthāpatti), inference (anumānā) and through direct perception, śrutis are found to be the cause that manifests svaras”:

arthāpattyānumānena1 pratyakṣajñānato'pi vā gṛhyante śrutayas tāvat svarābhivyaktithetavạḥ

Arthāpatti and anumāna were two standard pramāṇas or valid means of knowledge. The classic example of arthāpatti is the proposition 'pino devadatto divā na bhuṅkte' ('fat Devadatta does not eat during the day') leading one to deduce that he must be eating at night. The classic exampe of anumāna is the inference of fire at seeing smoke ('yatra dhūmastatrā-gniḥ'). Such inference was presented in logic through a five-step syllogism.

208 A Study of Dattilam

We see that Matañga presents his stand in standard philosophical terminology.

He gives many alternate views expressed by different ācāryas regarding the manner

in which śrutis became svaras. These views take into account different doctrines

regarding the relation of cause and effect current in Indian philosophy. Matañga

mentions five theories : tādātmyavāda, vivartavāda, kāryatvavāda, pariṇāmavāda and

vyañjanāvāda.1 Let us discuss these, in due order :

1. Some argued that śrutis were identical with svaras (svaraśrutyostu tādāt-

myam) : both svara and śruti, they reasoned, are grasped through the same organ of

hearing and our perceptive 'feeling' or experience (sparśa) of śruti and svara does not

cognise any thing which distinguishes them. Śruti and svara are thus identical, just

as a 'class' (jāti) is indistinguishable from the individual objects of that class (vyakti).

viśeṣasparśasūnyatvāccchravaṇendriyāgrāhyatā

svaraśrutyostu tādātmyam jātivyaktīyorivānayoh

(Br. 32).

This position cannot be said to express the primariness of śruti over svara.

2. According to others, svaras were a 'vivarta' of śrutis (just as Śankara

Vedantists believe the world to be 'vivarta' of Brahma). When the image of a person

is seen in the mirror, the reflected image is a vivarta ; similarly svaras are perceived

in the śrutis as their vivartas :

naraṇām ca mukham yadvad darpaṇe tu vivartitam

pratibhāti svarastadvacchrutiṣveva vivartitaḥ

(Br. 33).

The standard image employed by the Śaṅkara Vedantists in expressing vivarta

is the illusory perception of a snake on suddenly beholding a piece of rope—the

snake being the vivarta of the rope. In Mataṅga's example, however, the mirror

reflection, though unreal, is not totally illusory; it is not a false perception entirely in

the mind of the perceiver.

3. In the third view, śrutis 'cause' the svaras just as a lump of clay and a

turning stick—which is used to turn the potter's wheel and is hence an instrumental

cause—cause an earthen pitcher. This was kāryatvavāda :

śrutīnām śruti (? svara)2 kāryatvamiti kecid vadanti hi

mṛtpindadaṇḍakāryatvam ghaṭasyeha yathā bhavet

(Br. 34).

4. According to the pariṇāma theory, svara was an evolute of śruti : svaras

resulted from śrutis just as curd results from milk. Pariṇāmavāda was a Sāṅkhya

doctrine :

śrutayaḥ svararūpeṇa pariṇamanti na saṃśayaḥ

pariṇameta yathā kṣīram dadhirūpeṇa sarvathā

(Br. 35)

1 तादात्म्यं च विवर्तञ्च कार्यत्वं परिणामिता । अभिधेयैकता चापि श्रोतॄणां परिकल्प्यते ॥

—Br. 31.

2 Br. reads श्रोतॄणां श्रुतिकार्यत्वमिति etc., obviously a faulty reading.

5. According to the fifth view, svaras were manifested or revealed through the śrutis just as objects like ghaṭa (pitcher) etc. placed in the dark are revealed by the light of a lamp :

ṣad jādayah svarāḥ sapta vyajyante śrutibhiḥ sadā andhakārashitā yadvat pradīpena ghaṭādayah

(Bṛ. 36).

The light does not bring into existence the objects in the dark but without light these objects are bound to remain uncognised ; so from the point of view of perception and cognition, light ‘causes’ these objects by revealing them. Similarly, without śrutis, svaras could not be revealed. Mataṅga himself subscribed to abhivyañjanāvāda, though he also favoured pariṇāmavāda (pariṇāmābhivyaktistu nyāyyah pakṣaḥ satāṃ mataḥ ; Bṛ. 45).

The other views (vikalpas : alternate hypothesis) are refuted by him.1 The tādātmya hypothesis does not hold ground, he says, because svaras and śrutis are objects of distinct and separate perception : they are clearly cognised as different. Moreover, the one depends upon the other and hence they cannot be identical.2 The vivarta theory is also erroneous because if svaras were vivartas of śrutis, then the perception of svaras would have been an illusory perception because the vivarta resulting from a real cause is itself a nonexistent entity.3 The kārya doctrine is also inapplicable, because although a causal relation does exist between śrutis and svaras, it is not of the kind that exists between a lump of clay and a pitcher. When a lump of clay turns into a pitcher, the clay becomes non-existent and we have only the pitcher, but such is not the case with śruti and svara.4 This very argument can also be directed against pariṇāmavāda, a theory which Mataṅga seems to have approved of and does not refute. Mataṅga’s complete approval, however, was in favour of abhivyañjanāvāda and he concludes the discussion by stating that the theory which propounds that śruti manifest the svaras is, from all angles, the most satisfactory one (Bṛ. 53, quoted above).

Abhinava does not go into philosophical theories of cause and effect, but contrary to Mataṅga, argues that svara was the more primary entity and it was, indeed, svara which was the musically significant interval within an octave. It is svara, he reasons, which has the inherent quality of charm and musical appeal, not śrutis which are dependent on svaras.5 Svara, according to Abhinava, had the quality of anuraṇana, which śruti lacked.

1 इति तावन्मया प्रोक्तं श्रोतृणां च विलक्षणम् । इदानों समप्रवक्ष्यामि विलक्पस्य च दूषणम् ॥

— Bṛ. 37.

2 नानात्वाद्‌विप्रसाधितत्वात्‌ स्वरश्रुत्योः सुथ भिन्नता । आश्रयाश्रयिमेदाच्च तादात्म्यं नैव सिध्यति ॥

— Bṛ. 39.

3 यद्‌मात्रि विवर्तत्वं श्रुतीनां तद्‌वसतम् । विवर्त्ते च स्वराणां ज्ञान्तिज्ञानं प्रजायते ॥

— Bṛ. 40.

4 कार्यकारणभावस्तु स्वरश्रुत्योः सम्भवेत् । श्रुतीनामिळ सद्‌भावे प्रमाणं नैव विद्यते ॥

कार्यकारणदोःस्ति यद्‌अपि श्रुटरपे हि । तथापि करणत्वं श्रुतीनां नैव सम्भवेत् ॥ कार्येपे विद्यमानेऽपि कारणस्योपलम्भनात् । घटादौ विद्यमाने तु मृत्पिण्डो नोपलम्यते ॥

5 For some relevant remarks of Abhinava on this point, see ch. I.

The notion of anuraṇana—‘resonance’—is not very clear in this context, but apparently it was a notion connected with the laws of accoustics : the harmonic relations existing naturally between certain sounds. After making a sound if there arises another sound which has any harmonic relation with the first, then the two sounds will acquire a resonant quality. This seems to be what Abhinava meant by anuraṇana.

The octave itself is born of the harmonic principle and has many possible graded intervals, the śrutis, but not all of these have the resonant qualities inherent in svara. Only intervals separated from each other by a certain measure of śruti-interval have a naturally pleasing effect : these are the svaras. Thus some theorists in Abhinava’s days opined that “when the intervening śrutis (between two sounds) consist of a certain definite number, then on the final śruti if a sound is made through the friction of the breath (in singing), this results in svara; this consists of the quality of charm and pleasingness which belongs to that particular śruti position. Svarā imparts colour (i.e., musical significance’ in this context) to the śruti on which it rests (tasyāśrayabhūtāyāḥ śruteruparajñakah)”1. Svarā according to this view, was the pleasing quality that certain śruti positions naturally possessed.

Abhinava was in favour of somewhat amending this view and himself considered anuraṇana as the basic characteristic of a note : “the sound, consisting of anuraṇana, charming and sweet, which is produced as an effect of that sound which results on striking a (specific) śruti position, is svara”2. Thus svara according to Abhinava was not just a sound but an echo-like resonant, secondary sound which was the characteristic of certain śrutis and it was this resonance which produced the pleasant and charming sensation that the musical notes arouse in us.

Śrutis according to this view were secondary to svara (they were svarāśrayāḥ : i.e., ‘dependent upon svara’) ; they were pitch-gradations within the interval of a svara : “śrutayo hyuccanīcatayā api svarāśrayā eva pratīyante” (A.B. on N.S. 28, 21).

Though, doctrinally, Mataṅga has given primacy to śruti over svara (believing as he did that śrutis gave rise to svaras) yet in his view, too, svaras and not śrutis were of primary musical significance. Śruti he defines as a sound that is audible (Br. 26-27), but svara is much more. Svarā, Mataṅga says, has the quality of dīpti—‘splendour’, ‘beauty’, ‘illumination’. Etymologically analysing and defining the word svara, he further says that svara is an entity which shines forth on its own : this is another way of saying that svaras are ‘svayambhū’ entities, born of the very laws of sound :

tatrādau svaraśabdasya vyutpattiriha kathyate rājñ dīptāviti dhātoḥ svaśabdapūrvakasya ca svayaṃ yo rājate yasmāt tasmādeśa svaraḥ smrtaḥ (Br. 63-64).

1. तत्रान्तरालश्रुतिषु नियतसङ्ख्यैकत्व ईह्यचो वायुमन स्पर्शयत्सु महिम्ना यः स्वरः श्रुतिस्थाने स तिङ्गत्स्वररक्षकलक्षणो धर्मः तस्याश्रयाभूतताया श्रुतेरुपरञ्जकः स एव स्वर इति केचित् । —A.B. on N.S. 28, 21.

2. वर्ण तु श्रुतिस्थानमाभ्राप्तप्रभवस्खलदप्रभवितोद्गुरणनादमा दिनङ्गमधुरः शब्द एव स्वर इति वद्यामः । —A.B. ibid.

The Vṛtti on these lines comments that “svara is the sound which produces rāga” (rāgajanako dhvanih svarah). Taking the pun on rāga to be intentional, the meaning would be that svaras are sounds that arouse our affections or emotions and also are the basis of organised melodic structures. The Vṛtti quotes Kohala who had defined svara as an ‘affective sound’ (dhvani raktaḥ svaraḥ ; Vṛtti on Br. 64A). A śruti does not have this significance.

The nature of svara—like that of śruti—had given rise to some metaphysical questions : was svara, in the ultimate analysis, one or many (eko’neko vā) ? Was it perishable or perennial (nityo’nityo vā) ? The Vṛtti on Brhaddesī raises these questions and answers them by stating that svara is both one and many ; it is pervasive and perennial. Svarā, it says, in its indivisible—niṣkala—form is one, but becomes many in the form of individual notes like ṣadja etc.1 Svarā, the Vṛtti continues, is perennial because it is indestructible ; in other words, the Vṛttikāra thought that svara was not a ‘thing created’ and thus could not be destructible like created things. It was something which existed in the very nature of things and was nitya—without a beginning and end. Man did not create svara but only perceived this everexistent phenomenon and with its help created music.

Svara is also pervasive because it is universal (sarvagataḥ). This perhaps means that svara is not a mere subjective phenomenon perceived differently by different people but a universal entity perceived the same by all. Svarā is, indeed, basically dependent upon the laws of accoustics and its perception has therefore an objective, universal basis. The Vṛtti quotes a verse by Kohala which describes svara as vyāpaka2 which in the context, evidently, means ‘universal’.

In gāndharva there were seven svaras in an octave (svarāḥ ṣadjādayah. sapta ; Datt. 11) : ṣadja, rṣabha, gāndhāra, madhyama, pañcama, dhaivata and niṣāda. There were in addition two intermediary or auxiliary—sādhāraṇa—notes, namely, antara gāndhāra and kākali niṣāda. These latter two were not accorded the full status of svara : antara gāndhāra was a subsidiary of gāndhāra and kākali niṣāda that of niṣāda. Abhinava, commenting on N.S. 28, 21 where the seven notes have been named,3 says : “notes are declared to be seven in number ; by this statement Bharata reveals that kākali (niṣāda) and antara (gāndhāra) are not separate notes—sapta iti vacasā kākalyantarā svarau na pṛthak svarā viti darśayati” (A.B. on N.S. 28, 21).

1 Ālaṅkārikas had an analogous motion regarding the nature of ‘rasa’. Rasa is ultimately one, they said, but becomes many in the form of śṛṅgāra, karuṇa etc.

2 ननु स्वर एव स्यादेको वा व्यपको वा नित्यो निल्यो वा (स्यात् ?)। अनोच्यते—एकोऽनेकको व्यपको नित्यश्चेति। तथो निकलरूपेणकः स्वरः—नादैकरूपत्वाल्मति कृतो योगरस्वरपि । नित्योऽपि विनाशो । व्यपकः सर्वगतं । तथा चाह कोहलः— ऊध्वान्च्चादिप्रयत्नेन सर्वेभतिनि:षट्टनादतु । मूर्छतों ध्वनिरमूर्छनः स्वरोऽप्यको परः ॥ —Br. Vṛtti on 64A.

तत् स्वराः—पड्जश् ऋषभश्चैव गान्धारो मध्यमस्तथा । पञ्चमो धैवतश् चैव सप्तमो निपादवान् ॥ —N.S. 28, 21.

In tuning, the pitches of the svaras were apparently arrived at through the ear as in current practice. Śruti, the minute interval within a svara, could not be thus arrived at. It required a more elaborate procedure. Bharata has described the process of tuning at some length through which śruti could be determined. The process involved two vīṇās, equal in all respects : in size, in the number and thickness of strings, in upavādana (the 'mizrāb' or plectrum for striking them ?) and the succession of notes. The two vīṇās were then identically tuned to the ṣadja-grāma: "dye vīṇe tulyapramāṇatantryupavādanadaṇḍamūrchane kṛtvā ṣadjagrāmāśrite kārye" (N.S. 28, 27).

Bharata gives no method or process for arriving at this initial tuning. Evidently, it was done through the ear and not through a process employing mathematical ratios (such as is given by later theorists like Abhobala, Śrīnivāsa etc.) Bharata's attitude was in this respect empirical. It was an attitude which prevailed in all early musical thinking. No early text makes an attempt to numerically assess the magnitude of the svaras in any way; all assume their positions as given. The Vṛtti on Bṛhaddesī in describing the measure (māna) or magnitude of a śruti also assumes the position of the svaras themselves as known and merely reproduces Bharata's passage on this subject with minor changes (Br. Vṛtti on 28).

In ancient Greece, early theorists, chiefly Pythagoras (6th Century B.C.) had arrived at numerical ratios for the position of various notes: tonal intervals were demonstrated on a single stretched string in terms of spatial divisions.1 However, other theorists, especially Aristoxenus (born 350 B.C.), favoured the empirical method. Introducing the subject of harmonics (the study of musical sounds) he observes that the subject of the study is the question: "in melody of every kind what are the natural laws according to which the voice in ascending and descending places the intervals ?" In reply he asserts that the voice (meaning, of course, the musical voice) spontaneously arrives at the right interval according to an inexorable law of sound. "For we hold", he says, "that the voice follows a natural law in its motion and does not place the intervals at random." Further, he observes that the ear is a judge which cognises the measure of true musical intervals. He states: "our subject-matter being all melody, whether vocal or instrumental, our method rests in the last resort on an appeal to the two faculties of hearing and intellect. By the former we judge the magnitudes of intervals, by the latter we contemplate the functions of the notes." Recognising that an inexorable natural law is the basis of 'svara', Aristoxenus remarks that "there is a certain marvelous order which belongs to the nature of harmony in general; in this order every instrument, to the best of its ability, participates under the direction of that faculty of sense-perception on which they, as well as everything else in music, finally depend."

In tuning an instrument, too, it was this sense-perception, the sensitive ear, which was to act as the guide. "No instrument", he asserts, "is self-tuned and that the harmonizing of it is the prerogative of sense-perception is obvious and requires

See the Pelican History of Music, Vol. I, p. 100.

no proof." A student of musical science according to Aristoxenus was not like a

physical scientist who has to depend upon extraneous measures for judging his

observations. A geometrician, for example, uses phrases such as 'let this be a

straight line', but he makes no use of his faculty of sense-perception in judging a

straight line as such: 'He does not in any degree train his sight to discriminate the

straight line, the circle or any other figure." Such a discriminative training belongs

rather to the craftsman, the artist : people such as "the carpenter, the turner or some

other such handicraftsman", whose vocation involves the actual construction of geo-

metric figures made through the trained sense of sight. The student of music, too, has

to be similarly perceptive. Aristoxenus says: "for the student of musical science,

accuracy of sense-perception is a fundamental requirement."1

Though Bharata and other early Indian theorists have not explicity stated the

idea in so many words, they, too, were evidently empiricists when it came to judging

musical intervals. They had the same student in mind whom Aristoxenus considered

as ideal : a student with a developed and descriminatively trained ear for musical

intervals. Manuals such as those of Bharata, Dattila, Matañga and others were

written and studied within sampradāyas—schools of musical training—and a basic

knowledge of intervals as well as a trained faculty for preceiving them was, evidently,

taken for granted in such students or musicians as were considered fit to study the

writings of the great ācāryas.

It was quite late in the history of saṅgīta literature that musical intervals were

given in terms of spatial measures on a string. Śārṅgadeva is the first to speak of

svara-vīṇās, on the cross-bar (daṇḍa) of which the location of each svara was marked

and the magnitude of the intervening śrutis was similarly indicated. Vīṇās such as

the ekatantri, nakula, tritantikā, 'citra, vipañci, mattakokilā, etc. were all svara-vīṇās.2

They seem to have been both of the lute and the harp varieties. In describing the

construction of vīṇās, Śārṅgadeva notes distances between frets on the basis of spatial

intervals between svaras at different string points (see S.R. ch. 6) ; but the

measurements and the svaras they represent are not very clear. The matter is more

clear in works like the Rasakaumudi of Śrīkaṇṭha (C. 16th-17th century) which gives

exact numerical ratios.3 Other such works are the Rāgatattvavibodha of Śrīnivāsa (see

verses 36-49 of this work) and the Saṅgītapārijāta of Ahobala.4

1 Quotations from Aristoxenus' Harmonic Elements have been taken from Source Readings in

Musical History by Oliver Strunk. The editor's introductory note on Aristoxenus has these

lines : "Aristoxenus' thought has a distinct empirical tendency. Aristoxenus also holds that the

notes of a scale are to be judged not by mathematical ratio, but by the ear."

2 वध्यते स्वरषड्जादि तस्यामपि विचक्षणा । वद्धिकृत्वा स्वरदेशानां भागानुदिशन्ते श्रुतीः; ।

तद्भेदवैचित्र्यस्थानेकुलश्र वीतनान्तिका । चित्रा वीणा विपश्ची च ततः स्याद्वाद्यकौतुका ।।

ब्रालापिनी किन्नरी च पिनाकी तत्ररा । निःश्रृङ्खलवेणुज्याघातैर्यान्ति देवेन कौन्तिताः ।।

—S.R. 9, 8-10.

3 For a study of these ratios see the monograph, A Comparative Study of the Leading Music

Systems of the 15th, 16th 17th and 18 Centuries by Pandit V.N. Bhatkhande.

4 ibid.

Let us now return to the two identically constructed vīṇās of Bharata. Both were tuned to the seven notes arranged according to ṣadja-grāma. One vīṇā was taken as the constant vīṇā and was left unticcuched (this was the dhruva or the 'still' vīṇā) while the pañcama string of the other vīṇā (termed the cala-vīṇā) was loosened so that the string now sounded just a little lower than the pañcama string of the dhruva-vīṇā; this lowered pitch was tuned to the pañcama of the madhyama-grāma which was on the third śruti from the madhyama while the pañcama of the ṣadja- grāma lay on the fourth śruti from the madhyama. This gave one the measure of a śruti. Then with this lowered pañcama as the fulcrum, all other strings of the cala- vīṇā were also lowered so that they were returned to the ṣadja-grāma, but at a pitch one śruti lower than the dhruva-vīṇā.1 This process has been outlined by Bharata for giving the magnitude of a śruti interval: "tayoranyatarasyāṃ pañcamasyāpakārṣe śrutimadhyamagrāmikīṃ kṛtvā tāmeva ca pañcamasyā śrutyutkarṣavaśāt ṣadjagrā-mikīṃ kuryāt. ekāśrutirapakṛṣṭā bhavati" (N.S. 28, 27).

One cannot but conclude that the śruti-interval like the svara-interval was also arrived at through the ear. The basis for lowering the pañcama of the ṣadja- grāma so as to make it the pañcama of the madhyama-grāma was clearly an empiri- cal one in which the musician relied upon his ear. Only a person who could empirically distinguish the two grāmas could have applied Bharata's process.

There were, however, certain distinct harmonic properties of the slightly different octaves of the two grāmas, which must have aided a musician in applying the process. The ṣadja and ṛṣabha (besides other notes) had an identical śruti-value in both grāmas but there was the difference of a śruti between the pañcamas of the two grāmas and as a result a samvāda or harımonical concordance existed in the ṣadja- grāma between ṣadja and pañcama ard not between ṛṣabha and pañcama, while in the madhyama-grāma, pancama was a samvādi not of ṣadja but that of ṛṣabha. The other samvādas in both the grāmas were identical (N.S. 28, 23). Thus, in arriving at the interval which was termed a śruti, all one had to do was to lower the pañcama of the ṣadja-grāma with its ṣadja-pañcama samvāda so that it now had a ṛṣabha- pañcama samvāda. The magnitude of pitch intervening between the two pañcamas constituted a śruti. But this assumed a knowledge of the correct pitch of the ṛṣabha which itself must have been arrived at through the ear.

This ṛṣabha—unlike our own śuddha ṛṣabha in most rāgas of Hindustani music—had no ṣadja-madhyama samvāda with the pañcama, but was apparently somewhat lower.2

1 स्वराणां कस्यचिदपि वैलक्षण्य यत्स प्रतिभासनात्तत्का ध्रुवकस्यान्योर्वा वीस्वामपरस्यां पञ्चमतन्त्र्यौ श्रुति- मात्नं शिथिलीकार्य । तदा मध्यमग्रामो जायते । अनन्तरं मध्यमग्रामोः समस्वराणां श्रुतिमेकां पातयेत्, येन षड्जग्राम एव भवति । एवं ध्रुववीणायां चालयीणायां र्श्रृङ्गारेणैष मध्यमस्य श्रुतिरदृश्यते । तनू च प्रयक्षेणानुसन्धानबाललक्ष्यीया भवति । -A.B. on N.S. 28, 27-28.

2 Acārya Bṛhaspati in his Bharata Kā Sangita Siddhānta has worked out an ingenious method on the basis of accoustics, to arrive at ite ancient ṛṣabha (p 10-11). It is, of course, impossible to really judge if the value he has arrived at is the correct one. For, as Bṛhaspati himself says : "the knowledge of śrutis, of their magnitude and succession is derivative of the knowledge of grāmas and cannot itself result in such a knowledge. Bharata has himself considered only such a person fit for studying his tunings (tāranaṅs) who is fully cognisant of the form and nature cf the two grāmas," (p. 15). The two grāmas themselves are now a closed book.

Further lowering of the cala-vīṇā in relation to the dhruva-vīṇā revealed the measure or pramāṇa of the two-śruti, three-śruti, four-śruti and still larger śruti-intervals. The process was as follows : After all the svaras of the cala-vīṇā had been lowered to the extent of one sruti, the gāndhāra and niṣāda of the cala-vīṇā, now only slightly above the ṛṣabha and dhaivata of the dhruva-vīṇā, were so lowered that they reached the same pitch as the ṛṣabha and dhaivata of the dhruva-vīṇā. The rest of the vīṇā strings were again lowered to conform with the new positions of ri-dha, This second sāraṇā—as this process of tuning was called—gave the magnitude of the two-śruti interval: “punarapi tadvadevāpakarṣet yathā gāndhāraniṣādavantāvitaraśyāmṛṣabhadhaivatau pravekṣyataḥ dviśrutyadhikatvāt” (N.S. 28, 27), This sāraṇā, says Abhinava, clearly reveals to the perception the magnitude of each of the two śrutis contained in the two notes ga and ni; because when lowered by two srutis, as shown, these notes become one with ri and dha respectively.1

In the third sāraṇā the cala-vīṇā was so lowered that ri and dha (both three-śruti notes) of this vīṇā, which were now only a śruti above sa and pa respectively of the dhruva-vīṇā, became one with these latter notes : “punarapi tadvadevapakrṣṭāyāṃ dhaivatarṣabhavitarasyāṃ pañcamaṣadjau pravekṣyataḥ triśrutyadhikatvāt” (N.S. 28, 27). The process now revealed the magnitude of three-śruti intervals in three steps of a śruti each.2 The rest of the strings of the cala-vīṇā were again lowered to conform to this new position of ṛṣabha and dhaivata.

In the final sāraṇā, the three notes pañcama, madhyama and ṣadja (all four-śruti notes) of the cala-vīṇā (which already stood three śrutis lower than these same notes on the dhruva-vīṇā) were further lowered and made one with the notes madhyama, gāndhāra and niṣāda, respectively, of the dhruva-vīṇā. Thus the four-sruti interval was revealed.

This sāraṇā process as given by Bharata has found an echo in numerous subsequent works. The Vṛtti on Bṛhaddeśī almost paraphrases Bharata.3 Śārṅgadeva, Kumbha and others have based their exposition on Bharata's passages.4

In the sāraṇā process we see that the svaras form the initial 'given' framework within which the position of the śrutis is then arrived at; hence it was held by certain theorists that śrutis depended upon svaras and could be observed as fractional positions within the svaras (śrutayo hyuccanīcatayā api svaraśryā eva pratīyante, A.B. on N.S. 28, 21).

1 गान्धारनिषादं प्रत्येकं यच्छु. तिद्रयं तचच धैवतर्षभंमिश्रीभावाभिज्ञानेन स्कुटमेवोपलभ्यधमिति तच्ु. लुतय: प्रत्यक्षीभूता भवन्ति। —A.B. on N S. 28, 27-28.

2 पुनरपि चलवीणायां श्रुतियंद्वापकर्ष्यते । सर्वस्वरैरमस्तदा च वीणागतो धैवतर्षभो ध्रुववीणागताम्यां पंचमषड्जाभ्यां यथाक्रमं साम्यं गच्छति। यतो ध्रुववीणायां स्वरास्ततः: श्रुतिभिरधिकाः। एवं स्वरद्वये प्रत्येकं श्रुतिद्वयं श्रुटीकृतम् । —A.B, ibid.

1 The description is introduced with the words : ननु श्रुते: किं मामू । उच्चते । —Vṛtti on Br. 28.

2 S.R. 1,3, 17-22 and the two commentaries ; S. Raj 2,1,1, 95-114. (Kumbha has a more detailed exposition).

The ancients apparently believed that all the twenty-two śrutis were of an

equal measure : each successive sāraṇā on the cala-viṇā is said to lower it by one

śruti-magnitude, the magnitude itself, supposedly, remaining unchanged. Had each

śruti been equal, the ancient octave would have constituted a tempered scale, divided

into twenty-two equal divisions (with the seven notes located on given śruti positions).

Yet the very sāraṇā process belies the notion of a tempered scale. The process was

effected empirically through the ear which is attuned to natural harmonies on which

the process relied.

Each successive ləwering of the cala-viṇā could not then have traversed an

identical interval in pitch, for this is not a natural division and can only be achieved

through the help of an instrument (like the tuning fork) or an externally imposed

temperate numerical ratio in terms of string lengths or a similar device. The śruti

was not arrived at in this way.

Mātṛgupta (5th century A.D.?), an early predecessor of Abhinava, had opined

that both svaras and śrutis were born ‘spontaneously’, that is, through a natural

harmonic process. Matṛgupta, like Abhinava, believed that śrutis depended upon the

svaras yet he thought that srutis were ‘sahaja’ (naturally born) intervals like svaras.

Abhinava quotes him as saying :

(jāyate) sahajenaiva samastaḥ śrutivistaraḥ

svatāddhiṣṭhānato vāti śrotrāpiyūṣasāratām

(A.B. on N.S. 28, 21)

"The entire range of śrutis is produced spontaneously. The ear comprehends

the beauty of their essence on the basis of svaras."

Bharata's process depended upon the tone-sensitive ear and the ear dces not

divide the octave into twenty-two equal intervals, for such a division is alien to the

accoustics of sound.1 Yet Bharata's process (which must have been used in actual

practical demonstrations) does reveal twenty-two gradations of the octave. Let us

try to arrive at the probable values of the sruti-intervals that his sāraṇās measured.

The ancient ṛṣabha was a three-śruti note and had no harmonic concordance

with the pañcama when tuned to the ṣadja-grāma ; but the ṣadja in this grāma had

an accoustic affinity with the pañcama. Now, the natural ṛṣabha, born of the second

harmonics of a series of ascending fifths, does have a very close affinity with the

pañcama that is harmonious with the ṣadja. The natural ṛṣabha consists of a major

tone (expressed through the ratio 9 : 8) : this was probably the four-śruti interval of

the ancients. The ancient ṛṣabha was a śruti lower. What could be the probable

value of the ancient ṛṣabha? It must have been an harmonic note because, as we

have argued, notes were arrived at through the ear. So the ancient ṛṣabha must have

consisted of an interval somewhat smaller than the major tone and yet it must have

been a naturally harmonic interval. The difference between the ṛṣabha consisting of

1 Alain Danielou in his Introduction to the Study of Musical Scales arrives at a division of the

octave into fifty-three intervals through a ha.monic process, ‘allowing us to play accurately, i.e.,

without beats, all the usual harmonic intervals’ (pp. 45-46). These 53 inter vals, too' are not

equal (see ibid, chart at pp. 237-241).

the major tone (harmonic to the pañcama) and the three-śruti ancient ṛṣabha should give us the measure of a śruti, if we can arrive at a probable value.

For expressing measure we shall use savarts. The French scientist Savart had divided the accoustic length of the octave on the basis of the decimal sytem into a measure which has been named savart after its initiator. The octave is divided according to this system into 301.03 savarts, or for practical purposes, 301 savarts.

Now, before proceeding, let us state that we consider the ancient three-śruti ṛṣabha as consisting of the minor tone, a harmonic interval somewhat smaller than the major tone : this is very plausible since the minor tone is a very frequent harmonic interval in music (ratio to the tonic 10 : 9).1 If this be granted then the difference in measure between the major and the minor tone should give us the magnitude of a śruti.

The difference between these two tones comprises approximately 5 savarts, equal almost to the Pythagorean comma. But with this interval a division of the octave into 22 śrutis is patently impossible.

Yet a measure of less than a minor tone for the ancient ṛṣabha would not have formed a harmoniously satisfactory tone. Still, if we assume that the interval called the small tone (a not very satisfactory harmonic tone with a ratio to the tonic approximately 11 : 10) was the three-śruti ṛṣabha, then, too, the resultant śruti measure amounts to only about 10 savarts, still too small a measure for a 22 śruti scale.

An interval lesser than the small tone seems implausible, because natural intervals lower than the small tone are the two half-tones—large and small—which give an extremely improbable value for the ancient ṛṣabha.2 This only strengthens our inference that the śruti was not an equal measure.

We think that the interval called the pramāṇa śruti—the initial śruti measure— was indeed a very small one, equal to the comma, roughly 5 savarts. The second sāraṇā, however, encompassed a much greater interval, for now the gāndhāra and niṣāda—which, evidently, consisted of the harmonic half tone intervals—were lowered to the level of ri and dha respectively (gāndhāra, the harmonic minor third, has a ratio 6 : 5 with the tonic and niṣāda, the minor seventh, has a ratio 9 : 5).

In the first lowering, the minor third was, presumably, lowered by 5 (or may be 10 savarts) but the second sāraṇā through which gāndhāra became one with the ṛṣabha must have involved a lowering of at least 28 savarts : the distance between the minor third lowered by either 5 or 10 savarts (in the first sāraṇā) and the minor tone or the small tone (the probable values of the three-śruti ri) respectively.3 Other lowerings were similarly unequal.

1 This is the tone composed of the natural interval which occurs, for example, between the śuddha ri and śuddha ga of Hindustani music.

2 Most authorities are agreed that the ancient ṣadja-grāma was tuned to something like the ‘kāfi-scale if ṣadja be taken as the tonic; the ṛṣabha, then, could not have been too low.

3 In calculaing savart-distances we have followed the well-planned chart given by Alain Danielou in his Introduction to the Study of Musical Scales, pp. 237-241.

In Saṅgīta Cintāmaṇi (Hindī) by Acarya Bhaspati and Srimati Sumitra Kumari, the authors have propounded the theory (on the basis of Bharata’s sāraṇā process and Abhinava’s exposition of it) that there were three basic śruti measures: (1) the pramāṇa śruti; this according to the authors measured 5 savarts (these he calls ‘ghaṭakas’ 301 of which equal an octave), (2) the upamahati śruti measured 18 savarts and (3) the mahati śruti measured 23 savarts (see p. 105; the whole discussion is very interesting and fruitful).

TOPIC 3

GRĀMA

11A. grāmau dvau ṣadjamadhyamau

B. kecid gāndhāramapyāhuh sa tu nehopalabhyate

12A. ṣadjatvena grhīto yah ṣadjagrāme dhvanirbhavet

B. tata ūrdhvam trtīyah syāt ṛṣabho nātara samśayah

13A. tato dvitīyo gāndharaścaturtho madhyamastatah

B. madh yamāt pañcamastadvat trtīyo dhaivatastatah

14A. niṣādo'to dvitīyaḥ syāt tataḥ ṣadjaścaturthakah

B. pañcamo madhyamagrāme madhyamādyaṛtīyakakah

15A. evam dhvaniviśeṣan yah sarvān ṣadjādi (samȳinitān)

B. vyavasthitāntarān vetti sa vetti svaramaṇdalam

(There) are two grāmas : ṣadja and madhyama. Some speak of a gāndhāra [grāma] ; but that is not to be found here [in this world].

In ṣadja-grāma, commencing from the sound accepted as ṣadja (ṣadjatvena grhītah), the third higher sound (or śruti) is ṛṣabha ; this is beyond doubt.

From [ṛṣabha] the second [higher sound] is gāndhāra ; from gāndhāra (tatah) the fourth is madhyama. Similarly [four śrutis higher] from madhyama is pañcama, from pañcama (tatah) dhaivata is the third [higher śruti], from which niṣāda is the second. And from niṣāda (tatah) the fourth [higher śruti again] is the ṣadja.

In the madhyama-grāma, pañcama is the third higher śruti] commencing with madhyama.

Thus one who knows all the specific sounds beginning with the ṣadja with their established intervals, comprehends the svara-maṇḍala (the entire gamut of notes).

NOTE :

Nārada in his Nāradī Śikṣā speaks of the gāndhāra-grāma. He, too, like Dattila says that the gāndhāra-grāma does not exist in this world but is to be found only with the gods.4 Perhaps it was once current during a very early period. It

1 Datt. T. ed reads नात् संशय:, evidently a misprint. Sudhā reads नात्र संशय:.

2 Datt. T. ed reads (द्विती)य: (स्त ? स्वात्) ।

Sudhā on S.R. 1,4, 15-16 quotes Datt, 12-14A ascribing them to Dattila. The reading givenis the same as above. Abhinava quotes the major part af Datt. 12A with the words:दत्तिलाचार्योऽपि घ्वनिविशदेनानुसृत्यनमेवाऽऽङ्ग्यधात् । 'पडजत्वेन गृहीतो य: पडजग्रामे घ्वनि:' ॥

-A B. on N.S. 28,21,

4 स्वर्गोंऽनुग्राम्यतां धारो ।

Nāradī Śikṣā 1,2,7.

Mataṅga attributes the mention of gāndhāra-grāma to Nārada and adds that this grāma is not to be found in the world of men ; Br. 91.

had definitely gone out of use by the age of Dattila and Bharata. Bharata does not even mention the gāndhāra-grāma. Of grāmas he says : "atha dvau grāmau ṣadjagrāmo madhyamagrāmaśca" (N.S. 28, 24).

The notion of grāma is difficult for us to understand for we no longer have grāmas. The two grāmas, says Mataṅga, were two 'groupings' consisting of certain specific arrangement of svaras measured in śrutis : "samūhayācinau grāmau svaraśrutyādisamyutau" (Br. 89).1 Let us consider how svaras and their śrutis were grouped in the two grāmas.

In the ṣadja-grāma the arrangement of svaras and śrutis have been indicated by Dattila in detail. He then points out the slightly variant madhyama-grāmic arrangement. Beginning with ṣadja as the initial sound, the note rṣabha occurred on the third higher śruti. Gāndhāra occurred on the second higher sound from rṣabha, madhyama on the fourth higher sound or śruti from gāndhāra. Pañcama was on the fourth śruti beginning with madhyama in the ṣadja-grāma, but in the madhyama-grāma, the placing was different ; here pañcama occurred on the third śruti beginning with madhyama. Consequently, dhaivata in the madhyama-grāma was on the fourth śruti beginning with pañcama while in the ṣadja grāma dhaivata was placed on the third śruti from pañcama. This difference in the position and śruti-value of dhaivata occurred because in the madhyam-gram, pañcama, was lower (relative to the ṣadja-grama) by a śruti though dhaivata remained at an identical position in relation to ṣadja, rṣabha, gāndhāra and other notes.

Dattila has located the notes in terms of the distance in śrutis between one note and another; rṣabha, he says, occurs on the third śruti beginning with ṣadja and so on. Bharata expresses the fact in a somewhat different language. He recounts the positioning of notes in the two grāmas in terms of the number of śrutis 'contained in' or 'possessed' by a svara. Ṣadja, according to Bharata, contained four śrutis, rṣabha three, gāndhāra two, madhyama four, pañcama likewise four (three when in madhyama-grāma), dhaivata three and niṣāda two.2

The different ways of expressing the śruti-svara relation found in Dattila and Bharata is evidently connected with the different doctrines they held regarding the primacy of śruti or svara. Dattila thought that śrutis gave rise to svaras; hence he says that particular svaras occur on particular śruti located at certain śruti-intervals from each other. Bharata, on the contrary, held that svaras were the primary intervals and śrutis were secondary pitch positions one could perceive and demonstrate within a svara. He, therefore, describes svaras as containing or possessing a stipulated number of śrutis.

1 Abhinava puts this more succinctly in remarking: ग्रामोनाम विशिष्टश्रुतिकस्वरसमूहः — grāma is a group of notes each with a specified number of śrutis —A.B on N.S. 28, 64-65.

2 चतुःश्रुतिमंवेतु षड्जः शुयवमस्ति श्रुतिः स्मृतः । द्विस्रुतिशश्चैव गान्धारो मध्यमश्च चतुःश्रुतिः ॥ पञ्चमस्तद्वत् त्रिश्रुतिकः स्यात् तिथ्युतिश्र्च निषादः स्वरात् पड्जग्रामे विशिष्टभवेत् ॥ मध्यमग्रामे तु पञ्चमः श्रुत्यङ्कष्टः काव्यः । —N.S. 28, 25-26.

Viśākhila was of the same opinion as Bharata and he seems to have expressed his view on this point more explicitly. Abhinava quotes phrases from him where he clearly says: “śrutayah svarāntaragatāḥ: śrutis are contained within svaras” (A.B. on N.S. 28, 27-28).

Dattila has described grāma after śruti and svara, implying an hierarchical sequence of inter-relatedness in the same order: svara depended upon śruti and grāma upon svara. Mataṅga was evidently of the same opinion as Dattila. We have seen that he believed in the primacy of śruti over svara. Regarding grāmas he says that they are born of svaras : “svarebhyo grāmāsambhavāḥ” (Br. 92).

Bharata implies a different sequence of inter-dependence. He describes svaras first, then names the two grāmas and then says: “dependent upon these (two grāmas) are the twenty-two śrutis which are effected through the gamut of svaras.”1 Viśākhilācārya had also listed śruti after grāma2 and presumably expounded it in the same order, with evidently the same doctrinal scheme in mind. According to this view the primary function of the śrutis was to distinguish one grāma from the other : “grāma-vibhāgārthamēva śrutikīrtanam” (A.B. on N.S. 28, 21). Svaras were self-evident entities, and they were arrayed in two grāmas on the basis of a slightly different śruti arrangement; the svara pañcama in one grāma had a slightly different tonal value from its counterpart in the other grāma (so did, resultingly, the dhaivata), and one of the chief functions of śruti was to measure and analyse the exact distinction between the two pañcamas in the two grāmas. Thus the notion of śruti was secondary to that of grāma.

We have noted (see ch. I) that some commentaries were written on the Dattilam—a fact not surprising for such an important and pithy work as this. The commentary called the Prayogastabaka is quoted by Simhabhūpāla on (S.R. 1-4, 15-16). The quoted portion concerns Datt. 12 and is significant here: “ṣadjātvena ṣadjasvarābhavēna gṛhitāḥ parikalpitō buddhyā vyavasthāpitō yaḥ kāścid dhvaniviśēṣaḥ ṣadja-grāma iti syādṛṣabhaḥ iti.” This may be rendered as: “whatever the sound which is wilfully chosen and established as the ṣadja svara, from that specific sound in the ṣadja-grāma the third higher (śruti) is the ṛṣabha”.

Evidently, (as Datt. 12 implies and this commentary elucidates) Dattila did not believe in the notion of absolute pitch as it has conventionally and customarily come to be fixed for each note in the Western system. His system was like the one which still persists in India where ṣadja is fixed at any convenient pitch and other notes are arrived at in relation to it. However—as will become further clear—there was one basic difference with the modern system: ṣadja once established was not the constant tonic, the fixed centre round which all other notes evolved and on the basis

1 अथ दो ग्रामौ षड्जग्रामौ मध्यमग्रामश्चेति । अवस्थितिता हि ध्वनिव्यतिस्वरुतयः स्वरमण्डलसाधिताः ॥ -N.S. 28, 24.

2 विशाखिलाचार्यों ग्रामांतरं श्रुतीरदधाति स्म। -A.B. an N.S. 28, 21.

of which they had their being. Sadja in the ancient system was no more important

than any other note and could also be dropped if necessary (see the topic ‘jāti’).

The grāmas with their twenty-two śrutis and the location in them of svaras in

terms of specific śruti-distances, were illustrated by ancient theorists in terms of certain

graphs. The Brhaddeśī is the first work where such graphs are described and drawn,

though they must have existed earlier. We do not know if graphs were current during

the period of Bharata and Dattila. They do not mention any. These graphs can help

us in understanding the notion of grāma.

The Vṛtti on Bṛhaddeśī speaks of three ways of representing śruti and svara

positions within a grāma. These were known as the three prastāras: (1) the daṇḍa-

prastāra, (2) the vīṇā prastāra and (3) the maṇḍala-prastāra.

1. The daṇḍa-prastāra is cryptically described as consisting of twenty-two

lines each representing a śruti: “tatra kecid daṇḍaprastāreṇa darśayanti dvāviṃśatay-

aḥ śrutayo rekhānāmiti” (Vṛtti on Bṛ 53). In this prastāra, apparently, the octave

was represented as a straight line (daṇḍa), across which twenty-two lines were drawn

at a right angle to it, each equidistant from the other. These represented the twenty-

two śrutis. Svaras were placed on the śruti-lines according to the prescribed śruti

distances. The two grāmas were represented on two different daṇḍa-prastāras. The

resulting illustration may have been somewhat as follows:

Fig. 1

Fig. 2

We have follwed Dattilla’s cue in taking ṣadja as the initial śruti. Theoretically,

however, any note could evidently have been placed on the first śruti and the prastāra

drawn accordingly. We have shown the daṇḍa in a horizotal position and śrutirekhās

as vertical. The position can be reversed.

2: The nature of the vīṇā-prastārā is not at all clear. Perhaps it was a method

by means of which the śruti positions were indicated through a chart drawn up like

the frame of a vīṇā with the twenty-two śrutis stretched across it as straight lines like

stretched strings. But we do not know. Some theorists illustrated the grāmas through

the vīṇā-prastāra: “anye tu vīṇāprastāramāhuḥ” (Vṛtti on Bṛ 53).

3. The Vṛttikāra himself had a preference for the maṇḍala-prastāra which

he describes at length: “vayaṃ punar maṇḍalaprastāraṃ brūmaḥ” (Bṛ. Vṛtti on 53).

In this prastāra five equidistant horizontal lines were crossed at right-angles by six

vertical lines, also equidistant. The resultant graph had twenty-two free points on lines

222

A Study of Dattilam

jutting out of a group of squares: “tathāhi-tiryagūrdhvaṃ ca pañca ṣaḍ rekhā ityek-ādaśa; ubhayato dvāviṃśatih.”1

On such graphs the two grāmas could be illustrated as follows:

(ṣaḍja-grāma)

Fig. 3.

(madhyama-grāma)

Fig. 4.

The Vṛttikāra does not say that the lines should be equidistant but the graph found in the Br. implies this-however faulty it might be in other ways.

The Bṛhaddeśī has described the placing of svaras on the graphs of the two grāmas at length (Br. 54-61). The Vṛttikāra has sketched the actual graphs for the two grāmas. As extant in the available manuscripts, however, the graphs are patently faulty. But they can be reconstructed through the textual description.1

Abhinava also mentions the prastāras. He does not name the vīṇā-prastāra but only the other two. "The forms (svarūpa)" he says "of two-śruti notes (gā, ni), three-śruti notes (ri, dha and, in madhyama-grama, also ma) and four-śruti notes (sa, ma, pa) becomes clear through prastāra, where alone they can be readily perceived. Some (theorists) show (svaras and śrutis) through the daṇda-prastāra with the help of twenty-two lines (rekhās). Others describe the maṇdala-prastāra which consists of five horizontal and six vertical lines and counting both ends (of each line) they together make twenty-two points".2

Abhinava has not drawn the diagrams.

Another theorist, Nānyadeva, who followed Abhinava within a century, describes the maṇdala-prastāra in detail (B.B. 67-80). He has described the gāndhāra-grāma along with the other two. His description of making the prastāra-graph is the most succinct that we have. According to his description the placing of the notes on the ṣadja and madhyama-grāma graphs is as follows :

(sadja-grāma)

ni

dha

sa

pa

ri

ga

ma

Fig. 5.

1 Ācārya Bṛhaspati has made a correct reconstruction in his Bharata kā Sangīta Siddhānta, p. 29, also p. 40.

2 द्विश्रुतिभिस्त्रिश्रुतिचतुःश्रुतिस्वरस्स्वरूपं च श्‍रुतेरियत्तā (निर्धāरणā) श्रुटीकृरणार्थ प्रस्तारादेव दृश्यते । तत्र दण्डप्रस्तारेण केचिद्विंशत्यधिकद्विश्रुतिरेखāणाम् । अन्ये तु मण्डलप्रस्तारमाहुः । तिर्यग्गूणैश्च पंक्तचतुपुरेखā इति येकार्श्रोभयतो द्वाविंशति : ।

-A.B. on N.S. 28, 27-28.

(madhyama-grāma)

ga

ri ma

sa pa

ni dha

Fig. 6.

We notice that the placing of the notes is on somewhat different points than in the previous graphs.1 Fundamentally, this makes no difference because starting with any point as representing any note whatsoever, we shall arrive at an identical śruti-ratio between notes in a given grāma. That there were no fixed, universally agreed points for particular notes on the graphs, is, in fact, indicated by a remark in the Vṛtti on the Brhaddesī. The Vṛttikāra, after sketching the the maṇḍala-prastāra, as described by Mataṅga—where the ṣaḍja-grāma begins with the initial establishment of the note sa and madhyama-grāma with ma—says that Bharata has shown the śruti-maṇḍala as beginning with the svara ri : “bharatas tu punarṣabhādi śrutimaṇḍalam darśayati” (Vṛtti on Bṛ. 61).2

The maṇḍala-prastāra graph reveals an almost circular allocation of the śruti-points—an arrangement which by its very nature does not give predominance to one point over another. Nānyadeva, in fact, uses the expression svara-cakra—‘a circular arrangement of notes’—to describe the maṇḍala-prastāra graph. He further qualifies the expression svara-cakra by saying that ‘it should be like an alāta-cakra’. An alāta-cakra, as we have seen (ch. 1) was a circular ring of fire formed by skilfully rotating a burning log of wood (alāta).3

Thus, in principle, any point could initially be taken to represent any note and the others then could be arranged according to the fixed śruti-distances between the svaras of a grāma. This principle of arrangement suggests that a grāma was

1 The graphs as described by Nānyadeva have been ably reconstructed by Sri Chaitanya P. Desai, editor of the B.B. (chs. 1-5) who has written extensive notes in Hindi, on the text.

2 We do not know what remark in the Nāṭyaśāstra led to this conclusion. Yet, though this observation in the Vṛtti may be said to be unwarranted on the basis of Bharata, our point substantially remains unaffected.

3 See also ch. I, p. 10 where, in fact, we have shown the two grāmas through circular diagrams.

basically a particular grouping of seven notes (the sapta-svara) which together constituted a distinct grāma octave. In this grouping, no particular svara was, apparently,

the constant tonic around which the other ncies were grouped, unlike our system where the ādhāra-śruti-the basic tonic-is always called the sa and the other notes are arranged around it.

On a vīṇā of the ancient harp type, if the seven svaras of a particular grāma were to be tuned according to the given tonal distances expressed in terms of śruti-intervals and if this vīṇā had a range of at least two octaves (three octaves constituted

the standard range), then, after the svaras were arrayed within a grāma, any note could, evidently, be given the status of the tonic. None of the notes appears to have been a constant ādhāra-śruti like the modern droned sa.

In the 15th verse, Dattila calls the arrangement of śrutis and svaras with fixed sruti-intervals (vyavastitāntarān) within a grāma, the ‘svara-manḍala’. The word svara-manḍa'a has also been used by Bharata and with a similar meaning (N.S. 28, 24). The term svara-maṇḍala, in these two works, embraces the three topics : śruti, svara and grāma; it signifies the svaras with their specified śruti distances and their dual scheme of arrangement in the two grāmas. The svara-manḍala constituted the basic musical Iaw material, so to say, with which the gāndharva melodic structures were composed.

The Nāradi Śikṣā also uses the term svara-maṇḍala. Here the term embraces mūrchanā and tāna besides svara and grāma,

sapta svarāstrayo grāmā mūrchanāstvekaṿimśatih tā nā ekonapañcāśadityetat svaramaṇḍalam

The Two Auxiliary Notes

16A. niṣādauḥ kākalisaṃjño dviśrutyutkarṣaṇād1 bhavet

B. gāndhārastadvadeva syādantarārasavarāsaṃjñitaḥ

17A. anāṃśatvāttu bhedena svaratā nocyate tayoḥ

B. ato niṣādgāndhārāvetāvāptairudāhrtāu2

Niṣāda elevated by two śrutis, becomes the (niṣāda) called kākali, and gāndhāra, likewise [elevated], is called antara-svara (or antara gāndhāra).

Since these two cannot become amśas (anāṃśatvāt), they are not designated as separate svaras. This is the reason why they have been described as gāndhāra and niṣāda by the authorities.

NOTE :

Kākali has been defined by Bharata as“kalatvācca kākalisaṃjñāḥ” (N.S. 28, 35), which Abhinava explains as “a note slightly raised in pitch : īṣatkalastīrabhāvo’ smiṃ” . Kākali ni was a note higher than the regular niṣāda. Antara-gāndhāra was similarly related to gāndhāra.

Besides the seven svaras these two were the only other notes admitted in the gāndharva system. However, they were not given the status of independent notes but remained auxiliary to the notes of which they were considered as modified forms.

They were called vikṛta notes and were both two śrutis elevated from the normal positions of ni and ga respectively. The two different epithets, kākali and antara, attached to these two notes were a mere matter of usage and custom; in reality (paramārthatāḥ), as Abhinava on the basis of Bharata points out, both shared the trait of being sharpened notes and any of them could be termed either kākali or antara : “tattra saṃjñāvagrahaṇe bhedamāha vyavahārārthamime bhinne sañjñe para-mārthatas tu dvayorapi tīvrataratvāt kākalitvaṃ yadvakṣyati ‘kalatvācca kākalisaṃjñā’ iti dvayorapi cāntarasvaratvaṃ yaduktam—‘sādhāraṇaṃ nāmaṃtarasvaratā’ iti” (A.B. on N.S. 28, 35). Both these altered notes were also known by the generic term sādhāraṇa-svara (see the topic sādhāraṇa).

1 T. ed. reads : द्विश्रु (श्रुतिं) करषणाद्

2 The B.B. twice quotes Datt. 16, ascribing it to Dattila : (i) B.B. 3, 151-152; here the verse is quoted along with Datt. 46, as an authorititative pronouncement parallel to N.S. which pertains to connected matter concerning sādhāraṇa (see topic sādhāraṇa); (ii) B.B. (1)ch. VII quotes the verse again. Suḍhā on S.R 1, 3, 51-59 quotes the entire four lines given above, and in the same order, with the words तथा चोक्तं दत्तिलेन. Line 17A in Sudhā reads स्वरत्वं नोच्यते and 17B reads 'आख्यौ सदाहृतौ instead of the above.

Discussing the nature of kākali niṣāda, Bharata says: “vikṛtatvācca nāyamaṃśaḥ āptopadeśācca saptabhya nānyo niṣādavāneva” (N.S. 28, 35): “it cannot become an aṃsa for it is an altered note and has not a separate status apart from the seven notes ; it has been described by the authorities as nothing but niṣāda.” The same was true of antara ga as Dattila states (Datt. 17). Abhinava, elucidating Bharata's point, remarks: “on hearing kākali niṣāda, the idea of the note being niṣāda is first generated in the mind, only later is it distinguished as kākali, hence the note has been termed niṣāda and has not been given an independent status : pūrvam hi niṣādadhīprarūḍhāyāṃ tadbuddhirjāateti prāthamyāttenaiva vyapadeśaḥ” (A.B. on N S. 28-35).

The aṃsa note in gāndharva melodies, that is, jātis, was the predominant note. Kākali ni and antara ga could not become aṃsa notes. These two auxiliary notes had, in fact, very few and limited occasions of employ. Bharata gives the following general maxim governing their use:

antarasvarasamyogo nityamārohisamśrayaḥ kāryo hyalpo vieṣeṇa nāvarohī kadācana kriyamāno'varohī syādalpo vā yadi vā bahuḥ jātirāgaṃ śrutim caiva nihanyādantarasvarah (N.S. 21, 35-36)

“The antara-svara should always be associated (with a jāti) when making an ascending movement; its use should be exceedingly spare and never in making descending movements. If the antara-svara be used in descending movements, whether sparingly or with profusion, it destroys the śruti and the jāti-rāga.”

Abhinava points out that the word antara-svara in these verses denotated not only the auxiliary ga but also the kākali ni and the maxim applies equally to both the auxiliary notes : “antarasvarāśabdena cātra kākalyapi samgṛhīta iti kṛtopyayameva kramah” (A.B. on N.S. 28, 36).

Regarding the nature and mode of use of the antara-svara in the ārohī, Abhinava gives an illustration: if from the note madhyama the antara-svara (i.e., antara ga) was touched, then one was not permitted to descend straight away to the ṛṣabha but had to ascend back to madhyama or another higher note : “antarasvareṇa samyogo yasya madhyamasya tādrgyadlā prayuktastadā ārohaṇameva kāryaṃ madhyamādan-tarasvaram prayujya madhyama anyatra vā ārohaṇameva kāryam” (A.B. on N.S. 28, 35). This shows a kind of usage somewhat similar to what we would now call (in Hindustani music) either a vakra prayoga—such as that of the tīvra madhyama in the rāga rāmakalī—or an alaṅkārikā prayoga -such as that of the tīvra ni along with sa in some rāgas (e.g. the rāgeśvarī as sung in some gharānās) where the tīvra ni is in fact to be omitted.

Bharata says that the use of the antara note in the descent will destroy the jāti-rāga and the śruti. The notion of the destruction of śruti is rather puzzling. Perhaps what Bharata meant was that the employ of an antara-svara was liable to have an adverse effect on the fixed scheme of śruti-intervals through which the svaras in a grāma were revealed. By laying stress on an antara-svara this fixed and decreed

śruti-scheme could be disturbed. Because the antara-svara was placed on a śruti which did not constitute a proper and permissible svara-interval or, at best, constituted only a weak svara, its use would have tended to transform the tonal scheme of the grāma.1 Yet, this explanation does not give a reason why only in the avaroḥī was an antarasvara totally forbidden but not in the āroḥī. This, apparently, was one of the decrees peculiar to gāndharva.

Not only was the use of antara-svaras totally forbidden in the descending movements in gāndharva, even in the ascending movements only some jāatis permitted the auxiliary notes and in a very limited way. Bharata thus states :

svarasādhāraṇagatāstisro jñeyāstu jātayaḥ madhyamīpañcamī caiva ṣadjamadhyā tathaiva ca āsāmaśāstu vijñeyāḥ ṣadjamadhyamapañcamāḥ yathāsvaṃ durbalataravyatyāsāatpañcame tathā.

(N.S. 28, 37-38; again repeated as N.S. 28, 44-45)

"There are three jāatis which are connected with the use of the sādhāraṇa svaras (i.e., antara ga and kākali ni), namely, madhyamī-pañcamī and ṣadjamadhyā.

"The amśas in these jāatis (which permit the use of sādhāraṇasvaras) are known to be respectively (yathāsvaṃ) ṣadja, madhyama and pañcama; in case of pañcama the application is to be made as an alternative to or in exchange of (vyatyāsāt) the extremely weak note."

Bharata's language is cryptic and we must rely on Abhinava and others for an explanation.

The amśa in a jāati was its predominant note and Dattila equates it with the vādi (i.e. 'the dominant': "yo' tyantabahuulo yatra vādī vāṃśas cataatra saḥ.", Datt. 18).

Many jāatis had not one but multiple amśas. In the jāati ṣadjamadhyamā, for example, all the seven notes were potential amśas. The multiple amśas of a jāati were known as paryāyāmśas (see topic 'jāti') or alternate amśas. Only one of the alternate, amśas was, evidently, given the role of the acting amśa at a time. All the three jāatis named above by Bharata had multiple amśas : madhyamī (more often called madhyamā) had five amśas, sa, ri, ma, pa and dha; pañcamī had two amśas, ri and pa and ṣadjamadhyā (also called ṣadjamadhyamā) had all the seven notes as possible amśas.

Abhinava, explaining Bharata's dictum, says that only when these three jāatis had either sa, ma or pa as the ruling or acting amśa could the sādhāraṇa svaras be employed. Ṣadjamadhyamā, he says, was a seven-amśa jāati ; in it the auxiliary notes

1. Abhinava remarks :

तेन काकल्यनिप्रयोगे षड्जादीनां न श्रुतिप्रहारः; कार्यं । अत एवाल्पश्रुतिम् ।

• A.B. on N.S. 28, 23.

We have seen that the interval between ni and sa was a four-śruti interval, most probably that of a major tone : but if the kākali ni were to become too prominent the ṣadja interval would be reduced to a half-tone ; it would suffer a 'śrutyapahāra' by losing two śrutis. This would destroy the entire grāma-scheme. The same was true of the antara ga. Thus these two auxiliary intervals were to be used sparingly and in such a way that 'śrutyapahāra' and its resulting adverse effect on the grāma scheme did not take place.

could not be used when ga and ni, the two two-śruti notes, were the amśas. Only with sa, ma or pa as the amśa could the sādhāraṇa svaras be used. The same rule applied in madhyamā which had five possible amśas including sa, ma and pa : “yathāsvamiti” śadjamadhyamā śaptamśā, tatra niṣādagandhāryostāvadamśayoh nāsti sādhāraṇam; samapeṣu sātiśayastatprayogah, madhyamāyāḥ apyeṣa vidhiḥ, sā hi dviśrutivarjaṃ pañcāmśāmśāḥ” (A.B. on N.S. 28, 45).

In the jāti pañcamī, with two possible amśas, pa ri, the kākali and antara-svaras could be used only in case pa was the ruling amśa. There was a further condition: this jāti was rendered as śāḍava by dropping the note ga (ga lope śāḍavam ; A.B. on N.S. 28, 125), and rendered as auḍuva by dropping both ga and ni (ganilope auḍuvam ; A.B. ibid.). Abhinava suggests that when this jāti was rendered śāḍava by omitting ga and with pañcama as the amśa, then in the place of ga which was thus rendered weak,1 an exceedingly weak antara ga was to be employed. Similarly, in the auḍuva form of this jāti, kākali ni in place of the regular two-śruti niṣāda, too, was used as a very feeble note.2

Śārṅgadeva and others have also discussed this rule regarding the permitted employ of sādhāraṇa svaras. Śārṅgadeva expounds the rule mentioned by Bharata in this connection (S.R. 1,7, 21-22; see also Kalā). Kambala and Aśvatara, had also expounded this rule, phrasing it in more general terms (etadalpanigās vāhuḥ kambalāś-vatarādayah; S.R. 1, 7, 22); instead of naming the jātis where the rule applied, they had said that the rule obtained in all jātis where the notes ni and ga were to be rendered weak—or in other words were omitted svaras. As an implication of this general rule, says Källinātha, not only the three jātis specifically mentioned by Bharata, but also a fourth, the śāḍjī, becomes included in the fold.3

Śāḍjī had five possible amśas: sa, ga, ma, pa, dha. Källinātha's interpretation of Aśvatara and Kambala implied that in this jāti when ni was to be omitted and when either sa ma or pa was the ruling amśa, the kākali ni could be used in the approved manner of kākali-prayoga. However, we do not know if this interpretation of the dictum propounded by Kambala and Aśvatara is correct; the dictum states that sādhāraṇa is to be used in jātis which are 'alpanigāḥ'; the implication could be that only in jātis where both ni and ga can become alpa or weak can sādhāraṇa svaras be used. This would give us only the three jātis mentioned by Bharata, for in śāḍjī only ni could become weak.

1 A dropped note was often not totally omitted but rendered weak.

पञ्चम्यास्तु क्रियते यस्माद्वैषम्यं चाप्यन्वयो भवेत् .....विति येन पञ्चमं चैवास्यां काकलतरयोग(?)ः प्रसङ्गादस्मान्नयदादु दुर्बलतरस्य-व्यथासे' दुर्बलागात्वं कर्तव्यम् । तेन वाडवकर्णो गान्धारस्य स्थाने दुर्बलेनान्तरस्वरेण यत्न भवेत् तस्यासी दुर्बलतरः

—A.B. on N.S. 28, 45.

2 पञ्चम्यां मध्यमा इति भरतमतानुसारिणा वचननेत 'एतदल्पनिगासु' इति कम्बलाश्वतरादिमतततनुसारिणा वचननेत चौपात्तासु

पञ्चम्यां मध्यमावैपञ्चम्यांष्वड्जीग्रमध्यमावापञ्चमं वाडूजिं चतुष्प्रभू...

—Kalā on S.R. 1, 7, 21-23.

Kumbha, probably on the basis of Källinātha, explicitly states that Kambala, Aśvatara and theorists of this group had included śāḍjī as a jāti where sādhāraṇa svara could be used; he does not speak of the general dictum which Śārṅgadeva has noted but only the conclusion drawn from it by Källinātha :

कामचारम्महोभयतो साधारणमुदीरितम् । कम्बलाश्वतरदृष्टयस्तु षाड्ज्याद्या जातयोपि स्मृतम् ॥

—S. Raj, 2, 1, 4, 54.

The Four Kinds Of Notes Within A Melodic Structure

18 A. yo'tyantabahulo yatra vādi vāṃśaśca totra saḥ

B. mithah samvādinau jñeyau trayodaśanavāntarau

19 A. ato' nuvadinau śeṣā dvyantarau tu vivādinau

B. svarāṃścaturvidhānena jānīyāt svarayogavit.1

Wherever a [svara] occurs copiously, it is [called] vādi or amśa. Pairs of notes with an interval of nine or thirteen [śrutis] are consonant (samvādi) to each other.

The rest [of the notes] are [called] anuvādi. Pairs of notes having an interval of two [śrutis from each other] are discordant (vivādi). One who has grasped the application of svaras knows them to be of [these] four kinds only.

NOTE :

Notes were said to be of four kinds: amśa or vādi, samvādi, anuvādi and vivādi. Some scholars have taken amśa to mean the tonic. Alain Danielou's translation of verses 16 and 17 from Dattilam reads: "The note called kakali is obtained by raising ni (modern Si b); the note called antara is obtained from ga (modern ga k mi b) in the same way. They are not properly considered notes (svaras) because they cannot be taken as tonic (amśa). So niṣāda (ni k, si b) and gāndhāra (ga k, mi b) are given prominence over them (Dattila 16-17)" (Northern Indian Music, Vol. I, p.44).

This seems to us misconceived. We find Dattila describing amśa as a synonym of vādi, the predominant note which need not necessarily have been the tonic. Bharata and later authors as well, have interpreted amśa as synonymous with vādi.

Bharata explicitly asserts: "the (note) which is taken as the amśa is the vādi" (tatra yo yadāṃśaḥ sa tādṛśī vādi ; N.S. 28, 22). Bharata's statement implies that of all the possible amśas in a jāti (some of them having as many as seven) the note chosen as the ruling amśa in a particular rendering was the vādi or the predominant note. Abhinava, commenting on this, states that Dattila and others had unambiguously called amśa as identical with vādi, (amśa eva hi vādiiti dattilādyāḥ ; A.B. on N.S. 28-23). Simha-bhūpāla is still more explicit in affirming: "the note which is most recurrently sung

1 Line 18B in the Datt. T. ed. has a lacuna and reads मिथः संवदह + जनो जयो, obviously a reading that makes no sense. The line, however, is quoted in its correct form by Sudhā on S.R. 1, 3, 46-56 where it occurs as an acknowledged quotation : दत्तिले नाप्युक्तस्तम्: "मिथः संवादिनौ जेयो त्रयोदशनवांतरौ" ।

B.B. 5, 129-130 quotes Datt. 18-19A as supporting evidence of matter quoted from N.S. as sūtra (श्रुतिं गृण्वह). The quoted lines are introduced with the words : अयमेवार्थो दत्तिलाचार्येणाप्युक्तः ।

The lines are quoted again in B.B. (I) ch. VII in a similar manner supporting the testimony of N.S. with the words : 'अयमेवार्थो दत्तिलाचार्येणाप्युक्तः.' In the same chapter Datt. 18A is again quoted.

jātis, rāgas, etc. is called vādī, another synonym for which is amśa* (prayoge jatirā-gādau bahulo bāhulyena ya uccāryate so'mśavarāparaparyāyo vadi ; Sudhā, S R. 1, 3, 47). For further discussion on amśa in jātis, see note on verses 96-97.

Samvādīs depended on samvāda. The notion of samvāda between svaras was, evidently, based on the harmonic law of a natural acoustic affinity which exists between certain intervals in an octave. The ancients speak of two samvādas, one between notes separated by nine śrutis and another between notes separated by thirteen śrutis. These are what we now call the ṣadja-madhyamabhāva and the ṣadja-pañcamabhāva : the harmony of the natural fourth and the natural fifth. Another basic harmony which plays a very eminent role in the harmony-dominated Western music is the harmony of the natural third : the ṣadja-gāndhārabhāva. This the ancients do not speak of, though musicians must have been aware of it. The interval between ma and dha, in either grāma, i.e., the seven-śruti-interval, probably was empirically arrived at on the basis of the ṣadja-gāndhārabhāva though it did not enter into theory.1

The structures of the two grāmas were defined not only by the different śruti-intervals existing between some svaras but also by their slightly differing samvādas. This is, evidently, the reason why Dattila notes the samvādī relations as an aspect of the topic grāma.

Dattila has given only a general dictum in terms of śruti-intervals in describing the very pairs of notes that have the relation of samvāda with each other. Bharata names the pairs of notes that have the relation of samvāda with each other. In ṣadja-grāma the samvādī-pairs were : ṣadja-pañcama, rṣabhadhaivata, gāndhāra-niṣāda and ṣadja-madhyama. Madhyama-grāma, he says, has the same pairs of samvadis except that ṣadja-pañcama samvāda is missing here and is replaced by pañcama-rṣabha samvāda (N.S. 28, 22-23).

Looking back at the maṇḍala-prastāra of the ṣadja-grāma we see that pa is located on the thirteen śruti counting from sa, so is ni from ga and dha from ri. This is what Bharata and Dattila have called the thirteen-śruti interval (trayodaśakamantaram). Madhyama is situated on the ninth śruti from ṣadja and thus has the samvāda of the nine-śruti interval (navakamantaram).2 In this grāma pa is located on the tenth śruti from ri, hence there is no samvāda. In the madhyama-grāma, likewise, pa in relation to sa is on the twelfth śruti, hence the ṣadja-pañcama samvāda does not obtain ; but ri

1 Acārya Bṛhaspati calls this harmonic interval the ṣadjantarabhāva since it was the interval between ṣadja and the antara-gāndhāra—a seven śruti interval. He has used this interval as a basic harmonic interval in his suggested process for arriving at the tone values of the ancient svaras : Bharata kā Saṅgīta Siddhānta, pp. 12-13.

2 Trayodaśakamantaram and navakamantaram (=navatrayodaśāntaram, Datt. 18) were the conventional expressions used to denote these intervals. The actual 'antara' as some had indeed argued was of 12 and 8 śrutis respectively. Abhinava's teacher had attempted to resolve the disparity by remarking that the word 'antara' in this context denoted not 'interval' but 'svarūpa' or 'form' : thus the expression means 'a svara which forms the thirteenth śruti' and 'a svara which forms the ninth śruti' :

उपाध्यायैस्त्वतः— अत्रर्मिति स्वरुपं नान्तरालं तेन नवश्रुतिकं यस्य स्वरस्य स्वरस्य, यस्य च त्रयोदशश्रुतिकत्वं स्वरुपं

तो स्वरो परस्परसंवादिनो ।

—A.B. on N.S. 28, 23.

in this grāma has the required nine-śruti interval with pa and hence the pair were samvādīs. This was the only samvāda which differed between the two grāmas.

The Vṛttikāra on the Bṛhaddeśī adds a qualifying factor to the notion of samvāda. He says : “samvādinastu punah samaśrutikatve sati trayodaśanavāntaratve-nāvaboddhavyāl” (Bṛ. Vṛtti on 63).1 The qualifying phrase ‘sama śrutikatve sati’ means that any two notes situated at the proper interval could be samvādīs, provided the two contained an equal number of śrutis : thus the two-śruti ga could be a samvādī of the two-śruti ni and the three-śruti ri could be a samvādī of the three-śruti dha (and the three-sruti pa of madhyama-grāma). Similarly the four-śruti sa qualified as a samvādī of the four-śruti ma and also of the four-śruti pa (of ṣadja-grāma). But, ma could not be a samvādī of ni, although the ‘navaka-śruti antara’ rule applied, because ma was a four-śruti note whereas ni was a two-śruti note. The ma-ni samvāda has, indeed, not been noted by Bharata either, though he does not cite a rule to explain this omission.

The four-fold division of notes as vādī, samvādī, anuvādī and vivādī, was an ancient classification and is found in Bharata as well as in Mataṅga ; it continued in later saṅgīta literature and forms a part of our own present-day musical thinking.

Vādī has been defined as the predominant note. In fact, samvādīs, anuvādīs and vivādīs were all dependent on the vādī. The notion of a vādī can only be understood in terms of ancient melodic structures. Unlike the samvādī-pairs, which were natural pairs depending upon an inexorable acoustic law, no note was a ‘natural’ vādī in itself. The Vṛtti on Bṛhaddeśī, we find, lists all the seven notes of an octave as vādis : “vādimavaḍalaṃ yathā-sarigamapadani” (Bṛ. Vṛtti on 63). The notion of vāditva—the state of being a vādī—had no meaning apart from musical practice in a given melodic context : that note which was copiously used in a particular jāti, rāga, bhāṣā etc. was its vādī. This, depending upon the melody in question, could be any of the seven notes. This is a usage still current. “The rāga-ness of a rāgā”, says the Vṛttikāra, “is created by the vādī” (yad vādisvarena rāgásya rāgatvaṃ janitam; Bṛ. Vṛtti ibid.). Dattila and Bharata have clearly related the notion of a vādī note to actual jāti-structures in saying that the vādī is the aṃśa, the aṃśa being defined as the predominant note in a jāti.

Samvāda depended upon harmonic affinity, but a specific samvādī note was defined only in relation to the vādī ; samvādī was the note which had the samvāda relation with the vādī; it was the second important note in a melody. Like the vādī-note, a samvādī-note also could never be dropped from a jāti. Vādī and samvādī, Abhinava consequently says, are not to be omitted: “samvādino vādināścānupalapanam” (A.B. on N.S. 28, 22).

Vivādīs were svaras two śruti apart, namely, ṛṣabha and gāndhāra; dhaivata and niṣāda: “vivādinastu yeṣāṃ dviśrutikamantararam, tadyathā-ṛṣabhagāndhārau, dhaivata-niṣāda” (N.S. 28, 24). The notion of vivāda, as the word itself reveals, indicates ‘dissonance’. Vivādī was a note which, if used indiscriminately, was liable

1 The Vṛtti on Bṛ., has in this context sketched a samvādī maṇḍala and also an anuvādī and vivādī maṇḍala, somewhat resembling the maṇḍala-prastāra. The sketch in the text is erroneous and we do not know the nature of the genuine article.

to distort a melodic structure.1 Abhinava, using an often repeated old analogy, (cf. Br. Vṛtti on 63) says: “vādī is the king, samvādī is the minister who follows him, vivādī is like the enemy and should be sparingly employed (arivadvadītvalpaḥ); anuvādī denotes the retinue of followers.”2

In current theory, too, the notion of vivādī is of a note which distorts a rāga, but in present practice it is not defined as a note two śrutis apart from any other note. In fact, all notes or śrutis not included in the structure of a rāga are conceived of as its vivādīs. Moreover, in the more flexible form of current classical music like ṭhumrī, vivādīs are deliberately introduced, often quite profusely, in order to create dazzling effects.

In ancient theory, vivādīs have been described in pairs; there being two such pairs: ri-ga and dha-ni. These notes were two śrutis apart from each other. No single note as such has been described as a vivādī to the entire general structure of a jāti or a rāga.

From available accounts, it is difficult to reconstruct the position and function of a vivādī in an ancient musical structure. Two pairs, ri-ga and dha-ni have alone been noted. Did this imply that the relation of vivāda existed only between the two notes of these two pairs, that is, with ri as vādī, ga became the vivādī and vice-versa, and the same applied in the case of dha-ni? Abhinava’s remarks here though not quite clear seem to belie this idea. Pointing at the use of the plural in “vivādinastu” in the Natyāśāstra, he suggests that the two-śruti notes ga and ni are vivādīs of all other notes (dviḥśrutiḥ gāndhāro niṣādaśca svarāntarānām vivādinau sarveṣāmeva, tathā bahuvacanam). The fact that Bharata has mentioned only ri and dha as the respective vivādīs of ga and ni should not, he suggests, be made much of; these are named merely because of their proximity to ga and ni (ṛṣabhagāndhārau dhaivata-niṣādāviti caturvidhayorupadānāṁ naikatyāt). With ṣadja as the amśa (i.e., the vādī), he says in exemplifying his point, ma and pa are obtained as its two samvādīs (evidently, he has the ṣadja-grāma in mind); ri and dha as the two anuvādīs ; ga and ni, by implication, are the vivādīs (though Abhinava does not overtly say so). With gāndhāra as vādī, he remarks, ni becomes the samvādī and there is no vivādī.3

1 वादादिषु: स्वरेऽन्तु रागस्य वादित्वं संवदित्वमनुवादित्वञ्च प्राप्तं, तद्विनाशकत्वं नाम विवादित्वं । —Br. Vṛtti on 63.

2 गन वादी स्वामी, अमात्य इवेतरोज्ञप्तिसंवादी, अरिवद्विवादितवल्पः; अनुसारी परिजन इव योग (?ग्राम) वादी चेति विभागः । —A.B. on N.S. 28, 24.

3 यत्र विवादिनो लक्षणं ‘विवादिनस्तु त’ इति । अन्तर्हितस्वरूपतो गनिदृष्टान्ति श्रुतितरङ्गं तेन हि:श्रुति(:गन्धा)रो निषादश्च स्वरान्तराणां विवादिनौ सर्वामेव तत्र बहुवचनं । ‘ऋषभ-गान्धारो धैवतनिषादाविति’ चतुर्विधयोरुपादानं नैकट्यात् । अन्ये तु द्वाहर्त्तनललामभूतापेक्षया श्रुत्योर्विवादित्वमाहुः । तच्चासत् । सर्वस्वरस्स विवादिन: सतो मध्यसपक्षविषपक्षमहर्ष्य:; तथापि दृढयेऽनुस तस्य मपो संवादिनो, ऋषभानुवा-दिहेनो, ऋषभे धर्मनाश: पमसा; (?), गान्धारे नि: वादी (?संवादी) येतु, नास्ति विवादी निषादस्य संवाधिक्वात् । एवमन्यदनुसर्तव्यम् । —A.B. on N.S. 28, 24.

All notes in a melody which did not obtain as either vādī, samvādī or vivādī were anuvādīs: “vādisaṃvāddivivādīṣu sthāpiteṣu śeṣastvanuvādinah” (N.S. 28, 24). Bharata, in a prose passage, enumerates the anuvādīs pertaining to each individual note in an octave: the anuvādīs of ṣadja are listed as ṛṣabha, gāndhāra, dhaivata and niṣāda; of ṛṣabha as madhyama, pañcama and niṣāda; of pañcama as ṣadja madhyama and dhaivata (?).1 This is the list for the ṣadja-grāma. The anuvādīs of madhyama, in the madhyama-grāma , were dhaivata, niṣāda, ṛṣabha, ṣadja, and gāndhāra; of pañcama were madhyama, niṣāda ṛṣabha and gāndhāra; of dhaivata, ṣadja, ṛṣabha and gāndhāra and of niṣāda, ṣadja and ṛṣabha.2

Abhinava did not accept this passage as a part of the Naṭyaśāstra and has not commented upon it.3 The reason is evident: the list above shows that the two-śruti notes ni and ga were not universal vivādīs but vivādīs only of dha and ri respectively. We do not know if this passage is really an interpolation or what status it should be accorded.

The Vṛtti on Bṛhaddeśī has given its own interpretation of vādi and the other three categories of svaras in a melody. Echoing Bharata (N.S. 28, 24), the Vṛttikāra observes: “the notes like vādī etc. establish the sonance (vāditva), the consonance (samvāditva) and prosonance (anuvāditva) (of a rāga); the nature of vivādī is to disturb it.”4

Regarding the function and relative use of these notes within a melodic structure, the Vṛttikāra remarks that such is the sympathetic relation of samvādī to the vādī that in a melody, if the samvādī is used in place of the vādī, the rāga is not distorted. He gives examples, naming actual melodies. The relative function of the anuvādī in his opinion was similar; if instead of a stipulated note its anuvādī was used, the melody remained intact. Again he gives illustrations: for example, he says that if ri is used instead of a stipulated sa or vice-versa, the form of a jāti-rāga is not distorted. But the use of a vivādī distorts. The Vṛttikāra illustrating this point says that ri and dha, if employed instead of ga and ni respectively and vice-versa, have an adverse effect on a jāti-rāga (Vṛtti on Bṛ. 63). Evidently, the Vṛttikāra, unlike Abhinava, believed that dissonance occurred only between the two notes ri-ga and dha-ni.

1 The meaning here is not very clear. See the following footnote for the original passage.

अनुवादिसंज्ञकाः यथा—पडजस्यपञ्चमभौष्मागारहेवतनिषादाः, ऋषभस्य मध्यमपञ्चमनीषादाः,. गान्धारस्यापि मध्यमपञ्चम-धैवतौ:, मध्यमस्य धैवतपडजपञ्चमनिषादौ:, पञ्चमस्य धैवत(?)पडजमध्यमनीषादाः,. धैवतस्य पडजर्षभगान्धारनिषादौ:, निषादस्य पडजर्षभगान्धारौ ॥ –N.S. 28, 24.

The passage contains some faulty readings. It is not clear whether the aim is to list the anuvādīs of ṣadja-grāma and madhyama-grāma separately, note by note; if so, some notes have been left out. If only madhyama and pañcama were meant to be doubly listed on the basis of the two grāmas, then the repetition of ṣadja with a different set of anuvādīs, is inexplicable.

The passage has been marked off by a bracket in the G.O.S. edition.

4 बदनाद् वादी स्वामिवत् । संवदनात् संवादी अमात्यवत् । अनुवदनादनुवादी परिजनवल् । विवादनाद् विवादि शत्रुवत् ।

The Indispensable Svaras

20 A. pañcamaṃ madhyamagrame ṣadjagrāme tu dhaivatam

B. anāśinaṃ vijānīyāt sarvatraiva tu madhyamam1

Pañcama in the madhyama-grāma, dhaivata in the ṣadja-grāma and madhyama in both the grāmas (sarvatraiva) should be known as indispensable [notes].

NOTE :

Kallinātha has recorded two explanations held by different theorists regarding the indispensability of madhyama. The first view was simply a recitation of tradition and its spokesmen stated that madhyama was indispensable because Bharata and others had described it to be so while listing notes to be dropped in tānas and in rendering jātis ṣāḍava and auḍuvita. The other view had a more rational appearance. It held that madhyama is the central note dividing the octave into two parts : sa ri ga and pa dha ni—the lower three being consonants to the upper three, sa to pa, ri to dha and ga to ni (yathāsaṅkhya). Between the two divisions, madhyama stands solitary as the central point and thus has no consonants left for it; it has not to be dropped because of its solitary and focal position.2

Abhinava, who may have influenced Kallinātha on this point, had put the same idea more neatly. “In a single register”, he argued, “there are seven notes in succession, each containing a different number of śrutis. The first note, sa, contains the maximum number of śrutis assigned to a note, i.e., four. In the next two notes the number of śrutis is decreased successively by one. After reaching the minimum number of śrutis for a note, madhyama is then formed, again with the maximum of śrutis, i.e. four. Then in the same sthāna, i.e., the same vocal-register, such as the chest, throat and head, each of which were the anatomic seats of one single octave) or, in other words, the same octave, follows pañcama with four śrutis, dhaivata with

1 A.B. on N.S. 28, 33 quotes Datt. 20 and has a reading identical with the above text. The verse is ascribed to Dattila and quoted with the words : तदुकं दत्तिलाचार्यण्. Sudhā on S.R. 1, 4, 6—8 quotes the verse with the words: मध्यमस्याविलोपित्वमुक्तं दत्तिलेन. Datt. 20B in Sudhā has ‘अलोपिनं विजानीयात्’ instead of above. B.B. 4, 58 quotes the verse with ‘दत्तिलस्याह’. Line 20B reads ‘सर्वदैव तु’ instead of above. Dattila is quoted as testimony parallel to N.S. 28, 65.

2 ‘अविलोपित्वादिति’ । मध्यमस्याविलोपित्वं चाघस्तनानां सरिगाणामुपरितनानां पधनीनां च यथासंख्यं द्वयोर्द्वोरे कान् तन्त्र्यां संवदनं संवाद इति मतानुसारिकण्किनो मध्यमस्याव्यनं संवदभावात् परिशेषादवि लोपो नेष्यत इति केयाचिन्मतम् । अन्येषां तु शास्त्रान्तरसारेऽपि दृश्यते । नामेव पाङवोद्धकारित्वेन लोपविधानमध्यमस्याविनाशित्वर्मिति । —Kalā on S.R. 1, 4, 6-8.

three and niṣāda with two. With the upper sa we come to another sthāna or octave. This very arrangement occurs in all the three sthānas, namely, that to which the mandra-saptaka belongs (i. e., the chest), the throat and the head. The position of amśa and its samvādī is accorded only to notes which are similar (i.e., which have a similar number of śrutis). Thus, in the last analysis (paramārthataḥ), it can be said that there are only three notes: sa (four śrutis) ri, (three śrutis), ga (two-śrutis): pa (four śrutis), dha (three śrutis) and ni (two śrutis) are their complementaries. With madhyama occupying the cenral position and remaining constant, the lower three are samvādis of the upper three."1

The explanation offered by Kumbha, the author ef Sangītarāja, is more or less on the same lines. He has used an interesting term “tṛka” (group of three) in this context, and has pointed out that ma is the central note between the lower and the higher tetrachords (tṛkas), i.e., the tṛka ‘sa ri ga’ and ‘pa dha ni’. Hence its indispensability (S. Raj 2, 1, 1, 235).

Kumbha's statement makes the parallel we have drawn earlier (see ch. II) between Abhinava's concept on this point and the Greek system of arranging tones, very explicit. Yet, we must remember that in the Greek system, the tones of the two tertrachords in an octave changed values and the interval between sa ri ga and pa dha ni was not rigidly fixed to one single possibility as in gāndharva. We must also remember that the explanation of the centrality of madhyama as given above could apply only to the ṣadja-grāma.

It should be noted that though madhyama could never be altogehter dropped from a jāti, yet it could be rendered as quite a weak note. In ṣadjakaiśikī jāti, for example, Bharata (according to the reading accepted by Abhinava) describes madhyama as a weak note (see under jāti). Again, in pañcamī jāti, madhyama has been recounted as one of the weak notes. Abhinava adds that its position here was excessively weak (alpatara).

Apart from madhyama, which could not be dropped from either of the grāmas in gāndharva, Dattila has also mentioned two other notes, pañcama and dhaivata, as indispensable in madhyama and ṣadja grāmas respectively. Viśākhila, among others, also subscribed to this opinion (A.B. on N.S. 28, 64-65). But Abhinava has observed that this opinion was not upheld by many on the ground that madhyama alone and no other note has been described as indispensable by Bharata (A.B. on N.S. 28, 64-65). However, we find in Bharata's exposition, too, that in the jātis and tānas of ṣadja-grāma, dhaivata was a note which was never actually dropped, and similarly in the jātis and tānas of madhyama-grama, pañcama was never omitted.

चतुःश्रुतिस्विːश्रुतिद्विːतिरिति । तथाहि - एकरस्मिन्नेवोच चतुःश्रुतिरतया सप्तस्वराः; तथाहि चतुःश्रुतिकाल्पूर्वाद्वयात् तदूते (?) श्रुत्येकं निष्कास्य प्रक्रमेण पुनरुच्च प्रकृतिमधूःपूर्वकःस्पर्धावल्वारः सरिगमः । तस्मिन् स्थाने उभयवस्पर्शं पर्यनिसा:, केवलम् ऋज्जनं हृदिरीयं स्थानं सुप्रसिद्धं एव । मन्त्रःस्पश्वत; स्थानं स्वरानाम् । एवं कण्ठस्थाने शिरसि च तेषु स्वस्वेता(?)स्वद्वयस्य(?)मना(?)स्वादिव्यपदेश एव, तत्वस्थानगतानां तु साधुश्वरपरत एवं संवादत्वं वद्यते । तेन परमार्थतः; स्वर एव स्वराः-सरिगः:, पधनयः । मध्यमस्तु ध्रुवकस्थानीयो मध्यमत्वादेव!

-A.B on N.S. 28, 21.

TOPIC 4

MŪRCHANĀ

21 A. svarau yāvatithau syātam grāmayoḥ ṣadjamadhyamau

B. mūrchanā tāvatithyeva tadgrāmāvata eva tau1

In the two grāmas the serial number of a mūrchanā is [determined by] the numerical order of the notes ṣadja or madhyama [respectively]. Therefore are these two grāmas [called after] these two notes.

NOTE :

With the above verse, the topic grāma is concluded and the fourth topic, namely, mūrchanā is initiated. Dattila in this verse produces a reason on the basis of the mūrchanā process, as to why the two grāmas were so called. Bharata makes no parallel statement. Followers of Bharata—if Abhinava’s criticism be any indication—seem to have considered the reason Dattila gives for the nomenclature of the two grāmas, as insignificant (see chapter I).

The Saṅgītaratnākara has a verse parallel to Datt. 21 : “yasyāṁ yāvatithau ṣadjamadhyamau grāmayoḥ kramāt/mūrchanā tāvatithyeva sā niḥśañkena kīrtitā” (S.R. 1, 4, 18). Simhabhūpāla explains this verse and describes the process (see Sudhā on S.R. ibid.). We have described the process in chapter I.

This verse has also been an oft-quoted verse. Abhinava, we have seen (ch. I. ), quotes and criticises the idea postulated (A.B. on N.S. 28, 24). Sudhā on S.R. 1, 4, 18-26 quotes this verse along with Datt. 25B with the words: ‘दत्तिलोऽप्याह’. Sudhā reads the last part of 21B as ‘तद्ग्राममदितये तथा’, a reading not found elsewhere. S. Raj 2, 1, 1, 304-305 also quotes the verse commenting: एवं यद्वत्तिलो बूते मूर्छनागणनाधिप: । हेतुत्वेन तयोर्मध्ये युक्तिं दूषयते । He seems to have been influenced by Abhinava in making this remark.

This numerical process of arriving at the serial number of a mūrchanā is not found in Bharata. Later authors, however, eagerly adopted it and Dattila seems to have been their guide. The resemblance of Sārṅgadeva’s verse with that of Dattila is a clear indication of this point. Other authors also have almost echoed Datt. 21. For example : Śrīkanṭha says :

यस्यां यावतिथो पद्मधरमो मूर्छनाक्रमे । मूर्छना तावतिथ्येव ग्रामयोः परिकीर्तिता ॥

—R. Kau 1,71.

also : यस्यां यावतिथः(पद्जमस्तावदित्ये(चतिये)व मूर्छना ।

—Bharatasāśrarāgādisvaraniṛnaya of Raghunātha Prasāda, 1, 43 (ms no. B 6643/D10669-Tanjore Library).

यस्यां यावतिथः पद्जजस्तावदित्येव मूर्छना:

—Rāgamālā of Puṇḍarīka Viṭṭhala, 1, 30 (ms. no. 1985 in Oriental Institute, Baroda.).

238

A Study of Dattilam

Names and Variations

22 A. uttaramandrā rajanī ṛtīyā cottarāyatā

B. caturthī śuddhaṣadjā tu pañcamī matsarīkṛtā

23 A. aśvākrāntā tu ṣaṣthī syāt saptamī cābhirudgatā

B. svarakramagatā vidyāt saptamitāḥ ṣadjamūrchanāḥ

24 A. sauvīrī madhyamagraṃe hariṇāśvā tathaiva ca

B. syāt kapolanatā caiva caturthī śuddhamadhyamā

25 A. māndī ca pauravī caiva hṛṣyakā (ca yathākramam)

B. sarvāstatāḥ pañca ṣat pūrṇasādharāṇakṛtāḥ smṛtāḥ.1

The seven mūrchanās of the ṣadja-[grāma] in due sequence of notes, should be known as uttaramandrā [the first], rajanī [the second], uttarāyatā [the third], śuddhaṣadjā [the fourth], matsarīkṛtā [the fifth], aśvākrāntā [the sixth] and the seventh being abhirudgatā.

Sauvīrī, hariṇāśvā, kapolanatā and śuddhamadhyamā [which is] the fourth ; māndī (usually called mārgī), pauravī and hṛṣyakā, these, in due order, are [the mūrchanās] in the madhyama-grāma. It is well known that all [the mūrchanās] can be rendered [either] with five notes (pañca), with six notes (ṣat) [or] with the seven notes (pūrṇa), [and also] with the sādhāraṇa (notes, i.e., with kākalī ni and antara ga).

NOTE :

Mūrchanā was a straight sequence of the seven svaras, e.g., sa ri ga ma pa dha ni. Dattila gives an extremely short aphoristic description of the mūrchanā-process ; he says, are ‘svarakramagatāḥ’. To know the nature of this svara-krama, we must take recourse to other texts. The krama-process was two-fold : (1) each mūrchanā was itself a regular sequence of seven notes, and (2) each successive mūrchanā also depended upon a krama ; this was as follows. The first mūrchanā (called uttaramandrā) began on the middle sa and ran the gamut of notes in a straight

1 The Vāyu (and the Brahmāṇḍa) Purāṇas, in naming the mūrchanās, contain a text remarkably akin to the Dattilam. In the Purāṇic account, the madhyama-grāma list precedes the ṣadja-grāma, but we quote them in the order obtaining in Dattilam :

उत्तरमन्द्रा रजनी तृतीया चोत्तरायता । शुद्धषड्जा तुरीयाच (पञ्चम्यौ मत्सरीकृता ऽथवा पष्ठी) ऽश्वक्रान्ता तथा ऽभिषुद्गता ।

See also Brahmāṇḍa 3, 61, 33-34 (Venkateshwara ed.). The madhyama-grāma reads as : सौवीरी मध्यमग्रामे हरिणाख्या तथैव च । स्यात् कलोपनतप्ता चतुर्थी शुद्धमध्यमा। मार्गी च पौर्वी चैव हृष्यका च यथाक्रमम् ।

—Vāyupurāṇa 86, 16-17 see also Brahmāṇḍa 3, 61, 31-32. Datt. T. ed. 25A has a lacuna of 5 letters at end of line. Our suggested reconstruction is based on the Purāṇic reading. Datt. T. ed. 25B has three letters missing in the beginning of the line. Sudhā on S.R. 1, 4, 21-26 quotes this line in full ; the first word is सर्वस्ता:,

line to ni; the second began a step lower on the mandra ni and extended to the middle

dha ; the third began on the mandra dha and extended to middle pa and so on. Such

was the krama for successive mūrchanās. Bharata says : “āsāṃ ṣadjaniṣādhaivata-

pañcamamadhyamagāndhāraṣabha ānupūrvya ādyāḥ svarāḥ” (N.S. 28, 29).

Like Dattila, Bharata, too has defined mūrchanā as a sequence of the seven

notes : “kramayuktāḥ svarāḥ sapta mūrchanetyabhisamjñitāḥ” (N.S. 28, 32). Abhi-

nava, expounding the definition, states that mūrchanā is an octave (actually a svara-

saptaka) consisting of a gamut of seven notes sounded in a regular ascending order,

without the natural sequence of notes being changed. Each new mūrchanā, he adds,

begins on a successively lower note. He stresses the point that it is necessary to adhere

to the proper sequence of notes in every mūrchanā for, if the sequence is disturbed, a

mūrchanā cannot be obtained.1

The Vrtti on Brhaddesī has represented the mūrchanās through a chart, calling

it the mūrchanā-maṇḍala (Vrtti on Br. 95-96) The format of the chart is adequately

described but in the extant text the chart itself is extremely distorted.2 It is, however,

found intact in the Saṅgītarāja (S. Raj. 2, 1, 1; see the entire mūrchanā-prakaraṇa).

The ṣadja-grāma chart is as follows :

“jayeti samppiditam” kramānantarikrameṇāpi yacchṛu yamaṇamvarohaṇaṃ tadupaniṣadādisvarasaptakasvaraih (Ṭhā). mūrchanāḥ mūrcchanā iti

samucchāye paṭhyate . ata eṣopasamhāre svarāḥ : kramayuktā iti viśeṣanīyati . akramalye mūrcchanāsu mā bhūdit . tathā hi

pañcādiarohaṇaṃ nipādatām, dhāivatasadṛśamāntam, pañcamamādhyamamāntam, madhyamādgāndhāraṃbhāntamupabhāntamup pañcāntī .

—A.B. on N.S. 28, 27-28.

cf. also kramayukta iti krameṇārohaṇaṃ tathā svarāṇāmavarohaṇeṇa ca punarapyāharaṇamiti kramaḥ .

—ibid., 28, 33.

इदानों मूर्छनागणडलमुख्यते । तत्र परिपाट्याहि‌तत्वेन प्रवर्त्तत्वात् सप्तस्वराणां मूर्छनानां प्रतिप्रथममेकोनपञ्चाशत् स्वराः

कोष्टकाच्च भवन्ति । तथाहि—

एकोनवत्वं (?) पञ्चाशात् कतेभ्यः स्वरसंख्युलः ॥८५॥

तिर्यगूर्ध्वं च रेखामयरसङ्ख्यभिश्चैव कोष्टका ।

(Cond. on page 240)

240

A Study of Dattilam

The mürchanās of the madhyama-grāma were formed with the same scheme, except that the initial mürchanā in this case began on the note ma. The madhyama-grāma mürchanā-maṇḍala was represented as follows :

The word mürchanā has been derived by the Vṛtti on Bṛhaddeśī from 'mürchā,' which has two meanings: moha or delusion of mind, and samucchrāya which means to build or erect something (mürchā mohasamucchrāyayoh, Vṛtti on Bṛ. 93). Obviously, it is the latter meaning which applies here. Mataṅga in defining the nature of mürchanā states that 'mürchanā is the basis on which a rāga is built or erected (mür-chyate). He further says that a mürchanā consists of the seven notes in the order of ascent and descent.1 Kohala as quoted by the Vṛtti on Bṛ. 118 similarly defines mürchanā as the basis for creating jāti, rāga, bhāṣā etc.2 On the strength of such statements, it is generally believed that mürchanā was the ancient scale. It is, in truth, of all the concepts in ancient texts the one that bears the strongest resemblance to the notion of a

(Contd. from page 239)

तत्र स निघ प म रीत्याख्या: सप्त पडि॒जग्रामे । मर्गरत्ननिधिप्रपञ्च: सप्त मधयमग्रामे त्रियङ्गूषकेगा अपि स्वरासित्यककार्य: ।

— Br. Vṛtti on 94 and Br. 95-96 with Vṛtti.

The murchanā-maṇḍalas of both grāmas follow after a few lines. They are not constructed as the text suggests.

मूर्छना: कथ्यते । ननु मूर्छनाशब्दस्य व्युत्पत्ति: किय॑षी । लक्षणं च कीदृशं तस्य । उच्यते । मूर्छना व्युत्पत्ति: मूर्छा मोहसमुच्छ्रायोः;

मूर्छत् (√ मूर्छ्) ते येन रागो हि मूर्छनेतयाभिसंज्ञिता । प्रारोहणावरोहणक्रमेण स्वर सप्तकम् ।।

—Vṛtti on Br. 93 and Br. 94-95.

तथा चाह कोहल :-

योजनायोग्यतां व्रजेतिर्यं कमो ल‍क्ष्यादनुसारतः । संस्थाप्य मूर्छना जातिरागभाषादिसिद्धये ।।

ścale (or thāṭ),1 seven arising from each grāma with a different initial note. Indeed, in the ancient musical system with its rigidly fixed scheme of determined śruti-intervals between svaras, there could be no other way of obtaining different scales. Besides the seven-note mūrchanā, Mataṅga also speaks of twelve-note mūrchanās (Br̥. 118). It would be worth examining the sense in which this statement has been explained by the Vṛtti kāra, who takes the twelve-note mūrchanā to be the basis of rāga-enfoldment.

Commenting on Mataṅga's notion of the twelve-note mūrchanā, the Vṛttikāra argues that a seven-note mūrchanā is not sufficient for erecting a melodic structure. So in order that a melody may encompass the range of all the three octaves, a mūrchanā of twelve-notes must be accepted; and, this the Vṛttikāra adds, is in fact observed in practice. In support of his view regarding the twelve-note mūrchanās, the Vṛttikāra quotes Nandikeśvara, an ancient authority, who clearly stated that "in order that a (range) in both the higher and lower octaves may be obtained and in order that the (form) of a jāti, a bhāṣā, etc. be established (jātibhäṣādisiddhyartham), the wise must know the mūrchanā to contain twelve notes".2

The Vṛttikāra implies that the twelve-note mūrchanā was a relatively recent concept. Thus he states that although the ācāryas have propounded a scheme consisting of only seven-note mūrchanās, yet in order to obtain a melodic range over all the three octaves, a mūrchanā is in practice rendered with twelve notes : "yadyapyacāryaiḥ saptasvaramūrchanāḥ pratipāditāḥ, sthānatrītayāprāptyartham dvādaśasvaraireva mūrchanāḥ prayuktāḥ iti." (Br̥. Vṛtti on 118). The relevance of a twelve-note mūrchanā is not very clear but some idea of its form can be had from the Vṛtti on Br̥haddeśī. The Vṛttikāra also gives actual examples. He observes that the uttaramandrā which is formed, sa ri ga ma pa dha ni, should have five more notes added to it—two below and three above—the resulting form being : dha ni sa ri ga ma pa dha ni sa ri ga. Rajanī, which stands as, ni sa ri ga ma pa dha, should be, ni sa ri ga ma pa dha ni sa ri ga ma pa dha ni sa ri ga ma.3 Uttarāyatā, dha ni sa ri ga ma pa, should be, ri ga ma pa dha ni sa ri ga ma pa dha ni sa ri ga ma pa dha. Śuddhasaḍjā, pa dha ni sa ri ga, ma should be, ga ma pa dha ni sa ri ga ma pa dha ni sa ri ga ma pa dha.

1 A statement by Abhinava is here suggestive. A remark of his contains the phrase. मूर्छना-निदिष्टमध्यमसप्तक : this indicates that the extent and position of the middle octave—the pivotal octave in ancient as well as current Indian music— was determined by a mūrchanā (see A.B. on N.S. 21, 66).

2 द्वादशस्वरसम्पन्ना ज्ञातव्या मूर्छना बुधैः । जातिभाषादिसिद्धयर्थं तारमन्द्रानुसिद्धये ॥ - Br̥. Vṛtti on 118.

3 There could be a scribal error here. In the 12 note rajanī structure we do not obtain lower notes as in uttaramandrā, which precedes, but only 5 extra high notes. Perhaps the intended form was, pa dha ni sa ri ga ma pa dha ni sa ri ga. It is likely, however, that extra notes were not added to mūrchanās in accordance with any stan dard regular scheme; the other 12-note mūrchanās seem to indicate irregularity.

dha ni sa ri ga ma pa dha ni. Matsarīkṛtā, ma pa dha ni sa ri ga, should be, ma pā

dha ni sa ri ga ma pa dha ni sa and so on.1

The twelve-note mūrchanās were then apparently formed by adding extra

notes to the basic seven-note mūrchanā form. The seven-note mūrchanā thus still

formed the central foundational structure. Later theorists had appended more notes

to this structure so that a melodic range in the high and in the low octave could be

obtained (tāramandrādisiddhaye).2

A mūrchanā was fundamentally a septatonic structure, but it could be render-

ed hexatonic (ṣāḍavīta) or pentatonic (auḍuvita) by omitting one or two notes from

its seven-note form.

Dattila also speaks of mūrchanās which employed the auxiliary notes (sādhā-

raṇakṛtāḥ). The Nāṭyaśāstra and the Brhaddeśī also speak of such mūrchanās,

but their form is not clear from any ancient text. It appears that, in these mūrchanās,

the regular ga and ni were replaced by their altered forms: Kumbha gives a long list

of all the possible mūrchanās that can thus be formed (see S. Raj 2, 2, 1, 321-322,

where the list is given).

Transposition Of Mūrchanās

26A. gāndhāraṃ dhaivatīkuryād dviśrutyutkarṣanād yadi

B. tadvośāmasthitaṁsca niṣādādin vathāsthitān

27A. tato ’bhūd yāvatithyeṣā ṣadjagrāmasya mūrchanā

B. jāyate tāvatithyeva madhyamagrāma mūrchanā

28A. śrutidvayāpakarṣeṇa gāndhārikṛtya dhaivatam

B. pūrvavanmadhyamādyāśca bhāvayet ṣadjamūrchanāḥ3

1तद्यथा-- धनिसरिगमपधनिसरिंग । उत्तरमन्द्र । निसरिगमपधनिसरिमप । रजनी । सरिगपधनिसरिगमप । उत्तरायतु

(उत्तरायता) । रिसरिगमपधनिसरिगमपध । शुद्धषड्ज । गमपधनिसरिगमपधनि । मत्त्सरीकृता । मपधनिसरिगमपधनि ।

अभिनवकान्ता । पधनिसरिगमपधनिसरि । अभिरुद्गता । इति पद्मग्रामे । मध्यमग्रामेऽप्येवमेव । तत्र षड्जग्रामे मूर्छनानां अवन्तिगता

निसरिगमपधनिसरिमप । रागकाली द्वादशस्वरसमूहेन अवन्तिगता । ताम्रव ।

तद्यथा । सोमेरे (?शौरी) । सरिगमपधनिसरिगमप । हरिणाख्या । रिसरिगपधनिस-

रिसमपध । कावेलनता (? कलोपनता) । गमपधनिसरिगमपधनि । तुष्टमध्या (? शुद्धमध्या) । मपधनिसरिगमपधनिसा ।

मार्गी । पधनिसरिगमपधनिसरि । पौष्करी । धनिसरिगमपधनिसरिमप । हूणक ।

-Vitti on Br. 118.

2One big query the twelve-note murchanā poses is : was a range of twelve notes sufficient for

rendering all andient melodies? Many structures might have had a limited range, but certainly

not all. The very notion of having three octaves as the possible compass of musical develop-

ment indicates that many melodies must have utilised more than just a twelve-note range.

Kumbha has, in fact, criticisd the 12-note n̄ūrchanā theory on this and other grounds.

-see S. Raj. 2,1,1, 352-364.

3Sudhā on S.R. 1,4,15 quotes Datt. 26-28 (with a reading identical with above) introducing the

lines with the words : दत्तिलः षड्जग्राममूर्छनानां मध्यमग्राममूर्छनानां च किन्चिद्विक्रियारोणनयमुक्त्तवान । यदाह...

R. Kau. 1. 62 does not quote but refers to the process here described as being set forth by

Dattila : षड्जमध्यमजातीनां मूर्छनानां परस्परम् । किन्चिद्विक्रियोद्वापक्रमुक्त्तवान् दत्तिलः स्फुटम् ॥

29A. ityetā mūrchanāḥ proktāḥ sāraṇāścaiva vainikaiḥ

B. samsthāpya mūrchanāmevaṃ (anyān vaksyāmyanukramam)1

If gāndhāra [in ṣadja-grāma] be raised by two śrutis [and thus] rendered as dhaivata [of the madhyama-grāma, and] concurrently (tadvaśāt) madhyama and other [notes of the ṣadja-grāma] be adjusted (yathāsthitān kuryāt) [to become] niṣāda and other [notes of the madhyama-grāma], then, any mūrchanā in the madhyama-grāma will have the same serial number (tāvatithyeva madhyama-grāma mūrchanā) as it had in the ṣadjagrāmasya mūrchanā).

Lowering the dhaivata [in madhyama-grāma] by two śrutis [and] rendering it as the gāndhāra [of ṣadja-grāma] one should produce the ṣadja [grāma] mūrchanās as before, beginning [however] with madhyama [of the madhyama-grāma].

These, then, are the mūrchanās and sāraṇās declared by adepts on vīṇā (vainikaiḥ). Having established the mūrchanās in this way, I shall now expound the other [topics] in due order.

NOTE :

Dattila here puts forward a process whereby mūrchanās of the two grāmas could be obtained on vīṇā by tuning it to ṣadja-grāma and then making slight variations to obtain the madhyama-grāma on the same tuning (sāraṇā).

The conventionally accepted initial note of the ṣadja-grāma was sa, whereas that of the madhyama-grāma was ma. If on a vīṇā tuned to the ṣadja-grāma one wished to obtain the śruti-intervals pertaining to the madhyama-grāma, then taking the initial sa of the ṣadja-grāma as the initial note ma of the madhyama-grāma one gets pa (of three śrutis) on ri of ṣadja-grāma, which is also of three śrutis. The next note ga of the ṣadja-grāma is of two śrutis and, therefore, to render it as the dhaivata of the madhyama-grāma it needs to be raised by two śrutis (Datt. 26A). As a result, the note madhyama of the ṣadja-grāma which was on the fourth śruti from the original gāndhāra will now be on the second śruti from this altered position. Consequently the new two-śruti madhyama will now become the two-śruti niṣāda of the madhyama-grāma (Bharata here, like Dattila, says: “tadvaśāt madhyamādayo yathāsaṅkhyena niṣāditvaṃ pratipadyante” N.S. 28, 34). The other notes get automatically adjusted at their proper śruti-intervals (see graph 1 on next page). Now the serial number of the mūrchanās in the two grāmas would also be identical.

The process can also be effected with the madhyama-grāma as the base; this Dattila indicates in verse 28. The only difference will be that in this case the third svara, dhaivata, of the madhyama-grāma, in order to be rendered as ga of the ṣadja-grāma will have to be lowered by two śrutis (see graph 2 on next page). Thus Bharata states: “tadvanmadhyamagrāme dhaivatamārdavād dvaividhyaṃ tulyaśrutya-ntarattvācca samjñānyatvaṃ” (N.S. 28, 34).

1 T.ed 29A reads : सरणाश्चैव वैनिकैः and 29B reads : न्यया वद्यानुगः (?) क्रमः

244

A Study of Dattilam

GRAPH 1

sa

(ma)

ri(pa)

(ga)ni

ga

(ri)dha

raised ga

(dha)

pa

(sa)

ma

(ni)

ṣadjā-grāmic graph

(the notes in brackets denote madhya-grāma svaras)

GRAPH 2

(sa)

m

pa(ri)

(r)i(ga)

lowered dha

(ga)

dha

(dha)ri

sa

(pa)

ni

(ma)

madhyama-grāmic graph

(the notes in brackets denote ṣadja-grama svaras)

Bharata has discussed the process at a great length (N.S. 28, 34).

We see here that one grāma could be transposed upon the other with ease and

this apparently rules out the possibility that the two grāmas were conceived as cons-

nine śrutis higher on ma. The process of transposition also graphically reveals that

the structural difference between the composition of the grāmas did not amount to

much.

Yet, the two grāmas in ancient texts have been carefully separated and the

importance of the distinction in ancient practice can in no way be belittled. Tāna,

mūrchanā, even the jātis and the jāti-born rāgas were all classified on the basis of

grāmas. Indeed, the entire range of musical conception, it seems, was classified into

two distinct compartments on the basis of the two grāmas. This obviously leads to

the conclusion that there was some core of difference between the grāmas which was

basic and central to ancient music. This core, however, eludes us.

Abhinava gives us a pointed hint concerning the fundamental importance of

the grāma division in ancient times. He says that when the pañcama of the ṣadja-

grāma is lowered by a śruti (thus giving madhyama-grāma), the process gives rise to a

great deal of distinction (bhūyāneva bhed o jāyate); the saṃvādī scheme becomes dis-

uprooted and, in fact, the entire scheme of svaras is changed.1 Elsewhere, defining

grāma, Abhinava says: “grāma is a collection of svaras containing a specific number

of śrutis; it forms the basis of a group of mūrchanās (mūrchanātmā) and of a distinct

scheme of possible septatonic, hexatonic and pentatonic svara-structures, and of a

distinct plan of graha (initial note in a melody), aṃśa (predominant note) and other

notes: this plan engenders a specific group of jātis (the melodic form specific to

gāndharva) which are encompassed by these distinct characteristics of a grāma.”2

Mataṅga has also connected each grāma to a distinct śruti-svara scheme and

a distinct group of jātis: “jātibhiḥ śrutibhiścaiva svarā grāmatvaṃāgataḥ” (Br. 93).

The Vṛttikāra comments that “the purpose or function (prayojana) of grāma is to

establish svara, śruti, mūrchanā, tāna, jātis and rāgas” : prayojanaṃ ca yathā—svara-

śrutimūrchanātānajātiragānāṃ vyavasthāpanatvaṃ nāma prayojanam” (Br. Vṛtti on 92).

Dattila, in verse 29 (line one), says that the process by which the two grāmas

were correlated to each other was called sāraṇā by the vainikas.3

1 पञ्चमं चापकृष्टं संवाधनवैदिव्यवस्थास्य लोप्य, लोप्या स्वरव्यवस्थां चान्तीयव सम्प्राप्ते इत्थं भूयिष्ठं मतो जायते

—A.B. on N.S. 28, 27-28,

2 ग्रामो नाम विशिष्टश्रुतिकस्वरसमूहो मूर्छनात्मा पूर्णपुंस्त्वसभावस्वरप्रग्रहादिवशेषसमूहकृतसमूहश्च

—ibid., 28, 64-65.

3 Datt. text reads vaiśikaiḥ—‘those versed in the arts of courtesans’—instead of vainikaiḥ, a

reading we think more suitable. Vaiśikaiḥ, too, through lakṣaṇā, can denote ‘an expert in the

arts including music’ but, considering the context, ‘vainikaiḥ’ is clearly a more happy reading.

Kumbha, in fact, after referring to the process of mūrchanā-transposition in S.Raj 2,1,1, 370-74,

remarks:

वैनिकानामयं पन्थाः सुगमः; श्रुत्यादिलक्षणम्

—S.Raj 2, 1, 1, 374B.

It seems quite probable that Kumbha in making this statement had Datt. 29 in mind and the

reading he had before him was ‘vainikaiḥ’.

The term sāraṇā (form √sṛ, to shift, to arrange: to alter a note by raising or lowering it on the vīṇā) here evidently connotes the transposition, through a slight change in tuning, of the mūrchanās of one grāma on a vīṇā tuned to the other grāma in the way indicated by the preceding verses. That such a practice was prevalent among the vīṇā players is indicated by Abhinavagupta who commences his commentary on the relevant passages in the N.S. (28, 34), with the words: “atra vainikasyopadeśārthamaha—this (process) is described for the benefit of the vīṇā players.”

We note that Bharata does not call this process ‘sāraṇā’. ‘Sāraṇā’ in his exposition is the name given to the process of tuning by which the measure of śrutis was determined. However, sāraṇā appears to have been the general term in music for tuning of vīṇās and, as such, both processes were sāraṇās.

Line 2 of verse 29 has a lacuna. We have tried to reconstruct the line on the basis of the context. With this line the exposition relating to mūrchanā ends and a new topic begins.

TOPIC 5

TĀNA

30A. pañcasvarāḥ ṣaṭsvarāśca mūrchanā yāḥ prakīrtitāḥ

B. tānāścatraścitistutā1 tā evāptairudāhṛtāḥ1

31A. agniṣṭomādināmnāstā uktā nāradādibhiḥ

B. devārādhanagena tatpungyotpādakā yataḥ

32A. ṣadjarṣabhanīṣadaiśca pañcamena ca ṣadjagāḥ

B. ṣaṭsvarā madhyamagrāme gāndhārāntaistribhirvinā

33A. svagrāmamūrchanā hyeṭhā kriyamānā nāḥ pṛthak pṛthak

B. bhavantyekonapañcāśadevam pañcasvarā api

34 A. ṣadjapañcamayorvinā dviśrutibhyāṃ tathaiva ca

B. pañcamarṣabhayorvināśca ṣadjagrāme trayaḥ smr tāḥ

35A. dhaivatarṣabhayorvināśca dviśrutibhyāṃ tathaiva ca

B. dvāvetau madhyamagrāme pañcatrimśadamī smr tāḥ

The aforementioned mūrchanās [when formed] with five notes and with six notes, have, indeed, been declared by the wise as the tānas; eighty-four [in all].

Agniṣṭoma and so on are the names given to them by Nārada and others; because consecrated in the worship of gods they produce merit.

In ṣaḍja [grāma] six note [mūrchanās] can be obtained by dropping ṣadja, ṛṣabha, niṣāda or pañcama; [similarly] in madhyama-grāma [they can be produced] by dropping the three notes ending with gāndhāra, (i.e., ṣadja, ṛṣabha and gāndhāra).

In this way are obtained the 49 [ṣāḍava mūrchanās] by separately dropping the prescribed notes in the respective grāmas. The five-note ones are obtained in a similar way (evam) by dropping three [pairs of notes]: ṣadja and pañcama; the two-śruti notes (i.e., gāndhāra and niṣāda) and [lastly] by dropping pañcama and ṛṣabha.

In madhyama-grāma, the two [pairs] to be dropped are these: dhaivata and ṛṣabha, and the two two-śruti notes: in this way there are said to be thirty-five [auduvita mūrchanās] in all!

Note:

Mūrchanā was conceived as a sequence of all the seven notes in their due order. Mūrchanās, Dattila had said, could be rendered as hexatonic or pentatonic (Datt. 25B). Such mūrchanās were the tānas. What then was the difference between a hexatonic or a pentatonic mūrchanā and a tāna? The view of the Vṛtti on Bṛhaddeśī

1 Datt. T.ed. reads : (तनाना ? ताना)

regarding the difference between tāna and mūrchanā is interesting. Ācārya Viśākhila, the Vṛttikāra states, held that there was no difference between the two. But in the opinion of the Vṛttikāra himself such a position was incongruous. Had there been no difference, why should uddeśas (topic lists) enumerate them seperately? He posits that a mūrchanā was a series of notes in the ascending order (ārohakrameṇa), whereas tāna was conceived in a descending order (avaroḥakrameṇa).1 This observation is not overtly supported by any ancient authority who are all silent on the point. Dattila, it seems, subscrided to the view held by Viśākhila for he states that mūrchanās themselves are tānas (tā eva tānāḥ). Bharata, on the other hand, says that “tānas depend upon mūrchanā: tatra mūrchanāśritāstānaḥ” (N.S. 28, 34). This statement relates the one concept to the other in a causal or kārya-kāraṇa-bhāva relation and rules out identity.

The names agniṣṭoma etc. given to tānas are not to be found in the extant Nāradi Śikṣā (ascribed to Nārada) but that they were known through an ancient and unbroken tradition is testified by the fact that roughly the same list of names is obtained in the Vāyupurāṇa, the Brhaddesí, the Saṅgītaratnākara, the Saṅgītarāja and other texts.2 These names suggest (as Dattila and others have pointed out) that the tānas served a sacred purpose during sacrifices and had thus a sanctified status.

In gāndharva structures the rendering of śāḍava and aḍuḍuvita was regulated by very definite rules. Thus to render any ṣaḍja-grāma mūrchanā as śāḍava only four notes (sa ri ni pa) could be dropped. A similar restriction prevailed in madhyama-grāma; sa ri and ga were the only three notes that could be omitted. The process of aḍuḍuvita was still more restricted. In the ṣaḍja-grāma there were only three given pairs of notes that could be omitted at a time and madhyama-grāma had only two such pairs.

Dattila in describing this process has used a very compressed and aphoristic language. Bharata has described the same process in a prose passage where the details are more clear (N.S. 28, 34). The total possible number of hexatonics works out to be 49. The possible number of pentatonic mūrchanās was 35. The total number of both śāḍava and aḍuḍuva tānas was thus 84.

1 ननु मूर्छनातानयोः को वेदः । उच्यते । मूर्छनातानयोनुत्तया (रन्तेन) रत्वमिति विश्रान्तिः (विशाखिलः) एतच्चास्ततम् । संप्रदायोऽपि तु मूर्छनासंतानयोगेन्यैः प्रतिपाद्यतेव । तत्र कथम् । मूर्छनारोहक्रमेण तानोत्तररोहक्रमेण भवतीति भेदः । —Vṛtti on Bṛ. 105A.

2 The Vāyupurāṇa list is the earliest extant one. See 86, 20-27 (Vāyu). The list contains gāndhāra grāma tānas also Ṣaḍja and madhyama-grāma tānas here amount to only 34 : विष्णुस्तस्मैध्यमग्रामे पद्मजाग्रे चकार तु । 86, 30 (Vāyu). Gāndhārā-grāma had 15 tānas, thus totalling 49 for the three grāmas. The Bṛhaddeśī list has 84 tāna names, Bṛ. 106-117 ; but it is faulty. The Saṅgītaratnākara also contains a list of all 84 tānas including śāḍava and aḍuḍuva for both ṣaḍja and madhyama grāmas, S.R. 1, 4, 72-90. The Saṅgītarāja has an extensive tāna-prakaraṇa, S. Raj 2, 1, 1, 412-599. Kumbha, following his predecessors, lists the tāna names ; he also gives extensive charts illustrating the note structure of every individual tāna and indicates the mūrchanās to which they belong. He has a long passage, S. Raj 2, 1, 1, 476-493, where he discusses the relation of tānas to Vedic sacrifices. He says that the merit that accrues from a specific sacrifice can be acquired also through singing the tāna pertaining to that sacrifice : यदयज्ञनामकस्तानो यो यज्ञः परिकीर्तितः । तं तं सम्यग्विदन् गानात्तद्यज्ञफलमश्नुते ॥ —S.Raj. 2, 1, 1, 376.

It should be noted here that in gāndharva ṣadja was not the all-important basic note as it is today. We observe that in both the grāmas, ṣadja could be omitted for the tāna process. Madhyama, we see, was never dropped : it was 'anāśin'. Dattila (and Viśākila as referred to in A.B. on N.S. 28, 64-65) considered that pañcama in madhyama-grāma and dhaivata in ṣadja-grāma were also notes which could not be omitted in the two grāmas respectively. The tāna process bears this out.

The Method Of Dropping Notes

36A. tānakriyā dvidhā tantryāṃ praveśān nigrahāt tathā

B. tatra praveśo dhvanyaikyamasamśparśastu nigrahah1

The process of producing tāna on stringed instruments is two-fold : by praveśa and by nigraha. Praveśa in this context means rendering of a sound as identical (dhvanyaikyam) and nigraha means avoiding [a sound].

NOTE :

On a stringed instrument (tantryām) the process of omitting notes was accomplished in two ways : by praveśa or by nigraha. Dattila's compressed exposition hardly makes the nature of the process clear. Bharata's statements are also not very clear (N.S. 28, 34). The Vrtti on Brhaddesī, however, elucidates the process at a somewhat greater length, quoting Bharata as well as the above verse of Dattila in order to support its contention.

Praveśa, says the Vrttikāra, was effected in two ways. Taking ṣadja as the note to be dropped (lopanīya), he explains the two-fold process as follows : "either tighten (pīdana) the ṣadja, which is lower than the ṛṣabha, so that it is raised (vipra-karṣa) and is rendered as ṛṣabha (cf. 'dhvanyaikyam', Datt.), or lower it through loosening (mardavam śithilīkaraṇam) and render it as niṣāda". The use of the word pīdana for raising the note and śithilīkaraṇa for lowering is suggestive of a harp, which, as we have argued, was the standard ancient vīṇā. On such an instrument each string must have been tuned to a single note, so that in order to drop a note it was necessary to tighten (pīdana) the string tuned to that note and tune it to the next higher note; or, conversely, loosen (śithilīkaraṇa) the string and tune it to the next lower note. Nigraha is explained by the Vrttikāra as simply avoiding the note (or, in other words, the string tuned to that note) that has to be dropped.

Abhinava gives a similar explanation of the process of praveśa. Suppose, he says, ṣadja has to be dropped, then it should be raised and rendered as ṛṣabha. Alternately, in the mūrchanā uttaramandrā (the ṣadja-grāma mūrchanā beginning

1 Br. Vrtti on 117 quotes Datt. 36 with the words : 'दत्तिलेनोक्तम्'.

The quoted reading agrees with the Datt. text. Dattila is quoted by the Vrttikāra as a testimony parallel to Bharata, part of whose prose passage on tāna, N.S. 28, 34, is quoted with the words 'तथा चाङ्‌ह भरत:.' The Vrtti passage which contains the above quotation is quoted by Sudhā on S.R. 1, 4, 29-31.

with sa) sa, when dropped, could be lowered and tuned to ni. The general maxim was

that the note to be dropped was to be tuned to that neighbouring note, lower or

higher, which happened to be a strong note in the melody being rendered so that

further strength was imparted to the note decreed as powerful.1

The Method For arriving At The Source Mūrchanā Of A Tāna

37A. evamkrte'pi tānatve ganayitvā vināśinam

B. vidvānetāvatithyeṣā mūrehanetyavadārayet2

Thus having produced a tāna, the expert can again arrive at the serial number

of the mūrchanā [from which the tāna was derived] by counting (ganayitva) the note

[or notes] that had been dropped [to render the mūrchanā into a tāna].

NOTE :

Dropping certain notes in order to produce a tāna from a mūrchanā, could

lead to uncertainty regarding the mūrchanā from which a particular tāna was derived.

The process described in verse 37 is to help recognize the originating mūrchanā of a

tāna. The maxim recorded is that by replacing the left-out notes in their due serial

order, one arrives at the originating mūrchanā. Then (depending upon the grāma to

which the mūrchanā belonged) one could know the serial number of the originating

mūrchanā by observing the place occupied by ṣadja or madhyama in it (see Datt. 21).

On the face of it, this process appears to be so obvious that it could easily

have remained unmentioned by Dattila, who, on his own admission, seeks brevity.

Bharata, indeed, makes no mention of it. But its importance is realised when one

1 तानार्थक्रिया उपाय इत्यर्थः। 1 अपरस्य ऋषभादिपेक्षया पदजातस्य विप्रकर्षः पीडनमृषभादिमतापादनम् । तस्यैव निपादादिपेक्षयोत्-

मन्द्राय निषादतापादन्, तदनु यो बलवान् प्रयोगे भवति तत् तु अवलोच्योन्तर्मनः स हि दृष्टतामेव न्रजेत् ।

-A.B. on N.S. 28, 34.

The Vrtti on Br. 117 in this context reads :

प्रदेशो ऋषभादिपेक्षया पदजातस्याधरूढत्वस्य लोपनीयत्वे'पि विप्रकर्षपीडनम् ऋषभादावदानम् इति यावत् । 'ऋषभादावदानम्'

is obviously a wrong reading. ऋषभादौ निधानम्, the phrase used by Abhinava who may have,

indeed, borrowed it from the Vrtti, is the correct phrase ; it means : 'to make it (i.e., the ṣadja),

attain or reach the note rṣabha (through viprakarṣa).

Nijenhuis (see Dattilam—A Compendium of Ancient Indian Music, p. 151) has become rather

unnecessarily confused with the meaningless word ऋषभादानम् found in the Vrtti. The Vrtti in

the original form may have read ऋषभादौ निधानम् or ऋषभादावनिधानम्, both phrases carrying the same purport. Nijenhuis, indeed, quotes ऋषभादानम् and निषादापादनम् used by Kala in this context

on S.R. 1, 4, 17 ; but she gives more credence to the Vrtti reading than it deserves.

2 The Vrtti on Br. 117 quotes the above verse with the words :

'ननु वादविवादिते क्रियामाणे मूर्छनाप्रत्यभिज्ञानमस्ति वा न वा । अस्त्येव मूर्छना प्रत्यभिज्ञानमिति । तथा चाह दत्तिलः'.

The Vrtti reads Datt. 37B as 'विचिन्ते तावतिथ्येपा......'.

Abhinava, too, quotes Datt. 37 reading the same as our text. He quotes the verse with the

words :

'कुतः स्थानात् पड़जे लुप्ते हि तत्मध्येऽपिस्थि तुयेन वह्जेन भवितव्यमिति न तानलाभः । तदाह दत्तिल;

-A.B. on N.S. 28, 34.

finds that certian tānas have an identical structure, though they originate from different mūrchanās. For example, in the ṣadja-grāma the following seven tānas can result by dropping the note ṣadja :

ri ga ma pa dha ni

ni ri ga ma pa dha

dha ni ri ga ma pa

ma pa dha ni ri ga

ga ma pa dha ni ri

ri ga ma pa dha ni

It is clear from this chart that the first tāna (ri ga ma pa dha ni) is the same as the last tāna. Thus confusion can occur as to the originating mūrchanās of these two identical tānas. This can only be dispelled by considering the position of the dropped note (i.e., the ṣadja) which, in our illustration, is the first note of the last tāna but the last note of the first tāna. Even more confusing examples exist where serially two consecutive tānas have the same note structure (see S. Raj, charts following 2, 1, 1, 445). In such cases the only way to arrive at the originating mūrchanā could be by considering the place which the omitted note would have occupied.

Abhinava has discussed the process at some length and has done so on the basis of Datt. 37 which he quotes in this context (A.B. on N.S. 28, 34). The example given above has been taken from him.

The Kuṭatānas

38A. kramamutsr.jya tantrīnām tananairmūrchanāstu yāḥ

B. pūrnāascaivāpyapūrnāśca kuṭatānāstu te smṛtāḥ

39A. pūrnāḥ pañcasahasrāṇi trayastriṃśacca saṅkhyayā

B. kathayanti pratigrāmamupāyo ganane' dhunā

40A. hanyādanantarāyeṇa pūrvā yasya kramotkramāt

B. guṇakārasamaśātra kramāḥ śeṣāḥ syurutkramāḥ

41A. sāptasvaryam tu saptānāmekaikā bhajate yatah

B. ata ekonapañciśat kaiścit kūṭaih sahoditāḥ1

1 Br. Vṛtti on 117 quotes both Datt. 39 and 40 and expounds the process contained therein. The Vṛtti quotes Datt. 39 with the words: ‘कूटतानानां महासख्यापि पञ्चभिस्त्रिंशदधिकानि निष्पाद्यन्ते। दत्तिलतायुक्तिसत्तम्’. Datt. 39B reads : ‘प्रतिग्राममुपायोगो नैवस्ति:—obviously a faulty manuscript reading. Next follows Datt. 40 which is quoted, with the comment : गणनामुपरेण तानप्रयोगेण पूर्वस्तातानयोगो हन्यते । तथा चोह दत्तिलः

Line 2 Datt. 40 reads ‘kramācleṣaṁ’ ; rest as in text above. The Vṛtti then gives in detail the process of ‘hanana’ or multiplication indicated by Dattila. Sudhā on S.R. 1, 4, 32-36 also discusses the process of arriving at the number of kūṭatānas. Simhabhūpāla quotes the passage of the Vṛtti on Br. where Datt. 39 is quoted with the words : ‘नतु मत्क्षेप्त द्रविंशदधिकानि पटसहस्राणि (5033) प्रतिमूर्छनान्तं कूटतानसंख्योच्यते; यदाह’. ‘नतु मत्क्षेप्त द्रविंशदधिकानि पटसहस्राणि (5033) प्रतिमूर्छनान्तं कूटतानसंख्योच्यते; यदाह’.

Here the Br. passage has a correct reading. The phrase पञ्चभिस्त्रिंशदधिकानि निष्पाद्यन्ते reads पञ्चत्वस्थितिं मदाधिकानि Here the Br. passage has a correct reading. The phrase पञ्चभिस्त्रिंशदधिकानि निष्पाद्यन्ते reads पञ्चत्वस्थितिं मदाधिकानि

The quoted Datt. 39, in its second line reads as in text above. Clearly Simhabhūpāla had at hand a superior manuscript of the Bṛ.

If strings are so tuned (tantrīṇāṃ tananaiḥ) that, in the resulting mūrchanā, their (i.e., of the strings) proper [svara] sequence (krama) is discarded, then [one obtains] kūṭatānas [which are] known to be pūrṇa and apūrṇa (i.e., those containing all the seven notes and those containing less).

The pūrṇa [kūṭatānas] are five thousand and thirty-three in number. Now, the method of counting [the kūṭatānas] for each of the grāmas is being narrated.

One should multiply the last number with the ones preceding in a series progressing [arithmetically] (hanyādanantarāyeṇa pūrvāḥ). The resulting aggregate will include both the krama and the vyutkrama tānas of which the [number of] krama [tānas] will be equal to the end number in the series (guṇakāra samāstatra kramāḥ); and the remaining will be vyutkrama [tānas]. Since each of the seven [mūrchanās] can [further] be rendered into groups of seven notes (sāptasvaryam bhajate), therefore some also include [these resulting] 49 with the kūṭa [tānas].

NOTE:

The phrase ‘tantrīṇāṃ tananaiḥ’ literally means ‘through spreading or arranging of the strings’ - and again strongly suggests that the instrument Dattila had in mind belonged to the harp group. It is only on a harp or on a similar instrument that the tuning of the strings--as indicated by Dattila--can be obtained without keeping the proper note sequence, i.e., discarding the sa ri ga ma pa dha ni order.

A kūṭatāna, as Dattila explains, was a mūrchanā but with this basic difference that unlike the mūrchanā its note sequence was in disarray. A kūṭatāna when rendered with all seven notes was called a pūrṇa kūṭatāna. When rendered with less than seven notes, i.e., with a hexatonic or pentatonic structure, it was called an apūrṇa kūṭatāna. Apūrṇa kūṭatānas were rendered also with four notes and less.

Bharata, we have observed (see ch. 1), has not described the kūṭatāna. Abhinava does. It was a favourite subject with all later theoreticians and came to be known as khaṇḍameru in many works.

After expounding the nature of a kūṭatāna Dattila mentions the total possible number of pūrṇa kūṭatānas and describes a simple mathematical method by which the kūṭatānas could be calculated. Dattila’s language here (in Datt. 40) becomes extremely cryptic. Fortunately, the Vṛtti on Brhaddeśī quotes Datt. 40 and elucidates the method in detail. The readings in the Brhaddeśī are themselves confusing at times, but as the method itself was very popular with theoreticians and has been described in a number of texts, it will not be difficult to arrive at Dattila’s meaning.1

Let us then attempt to elucidate the purport of verse 40. We shall first deal with the two terms krama and vyutkrama.

Krama in this context means the proper sequence of notes in an ascending order; thus sa ri ga ma pa dha ni is a krama sequence obtained with a pūrṇa or

1 The Bṛ itself has been quoted by Sudhā on S.R. 1, 4, 32-36 with a better reading,

seven-note structure. In the case of an apūrṇa krama, the notes are in a proper serial order, but contain less than seven notes; sa ri ga ma pa, for example, is an apūrṇa krama employing five notes. A pūrṇa krama sequence is nothing but a mūrchanā. A vyutkrama arrangement, on the other hand, is one in which the proper sequence of notes is either disturbed or reversed. Thus ni dha pa ma ga ri sa, sa ga ri pa ma ni dha, dha pa ma ga ri sa, ri ga sa dha pa ma, etc. are vyutkrama tānas. The kūtatānas were such vyutkrama arrangements.1

The method of calculating the tānas is the same as the one used for calculating the possible permutations and combinations of any given number of entities. If seven be the number (the number of svaras in the octave), then the numbers are arranged in an arithmetic series from 1 to 7—1 2 3 4 5 6 7—and the last number (in the words of Dattila) is to be multiplied by all the preceeding ones (pūrvāḥ) in an unbroken sequence (hanyādanantarāyeṇa); thus: 7×6×5×4×3×2×1=5040. The same process will apply in case of any number less than 7. The resulting aggregate includes all possible permutations of a given number of svaras and thus contains both krama and vyutkrama sequences of notes. As the krama sequences are not kūtatānas but mūrchanās, they should be subtracted from the total. Thus in each grāma, with a seven note octave, we have 5040−7=5033 vyutkrama arrangements or kūtatānas; this is the number noted by Dattila in verse 39.2 Similarly with any given number of notes which are multiplied in an arithmetic series, the number of krama sequences will amount to the last number in the series; the rest will be vyutkrama arrangements—as stated by Dattila in the second line of verse 40. For example, with a five-note series there will be five krama tānas and 1×2×3×4×5=120−5=115 vyutkrama tānas.

The above is in essence the process described in the Vrtti on Brhaddeśī (see Vrtti on Br. 117). A detailed list of all possible combinations with 7 notes and less can be found in the appendix to Volume I of the Adyar edition of the Saṅgītaratnā-kara text.

Verse 41 refers to a further list of 49 kūtatānas given by some theorists. The idea appears to be somewhat as follows: we have seen that mūrchanās were seven in each grāma. Supposing any of the mūrchānas, say, ni sa ri ga ma pa dha be itself taken as the basic frame for further mūrchanās, then seven mūrchanās will result. In this way the seven mūrchanās of ā grāma will give rise to 49 such groups of seven notes.

1 Later theorists describe kūtatānas on the lines of Dattila. As Bharata has not expounded the matter, there is in this perhaps a direct influence of Dattila. See for example:

असंपूर्णाश्च संपूर्णा व्यूहमोचचारितस्वरा: । मूर्छना: कूटताना: स्युसततसंख्या मभिद्वधम्मह्ये ॥

-S.R. 1, 4, 32.

व्यूह क्र मस्वरसंस्थाना: सवरिता एव मूर्छना: । कूटतानाभिधाना: स्यु: स्थिरियुक्तं पृथिवीसुजा ॥

-S.Raj 2, 1, 1, 494.

2 Thus Simhabhūpāla, justifying the number given by Dattila, says:

प्रतिमूर्छनं प्रस्तारे क्रियामाणे सप्त क्रमा: संजायन्ते ; तै: सह चत्वारिंशदुत्तराणि पंचसहस्राणि भवन्तीति तात्पर्यं-दत्तिलादिभिस्तु क्रमानिवहाय कूटतासंख्येन निर्दिष्टेत्यविरोध: ।

-Sudhā on S.R. 1, 4, 32-36.

254 A Study of Dattilam

Dattila says, were included by some with the kūṭatānas. Dattila evidently did not consider these arrangements as kūṭatānas because these were all krama arrangements which he excludes from the fold of kūṭatānas by definition.

This above notion of what may be called 49 intermūrchāna arrangements is a rather obtuse one and seems to lie more in the field of mathematical subtlety than of music. Even the concept of the kūṭatānas—so dear to theorists—is not very important musically (probably the reason why Bharata excludes it from his exposition).

It must be admitted, however, that some great composers like Śyāmaśāstri and Muṭṭusvāmī Dīkṣitar of Karnatic music, who were proficient in saṅgīta-śāstra, have been able to draw genuine musical inspiration from their knowledge of kūṭatānas.

TOPIC 6

STHĀNA

42A. atha sthānāni yeṣākto dvāviṃsatividho dhvanih

B. vyastāni tāni ṣaṭṣaṣṭiṃ vidyānmandrādisiddhaye

Now, the sthānas (or the registers), [each of] which, as has been described, contains twenty two distinct sounds. To obtain the mandra and other (octave, i.e. the tāra) these [twenty-two sounds] should be known to extend to sixty-six.

NOTE :

Dattila has already given us an idea of sthāna in another context, as he himself says. The context was that of śruti where Dattila refers to the three anatomic seats of the musical octaves. Now he expounds the topic formally. Ancient music (as also much modern music) mainly employed three octaves: middle (madhya), lower (mandra) and higher (tāra). Registers lower than mandra and higher than tāra might have been utilized as in modern practice, especially in instrumental playing, but their description has not entered into regular theoretical consideration.

The term sthāna basically denoted the anatomic seat (sthāna)1 of a particular octave; by association it also came to mean the musical octave itself. Each sthāna and the octave which resided in it contained 22 śrutis. Dattila evidently takes the madhya octave as the basic one; sounds extending below and above this octave, he says, resulted in the mandra and tāra (vidyānmandrādisiddhaye). The madhya had 22 śrutis; the total range with two added octaves thus had a total of 66 śrutis. Dattila’s language is as usual a bit cryptic. The Vṛtti on Bṛhaddeśī puts the same idea more succinctly: “The twenty-two distinct śrutis of the three distinct registers, mandra, madhya and tāra, taken together result in sixty-six śrutis–such is the opinion of some.”2

Bharata does not expound sthāna in the geyādhikāra, for, as he says, he had already dealt with the topic earlier in speaking of kāku (voice modulation and intonation during dramatic speech) in chapter 17. He refers us to this chapter with the words: ‘sthānām ca trividham pūrvoktalakṣaṇam kākuvidhāvīti’ (N.S. 28, 33). In

1 The S.S.S. in a verse which seems to echo Datt. 42 clearly describes the chest, throat and head as the sthanas:

ūruṇi sthānani hṛkkuṭṭhaṇarāṃcit samāsataḥ | 1 ekaikamapi teṣu sthānad dvāviṃśantivibhidāyutam ||

— quoted by Sudhā on S.R. 1, 3, 7.

also cf. S.R. 1, 3, 7-10 where the three sthānas are described and 22 śrutis assigned to each. See also S. Raj 2, 1, 1 24-27 for similar matter.

2 इदानীं पट्सु पट्श्रुतिमिन्न; श्रुतय: कध्यान्ते । मन्द्रमध्यतारेषु उरु.कण्ठशिरस्सु विषु स्थानेषु प्रत्येकं द्वाविंशतिप्रकारतया

भिद्यमानेषु श्रुतयो हि पट्सुष्टि श्रुतय: भवन्तीत केचित् मन्यन्ते ।

—Br. Vṛtti. on 28.

chapter 17 Bharata speaks of the three sthānas—chest, throat and head (N.S. 17, 104). He also associates these sthānas with the three octaves—mandra, madhya and tāra—while describing the details of the nature and variety of kāku.1 Bharata, in this connection, also speaks of tāratara, anumandra and mandratarā (N.S. 17, 114), terms we do not find in the Dattilam. But these, Abhinava explains, are not octaves beyond the normal three but relatively lower or higher positions within mandra and tāra.2 It should be noted that Bharata like Dattila uses the analogy of the vīṇā in the context of sthāna for the human frame in saying that ‘kāku arises from the three sthānas of the śārīra-vīṇā: chest, throat and head (N.S. 17, 106 ; see fn. below). But he does not, like Dattila, make the point that on the vīṇā frame a relatively higher pitch was obtained by striking on positions relatively lower while in the case of the human frame the pitch positions were reversed.

1. त्रीणि स्थानानि—उरः कण्ठः शिर इति । भवत्यपि च शारीयाणाम वीणानां विभ्यः स्थानेभ्य एव तु । उरसः शिरसः कण्ठात्स्वरः: काकु: प्रवर्तते । —N.S. 17, 106. Also : सर्वेषांप्येषां मन्त्रमध्यात्करत्वात्: प्रयोर्गस्थस्थानेमगत्: —N.S. 17, 130.

2. स्वरकाकोरव पृथकाकुच्चते भेदेन शिरस्येव ऊध्वंस्वरातरम् आख्यातपो, तारतमः; प्रकारोपलक्षणम् । बभूव ह्याक्षेपादौ यथावश्रुतिर्विमागवरकाकुमेवेदयन् विभागनीयम् । उरस्यूर्ध्वभागे समगम्;, उरसयैव नीचभागे मन्द्रतमो नीचः । —A.B. on N.S. 17, 114.

TOPIC 7

VRTTI

43A. dakṣiṇā vrtticiträśca vrttayastās-vayam vidhih

B. pradhānam gītamubhayam adyām ceti yathākramam¹

44A. yad vrttiṣūktamācāryair-vīṇā-vādyā-dilakṣaṇam

B. tad granthavistarā-bhyā-dīha nodāhṛtaṁ mayā

The [three] vṛttis are: dakṣiṇā, vṛtti and citrā. In them the method is [as follows]: in the first, song (gīta) dominates, in the second both [song and] instrumental playing (vādyam) are balanced equally, and in the third, instrumental playing dominates.

I shall refrain from expounding the characteristics of vīṇā and other instrumental [playing] in [various] vṛttis as laid down by teachers, for I am afraid my work will then become too lengthy.

NOTE :

The term vṛtti (from √ vṛt, ‘to be’) in general means the mode of existence or being of any object. In music and dance it has been used in various contexts with different connotations. In gāndharva, it denoted three general modes or ways of instrumental playing when accompanying a song.² Bharata, after delineating instrumental mental techniques (dhātu), states that the three vṛttis form the basis of instrumental playing (yeṣu vādyam pratiṣṭhitam, N.S. 29, 70).

The gīta or the song-form seems to have been the primary form of ancient gāndharva: only song contained all the three necessary attributes of svara, tāla and pada. Instrumental playing was apparently understood in relation to song. Bharata has described the vṛttis in a verse very similar to that of Dattila:

tisṛāstu vrttayaś-citrā-dakṣiṇā-vṛttisamjn̄itāḥ

vādyagītobhayaguṇā nirdiṣṭā-stā yathākramam

Abhinava, commenting on Bharata, elucidates the vṛttis in an interesting passage at the end of which he quotes Datt. 43 in support of his explanation of Bharata’s meaning. Vṛtti, Abhinava says, is in musical practice the notion of dominance or subordination of song in relation to the accompanying instruments. When

¹ T.ed. reads the first word of line 43A as दक्षिणो. Our amendment is based on Abhinava who quotes the entire verse with the words: तथा च दत्तिलाचार्यः (A.B. on N.S. 29, 70-71). His reading of Datt. 43 is the same as given above.

Abhinava says:

गीतविपयो वाद्यस्य मेलनप्रकारः स विधिरिति सामान्यलक्षणस्य वृत्तिः

— A.B. on N.S. 29, 27-28.

The language is somewhat faulty but the meaning is clear enough.

string instruments dominate and song is subordinate, then the vṛtti which occurs is citrā. In this vṛtti the melodic line as rendered on an instrument does not strictly follow the song (gītamukhāpekṣāvirahitam hi vādyam, as Abhinava puts it idiomatically); but instrumental playing strives to create an independent effect of variety, though it does so in accordance with the stipulated procedure laid down for such a mode of accompaniment. Dakṣiṇā vṛtti is the converse of citrā; in it song occupies the position of honour and dominates over instrumental playing which seeks merely to follow and echo the song. The vṛtti called vṛtti occurs when both song and the instrument maintain an equal balance.1

In his zeal for brevity, Dattila has not described instrumental techniques. Bharata describes them under the topic dhātu. (See our discussion in ch. I). The nature of playing will throw further light on the nature of the vṛttis. There were three styles of instrumental playing. These were related to the three vṛttis. The styles were called tattva, anugata and ogha. Tattva was defined as instrumental playing in which the song-form was strictly adhered to. In the words of Bharata, the tattva style echoed all the elements -such as laya, tāla, varṇa, pada and yati-of the structure of a song being rendered (layatālavarṇapadayatigītyakṣa rabhāvakam, N.S. 29-77). Abhinava, giving clarifications, says that tattva was instrumental rendering (vādyam) which aimed at reproducing the song completely in all its structural details comprising elements like tempo, tāla, melodic varṇa (like sthāyī, etc.), pauses and divisions (‘padam’, virāmārūpo vicchedah) and yati (the regulation of tempo), as well as the very syllables (akṣarāṇi) being sung such as ‘devam śarvam’ etc. Such a manner of playing (vādya vidhiḥ) provided a complete accompaniment to the song without the least departing from it.2 Tattva was related to dakṣiṇā vṛtti.

In the anugata style, instrumental accompaniment followed only certain elements of the song-form while departing from others. For example, the tempo maintained on the instrument could be medium while the song had a slow tempo ; two syllables of the song could be rendered by three or four strokes (na sarvaṃ tadrūpamanuharatyapi tu kincidyathā vilambite’pi laye nūnaṃ prayogaṃ madhye

1 The term vṛtti, as Abhinava points out, is used here in both a general and a particular sense ; as both the name of a class and of an individual entity within the class. It is like the case of a brāhmaṇa being also named Brāhmaṇa. The passage on vṛttis in A.B. reads:

उक्तां वृत्तिं निरूपयति । तत्रस्तु वृत्तिर इति । वृत्तिर गीतप्रधानभूता तत्र व्यपहार इति सामान्यलक्षणम् । तत्रवाध्यप्राधान्ये चित्तेर्वृत्तिः । एतदपययेऽस्तु दक्षिणा वृत्तिरिति । गीतं मुख्यं दक्षिणासु गीतेषु प्रधानं तदनुकृतिगौणं व्यावहारः; गीतं हि दक्षिणासु गीतप्रधानासु वाद्यमनुसंवद्धते । वृत्तिरिति सामान्ये विशेषे च प्रयुक्तोऽत्र । ब्राह्मणार्यो ब्राह्मण इति यथा । यथा च दक्षिणासु: । (Quotes Datt. 43)

—A.B. on N.S. 29, 71.

2 लया द्रुतमध्यविलम्बिता वाद्यमाणलक्षणा: । तालश्रचर्यादिकृतः साम्यविरोधो वा शम्यतालादिसहितवीणया शङ्क्य-प्रयोगः । I वर्णः: रशायादि: । पदे विरामरूपो विच्छेद । यति: लोलोग्रादिः । गीतस्मरविभागतः । जात्यंशविशेषोचिततद्-शकभेदंश्चिदयसम्बद्धिता वा गीतिः । रक्षराणि देवन्त शाब्दमिल्यादि (द्रुति) । एतच्च समस्तं गीतं वीणायाः शङ्क्यप्रयोगं स्वरपद ताल निरीक्ष्यप्रकटं दर्शन्ति । मु प्राप्तो । अन्र तेन तयादिमेदं भाावयते प्राग्नोति स्वीकरोति यथायमतत्, न सभावं वाद्यतिनिगोतेन सह समवाय: समवेता स तत्त्वाद्यो वाद्यविधिः ।

—A.B. on N.S. 29, 77-78.

karoti, varṇadvaye gīyamāne tricaturān prahārān karoti, A.B. on N.S. 29-77).1

Anugata pertained to the vṛtti called vṛtti.

In ogha, instrumental accompaniment was comparatively free of the song form. The instrumentalist often took no heed of pauses and divisions in a song, so necessary to communicate its meaning ('anapekṣitagītārthaṃ gītaṃ tasya ca yo'rthaḥ pravṛttividarilakṣaṇo vicchedaḥ so'napekṣito yatra, N.S. 29-78 with A.B.).

This style was related to the citrā vṛtti.

Some other characteristic details of the vṛttis such as different regulations of tempos (graha), etc. to be observed in them are found in the Saṅgītaratnākara and later texts—the basis presumably being the Natyāśāstra (N.S. 29, 72). Citrā vṛtti was, according to Bharata, characterised by a general ‘brevity’; in it the tempo was uniformly fast, the flow of tempo regular (samā yati) with stresses occurring before the beat (anāgata graha). The vṛtti called vṛtti was twice in expanse as to tāla (dvikala) in comparison with citrā; the basic tempo was medium but it underwent a gradual acceleration (srotagatā yati); the stresses coincided with the beat (sama graha). In dakṣiṇā vṛtti, tāla was twice in expanse to that in vṛtti (catuṣkala tāla) ; the basic tempo was slow with a gradual slackening (gopucchā yati) ; the stresses came after the beat (atīta graha).2

These descriptions of the vṛttis give rise to interesting inferences. The nature of instrumental accompaniment in ancient times was, we note, often quite different from what obtains in classical music today. What we have today may be likened somewhat to the dakṣiṇā vṛtti where the instrument follows the melodic line of the vocalist. The other vṛttis appear quite alien to us.

Instrumental playing in the manner of the citrā vṛtti apparently aimed to impart tonal colour and variety to the vocal melody : it did not to follow it in all its contours. This is a manner of accompaniment found in Western music and also perhaps in classical Persian and Arab music where the accompanying instruments do not always copy the song. It exists also in modern Indian forms of light music—but the inspiration in this case is certainly foreign. We are told tbat in citrā vṛtti instruments dominated over song. This sounds quite strange to us today. Even the nature of the vṛtti termed vṛtti, where song and instrumental playing were equally important, is difficult for us to gauge in the light of existing musical forms.

1 Compare Sāṅgadeva :

किचिद्गीतानुहरणाद्द्रै त्वनुगते मतम् । यथा विरतरसङ्कीर्णस्थितिः स्थागतरेखरच॥

विलम्बिते गीतलये वाद्यगद्याद् द्वृतवदनम्

—S.R. 6, 173-174.

2 तिस्रो गीतवृत्तयः प्राधान्येन ग्राह्याः । इ चित्ता वृत्तिदर्शिता चेति । तासां तालगीतिलययदिमाग्रप्रधान्यानि यथास्वं व्यवच्छकानि भवन्ति । तत्र चित्तार्ँ सदृक्षपतवाघुतलय समायातिरनागतपरा॑ प्राधान्यम् । तथा द्रुतो गीतवादित्र-

द्विकलतलामध्यलयगोतोत्कर्षात् समप्रगृहमागाणि प्राधान्यम् । दक्षिणायां गीतचतुष्कलतलविलम्बितलयगोपुच्छायति :।

अतीतप्रगृहमागाणि प्राधान्यम् ।

—N.S. 29, 72.

Abhinava has not accepted this passage as part of the N.S. In the G.O.S. edition it has thus been marked off with a bracket. For passages parallel in later texts to the above in N.S. see : S.R. 6, 168-170 ; S. Raj 2, 1, 4, 380-94 and other texts.

TOPIC 8

ŚUṢKA

45A. vadyaṁ yad gītavṛttisthamagītam samBrayuīate

B. vacitryārthaprayogajñaiḥ śuṣkam tadabhidhīyate

Instrumental playing according to the vṛttis [which depend upon] song-form, when rendered without a song (i.e., by itself), is called śuṣka by those versed in the varied forms of [musical] practice.

NOTE :

The three modes (vṛttis) of instrumental playing, discussed in the previous note, have been termed gītavṛttis by Dattila in the verse above. Bharata uses the same term in a slightly different form as gītivṛtti: "tisso gītivṛttayah", (N.S. 29-71). Vṛttis were called gītavṛttis because their nature was determined by the relative dominance of the song-form which was being accompanied (prādhānyena grāhyā, N.S. 92 , 72).

Śuṣka, as defined by Dattila, was also related to instrumental playing, but it was instrumental music in itself. Yet in śuṣka playing, the instruments, according to Dattila, still maintained the gītavṛtti; they remained, as he says, 'gītavṛttistha'. Apparently, śuṣka had no mode of playing peculiar to it. In śuṣka, too, it was a song that one played, with the difference that there was no vocal melody with which the instrumental melody was being synchronised (agītam samprayujyate).

Bharata has the term śuṣka in his uddeśa (N.S. 28, 13), but expounds the topic under the name of nirgīta and bahirgīta (N.S. 29, 79–113) ; these, Abhinava points out, were the same as śuṣka: "nirgītam bahirgītam śuṣkamiti" (A.B. on N.S. 29, 79).1 Nirgīta played an important role in the ancient pūrvaranga and Bharata has dealt with it in his chapter on pūrvaranga (N.S. ch. 5). He expounds the details pertaining to its instrumental techniques in chapter 29.

From Dattila's short, single-kārikā exposition it appears that śuṣka was instrumental music without song, but Bharata relates nirgīta to singing also : nirgīta, he says, is also known as bahirgīta because it is sung with a string of syllables without using meaningful words 2. He connects Bahirgīta with the vīṇā too : "bahirgīta

1 Elsewhere, 100, Abhinava makes a similar statement : (nirgītamiti ) vādyante tadgate śabdaṅye vīṇāvādyaṅmu

—A.B. on N.S. 5, 31-41.

2 निर्गीत गीयते यस्मादपदं वर्णयोजनात् । असूयया च देवानां बहिर्गीतिमिति स्मृतम् ॥

—N.S. 5, 43.

Abhinava remarks :

गीयमानेऽपि सारङ्केऽपि पदेभ्यो बहिर्गीतिमिति

—A.B. on N.S. 5, 42-44.

is played on the vīṇā with appropriate techniques (dhātubhīḥ) employing varṇa, alañkāra, short and long syllables."1

The manner of rendering nirgīta (nirgīta-vidhāna-samavāya), during pūrva-raṅga, consisted of parts (or modes) called āśrāvaṇā, ārambha, vaktrapāṇi, saṅkho-ṭana, parighaṭṭana etc. (N.S. 5, 9-10 ; also N.S. 29, 80-81). Describing these Bharata notes details of instrumental techniques used in them as well as a group of nonsense syllables related to each. For example, the nonsense syllables connected with ārambha were :

jhaṇṭuṃ jhaṇṭuṃ jhaṇṭuṃ jhaṇṭuṃ jagatiya valitaka diginigicā diṅgle diṅgle titijhala kuajhala jicambuka titicā gaṇapati (jhaṇṭuṃ jhaṇṭuṃ) surapati paśupati cā.

(N.S. 29, 91)

These syllables were apparently related to vīṇā playing, for the method of rendering them on the vīṇā is introduced by Bharata with the words : “asya tu vādyam” (N.S. 29, 91). The syllables perhaps formed the basis for particular strokes such as do dā rā dī rā etc. as in sitar playing today.

Śārṅgadeva classifies instrumental playing as of four kinds: śuṣka, gītānuga, ṇṛttānuga and dvayānuga. Śuṣka he defines as instrumental playing independent of both song and dance. In gītānuga, the instruments accompanied the song whereas in ṇṛttānuga they accompanied the dance. In dvayānuga the instruments accompanied both song and dance (S.R. 6, 16-46). Following Bharata, Śārṅgadeva has also named śuṣka, nirgīta and has described āśrāvaṇā, ārambha etc. in detail (S.R. 6, 178-240). He mentions Viśākhila in connection wiih a certain matter of detail pertaining to āśrāvaṇā (S.R. 6, 183).

1 देवानां बहुमतेन बहिर्‌पितमिति स्मृतम्‌ । धातुभिः विशुद्धवर्णाभिः गुरुलघ्यक्षरान्वितम्‌ ॥ वर्णालङ्‌कारसंयुक्तं प्रयोक्तव्यं बुधैः रथ ॥

—N.S. 5, 41-42.

also धातुवाद्याश्रयकृत्‌ निर्गीतं’ मा प्रणश्यतु

—ibid., 5,33.

TOPIC 9

SĀDHĀRAṄA

46A. sādhāraṇe tu vijñeye svarajātīyupalakṣite

B. svaramadhye tayoḥ pūrvam tat kālyantarau svarau1

47A. jātyantaraṇa sadṛśam yajjātāvupalabhyate

B. ekagrāme ca bāhulyājjātisādhāraṇam tu tat

The two sādhāraṇas are distinguished as the svara-[sādhāraṇa] and the jāti [sādhāraṇa]. The former occurs between two notes : the two svara-sādhāraṇas are kākali and antara svaras.

The abounding resemblance (bāhulyāt sadṛśam) observed in a jāti with another jāti of the same grāma (ekagrāme) is [called] jāti-sādhāraṇa.

NOTE :

We have noted that the two auxiliary notes, antara ga and kākali ni, were called by the generic name sādhāraṇa svaras. Sādhāraṇa was a generic term applicable in certain cases to another major aspect of gāndharva, namely, its melodic-structures, the jātis.

The notion of sādhāraṇa, according to Bharata, expressed something which was intermediary, which fell in-between two major entities and was thus common (sādhāraṇa) to them both. He explains the idea through an analogy : “there is a time of the year”, he says, “when one feels cold in the shade but sweats in the sun ; this is the period when spring cannot be said to have not arrived but winter is not yet fully over” :

chāyāsu bhavati śītaṃ prasvedo bhavati cātapāsthasyā

na ca nāgato vasanto na ca niḥśeṣaḥ śiśirakālāḥ

(N.S. 28, 34)

This Bharata calls ‘the idea of sādhāraṇa as a period of time’ (kālāsādhāraṇatā). Svara-sādhāraṇa was similarly a note which fell between two notes—it was common to both.

Jāti-sādhārana occurred when two or more jātis happened to have the same aṃśa or predominant note in common.2 The aṃśa was one of the chief factors that characterised a jāti and when two jātis happened to be predominated by the same

1 Nānyadeva quotes Datt. 46 at two places in his B.B. First in ch. III with the words दत्तिलाचार्यो-

प्याह (B.B. 3, 151). Datt. 46 is quoted here alongwith Datt. 16 which also deals with sādhāraṇa svaras. In ch VII the line is quoted again ; this time it occurs with Datt. 16A only and is introduced with दत्तिलाचार्योऽप्याह (B.B. (1) ch. VII).

2 जातिसाधारणमेकाङ्गेन समवायाद्वर्त्यं लङ्कणसंज्ञानमिति

-N.S. 28, 35.

amśa, they tended to resemble: a commonness of structure could be observed. Dattila

does not mention the amśa in this connection, but makes a general observation stat-

ing that when many common features are shared by jātiṣ of the same grāma (ekagrāme),

then occurs the jāti‐sādhāraṇa. The sameness of the grāma appears to have been an

important factor in jāti‐sādhāraṇa. This meant that the jātiṣ which had the sādhāraṇa

relation had to share the same svara‐scheme with its identity of samvāda etc.—an

essential factor in any structural likeness.

The reading in the Kāśī and the Asiatic Society editions of the Nāṭyaśāstra

suggests that a resemblance between amśa and other features of two or more jātiṣ

belonging to different grāmaṣ could also form jāti‐sādhāraṇa.1 The reading accepted

by Abhinava is silent regarding the grāma aspect.

Later theorists were also not quite unambiguous in their opinions regarding

the nature of jāti‐sādhāraṇa, evidently due to the differences found in ancient accounts.

Kumbha mentions three alternative views : (a) jāti sādhāraṇa takes place between jātiṣ

of the same grāma having the same amśa ; (b) jāti‐born rāgaṣ are designated as jāti

sādhāraṇa (Kumbha attributes this view to Dattila among others.2 We do not know

his source of information. No indication of such a view is to be discovered in the

Dattilam) ; (c) similarity of amśa and other features between jātiṣ belonging to different

grāmaṣ is called jātiṣ‐sādhāraṇa (Kumbha adds that this last view is not preferred by

the wise : “na tatsujñajanapriyam”, see S. Raj 2, 1, 2, 36‐38).

Sārṅgadeva lists two opinions: he does not mention the last view and does not

associate Dattila with any of the views (S. R. 1, 5, 9‐10). Siṃhabhūpāla ascribes the

first view to ‘Bharata and others’ (bhāratādyah).

In spite of common features, the ‘sādhāraṇa’ jātiṣ could yet be recognised as

distinct. Abhinava reads a short phrase as part of the Nāṭyaśāstra text in this context.

This phrase points out factors that differentiated sādhāraṇa jātiṣ. The phrase is :

“nyāsāntaramārgau tu viśeṣakau.”3 Its purport is that sādhāraṇa jātiṣ could be dis‐

tinguished through their nyāsaṣ and their antaramāargaṣ. Nyāsa in a jāti was its

final note and antaramārga waṣ its characteristic melodic movement (what in Hindu‐

stani music is known as the ‘calana’ of a rāga). Abhinava, elucidating Bharata's

phrase, remarks that jātiṣ were formed through a substantial number of distinct

characteristics (numbering ten as listed in texts on gāndharva) ; thus even when the

1जातिसाधारणमेकग्रामांशानां जातीनां जत्योर्वान्यस्मिन्‌ग्रामे प्रत्यङ्गदर्शन स्वराणामवमवादगमात्‌।

‐N.S. 28, 35 (Asiatic ed.); also Kāśi ed. 28, 33.

2दत्तिलाख्याः पुनरिदं रागानां प्रचक्षते ।

‐S. Raj 2, 1, 2, 37.

3Abhinava reads this sentence after :

जातिसाधारणमेकांशानामविशेषो$प्यजातीना समवायात्प्रत्यङ्गं लक्षणसंज्ञानमिति ।

‐N.S. 28, 35.

In his opinion लक्षणसंस्थानम्‌ in N.S. was the indicātor of the fact that non‐common elementṣ

between common jātiṣ could be readily recognised : लक्षणमसाधारणं शक्यं ज्ञातुम्‌ । The A.B. text

readṣ न्यासान्तरभागो तु instead of न्यासान्तरमार्गो तु but the former makes no senṣe in the context

and the latter reading—which we have adopted—seems to have been the intended one,

264 A Study of Dattilam

amśa between two or more jātiś happened to be common one could distinguish them through other features which were not common. In this way sādhāraṇa jātiś could be made out as ‘not-identical’ (tena nābheda iti bhāvah).1

In our present system many rāgaś belonging to the same ‘thāṭ’ or ‘mela’ and sharing the same svaraś (whether ṣādava, auḍuva or sampūrṇa) tend to resemble each other strongly, but there are factors such as melodic movements, stresses on particular noteś and characteristic phraseś which clearly reveal them to be different. Jātiś which shared common traits were, it seems, similarly distinguished.

Bharata and Dattila, in the coatext of gāndharva, speak of only two sādhāraṇaś: svara and jāti. Later theoriśtś, however, extended the notion of sādhāraṇa to śruti, grāma and also tāna. They speak of śruti-sādhāraṇa, grāma-sādhāraṇa2 and tāna-sādhāraṇa. Rāṇā Kumbha haś discusśed theśe extra sādhāraṇaś at length. He argues that the notions of tāna-sādhāraṇa and grāma-sādhāraṇa were redundant and born of misconceptionś.3

1 महन्यासादिवक्ष्य यः समवायस्सलो हेतोरजातीनामसौ (s) विच्छिद्येऽपि लक्षणसाधारणं शक्यं ज्ञातुं, यदाहुर्न्यासान्तरमार्गों तु विशेषकविति । तुब्यंस्तिरेकमा हू तेन नाभेद इति भावः ।

—A.B. on N.Ś.28, 35.

The A.B. text readś हेतोरजातीनामसौ विच्छिद्येऽपि. Obviously a lupta-akāra haś been dropped after अंशो due to inadvertence of printing. Without the lupta-akāra, Abhinava's purport remainś contradictory and confuséd.

2 Abhinava, too, mentionś śruti and grāma-sādhāraṇaś.

—A.B. on N.S. 28, 35.

3 पणिडतं तन्मन्यमानेषु केष्चिद्वलोदपवर्णितम् । तानसाधारणं तत्र युक्तिलेख्यो न विच्यते ॥

—S. Raj 2, 1, 2, 47.

also : ग्रामसाधारण केष्चिद्वदिपुरतदिव्रिदः:

—ibid., 2, 1, 2, 13.

TOPIC 10

JĀTI

48A. jātayo aṣṭādaśa jñeyātāsāṃ saptavarākhyayā

B. śuddhāśca vikṛtāścaiva śeṣāstatsañkarodbhavāḥ1

The jātis should be known to be eighteen. Those named after the seven svaras are the śuddha [jātis] along with their vikṛtas (or modifications). All others arise out of the intermixture of these.

NOTE :

The nine preceding topics outline the basic scheme of ancient melodic structures in general and gāndharva in particular. They deal with matter both doctrinal and formal, in terms of which gāndharva was analysed and understood. Now we come to the actual melodies of gāndharva: the jātis. These were eighteen in number. Dattila, in expounding the jātis, first classifies them. There were basically two classes: (1) the śuddhas and their vikṛtas and (2) the saṅkaras. The saṅkaras arose out of the intermixture of the jātis belonging to the first class. Dattila takes up the saṅkaras after classifying jātis and names the source jātis of each of the saṅkara forms in the next six kārikas. He deals with the nature of the śuddha and its implied distinction with a vikṛta in a later kārikā (see Datt. 62).

Dattila says that the śuddha jātis were named after the seven svaras. These jātis, it follows, were seven in number; they were called: (1) ṣāḍjī (2) āṛṣabhī (3) gāndhārī (4) madhyamā (5) pañcamī (6) dhaivatī and (7) niṣādinī. Ṣāḍjī, ārsabhi dhaivatī and niṣādinī belonged to ṣadja-grāma, the rest to madhyama-grāma.2 Each śuddha-jāti could result in a number of vikṛtas, but these vikṛta jāti-forms had no separate nomenclature apart from their parent śudha-jāti: they were, in fact, not considered as separate jātis but modified forms of the śuddhas: "śuddhanāmeva hi vikṛtatvaṃ na tu jātyantaratatvamuktam" (A.B. on N.S. 28, 46).

The Saṅkara Jātis And Their Sources

49A. ṣadjāyā madhyamāyāśca samsargāt ṣadjamadhyamā

B. ṣadjāyāstvatha gāndhāryā jāyate ṣadjakaiśikī

1 B.B.(I) ch. VII quotes Datt. 48 with the words 'तथा च दत्तिलः'. The reading in B.B.(I) is slightly faulty. Datt. 48B (the last phrase) appears to read : 'शेष (पा) स्तु य (सं) करोद्भवा'

2 Bharata says : स्वरजातयः शुद्धा विकृताश्च । तत्र शुद्धाः षड्जी, आर्षभी, धैवतो निषादिनी च षड्जग्रामे । गान्धारे मध्यमा 'चमीति' मध्यमग्रामे । एतावानेवोक्तस्वरः स्वररागप्रहृत्यावाप्यन्वयः साङ्गः ।

50A. tayoreva sadhaivatye ṣadjodīcyavati bhavet

B. īsāṃ samadhyamānām tu gāndhārodīcyavā bhavet

51A. gāndhāryā madhyamāyāśca pañcamāyāścaiva saṅkarāt

B. sadhaivatīnāmāsāṃ tu madhyamodīcyavā bhavet

52A. īsāṃ syād rakta­gāndhārī naiṣādī cec­cartthikā

B. ārsabhyāstu bhavedandhrī gāndhāryāścaiva saṅkarāt

53A. anayostu sapañcamayornandayan­tīb prajāyate

B. saniṣādastvagāndhāryāḥ kuryuh kārmāravīmīmāḥ

54A. gāndhāri pañcami caiva tathā gāndhārapa­ñcamī

B. ārsab­hīdhaivatīvarjāḥ kaiśikīmiti saṅkarāḥ1

From the conjunction of ṣadjā (i.e., ṣādjī) and madhyamā [jātis, is produced] the [jāti] ṣadjamadhyamā. And from [the conjunction] of ṣadjā and gāndhārī is born ṣadjakaiśikī. These very two (i.e., the ṣadjā and the gāndhārī) along with the dhaivatī form the ṣadjodīcyavatī. These (i.e., the ṣadjā, gāndhārī and dhaivatī) combined with the madhyamā result in the gāndhārodīcyavā.

The intermixture of gāndhārī, madhyamā and pañcamī together with the dhaivatī produce madhyamodīcyavā. Along with these (i.e., gāndhārī, madhyamā, pañcamī) if the fourth is the naiṣādī [then arises] rakta­gāndhārī. Ārsabhī inter­mixed with gāndhārī becomes the andhrī. From these two (i.e., ārsabhī and gāndharī) together with the pañcamī is born the nandayantī. These (i.e., ārsabhī, gāndhārī and pañcamī) omitting the gāndhārī (agāndhāryah) [but] along with niṣadī make the kārmāravī. Gāndhārī and pañcamī [make] the gāndhārapañcamī. [Intermixture of all the jātis] excepting the ārsabhī and the dhaivati [make] the kaiśikī. Thus ends the [enumeration of] saṅkara [jātis].

NOTE :

After classifying jātis, Dattila enumerates the saṅkara or the intermixed jatis that arose out of various combinations of the seven pure or their vikṛta jātis. Ṣadjā has been called ṣādjī by Bharata.

It may be noted here that Bharata and later authors held that it was a combi­nation of the nāma-jātis (as these seven jatis were also called) in their different vikṛta

1 The Vṛtti on Br. 187 quotes eight verses pertaining to the jātis and ascribes them to Bharata with the words: ‘तथा चाऽह भरतः.’ However, except for the first two verses in this quotation, which correspond to N.S. 28, 46-47, the rest are quoted not from Bharata but from Dattila. They correspond almost word by word with Datt. 49-54. For our discussion on the value to be accorded to this testimony see ch. II. Line 50A in the Br quotation reads सदैवत्योः instead of सदैवल्योः. 53 B reads ‘सनिषादा स्वगान्धार्याः’ instead of above text. T.ed. reads: सनिषादस्तु गान्धार्यै; Our reading is based on considerations deduced from other accounts,

or modified forms that produced the intermixed (sañkara) jātis1 and not in their śuddha forms. Dattila has nothing to say here. Abhinava is categorical that the modified jātis alone (and not the śuddha ones) give rise to the sañkaras ('vikṛtā' iti na tu śuddhā ityarthah, A.B. on N.S. 28-46). This seems likely, since the structure of śuddha jātis was relatively rigid.

Dattila has recounted the combinations that gave rise to the 11 intermixed or sañkara jātis in an extremely terse and compressed language, verging at points towards the riddle. However, Bharata and others are more specific and precise and we can easily reconstruct Dattila's intended meaning on the basis of their exposition.2

Certain of these 11 sañkaras expressed their parentage through their very names: e.g., ṣadjamadhyamā and gāndhārapañcamī. O hers were named after the regions where they were supposedly more popular, such as andhrī (from Āndhra; Bharata indeed calls it āndhrī) and the three udīcyavās (from the North). Though born of certain mixtures, these jātis were evidently not heterogeneous structures but homogeneous wholes. Even today many of our better rāgas-born of combining two or more rāga forms-are by no means mere 'combinations'3 but have an individual entity of their own.

In gāndharva only 11 sañkara forms were accepted. These along with the 7 śuddhas (and their vikṛtas) were the sanctified forms taught by Brahmā himself (jātayo 'ṣṭādaśetvevam brahmaṇābhihitāḥ purā, N.S. 28, 39). Evidently, many more 'sañkaras' were theoretically possible. A number of these were, indeed, actualised into musical forms. They were, however, outside the fold of gāndharva. Abhinava, commenting on N.S. 28, 46 (where Bharata says: "punarevāśuddhakṛtā bhavant-yekādaśāśānyāstu") remarks that the expression 'punareva' in this context expresses a sense of exclusion; it denotes that the sañkara jātis born of the vikṛtas are exclusively 11, they are no more than that number: "punareveti śuddhābhirvikṛtāḥ kṛtāḥ tābhist-vekādaśeti; punaḥ śabdo vyatirekārthaḥ; ekādaśaiva nādhikā" (A.B. on N.S. 28, 46).

1 तत्र एकादश संसर्गजा विकृताः परस्परसंसर्गदोषादेकादश निषर्वयन्ति यथा— शुद्धा विकृतास्तश्च हि समवायाज्जातयस्त्वस्रु जायन्ते । पुनरेवाशुद्धकृता भवन्त्येकादशैव हि तादृशाः सन्तु ॥ —N.S. 28, 46

The use of the negative prefix 'अ' before परस्परसंसर्गजा seems to be an erroneous reading. Abhinava clearly remarks : तत्रां संसर्गात......

2 Bharata says : पञ्जीगान्धारमध्यमां निवृत्ता जेया पञ्जीगान्धारमध्यमा, पञ्जीगान्धारौर्ज्जयतीभिः पञ्जकैः शिखी, गान्धारौषधजर्ज्जरतीभिः पञ्जोदरीच्यवा, पड्जगान्धारोमध्यमार्ज्जवतीभिरन्जारोदरीचवा, धैवतीपञ्चममध्यमांश्यारीभिमार्ज्जयमोदरीचवा, गान्धारोमध्यमापञ्च- मीनिषदवतोभिरक्तगान्धारौरी, गान्धारौपञ्चमांश्यामान्ध्री, पञ्चचम्यार्ज्जिभीगान्धारौरीभिनन्दयन्ती, निषादवत्यार्ज्जीभिपञ्चमीभिः कामररखी, पञ्चचमीगान्धारोमध्यरी गान्धारपञ्चचमी, धैवत्यार्ज्जीभिपञ्चमीवरर्ज्जिभिः कैशिकी । —N.S. 28, 48.

Some recensions give the same matter in verse form. See also the jātiprakaraṇa in S.R. and S. Raj.

3 Some rāgas are, however, no more than mere combinations, for example, basanta-bahār, aḍāne ki bahār and similar other heterogeneous combinations made with bahār as the base in Hindustani music.

55A. grahāmśau tāramandrau ca śāḍabaudubhite kramāt

B. alpatvaṃ ca bahutvaṃ ca nyās'opanyāsau eva ca

56A. evametad yathājāti daśakaṃ jātılakṣaṇam

B. lakṣaṇaṃ daśakasyāsya saṅkṣepenābhidhīyate

57A. tatra grahastu gītaḍisvaro'mśaḥ pūrvakīrtitah

B. pañcasvaraparāstara uccairamadhyeseṣvānte

58A. āśadjānnandayanityāṃ tu varo nātaḥ praśasyate

B. mrdu ramśaparo mandro nyāsāntastatparo'pi vā

59A. ṣaṭpañcavarake gīte śāḍabaudubhite kramāt

B. alpatvaṃ ca bahutvaṃ ca prayogālpabahutvataḥ

60A. gītakāntyasvaro nyāso vidārīmadhyagastathā

B. nyāsavatsyādapanyāso yathājāti bravīmyaham

61A. yaṃ vinā hīnatā yasyaḥ syāccet tasyāṃ tu so'lpakāḥ

B. amśādyamanyannyāsattu¹ svarajātisu nāmakṛt

62A. tadgrahā tadapanyasā tadasmā ca yadā bhavet

B. mandranyāsā ca pūrṇā ca śuddhā jātistadocyate²

(1) Graha and (2) amśa, (3) tāra and (4) mandra, (5) ṣāḍava and (6) auḍubita,

(7) alpatva and (8) bahutva, and also (9) nyāsa and (10) apanyāsa—these, in due

order, are the ten characteristics of jātis [applicable] to them in accordance with their

form (yathājāti). These ten characteristics are now being expounded in brief.

Graha is the commencing note of a song (gīta). Amśa has already been

defined (see verse 18). Here [in the jātis], movement in the tāra ought to be as high as

the fifth note from the amśa.

1 Datt. T.ed. reads अंशाद्यमनन्याससस्तु…a reading which does not make sense.

2 Br. 195-197A. are almost identical with Datt. 55-56. Br. does not acknowledg : the verses to be

quotations. Datt. 56A in the Br. has a very corrupt reading, otherwise the passage conforms to

the Datt. text. Br. 251 is again same as Datt. 62. The Vrtti on Br. 251 quotes Datt. 61A as an

authoritative pronouncement without naming the source. The line is given in support of a

statement made by the Vrttikāra with 'कृतु:'-and then occurs the remark इति बचनात्.

B. B. (I) quotes quite a few of the above lines, acknowledging the source as Daṭtila. Part of

Datt. 57A occurs in B. B. (I) ch. VI with the words 'यथा च रसिक:' . The line reads as 'मुहुर्गीतादि

प्रथमालाप'. Same line occurs again in B.B. (I) ch. VII with 'वर्तिलोप्याह'. Here the reading is same as

ours. Datt. 58A is quoted in B.B. (I) ch. VII with 'तथा च वदति:'. The line is corrupt. The first

phrase reads ; 'आ वरत:' . Datt. 58B is quoted in B. B. (I) ch. VII as parallel testimony to N.S. 28,

71B with the words : 'दत्तिलोेप्याह'. The end phrase reads : 'मन्द्रो भवेत्'. Datt. 60A is quoted twice

in B. B. (I) ch. VI ; once it occurs along with Datt. 8 and Datt. 61B. The matter is introduced

with the words : 'द्विविधो मन्द्र:' . अधुन् (अत्र) परो न्यासपरकत्वं . यदाहु मन्द्रस्तु (?स्वरं) शपरो नासित त्वा (?न्या) तु हो व्यवस्थितो इति 'दत्तिलाचार्यैषां त्रितयमप्य महीतं'.

Datt. 60A in this quotation reads as 'गीतान्तवस्थितो न्यासो'. Datt. 60A is quoted again in B. B. ch.

(I) VII and reads : अंशभूतो लपन्यासी'. Datt. 61B in B. B. (I) ch. VI reads 'अंशस्तु मन्द्रन्यासस्तु…

It is quoted again with Datt. 62-64 in same chapter. A. B. on N. S. 28, 73, quotes part of 60 A

तथा च दत्तिलोकाम् : । द्वितारोन्मध्यगतस्य इति. Datt. 61B in T.ed. reads 'मन्द्रव्यासा', a reading which

does not make sense. We have adopted the reading 'मन्द्रन्यासा' as given in B. B. (I) where the line

reads : मन्त्रच्यासाह (?च) पूर्णाच्च. मन्त्रन्यासा is also the reading found in the Br. interpolation of the verse

(Br. 251). B.B. (I) thrice quotes Datt._62 each time readings, मन्त्रन्यासा, twice in ch. VI and once

with Datt. 48 in ch. VII,

In [the jāti] nandayantī [tāra] is granted as desirable (varah) upto the ṣadja, and is not commended (praśasyate) beyond it. [The limit of] mandra [in all the jātis, extends upto] the aṃśa in the lower octave (mṛduramśaparah) or alternatively it may end with the nyāsa, or even with the next [note] below the nyāsa (tatparo'pi vā). The gitas [consisting] of six or five notes respectively are [called] sāḍava and auḍuvita. Ample or scarce use [of a note in a jāti] results in bahutva or alpatva [of that note].

The note on which a gīta (song) terminates is [called] the nyāsa. [The note] which forms the nyāsa within a vidārī is [called] apanyāsa. I shall enumerate [these] while describing the different jātis.

That [note] is alpakā in a jāti, whose omission in it makes [the jāti] hīna (i.e., renders it ṣāḍava or auḍuvita). In jātis [named after the seven] svaras (svarajātiṣu), the svara after which the jāti is named (nāmakṛt) should (always) remain its nyāsa : aṃśa and other [features] can vary.

A jati is called [śuddha], if all seven svaras are used in it (pūrṇa), if the final note (nyāsa) is in the lower octave (mandra) and if [the svara after which it is named] is its graha, its apanyāsa and its aṃśa.

NOTE :

Like rāgas which have descended from them, jātis had a well-defined structure. The ten characteristics of jātis were the chief elements that governed their structure. Bharata gives the same list of characteristics as Dattila, calling it likewise ‘daśakaṃ jātilakṣaṇam’.1

Graha was the initial note of a jāti while aṃśa was the note most copiously used. The rule of tāra indicated the limit of movement in the higher octave while the range of movement in the lower octave was indicated by mandra. Ṣāḍava and auḍuvita as jāti lakṣaṇas denoted the stipulated rules and conditions for rendering a jāti hexatonic and pentatonic. Strict rules regarding the notes that were to drop-ped were observed. Alpatva (lit. the state of being diminished) was the principle of rendering certain notes in a jāti weak, while bahutva was the opposite. Bahutva could also be enjoined to operate on notes which otherwise need not have been strong notes such as the aṃśa and its samvādī necessarily were. Nyāsa was the final note in a jāti while apanyāsa was the last note of an intermediary melodic phrase.

The Aṃśa

Aṃśa, says Dattila, was the vādī, the predominant note in a melodic structure. He gives no further details. Bharata has, however, summed up the functions of aṃśa in a couple of kārikās :

tatrāṃśo nāma : yasmin bhavati rāgaśca yasmāccaiva pravartate

1 दशकं जातिलक्षणम्...

बहुलांशो तारमध्यो च न्यासोऽन्यासोऽथ वा तथा । षषड्वोऽपि च बहुल्वं च अल्पत्वं च षाडवोडुविते तथा ॥

• N. S. 28, 66.

270 A Study of Dattilam

mandraśca tāramandraśca yo'tyarthaṃ copalabhyate grahāpanyāsavinyāsasannyāsanyāśagogocarah anuvṛttaśca yasyeha so'mśaḥ syaddāśalakṣaṇaḥ (N.S. 28, 68-69)1

"The amśa has 10 characteristics : (1) it is the svara on which the rāga (the aesthetic charm) depends and from which it is generated; (2) it determines the mandra and (3) also the tāra and mandra; (4) it is the note most frequently heard; it determines the (5) graha, (6) apanyāsa, (7) vinyāsa, (8) sannyāsa and (9) nyāsa; (10) it is the note which the others follow."

Abhinava explains : amśa is the note, the presence of which creates rāga, i.e., charm, and it is the note which also formulates a jāti, just as the head formulates the form of a man : "amśamāhā 'yasmin bhavati rāgaśceti', yasmin vidyamāne rāgo raktirjātisvarūpam ca bhavati, śirasīva puruṣasvarūpam." It is the note which is used more often than any other note : "yaśca samastasvarāpekṣayā bāhulyena bhāti." In relation to amśa, other notes are established as samvādī, anuvādī etc., and on it depend the five factors, graha, apanyāsa, vinyāsa, sannyāsa and nyāsa2. It is the note which also governs the movements in the tāra and mandra octaves3 and is thus the main note.

The graha

Bharata defines graha as the first note of a song, a note with which any song begins (yatpravṛttam bhavegeyam, N.S. 28, 67) So does Dattila : "grahastu gītādisvarah." Some modern scholars have interpreted Bharata as having taken graha as synonymous with amśa" and amśa according to them is the tonic (Alain Danielou, Northern Indian Music, p. 111). One can, of course, read such a meaning in Bharata and interpret the verse where graha is defined (N.S. 28, 67) accordingly. Dattila, too, can bear a similar interpretation. By taking the clause "amśaḥ pūrvakīrtitah" as qualifying "grahastu," the first line of Datt. 57 can be interpreted to mean that "graha is the same as amśa, which has already been defined".

However, the actual words of Bharata, in this context, are "amśa and graha [can act] as alternatives" (amśo graha-vikalpitah, N.S. 28, 67). This does not mean that the two terms are synonymous, for it is clear that functionally graha was distinct from amśa, though in practice the same notes performed both the functions. It was only in this sense, as Abhinava indicates, that amśa could also be called graha: because this other function of acting as the graha was added to it (tenāmśa-syaiva dharmāntarayogannāmāntaram, A.B. on N.S. 28-67). It would, therefore, be erroneous to conclude that the two concepts were synonymous or that

1 Some manuscrips—not accepteḍ by Abhinava—insert another kārikā in between : अनेकस्वरसंयोगे योऽर्थप्रतिपादनक्षमे । अन्येषु च बलिनो यस्म संवादी चानुवाद्यपि । The G.O.S. ed. marks the kārikā off with a bracket.

2 यथ्च स्वयमेव स्वसंवादिनुबादित्वं वा ग्रहादिपञ्चकरूपं वा करोति न तु कदाचिद्विवादिनम् । ग्रहादयो गोचरो नियतस्व्यो विषयो यस्म तथैव । —A.B. on N.S. 28, 68-69.

3 तदस्थमेव तारावृत्तिमध्ये च वक्ष्यमाणपञ्चकमादित्यादिना तारमन्द्रस्थ्यति: । —A.B., ibid.

functions of graha and amśa were identical. Abhinava here quotes the case of the jāti nandayantī where the different functions of graha and amśa were obvious ; for in this jāti, pañcama was the amśa whereas the graha was gāndhāra.

Range in tāra and mandra

Tāra and mandra, in a general sense, meant the higher and the lower octaves, respectively. But when used in the context of jāti-characteristics, these terms acquired a somewhat different implication. The tāra of a jāti indicated the highest note in the higher octave up to which a jāti could ascend; similarly, mandra was the lowest note in the lower octave upto which it was permissible in a jāti to descend. To avoid confusion between tāra and mandra in general and these notions as jāti-characteristics, Bharata uses the expressions tāragati (movement in the higher octave) and mandragati (movement in the lower octave) instead of just tāra and mandra.

The general rule regarding the limiting extent of the melodic movement in the tāra octave was summed up in a formula: one could move as high as the fifth note from the amśa in the tāra octave. Thus when the note ri was the amśa in a jāti, the movement in the higher octave could ascend upto the note dha which is the fifth note beginning with ri. However, since only three octaves were conceived in the jāti system, jātis with ma, pa, dha and ni as amśas could only ascend in the higher octave upto the higher ni; the octave beyond, namely, the atitāra, being out of bounds (madhyama-pañcamadhaivataniṣādeṣu niṣādantā eva te grāhyā, A.B. on N.S. 28-70).

Dattila has specifically mentioned only the fifth note after the amśa as the limit of movement in tāra. Bharata states that the movement could either be upto the fourth or the fifth note after the amśa (amśāttāragatiṃ vidyādaca-turthasvarādiha āpañcamātpañcamādvā, N.S. 28, 70). The Vrtti on Brhaddesī states that in some cases one could ascend upto the sixth note from the amśa.1 Kallinātha (on S.R. 1,7, 35-36) states that the fifth note from the amśa was the limit in the ṣadja-grāma whereas in madhyama-grāma the limit was upto the fourth note. We do not know his source of information in this matter.

Abhinava's comments on tāra-movement are worthy of attention. Giving an example he says that if ṣadja be the amśa, then one could move in the tāra upto ri ga ma pa, provided one was skilled enough to reach such a high note, but not higher even if one was capable of doing so. No blemish, however, lay in moving only upto notes lower than the highest permissible one: "tatra yadā ṣadjo'mśo bhavati tadā tārasaptakaṃ rigamapa ityetatparyantaṃ grāhyaṃ yadi śaktirasti na tu tataḥ param śaktiyoge'pi, ūnagrahane tu na doṣa iti paraśabdena darśitaḥ" (A.B. on N.S. 28, 70). With ri as amśa, Abhinava adds, one could move in the tāra upto dha; with gāndhā-ra as amśa one could sing all the seven notes of the tāra octave upto ni (gāndhāre v amśe niṣādantā sapta). With madhyama the rule was to move only to the fourth

1 कदाचित् पञ्चस्वरोहणं वि (? रोहणमपि) तार; । —Br. Vrtti on 196.

higher note, i.e., tāra ni, if one could move so high.1 In effect, with the five notes gā ma pa dha and ni as aṃśa the whole of the tāra octave remained the potential field of movement; only with sa and ri was the range limited.

Jātis were sung to a hexatonic or pentatonic structure also. How was tāra movement to be calculated in such cases? With a structure like sa ri ma pa ni and with ri as the aṃśa, could one move upto ni in the tāra ? Dattila gives no answer but Bharata (in the text as followed by Abhinava)2 has specified the rule to be followed in such cases. The same rule had been stated by Viśākha also (as Abhinava quotes him).3 The rule was that in determining the upper limit of the possible movement in the tāra, the omitted note was also to be counted. Thus in the pentatonic structure cited as an example above, one could move upto dha, and since dha was omitted, the upper tāra limit remained pa.

These rules regarding tāra movements were a binding factor only in gāndharva and the jāti form. In grāma-rāgas they were not binding. Commenting on the word ‘iha’ used by Bharata in expounding the tāra rule4, Abhinava says : “ ‘iheti’, jātiṣu, grāmarāgādau na yaṃ niyamam iti vibhāvah” (A.B. on N.S. 28, 70).

The general rule regarding tāra-movement had an exception in nandayantī, for in this jāti the movement in the higher octave was confined upto the first tāra note, i.e., the tāra ṣadja. Thus the ancients allowed this jāti virtually no scope for movement in the higher octave (tāragateḥ saṃkocastvācāryaiḥ kṛtaḥ, Kalā on S.R. 1, 7, 36). Bharata has not noted the example of nandayantī as being an exception in the context of the general rule regarding tāra but in expounding the structure of nandayantī he points out its restricted tāra movement, decreeing that tāragati in this jāti does not stretch beyond the tāra ṣadja (tāragatyā tu ṣadjastu kadācinnātivartate, N.S. 28, 134).

The extent of movement in the lower octave also depended upon the aṃśa. One possible lower limit of the movement was the aṃśa itself in the lower octave. Thus in the jāti śuddha ṣadja (or ṣādjī) one could descend upto the note sa in the lower octave, the lowest possible note of the three octave range.

There was another alternative for the limit of the lower movement: the final note or nyāsa could be the limit. There was yet another alternative; it was permissi-

1 ऋषभोऽत्रे पुनर्धैवतस्तथा: स्वरभूस्तारसप्तकं ग्राह्या भवति । गान्धारे स्वभे निपादान्ताः सप्त । मध्यमपञ्चमयोर्मध्यवर्तिनियादेपु निपादान्ताः एव तु प्राग्भ्या इति । पञ्चस्वरस्थाप्यकृत: संगृहीतो भवति । मध्यम: पुनर्वेदि शक्तिः: चतु:स्वरप्रमाणोहणम्

—A.B. on N.S. 28, 70.

2 After N.S. 28, 70 a line is added in brackets: (नानिस्वरत्वाद्विधिस्वरसंनिवेशविधौ पुः)

(लोयिस्वररागपक्षे स्वरसंनिवेशविधौ पुः)

The editor of the N.S. text comments:

एतदंशव्युऋकf प्रन्यभागः न कस्याश्चिदपि मातृकायां दृश्यते । यद्वयानातस एव पतितः स्वार्दिति श्लोकाई रचयित्वा निवेशितम्

Though Abhinava gives only a word or two of the line, which in his ms. was part of the text, the purport of the line is well put together by the editor on the basis of Abhinava’s comments.

3 तथा च विशाखिलाचायं: । नानिस्वरस्तारविघ्नो संयायते इति ॥

—A.B. on N.S. 28, 70

4 अशास्तारगति विधादचवुंर्स्वरादिवि्ह । आपञ्चममातपञ्चमाढा नात:परमिहेष्यते ॥

—N.S. 28, 70

ble to descend to the note immediately below the nyāsa. Bharata, giving an illustration, says that with gāndhāra as nyāsa one could move down upto ṛṣabha1. According to the reading in the Asiatic Society edition of the Nāṭyaśāstra, there was still another alternative : the apanyāsa could also function as the lower limit for the movement in the mandra (tridhā mandragatih—amśaparā nyāsaparā apanyāsaparā ceti’ N.S. A.S. ed., 28,95). Kallinātha seems to have accepted this reading of the Nāṭyaśāstra.2

Bahutva

According to Kallinātha, bahutva or the ample use of a note could be effected in two ways: (1) by alaṅghana, i.e., by not transgressing it in rendering a melodic movement and (2) by abhyāsa, the frequent repetition of it.3 The notes which obviously had an ample use in a jāti were its vādī and samvādī (sonant and consonant); other notes on which bahutva operated were the notes called paryāyāmśas or the alternative amśas.

The term paryāyāmśa has not been used by Dattila or Bharata. Later theorists seemed to have coined it in order to explain a certain phenomenon present in the jāti-structure. The term is to be found in the expositions of Abhinava and others.

Many jātis are decreed as having more than one amśa; ṣadjamadhyamā, for example, had seven amśas. All possible amśas of a jāti could not become the vādi-note at the same time. But evidently they were all potential vādis and each could become a vādi in a different rendering. Amśas other than the operative one, remained paryāyāmśas, or alternate amśas. Thus Kallinātha explains paryāyāmśa as: “vādibhūt-āmśādvyatiriktāmśaḥ” (Kalā on S.R. 1, 7, 49).

Paryāyāmśas, when not acting as the main note, did not lose their importance and remained strong notes. Abhinava states this idea negatively: “notes which are not paryāyāmśas when used in antaramārga should not be emphasised by repetition, for example, the niṣāda and the ṛṣabha in the śāḍjī jāti.”4 In śāḍjī, which had five amśas, the only notes which were not amśas were niṣāda and ṛṣabha.

Alpatva

Alpatva was just the reverse of bahutva. It was effected by laṅghana and anabhyāsa. The first was effected by gliding over a note in a way that left it unempha-

1 निधा मन्द्रगतिः । अंशपरा न्यासपरा तल्पराचिति । मन्द्रस्वरस्थितपरो नासि न्यासो तु हि द्वयस्थितो । गान्धारन्यासतस्त्वस्य तु दृष्टमार्गभसेवनम् ॥ —N.S. 28, 70–71.

2 Kallinātha says : यथोक्तं भरतेन… “निधा मन्द्रगतिरंशपरा न्यासपराऽपन्यासपरा चैति । —Kalā on S.R. 1,7, 37–38.

3 He evidently takes his cue from the N.S. where alpatva, the converse of bahutva, has been defined as : द्विविधमपि लाघवदनभ्यासाच्च । —N.S. 28, 74.

4 येऽन्य तु स्वराः पर्यायौषा न भविष्यन्ति यदाज्ञातरमार्गे प्रयोक्तव्येते तदाऽन्य्याम्शासः । —A.B. on N.S. 28, 74.

sised and thus rendered it subordinate to the emphasised note (lañghanam spr̥tvaivā-

viśramyaiva tatsvarāntaragamane tasya svarasya svapradhānikaranam, A. B. on N.S.

28-74), or by avoiding to repeat it (anabhyāsa). Anabhyāsa operated on weak notes,

i e., the notes which were not potential amśas (anamśā). Curiously, it could also operate

on notes that were actually to be omitted (śādabaudubhitakārāṇām).1 Dattila too says

that the note which has to be omitted is an alpaka note; alpatva, in other words,

operates on it. Thus in making a jāti śādava or auḍuva and (in Dattila's words) ren-

dering it 'hīna'—diminished—the notes to be dropped were not sometimes totally

dropped but were made alpaka or weak; this was considered as amounting to omis-

sion (Datt. 61A).

Ṣādava and Auḍuva

Well-defined rules as to notes that could be dropped for forming ṣādava and

auḍuva had been laid. We have already discussed some general rules. We shall meet

with details in dealing with individual jātis.

Nyāsa and Apanyāsa

Nyāsa was the note on which a melody came to rest. Apanyāsa was a nyāsa

within a vidārī. Vidārīs were, in short, smaller units within the musical whole of a

melody.

Bharata gives a somewhat different definition of nyāsa and apanyāsa. Nyāsa,

he says, occurs at the end of a portion or part (aṅga); apanyāsa is similar but it comes

at the end of a smaller segment within a larger part or aṅga : “aṅgasamāptau nyāsah,

tadvadapanyāso hyaṅgamadhyē” (N.S. 28, 72). Abhinava defines nyāsa as the svara on

which the structure of a jāti finally comes to rest, that is, the note on which a jāti is

consummated.2 Apanyāsa, Abhinava remarks, is, as Bharata states, similar to nyāsa

(tadvad), but it occurs as an intermediate final note (avāntarasamāptau): it occurs at

the consummation of smaller parts within the jāti structure. Bharata's phrase 'aṅga-

madhyē' then evidently meant the same as Dattila's words, 'vidārīmadhyagah', which

Abhinava indeed quotes.3

We shall see later on that every jāti had one or more specified note as nyāsa

and apanyāsa. Bharata has given us the total count of all the nyāsas and apanyāsas

laid down for the 18 jātis. The total number of nyāsas were 21 and apanyāsas 56:

“atha nyāsa ekavimśatisañkhyāḥ...ṣaṭpañcāśatsañkhyo' panyāsah” (N.S. 28, 72). The

jāti ṣadjamadhyamā had two svaras as nyāsas, kaiśikī had three; the other 16 jātis had

1 द्विधा धमल्यतश्च लाघवात्स्वनभ्यासाच्च । तत् पाडवौडवितकारणैरमंशकानां च गीतादंतरमार्गमुपगतानां स्वराणामनभ्यास:

लघूनादनभ्यासात्स्वरा जातिलपर्वन् तदवबहुलतश्च व ।

—N.S. 28, 74.

2 अस्वां जातिशरीरसमाप्तौ कृतव्यतितैर्वा स्वरः सम्न्यासः तर्हि अस्यते प्रयोजो वेनेति न्यासः ।

—A.B. on N.S. 28, 73.

3 अपन्यासमाह तत्रदिति । समाप्तस्तर्हि न्यासात् कौटस्थ विशेषः । आह । अङ्गमध्ये इति । अवान्तरसमाप्तौ विशेष:

तथा च दत्तिलावार्यः: 'विदारी मध्यगतस्थया' इति ।

—A.B., ibid.

only one each; total number of nyāsas thus being 21 : “ṣadjamadhyamāyāṃ dve, kaiśi-kyāstrayah, śiṣṭānāṃ ṣoḍaśeti” (A.B. on N.S. 28, 73). The jātis of the ṣadja-grāma jātis of this grāma had 3 each; together they had 24 apanyāsas.1 The madhyama-grāma jātis similarly had 32 apanyāsas.

Other Jāti Characteristics

Bharata has once used the phrase: “samavayājjātāyastu jāyante”: jātis are born of a conglomeration—(samavāya) (N.S. 28, 46). This samavāya, as Abhinava remarks, consisted of śrutis and svaras grouped according to the basis of graha etc.2 The jāti-lakṣaṇas, in other words, together formed the ‘samavāya’ which resulted in a jāti. Yet any living and complex structure cannot be entirely summed up in a precise number of characteristics; nuances are bound to escape lakṣaṇa-formulations. Abhi-nava, in fact, observes that some old theorists had said that jātis resulted not only through aṃśa, graha etc. but a hundred such formal arrangements and combinations (yojanā): “aṃśagrahādiśatatayojanayā jāyanta iti jātiyā ityanye” (A. B. on N. S. 28, 46).

The Nātyaśāstra recounts one more prominent element of the jāti-structure besides the ten characteristics. This was the antaramārga (N.S. 28, 75). Though not included as part of the jāti-lakṣaṇa list which appears to have become conventionally limited to the 10 standard characteristics (daśakaṃ jāti-lakṣaṇam; N. S. 28, 66), antaramārga was an important element. Saṅgītaratnākara and later texts include antaramārga as a jāti-lakṣaṇa and add two more, namely, sanyāsa and vinyāsa. Śārṅgadeva thus enumerates not 10 but 13 jāti-lakṣaṇas (S R. 1, 7, 29-30).

Dattila defines and indicates the use of sanyāsa while pointing out some details of jātis when moulded to gītakas (Datt. 141). But he does not include sanyāsa among the list of lakṣaṇas. Both sanyāsa and vinyāsa appear to have been sub-categories of apanyāsa. Kallinātha states that the reason why Bharata, Maṭaṅga and others have not included these as jāti-characteristics, like apanyāsa, were dependent upon vidārīs (see verse 142), and as such were included in the general idea of apanyāsa.3

Giving reasons for exclusion of antaramārga from the ancient jātilakṣaṇa list, Kallinātha says that antaramārga was only a mode of combining the various elements of a jāti such as aṃśa and has, therefore, not been separately enumerated. But Kallinātha also justifies Śārṅgadeva who has included antaramārga as well as sanyāsa and vinyāsa as lakṣaṇas; he says: “Since sanyāsa and vinyāsa are undoubtedly two distinct components, they should be separately enumerated. Inclusion of antaramārga,

1 तत्र पञ्जग्रामे पञ्जमध्यगं यथा: सप्तापन्यास: । पञ्जोदोच्यवायां द्वे, पञ्चचानां प्रत्येकं द्वय द्वे इति चतुर्विध्यात: । —A.B. on N.S. 28, 73.

2 निह्ह्र्तो समवायात् श्रुतिस्वरग्रहादिसमूहाद्यो जायन्ते ततो जातय इति चतुर्विध्याति: निर्वचनम् —ibid., 28, 46.

3 भरतमतङ्गादिभि: सञ्यासविन्यासोभयादर्शितस्तवादपन्यै:सङ्कतभाज्चित पृथयुद्देशो नापेक्षित इति दशकं जातिलक्षणमुक्तम् । —Kalā on S.R. 1, 7, 29-34.

too, as a distinct element is necessary, for it is the mode by which various elements

of a jāti are related to each other and this is essential for rendering a jāti."1

Antaramārga

Bharata, in a verse, says : "movement (sañcāra) over the aṃśa and strong notes,

rendering the weak notes as unemphasised, nyāsa and antaramārga [are the factors]

that manifest jātiś:"

sañcārāṃśabalasthānāmalpatve durbalāśu ca

nyāsaścāntaramārgastu jātiīnāṃ vyaktikārakāḥ

(N.S. 28, 75).2

Abhinava comments that some theorists understood this verse as a definition

of antaramārga (antaramārga-lakṣaṇatvena ślokaṃ vyācakṣate). Taking the word

nyāsa not in its technical sense as the final note but in a general way (ni+√as : 'to

arrange' etc.), they interpreted the verse to the following effect : antaramārga was the

factor which manifested a jāti; antaramārga depended upon alpatva and bahutva; it

was rendered by emphasising the aṃśa, the anuvādī and other strong notes and plac-

ing or arranging (nyasanam) the weak notes through laṅghana in such a way that

they remained unemphasised.3

Abhinava himself explained the śloka somewhat differently as laying down a

rule for certain jātis which had a large number of paryāyāṃśas. Yet he seems to have

shared views regarding the general nature and function of antaramārga with the

theorists he refers to. However, he makes the reservation that in some jātis even weak

notes could be emphasised in the antaramārga; in kārmāravi, for example, ga was a

weak note being an anamsa (i.e., other than a paryāyāṃśa) but it was yet copiously

used and movements were made to it from every note.4

Antaramārga emerges to have been a term which denoted the characteristic

melodic movement over notes in a jāti. The term itself is suggestive. It denotes the

internal, the deeply individual path traced by the melody in a jāti (antaramārgaṇami-

1 तथाऽडपि संचारो विन्यासयोः पूर्वगयवबोधनान्तमार्गस्य तु सरसकादिवच्चरितेषु तेन विना प्रयोगासिद्धेरतरतस्याप्यधिक्यमव-

स्वलक्षणस्य पूरणाद्दृष्टं वैय्यत्यं ग्रन्थकारिणः

—Kalā, ibid.

2 Abhinava interprets sañcārāṃśa in this verse as meaning paryāyāṃśa : 'सच्चारेण पर्याय योऽङ्ग:'

Thus according to Abhinava's interpretation the notion of paryāyāṃśa has been overtly listed;

if not explained, by Bharata.

3 अन्ये त्वन्तमार्गस्य सर्वत्र प्रधानस्याल्पत्वबहुत्वाभ्यां विना स्वरूपलाभ एव (न) भवतीति तात्पर्यं मन्यमाना अन्तरमार्ग-

लक्षणत्वेन श्लोकं व्याचक्षते । कथम् । 'चो' हेतौ । यथामादन्तरामों जातीनां' व्यक्तिकृत् सु चाल्पत्वबहुत्वजोऽपि तस्य हि

लक्षणम् । 'सच्चरेऽति' सच्चारेण अनुवादिस्वरे ऽनङ्गस्बरे च प्रधानमुते दलकल्प्ये रिष्यते 'वलन्यानां स्वराणां संवादिनां

स्वराणां' बहवस्तत्वाल्पत्वे (दुर्बलासु)...... .. स्वराणां दुर्बलप्रकर्षे विलङ्घनेन न्यासं यः (करोति) सोऽन्तमार्ग-

स्तस्मादल्पत्वबहुलते व कथं

—A.B. on N.S. 28, 75.

4 अन्तरमार्गणमित्यन्तरमार्गस्य लक्षणम् । 'यथाजातीयोऽति' । वचिदन्वोषण नाल्प । । तथाऽ च कार्मारव्यां गान्धारस्य

सर्वस्वरसरस्य बहुलीकरणात्तस्मार्ग प्रयोक्त इति वद्यते ।

—A.B., ibid.

tyantaramārgasya lakṣaṇam, A.B. on N.S. 28, 74). It is significantly suggestive of our own term ‘calan’—individual movement of a rāga.

Dattila does not list antaramārga as a jāti-lakṣaṇa, but uses the word twice in a significant context. In defining aṅga, a topic in tāla, he relates the structure of a certain class of aṅgas to melodic structures and says that these aṅgas should be rendered according to antaramārga and further specifies that in case of śuddha jātis, antaramārga should depend upon aṃśa (Datt. 147).

Sanyāsa and Vinyāsa

Dattila defines sanyāsa as the note which was not a vivādī of the aṃśa and which served as the final note (nyāsa) in the first vidārī (Datt. 141). Dattila does not mention vinyāsa. Perhaps the use of vinyāsa was not a universal factor in the jātis. Kumbha argues that this was the reason why Dattila had not spoken of vinyāsa.1 Unlike Dattila, Bharata defines both sanyāsa and vinyāsa.2 Śārṅgadeva defines vinyāsa as the note that occurs at the end of each small melodic unit set to a single word within a vidārī (yo vidārībhāgarūpapadapraṅte’vatiṣṭhate, S.R. 1, 7, 48; see also Kalā). Śārṅgadeva, on this point, seems to have based his words upon Abhinava’s comments on Bharata’s definition of vinyāsa. Abhinava says that “vinyāsa is a note which is placed (vinyasyate) at the end of a single pada or word (padānte) within a vidārī-unit. Such a note could be either the samvādī or the anuvādī of the aṃśa: aṃśasya samvādyanuvādī vā kāpi vidārībhāva(?ga)rūpasya padaśya padānte vinyasyate tadā vinyāsaḥ” (A.B. on N.S. 28, 73).

After expounding jāti-lakṣaṇas, Dattila explains the nature of śuddha jāti in Datt. 62.

In a śuddha jāti the note after which the jāti was named was its aṃśa, nyāsa, graha and also its apanyāsa. (svasvarāṃśagrahan yāśāpanyāsaśca; N.S. 28, 45). Another necessary factor was that the nyāsa occurred necessarily in the mandra or lower octave. Bharata also states: “nyāsavidhāpyāp yāsaṃ mandro niyamah” (N.S.28, 46).

In vikṛtas there was no strict limiting factor except for the nyāsa which had to remain on the nāma-svara (the note after which a jāti was named) though not necessarily in the mandra(yā vikṛttāstu nāmakārī mandra bhavatītyamiyamah, Br̥. Vṛtti on 187, also quoted by Sudhā on S.R. 1, 7, 3). By successively dropping one or more essential features from a śuddha jāti, many vikṛtas could be formed. Śārṅgadeva

विन्यासस्य प्रयोगोऽत्र न सर्वत्र मतः सताम् । दत्तिलादैरततो नाम न हि शास्त्रेषु लक्षि‌त: ॥

—S. Raj 2, 1, 4, 98,

तत्र प्रथम विदारीमध्ये न्यासस्वरप्रकृष्टु । विवदनस्थीलं मुक्त्वा सन्यास: सौभिद्यतत्स: । क्रत्वा पदावसाने विन्यासातू

क्वापि विन्यास: ।

—N.S. 28, 73,

Abhinava, in introducing his comments on these lines, says that though unlisted, these notions were important being offshoots of the notion of amśa and have therefore been expounded ;

अर्थानुदृष्टंमर्थंश्रलक्षणंपेशितं प्रसङ्जादलक्ष्यतिति !

enumerates fifteen vikṛtas of the ṣāḍjī alone. (For details, see S.R. 1, 7, 4-7, and the

two commentaries). Abhinava enumerates the vikṛtas when describing śuddha jātis in

detail.

A question here arises as to why should nyāsa have been chosen as the cons-

tant invariable when the same result could have been achieved by taking the graha

or the apanyāsa or the aṃśa as the constant, for in a śuddha jāti, the nāma-svara was

aṃśa as well. Kallinātha says not only its nyāsa, but also its graha, apanyāsa and

that nyāsa was the main factor which distinguished two jātis, but does not give

reasons for this ascertion. If the nyāsa, too, he further says, were to be taken as a

variable, then it would not be possible to recognise the śuddha source of a vikṛta jāti.1

Kallinātha’s comment does not answer our question, for the problem remains as to

why nyāsa should have been taken as the constant invariable where other choices

seem to be equally eligible.

The Jāti Ṣādji

63A. aṃśāḥ syuḥ pañca ṣadjyāyā niṣādarṣabhavarjitāḥ

B. apanyāsāstu gāndhāraḥ pañcamamaśca sāṅgatīḥ

64A. ṣadjagāndhārayostu syāt ṣadjdhaivatayostathā

B. ṣāḍabatvaṃ niṣāde syānnityamaudubitaṃ bhavet2

In ṣādjī [jāti], with the exception of niṣāda and ṛṣabha, there are five aṃśas;

gāndhāra as well as pañcama are the apanyāsas. In it, association (saṅgati) occurs

between ṣadja and gāndhāra as well as between ṣadja and dhaivata. It is rendered

ṣāḍava by [dropping] niṣāda. It admits no auḍuvita.

NOTE :

Dattila now begins delineating the structural features of different jātis, noting

their aṃśas and other necessary characteristics. These in ṣādjī are clear enough. The

term saṅgati, however, needs, an explanation. The Vṛtti on Brhaddesí in this context

1 नामस्वरमेव न्यासः क्रत्वाडन्यस्यादौसर्वत्राणि कुर्यादिति । एवं क्रतु यद्, तदा विकृता, विकृतावस्थाप्यसन्नि भवन्ति,

न तु विकृतसंघातजातितदित्थादपदेशान्तरभङ्ग इत्यर्थः । यद् न्यासनियमस्य परिलङ्घनं नेद्᳚᳛; तस्मिन्नपि परित्यक्त᳚

सति विकृतातुस जायन्तरभेदकत्वेन प्रवृत्तमुत्पाद्यवदनुबुद्धो तासां तत्त्वज्ञाञ्जातिभेदप्रतिपत्तित᳚᳛ स्वात᳚

—Kalā on S.R. 1, 73.

2 Br. 201-230A is the same as Datt. 63-91A. The lines are not acknowledged as borrowed from

Dattila. For textual implication of this fact see ch. II. Br. 201-202 have the same reading as

Datt. 63-64. Apparently in Datt. 63-64 some words were missing and were filled in by the editor

from Br. These have been bracketed in the T. ed. from ‘pañcamaśca’ in Datt. 63B to पाङ्ग्, the first

word in Datt. 64A. B.B. (1) ch. VI quotes 61B to 64 with the words: ‘यदाह᳚ तत्र᳚.’ 63A here reads

‘अं᳚ श᳚ पञ्च᳚’ पाङ्ग्: स्प᳚᳛ष्टीभू᳚᳛ताद्व᳚भ᳚िजता᳚;’. 63B reads पञ्चम᳚ स᳚कृत्ति᳚; instead of above. 64 reads :

बहुगान्᳚᳛यर᳚पो᳚सु स्वात᳚ बह᳚᳛दैव᳚पो᳚सु ᳚᳛. पाङ्ग्᳚ तु निपाद᳚ स्वात᳚ पञ्चम᳚द᳚व᳚त᳚ भ᳚व᳚त᳚ ।

However it must be remembered that B.B.(1) readings are full of scribal and other errors and

are not collated readings.

defines saṅgati as a to and fro movement (parasparagamanam ca saṅgatih, Bṛ. Vṛtti on

251. between the two notes ṣadja and dhaivata and also between ṣadja and gāndhāra.

Bharata in N.S. 28, 95-97 uses a clearer term sañcāra, to describe this very movement.

Kallinātha in his commentary on S.R. 1-7-61, defines the nature of this movement

more spcifically. “Here in ṣāḍji”, he says, “the note ṣadja should be associated with

the two notes gāndhāra and dhaivata, both of which are one note removed from it.

The movement should be aesthetically satisfyng (yathārakti); the phrase rendered

should be either sa ga, sa ga sa dha, or ga sa, ga sa, dha sa.”1 This saṅgati was, it

appears, a typical movement of the antaramārga in this jāti.

Dattila has not named the weak notes in this jāti. But on the basis of the general

principles discussed above, we may infer that ni and ri, which are not possible amśas,

(i.e., paryāyāmśas) should have been weak. Bharata explicitly states: “alpau vai sapta-

marṣabhau” (N.S. 28, 97).

Bharata mentions the nyāsa here as ṣadja (N.S.28, 97). Dattila does not name

thenyāsa. Naming it was indeed unnecessary since in a svara-jāti like ṣāḍjī, ṣadja was by

rule the nyāsa.

The Jāti Āṛṣabhī

65A. āṛṣabhyāstu smṛtā amśā nisādarsabhau2vatāḥ

B. ṣadjapañcamahīne ca ṣādabaudu3bite kramāt2

Niṣāda, ṛṣabha and dhaivata are said to be the amśas of the āṛṣabhī (jāti). It

is rendered ṣāḍava and auḍuvita by omitting (hīne) ṣadja and pañcama in due order.

NOTE :

The last line of Dattila's description does not imply that this jāti can be ren-

dered ṣāḍava by successively omitting ṣadja and pañcama. What Dattila intends is

that the jāti is rendered ṣāḍava by omitting ṣadja alone and auḍuvita by dropping both

ṣadja and pañcama. Bharata has a statement similar to Dattila on this point.3 Abhi-

nava clarifies: “ṣadjahīne ṣāḍavaḥ ṣadjapañcamayorhīne auḍuvitatvam” (A.B. on N.S.

1

सगयोः सध्योः स्वान् संगतिरिति यत् पड्जस्य गान्धारेणै कान्तरितेन तादृशोभैव धैवतेन च व्यावृत्तिः सम्बन्धः सगस-

ग्धवेतो रसगसङ्गसेति वा कार्यं; -Kalā on S.R. 1, 7, 61.

It should, however, be remebered that the saṅgati here noted by Kallinātha was most likely an

imaginative reconstruction of what it might have been like. No available ancient text gives the

actual structure of the saṅgati.

2

T.ed 65 has a lacuna of five letters at the end of the verse. However both in Br. and in B.B. (I)

the full verse is found. In Br. the reading is faulty but B.B. (I) clearly reads ṣाडबौढवित क्रमात्.

Br. 203 is same as Datt. 65. B.B. (I) ch. VI quotes Datt. 65 as a parallel testimony to N.S. 28,

98 which is quoted with the words : .आर्षभी; तत्र सूत्रम्. Datt. 65 is then quoted with दत्तिलस्साह,

reading as above.

3

षट्पञ्चस्वरता चान्ते पड्जपञ्चमयोः हि दना

-N.S. 28, 98,

280 A Study of Dattilam

28, 98). We have translated the word ‘hīna’ as ‘omitting’ or ‘dropping’; but in this context the word was used in a rather qualified sense. For, a note rendered hīna was one on which alpatva operated and the note was not always totally dropped.

Bharata notes details about ārṣabhī which Dattila seems to have taken for granted (N.S. 28, 98-101). The apanyāsa notes in this jāti were the same as the amśa granted (N.S. 28, 98-101). The apanyāsa notes in this jāti were the same as the amśa granted.

ṛṣabha was the only nyāsa. The fact that ṛsabha was the nyāsa, however, followed from the general maxim that in śuddha jātis the nāma-svara was the nyāsa.

Abhinava notes two saṅgatis in this jāti: one between sa and dha and another between ri and ga (sadhar rigau ca saṅgacchete, A.B. on N.S. 28, 98). He further says that when this jāti was pūrṇa—septatonic—sa, ga, pa were alpa notes; when auḍuvita, ga ma were weak notes: “pūrṇāvasthāyāṃ ṣadjagāndhārāpañcamā alpatvaṃ bhajante, auḍuvite ca gamayoralpatā” (A.B., ibid.).

The Jāti Dhaivatī

66A. dhaivatyaṃ gurubhiḥ proktāṃsārsabhaivatau

B. samadhyamāvapanyāsau prāguktā hīnatotkramāt.1

Teachers declare ṛṣabha and dhaivata as the two amśas in dhaivatī [jāti]. These two, along with the madhyama, are [also] the apanyāsas. The [process of] hīnatā [occurs in the same way] as afore-mentioned, but by reversing the order (i.e., ṣādava is obtained by dropping pañcama and auḍuvita by dropping both ṣadja and pañcama).

NOTE :

Dattila (unlike Śārṅgadeva) has taken up the jātis according to the two grāma divisions which loomed large in actual ancient musical practice and not in their serial order. Thus, after ṣādjī and ārṣabhī, he deals with dhaivatī and other jātis of the ṣadja-grāma. Later he takes up the madhyama-grāma jātis as a separate group. Bharata has followed the same order as Dattila.

The phrase ‘prāguktā hīnatotkramāt’ evidently refers to the hīnatā in this jāti and this is described in relation to the previous jāti, ārṣabhī. Dattila’s attempt at brevity has made his language more precise than clear. However, his purport (as we have rendered it) is clear from the Nātyaśāstra which says:

ṣadjapañcamahīnaṃ tu pañcasvaryaṃ vidhīyate pañcameva vīṇā caiṣā ṣādavaṃ parikīrtitam (N.S. 28, 100)

“The five-note structure is rendered by dropping ṣadja and pañcama and ṣādava is rendered by dropping pañcama.”

1 Br. 204 is same as Datt. 66. B.B. (I) ch. VI quotes Datt. 66 with the words ‘dattilācāryoditaḥ vāh’. Line 66 B has scribal mistakes. The last portion of the line reads as ‘prāmuktāhīnetakramādīt’. The verse is quoted as authoritative testimony parallel to N.S. 28, 120–122.

2 In this connection S. Raj 2, I, 4, 168 reads गीध्व समग्रोनिसात्, an error either of the compositor or th: copyist, sin ce no other authority describes ‘m 1’ as hina. The correct reading evidently is ; गीध्व सपयोलौंपात्

We notice that in this jāti ṣadja could be dropped. In addition (a fact which further confounds us moderns) pañcama could also be dropped.

Bharata notes some more characteristics of this jāti. We can infer that, being amśas, ri and dha were strong notes; Bharata adds that gāndhāra was also strong.

He also states that when the jāti was pūrṇa, sa and pa were to be employed only in the ascent but were otherwise notes on which laṅghana applied (N.S.28, 100-101).

The Jāti Niṣādavatī

67A. amśau niṣādavatya̱stu dviśrutī sārsabhau smr̥tau

B. dhaivatīvad bhavecceṣo nyāso nāmakṛdeva tu1

Along with the ṛṣabha, the two two-śruti notes (i.e., niṣāda and gāndhāra) are said to be the amśas of the [jāti] niṣādavatī. The note after which it is named is indeed its nyāsa. The rest [of the characteristics] correspond to dhaivatī.

NOTE :

A word should be said here regarding features in this jāti which are said to correspond with dhaivatī. Bharata, enumerating these features, says: “The process of śādava and auḍuvita in this jāti is the same as in the dhaivatī (i.e., by dropping pañcama and by dropping ṣadja and pañcama respectively); moreover, these two notes are unemphasised (laṅghanīyau) and are also strong (balavantau) in a similar way as in the dhaivatī (jāti)” (N.S. 28, 103).

The latter statement of Bharata may appear contradictory and perplexing. For a clarification we shall consider the comments of Abhinava and observations found in the Vṛtti on the Brhaddesī. Ṣadja and pañcama in dhaivatī were notes that could be dropped (hīna); this implies that they were meant to be weak notes even when employed in the pūrṇa state of the jāti.

Bharata has said that these notes are laṅghanīya and are to be used only in the ascent. Abhinava comments that in a pūrṇa dhaivatī, ṣadja and pañcama should be used in the ascending movement and although these notes are hīna and omitted for the process of śādava and auḍuvita; yet they are powerful notes (when the jāti is pūrṇa). (lopyatvāllaṅghanam̐ siddhamapi punah̐ prakarṣalābhatvamuktam ; A.B. on N.S. 28, 101).

The Vṛtti on Brhaddesī remarks: “ṣādava is rendered by dropping pañcama and auḍuvita by dropping pañcama and ṣadja. These two, ṣadja and pañcama, should be made strong (balau), [although] they should sometimes be unemphasised by laṅghana (laṅghanīyau). When the jāti is pūrṇa the notes ṣadja, gāndhāra, madhyama, pañcama and niṣāda with the exception of the notes ṣadja and pañcama are rendered as alpa [last two] are alpa in the auḍuvita. The rest are strong.”2

1 Br. 205 is same as Datt. 67. Datt. 67B has छष्यो नामकदेव तु within brackets. Evidently the line was indistinct in Datt. and the editor supplied the words from Br. B.B. (I) ch. VI quotes Datt. 67 with ‘दत्तिलोय्याह’. The Datt. verse is quoted as parallel authority to N.S. 28, 102-103.

2 पंचमहीनं पादव् । (पंचमपडजहीनं पादव?) पंचमपडजहीनमौडुवितम । तावेव पहर्जपंचस्वरो तु बलौ कर्तव्यौ । वच्चित लघनीयो ।

-Br. Vṛtti on 251,

282 A Study of Dattilam

This similarity between the two jātis, dhaivātī and nīṣādavatī, (which certainly provides an example of jāti-sādhāraṇa) did not, however, extend to apanyāsas. The apanyāsas in the dhaivatī were dha, rī, ma, whereas those in nīṣādavatī were the same as its aṃśas, viz., ni, rī, and ga. There must have been other features which distinguished the two jatis (such as their differing antaramārgas) in actual musical practice.

The Jāti Ṣadjakaiśikī

68A. aṃśāḥ syuh ṣadjakaiśikīyāṃ ṣadjagāndhārapañcamāḥ

B. saniṣādādastvagāndhārā apanyāsāsta eva tu

69A. ṛṣabho’lpsaprayogāḥ syānnyāsso gāndhāra iṣyate

B. nityam pūrṇasvarā ceyamācāryaiḥ parikīrtitā1

In ṣadjakaiśikī [jāti], the aṃśas are ṣadja, gāndhāra and pañcama. These very notes including the niṣāda but excluding the gāndhāra are the apanyāsas. The application of ṛṣabha is not frequent (alpaprayogab). The nyāsa ought to be on gāndhāra. This [jāti] invariably consists of seven notes; so have the teachers declared.

NOTE :

We come now to the saṅkara jātis of the ṣadja-grāma.

Ṣadjakaiśikī jāti was born through a combination of ṣāḍjī and gāndhārī (Datt. 49). It had three aṃśas: sa, ga, pa; the other notes, the anamśas, were by implication relatively weak notes. Dattila overtly describes only ṛṣabha as weak, probably implying that this svara was weaker than the other anamśa notes. Bharata mentions another svara as weak along with rī; this according to Abhinava’s reading was ma, but dha according to another reading.2 “That ma and rī are weak”, says Abhinava, “follows from their being anamśas but they have been especially named to reveal their extreme weakness; ni and dha consequently are relatively stronger notes.3 The Vṛtti on Brhad-deśī also names ma and rī as weak notes; it, however, adds a qualifying remark: “ma was weak but rī was weaker: madhyamānāṃ (?) alpatvaṃ ṛṣabhasyālpataratvam” (Br̥. Vṛtti on 251).

Jātis were rendered ṣāḍava and auḍuva according to strict rules. In preceding jātis, notes which could be dropped are named. Further, we have seen that there were jātis, such as ṣāḍjī, which permitted no auḍuva. Here we meet with a jāti which permitted neither ṣāḍava nor auḍuva. Such restrictions were typical of gāndharva.

1 Datt. T.ed. 68B reads : Sanīṣādastu gāndhāra... which makes no sense. We have modified the text on the bsais of Br̥. 206 which reads the same line as ‘sanīṣādā gāndhārā ... Br̥. 206-207 are same as Datt. 68-69.

दौर्बल्यं चात्रैव धैवत (मध्यम) स्याद्र्भस्य च -N.S. 28, 105.

The word ‘madhyama’ in brackets indicates the reading accepted by Abhinava who say ‘मध्यमस्यैवोद्बलत्वं’

3 ‘मध्यमर्षभयोदीच्यम (नं) श्वतविसर्ग पुनराधिक्यार्थं ज्ञापयति । निधयोक्त किञ्चिद्देश्चितमिति । -A.B. on N.S. 28, 104-105.

A.B. text reads मञ्चत्वात्तसिद्ध्—an evident error since ma and rī were anamśas, not aṃśas.

The Jāti Ṣadjodīcyavatī

70A. syāt ṣadjodīcyavatyamsāīḥ ṣadjamadhyamadaivataīḥ

B. saniṣādairaparyāsau vijñeyau ṣadjadhaivatau

71A. ṛsabheṇa vihīneyam dvābhyām cet pañcamena ca

B. mandrasāgāndhārākkhāstavamasya nyāsastu madhyamaḥ1

Ṣadja, madhyama [and] dhaivata along with niṣāda are the amśas of the [jāti] ṣadjodīcyavatī. [In it] ṣadja and dhaivata should be known to be the apanyāsas. Ṛṣabha is dropped [for rendering ṣādavita] and if two notes are to be omitted pañcama as well is dropped. Madhyama is the nyāsa in this [jāti]. [The use of] gāndhāra is abundant in the lower octave (mandra).

NOTE :

Abhinava explains the name of this jāti by remarking that it was much liked and cultivated by the people of the North, hence it was called udīcyavā: ‘of the north’, or ‘from the north’. He adds the general remark that many compositions (gīta) are named after particular regions for this very reason. He then names a few: ṭakka rāga (from North-West Punjab); mālava-pañcama (from modern Malava in Central India); gauḍī (from Gauḍa or North-West Bengal); mālavī (again from Malava) and kāmbhojī (from a region in North India).2

The abundant use of ga in the lower octave was certainly a peculiar feature of the antaramārga in this jāti. Bharata mentions another movement to be rendered between the amśa notes sa ma dha ni: “parasparāṃśagamanamiṣṭatāśca vidhīyate” (N.S. 28, 107).

Ga was an anamśa and yet a strong note, used abundantly in the lower octave. Ri, another anamśa, was also strong; sa already strong by definition, being an amśa, was especially powerful : “ṣadjaścāpyaraṣabhaścaiva gāndhāraśca balī bhavet” (N.S. 28, 109).3

The Jāti Ṣadjamadhyamā

72A. saptāmśāḥ ṣadjamadhyāyā nyāso vai ṣadjamadhyamau

B. kramānniṣādagāndhārāvasyām hīnatvakāriṇau

1 Br. 208-209 is same as Datt. 70-71. Datt. 71B in Br. reads मपन्यासस्तु मध्यमः

This is evidently erroneous since apanyāsas have already been named in Datt. 70.

2 उदीच्या: प्रयोक्तारो बाहुल्येन स्तस्यामित्युच्यते । वान्ति मच्छल्नीतः । उदीच्येऽपि च दृश्यते । गीतं च देशनामध्ये-बाहुल्यादिति तत्त्वापेक्षया व्यक्तिदर्शयत एव । टककरागो, मालवपञ्चमो, गौडी, मालवी, काम्बोजीत्यादि

-A.B. on N.S. 106-109.

3 सरिग बलिन्; १ पड्जस्यां स्वरात् सिड्ड बलिन् पुनराधिक्यात्‍यामुख्तमम् । ऋषभस्याल्पत्वात् प्रायं प्रतिषेद्धु (पिद्ध) गान्धारस्य मन्त्रस्थाने बाहुल्यम् ।

-ibid. 28, 106-109,

73A. yathesṭam syācca sañcāro yathāgrāmāvirodhakṛt

B. ṣadjagrāme tu saptaitāḥ śeṣā madhyamajātayaḥ1

There are seven aṃśas in ṣadjamadhyamā. Ṣadja and madhyama are [its] nyāsas. Niṣāda and gāndhāra in due order effect the process of hīnatā, (i.e., it is rendered ṣādava by dropping niṣāda and auḍuvita by dropping niṣāda and gāndhāra). In it sañcāra is left to one’s discretion (yathesṭam) provided it is not contrary to the grāma.

These are the seven [jātis] of the ṣadja-grāma. The rest are madhyama-[grāma] jātis.

NOTE :

The term sañcāra here signifies a specific melodic movement, which was evidently included in the antaramārga.

Dattila’s description of the ṣādava and auḍuva in this jāti leaves the matter somewhat ambiguous. Bharata, however, clearly says that ṣādava was effected by dropping ni and auḍuva by dropping both ni and ga (N.S. 28, 111).

The Jātis Gāndhārī And Raktagāndhārī

74A. gāndhāryā dvāvanamśau tu heyāvṛsabhadhaivatau

B. kramānityamapanyāsau vijñeyau ṣadjapañcamau

75A. ṣadhihatādrṣabham gacchedévam syāt sarvameva tu

B. prāyaśo raktagāndhāryā apanyāsastu madhyamah

76A. bahuprayogah kartavyo dhaivato’tha niṣādavān

B. ṣadjagāndhārāsañcārah kāryaścāsyāḥ prayoktrbhiḥ,2

Ṛṣabha and dhaivata, [these] two [alone] in the gāndhārī [jāti] are non-aṃśas. They are also heya in due order (i.e., their dropping renders the jāti ṣādava and auḍuvita). Ṣadja and pañcama are the two constant apanyāsas. [In this jāti] movement should be from dhaivata to ṛṣabha. All these characteristics are largely the same in raktagāndhārī, though the apanyāsa [in it] is madhyama. In this jāti the application of dhaivata and niṣāda should be copious [and] the experts should render the typical movement (sañcāra) between ṣadja and gāndhāra.

NOTE :

Now we come to madhyama-grāma jātis.

Many features were common between gāndhārī and raktagāndhārī. They had the same aṃśas, the same nyāsas; moreover, the same notes were to be dropped

1 Br. 210-211 are same as Datt. 72-73. Datt. 72 in Br. however reads as:

सवांश्चान्नाकुत्रापि विलीयते पद्जमध्यमा । स्यातां निषादगान्धारावष्टावस्वां गीतत्वकारिणौ ॥

The line might have originally contained some significant pāṭhabheda with the Datt. text, but in the form that we have it the lines are undoubtedly corrupt.

2 Br. 212-214 are same as Datt. 74-76. Line 76B in Datt. T.ed. reads ‘kāryeścāsāṃ prayoktrbhiḥ. Our amendment is based on Br. where 214 reads का (यं चा ?यंश्चा) स्वाः प्रयोकृतृभिः.

in rendering the jātis ṣāḍava and auḍuvita. The distinctive features of raktagāndhārī, however, were that the apanyāsa in it was madhyama; the notes niṣāda and dhaivata (and dhaivata in spite of being an anaṃśa and a lopyasvara)1 were strong. Further, it had a special movement between ṣadja and gāndhāra. This special sa ga movement was, says Bharata, to be made without sounding ri: “gāndhāraṣadjayoścātra sañcāraścārsābhādvinā” (N.S. 28, 117). Abhinava remarks that ri should be skipped in moving from sa to ga and back, thus bringing these two notes together: “rṣabhaṃ collaṅghya sagetyanoranyonyanaikaṭyaṃ melanaṃ ca” (A.B. on N.S. 28, 116–117).

In gāndhārī jāti the two notes rṣabha and dhaivata were not included among the amśas ; further, they were dropped in the process of making the jāti ṣāḍava or auḍuvita. Ṣāḍava, says Bharata, was rendered by dropping ri and dha by omitting ri and dha (N. S. 28, 114–115). This implied that they were characteristically weak notes. Yet a typical movement in the jāti was made from dhaivata to rṣabha ; though, as Abhinava says, only when the jāti was pūrṇa, that is, it used all the seven notes (pūrṇāvastāyāṃrṣabhāddhaivatagamananam; A.B. on N S 28, 115). Bharata mentions another typical movement: “vihitastviti gāndhāryāḥ svaranyāsāṃśagocaraha” (N.S. 28, 115). This Abhinava explains as meaning that all notes which are not amsa or nyāsa should, in making a movement (sañcāra), be associated with the nyāsa and the amśa : “anyeṣāṃ svarāṇāṃ nyāsasvaraviṣayo’mśasvaraviṣayaśca sañcāraḥ, tadāha 'svaranyāsāṃśagocara iti '” (A.B. on N S. 28, 113–115).

The Jāti Gāndhārodīcyavā

77A. gāndhārodīcyavā prāyāḥ ṣadjodīcyavatīsamā

B. ṣadjaśca madhyamścāṃśau na caḍubitaṃśisyate1

The [jāti] gāndhārodīcyavā is largely similar to the [jāti] ṣadjodīcyavatī. Ṣadja and madhyama [in it] are the two amśas and [it] should not be [rendered] auḍubita.

NOTE :

We here meet with another possible case of jāti-ṣādhārana.

In ṣadjodīcyavati, Dattilla has described sa and dha as the apanyāsas, ma as the nyāsa and has said that this jāti employs ga profusely in the lower octave. These characteristics he obviously implies as being typical also of gāndhārodīcyavatī. Features which were not common, he names. Ṣadjodīcyavatī had three amśas while the present jāti had only two ; moreover, the auḍuvita was not permitted here.

1 धैवतो बलवदन दोषंल्य तस्य लोपतः —N.S. 28, 116.

2 Br. 215 is same as Datt. 77.

Ṣāḍava evidently was permitted. Ri as in ṣaḍjoḍīcyavatī was presumably the note dropped. Bharata explicitly says : “ṣaṭsvaryamṛṣabhaṃ vinā” (N.S. 21. 118). Regarding similarities between this jāti and ṣaḍjoḍīcyavatī, Bharata remarks that antaramārga, nyāsa and apanyāsa are here same as in ṣaḍjoḍīcyavatī (N.S. 28, 119). Abhinava elucidates : “antaramārga here comprises of a sañcāra between the two amśa notes—asyāmantaramārgaḥ parasparamaśasvarayoḥ sañcāraḥ”1 (A.B. on N.S. 28, 118-119 ; regarding ṣaḍjoḍīcyavatī Abhinava had similarly remarked : “amśānā-manyonyasaṅgatīḥ”). As in the foregone jāti, gāndhāra was used profusely in the mandra : “mandrasthāne gāndhārasya bhūyastvamiti sarvamatideśāt” (A.B. ibid.).

The Jāti Madhyamā

78A. pañcāṃśā madhyamā yāstu jñeyā dvisrutivarjitāḥ

B. kramāt tābhyāṃ ca hīnatvaṃ bahulau ṣadjamadhyamau2

With the exception of two two-śrutis notes (viz. the gāndhāra and the niṣāda) the madhyamā [jāti] has five amśas. These two [notes] are [also] omitted in due order [to render the jāti ṣāḍava and audubita]. [In this jāti] the use of ṣadja and madhyama is copious (bahulau).

NOTE :

A few words regarding ‘bahula’ in the context of madhyamā. In this jāti, ṣadja and madhyama were, as it is, among amśas, and thus bahutva by definition applied to them as to the other amśas. Why then repeat ? Abhinava, commenting on Bharata, who makes the same injunction, states the purpose : “the bahutva of sa and ma is self-evident for they are paryāyāṃśas. It should be understood from the special injunction that they are more than normally strong : “samayoḥ paryāyāṃśat-vāt siddhaṃ bahutvaṃ punarvacanāt prakṛṣṭaṃ mantavyaṃ” (A.B. on N.S. 28, 122).

Gāndhāra, in this jāti, was a weak note being a lopya svara. Yet Bharata remarks : “gāndhāralaṅghanāṃ cātra kāryaṃ nityam prayoktr̥bhiḥ” (N.S. 28, 122). Abhinava comments that gāndhāra was a svara that could be dropped and was thus by definition weak. To name it specifically as a weak note implied that even when the jāti was rendered as a pūrṇa, septatonic structure, ga was particularly weak: “gāndhārasya lopyatvāllabdhamalpatvaṃ, pūrṇāvastāyāmapī paścāllabhyate” (A.B. on N.S. 28, 120-122).

1 The sañcāra thus was between sa and ma. Br., however, says : “ṛṣyosarṅgatīyeṣāḥ; (Br. 257), a detail which diverges sharply from Bharata. Sārṅgadeva and Kumbha follow Mataṅga on this point: See S.R. 1, 7, 90 and S. Raj 2, 1, 4, 279.

2 Br. 215 is same as Datt. 78. B.B. (I) ch. VI quotes Datt. 78 with the words ‘dattilo pyāh’. The line is quoted in support of N.S. 28, 120-122 which ts introduced with ‘अथमध्यमा, तत् सूत्रं यथा’ B.B. (I) inserts an extra line in between Datt. 78 : ‘अन्य्यासास्त एवं स्यु: वृद्धद्र्मिः परिकीर्तिताः ।’

The Jāti Madhyamodīcyavā

79A. gāndhārodīcyavāttu madhyamodīcyavā bhavet

B. sāptasvaryam tu nityam syādasyāmamśastu pañcamah

The [jāti] madhyamodīcyavā resembles [the jāti] gāndhārodīcyavā. It is invariably of seven notes (i.e., it has no ṣādava and auḍuvita forms) and in it pañcama is the amśa.

NOTE :

This jāti resembled the gāndhārodīcyavā which in turn resembled ṣādjodī-cyavā. They were thus of a group. Their very name, as Abhinava indicates, implied that all these jātis were popular in the same geographical region. This was probably the reason for the many common features in their structure.

In gāndhārodīcyavā there were two amśas : ṣajña and madhyama. It had no auḍuvita but ṣādava was effected by dropping the rṣabha. This jāti madhyamodī-cyavā permitted neither ṣādava nor auḍuvita. Abhinava adds that in madhyamodī-cyavā there was copious use of ma and ga : “madhyamagāndhārabāhulyam” (A.B. on N.S. 28, 123). In gāndhārodīcyavā, on the other hand, only ga was bahula, especially in the mandra octave.

The Jāti Pañcamī

80A. pañcamyām gurubhih proktāvamśavrṣabhapancamau

B. saniṣādāvapan yāsau madhyamarṣabhasaṅgatiḥ

81A. ṣadjamadhyamagāndhāraḥ alpāstu parikīrtitāḥ

B. syānniṣādācca gāndhāro madhyamāvacca hīnatā1

In the pañcamī [jāti] rṣabha and pañcama have been declared to be the two amśas by the teachers. These two along with niṣāda are [also] the apanyāsas. [In this jāti] the saṅgati is between madhyama and rṣabha. The weak notes (alpāḥ) are ṣadja, madhyama and gāndhāra. [Movement] should also be [made] from niṣāda to gāndhāra. In it the process of hīnata (i.e., ṣādava and auḍuvita) is the same as in madhyamā [jāti].

NOTE :

There are two readings in the Nāṭyaśāstra regarding the saṅgati in this jāti. The Asiatic Society edition enjoins a sañcāra (i.e., saṅgati) between madhyama and

1

Br. 218-219 is same as Datt. 80-81. Br. 219B appears to have been repeated with slightly different readings. The line first reads : र्षानिषादावच्च गान्धारो अल्पस्य परिकीर्तितः. The very next line repeats the matter with a reading closer to the Datt. text. It has a lacuna of three letters : स्वानिषादाच्च++मध्यमा चकच हीनता. B.B. (I) ch. VI quotes Datt. 80-81 after describing pañcamī in स्तानिषादावच्च++मध्यमा चकच होनता. the words of N.S. 28, 124-126. The Datt. lines are introduced with 'यददाह दत्तिलः. 80A in B.B. (I) reads पञ्चम्यां मुनिभिः;; 80B reads मध्यमया यत्संगतिः.

ṛṣabha : “sañcāraṃ madhyamasyarṣabhasya ca” (N.S. A.S ed. 28, 133)—a reading which agrees with that of Dattila. Śārṅgadeva and Kumbha also supports this.1 However, the reading in the Gaekwad edition of the Nāṭyaśāstra records a sañcāra between pañcama and ṛṣabha, though the commentary of Abhinava seems to lend support to the reading in the Asiatic Society edition: “rima (pa) ityanayoranyonyasaṅgatih” (A.B. on N.S. 28, 126). The pa in brackets is not part of the manuscript reading but added by the editor. Indeed we see that commenting on N.S. 28, 127-128 Abhinava clearly remarks that Bharata has enjoined a saṅgati between ma and ri: “uktam ‘madhyamarṣabhasaṅgatirniṣādādgāndhāra iti pañcamyām’.”

Though the main saṅgati was between ma and ri, another movement has been mentioned between ni and ga; this might be called a secondary saṅgati, for Bharata says that its use should be less frequent (gāndhāra gamanam caiva kāryam tvalpaṃ ca saptamāt, N.S. 28, 126). Abhinava, logically enough, adds that this movement takes place when the jāti is pūrṇa; for, otherwise, these two were the notes omitted in the process of ṣāḍava and auḍuvita (see madhyamā).

Regarding the apanyāsas in this jāti Abhinava reads the N.S. text as ‘ṛniṣādā-vapanyāsau. But on the basis of Dattilam and considering Śārṅgadeva's remark, who recounts three apanyāsas like Dattila (ṛṣabapañcamanīṣādā apanyāsāḥ, S.R. 1, 7, 75), the correct N.S. reading seems to be: saniṣādāvapanyāsau. For, ṛ as short for ṛṣabha would be unusual in N.S. where no svaras are read in a shortened form.

The Jāti Gāndhārapañcamī

82A. jñeyo gāndhārapañcamyāṃ pañcamo’ṃśah prayoktrbhiḥ

B. sarṣabhaḥ syādapanyāso nyāso gāndhāra iṣyate

83A. gāndhāryāmātha pañcamyāṃ yat sañcārādi kīrtitam

B. tadasyāmapi vijñeyaṃ kintu pūrṇasvarā sadā.2

The experts should know that in gāndhārapañcamī, pañcama is the aṃśa [which] along with ṛṣabha is [also] the apanyāsa; gandhāra ought to be the nyāsa. The sañcāra etc. which were mentioned in gāndhārī and pañcamī should be known to [apply] here as well, though this [jāti] invariably remains pūrṇa (i.e., septatonic).

1 रिमयोःसगति: - S.R. 1, 7, 73. रिमयोः संगति: कार्यो - S. Raj 2, 1, 4, 258.

2 Br. 220-221 is same as Datt. 82-83. Line 220A reads : गान्धारपञ्चम्या:. B. B. (I) quotes Datt. 83B along with Datt. 80-81. The line in the B. B. (I) ms. has a corrupt reading: स्यात्तरालीस्वरपपदति च (?)1

NOTE :

This jāti was derived from an intermingling of gāndhārī and pañcamī (see Datt. 54). This seems to be the reason why it contained the sañcāra of both. The sañcāra in gāndhārī was a movement made from dha to ri. Pañcamī had two sañcāras, a predominant one between ma and ri and a secondary one from ni to ga.

Gāndhārapancamī included all the three characteristic movements of its parent jātis. This fact had led Kallinātha to remark that the sangati here should be effected with many notes: “evaṃ bhūribhiḥ svaraiḥ sangatiḥ kartavyāḥ.” (Kalā on S.R. 1, 7, 103-104). Bharata adds that, in this jāti, the movement in the tāra was not to reach beyond the tāragatyā tu ṣadjo’pi kadācinnātivartate” (N.S. 28, 127). This was a trait which Dattila records only regarding the jāti nandayantī.

The Jāti Andhrī

84A. andhryāmanamśā vijñneyāḥ ṣadjamadhyamadhaivatālḥ

B. ṣāḍabaṃ ṣaḍjahināṃ tu nyāso gāndhāra iṣyate¹

Sadja, madhyama and dhaivata in the[jāti] andhrī should be known as the non-aṃśas. Dropping of ṣaḍja [renders it] ṣāḍava. Gāndhāra should be the nyāsa.

NOTE:

This jāti has been called āndhrī by Bharata (N.S. 28, 129), perhaps a more correct name. Āndhrī, as his discription indicates, had no auḍuvita (nāsti cauḍuvitaṃ sadā”, N.S. 28, 131). The aṃśa notes were also apanyāsas (N.S 28, 129). A sañcāra is mentioned by Bharata between gāndhāra and ṛṣabha (N.S. 28, 130). He gives another intriguing detail: “saptamasya ca ṣaṣṭhāsya nyāso gayanupūrvaśaḥ” (N.S. 28, 131). Abhinava's elucidation is not very clear; he seems to be suggesting that a movement to dha ni should be made from the note which was accepted as the aṃśa and was hence predominant. Abhinava, however, adds that another explanation maintained that the aṃśa svaras were rendered in the order of their enumeration (pa, ri, ga, ni) till the apanyāsa was reached.² Kallinātha (on S.R. 1, 7, 105-106) quotes Bharata and an unnamed authority in explaining this matter. But he makes a statement so general in nature as to be hardly any explanation at all. He says: “the first note to be sung should be the one which is accepted as the aṃśa amongst the notes ni, ri, ga and pa. Subsequently should follow a paryāyāṃśa or a non-aṃśa and the process should culminate with the last note of the gīta, i.e., the nyāsa.³

1 Br, 220B-221A is same as Datt. 84.

2 ‘niṣādrīdhaivatayośca nyāso gayanupūrvaśaḥ.’ yo yadāṃśaḥ: pradhānabhūtastena sva (nya) sāgaman kāyam | taduktam | ‘aṃśānukramato yojyaṃ gītastayānuṣṭhānukramāt:, tamśo bhartṛ puṣpāṃ nyāsagamanmiti | pratye caḥḥu: yenaiva krameṇāśaḥ svāraḥ paṭhitavastaneva tat kāryābhyāsasvar iti | —A.B. on N.S. 28, 129-131.

3 अंशानुक्रमतो न्यासपर्यन्तं व्रजेदिति १ अंशानुक्रमतः. अत्र निरूपणां मध्ये यदा योगपीठतस्त्यानुक्रमात:, तमशोभर्तः पूर्वमुच्चार्यन्ते पर्यायो वा पश्चादौदुचोर्यन्त न्यासपर्यन्तः., गीतसमाप्तिकृत्स्वरपर्यन्तम्, आगीतरिसमाप्तौपरित्यर्थ:,., पूर्वमुच्चार्यन्ते पर्यायो वा पश्चात् (N.S. 28, 130B-31A follows) ग्रन्यान्तरे च—स्वांशानुपूर्वीं ब्याक्तिं च व्रजेद् गायेदिति नियमः: तथा चाङ्ग भर्तः (N.S. 28, 130B-31A follows) ग्रन्यान्तरे च—स्वांशानुपूर्वीं ब्याक्तिं च कुरुत्‍न्यसावधि ब्रजेत् । —Kalā on S.R. 1, 7, 105-106.

290 A Study of Dattilam

Bharata remarks that the ṣadja here should be unemphasised by langhana (N.S. 28, 131).

The Jāti Nandayantī

85A. nandayantyāmapanyāsau jñeyau madhyamapañcamau

B. graho nyāsasca gāndhāraḥ pañcamo’ṃśaḥ prakīrtitaḥ

86A. andhrīvat sāḍabam jñeyamanudubhitameva ca

B. syānmandrurṣabhasañcāro langhanīyaśca sa kvacit1

In nandayantī [jāti] madhyama and pañcama should be known as the two apanyāsas. Gāndhāra is [its] graha and the nyāsa; pañcama is said to be the aṃśa.

As in andhrī, [the process of] sāḍava should be known [by dropping ṣadja] and [likewise it] cannot be [rendered] auḍuvita. [In it] the sañcāra is that of mandra ṛṣabha, though at times the note may be [unemphasised by] langhana.

NOTE :

The phrase ‘sañcāra of ṛṣabha in the mandra’ here has been used more in the sense of bahula or in other words ‘copious use’. In fact, Mataṅga, Śārṅgadeva and Kumbha have used the word ‘bahula’ instead of ‘sañcāra’.2 A specific mention of the mandra ṛṣabha evidently implied that in the other octaves the use of ṛṣabha was not so copious.

Another characteristic feature of this jāti was that its movement in the tāra never exceeded beyond the ṣadja (Datt. 58).3

This jāti was born of pañcamī, āṛṣabhī and gāndhārī. It had similarities with the andhrī which arose out of gāndhārī and āṛṣabhī. To distinguish it from andhrī, Bharata enjoins that the sañcāra or the characteristic movements of andhrī are not here applicable (nāndhrīsañcaranāṃ bhavet, N.S. 28, 133). Abhinava regards this particular injunction as indicative of a general precept: jātis which arise out of a common source should not share a common saṅgati.4

Bharata and Dattila are unequivocal in considering gāndhāra as the graha in this jāti. But Mataṅga and later authors indicate that, according to some, pañcama was the graha in this jāti (Br̥. 275). In all jātis, graha and aṃśa were the same note. Whichever note happened to be the aṃśa was also the graha (Datt. 57; N.S. 28, 67). Nandayantī was the only jāti where, according to Bharata and Dattila, aṃśa

1 Br̥. 223-B225A is same as Datt. 85-86.

2 मन्त्रवषंस्य बाहुल्यं Br̥. 275. Śārṅgadeva uses the same expression: S.R. 1, 7, 108: ऋषभे मन्त्रदेशेग बाहुल्यं

See also S. Raj 2, 1, 4, 329-330.

3 Bharata has an identical injunction: तारगत्या तु पञ्जास्तु कदाचनातिवर्तते ।

-N.S. 28, 134.

4 नाम्नीसंचरणमिति । गान्धार्यर्षभीसंयामन्थी पंचभ्यार्षभीसंध्योऽधरीसंनतीति संसर्जप्रकरणं निरूपणं (1) जाति वयक(क)रणतुल्य द्वयेक(क)रणतुल्याद्()ध्रीसंचरणं निपीडयते । एतच्चवार्त्रक(क)तुल्यजातिकारणिकानां जातीनां सझीति(ति)रतुल्येन्व कथंक्षयेति सूचर्यति ।

• A.B on N.S. 28, 132-34.

and graha were separate notes. Abhinava says: “sometimes the amśa is not the graha, such as pañcama in nandayantī. In this jāti gāndhāra is the graha.”1

Regarding the tāra movement in nandayantī, Bharata says: “tāragatyā tu ṣaḍjastu kadācinnātivartate” (N.S. 28, 134). Abhinava explains this as meaning that in the tāra one should never reach beyond the sa in this jāti; no note above the sa in the tāra octave should be touched ; “tāragatyā saḍjo nātivartate nātiśayyate tata ūrdhvakam na śirṣyasaptakasvarāḥ praṣṭavyā ityarthah.” Others, he says, had explained the injunction as meaning that the tāra ṣadja itself also was not to be touched, thus leaving no range in the higher octave at all : “anye tvācakṣate ṣadja eva kadācinnātivartate nāroham bhajate tāram na kīñcit praṣṭavyamityartham.” Yet others understood the word ‘kadācit’—sometimes—in Bharata as indicative of an alternative; it was upto one’s discretion to move in the tāra octave beyond sa or to avoid this : “kadāciditi vaikalpiko’yaṃ tāra ityarthah.” Dattila, however, unequivocally indicates that no such alternative was permissible.

The Jāti Kārmāravī

87A. kārmāravyāmanamśāstu ṣadjagāndhāra pañcamāḥ

B. pūrṇatā pañcamo nyāso gāndhāragamanam bahu3

In the [jāti] kārmāravī the non-amśas are ṣadja, gāndhāra and pañcama. [This jāti] is pūrṇa. Pañcama is the nyāsa [and] there is a recurrent movement to the gāndhāra.

NOTE:

The movement to gāndhāra needs a few remarks. Dattila merely indicates that a recurring movement should be made to ga. Bharata, however, has more details regarding the nature of this movement. He says: “gāndhārasya viśeṣeṇa sarvato gamanam bhavet” (N.S. 28, 136). Abhinava explains: “one should associate or move to the gāndhāra from all the notes, as far as possible : sarvebhyo’pi viśeṣeṇa tu gāndhārasya yathāśakti saṅgatim” (A.B. on N.S. 28, 135-36). This movement was to be made also from the amśa notes. For in this jāti, according to some, it was held that non-amśas predominated. Abhinava quotes Viśākila who had stated : “since the nonjamśas here are copious (bāhulyāt) one should particularly make a movement from all [the notes] to the gāndhāra—yadāha viśākhilācāryaḥ bāhulyādnamśānāṃ viśeṣataḥ sarvato gāndhāragamanamanam” (A.B., ibid.).

In this jāti, contrary to the general rule, the non-amśa notes appear to have been rather strong. This fact is clear from the above quotation of Viśākhila as well

1 अंशोडपि हि कदाचिद् ग्राहो न भवति । नन्दयन्त्यामपिव पंचमः । तस्य हि ग्रहो गान्धारः ।

• A.B. on N.S. 28, 67.

2 Br. 225B-26A is same as Datt. 87. T.ed. reads गान्धार (गमनं) बहु. The word in the bracket has been obviously filled in on the basis of Br.

292 A Study of Dattilam

as from the views held by later authorities like Matanga, Śārṅgadeva and Kumbha.1

Kallinātha, in fact, quotes Bharata as having stated that “in the application of [this

jāti] the non-aṃśas are invariably strong”.2 Matanga somewhat qualifies this state-

ment in saying that “the anaṃśas were strong in the antaramārga” (Br. 268).

The antaramārga, we see, had a good deal of variety. Though rendered

usually by stressing the strong notes, normally the aṃśas, and unemphasising the weak

ones [usually the hīna notes and the non-aṃśas], yet, as in this case, the non-aṃśas, too,

could be given a strong character and emphasised in the antaramārga. We have

noticed other irregularities and peculiarities of the antaramārga earlier and here we

have a particularly out-of-the-way example. Our own experience of rāga-forms,

however, teaches us that such irregularities in movements are often the rule, for not

every aspect can be generalised. The ‘calana’ of many a rāga (though rāgas are

bound by many general rules as to strong and weak notes and principles of movement)

achieves its required form only outside the boundary of regularities and general

maxims.

Kallinātha in the context of kārmāravī raises a pertinent question : since in

this jāti both the aṃśas and the non-aṃśas were emphasised what, then, was to

distinguish them ? His answer was that in antaramārga the notes which were

emphasised in the sthāyi varṇa (sthāyitvena) were the aṃśas and those emphasised in

the sañcāri varṇa (sañcāritvena), the non-aṃśas.3

The Jāti Kaiśikī

88A. kaiśikyāṃrṣabho’ namśo nyāsau tu dviṣr̥tī smr̥tau

B. r̥ṣabho dhaivataścaiva heyāvasyāṃ yathākramam

89A. pañcamo’ pi bhavennyāso niṣāde’ me’ tha dhaivate

B. r̥ṣabhaḥ syādapanyāsaḥ kaiścidukto’ mśavat tathā

90A. pañcamo balavānasyāṃ syānniṣādasthathaiva ca4

1 Matanga says : बहवोऽन्तरमार्गा विवृतवन्तः; परिकोचितः

• Br. 268.

Śārṅgadeva repeats the same words.

• S.R. 1,7, 101.

Kumbha rephrases Matanga's words : तत्रैवान्तरमार्गावदानं बहुलता मता ।

• S. Raj 2, 1, 4, 350.

2 तथा च भरतः- अनाम्सस्तु नित्यमेव प्रयोक्तः ।

-Kalā on S .R. 1, 7, 101-102.

This line is not in N.S.

3 यः स्थायितवेन बहुप्रयोगः सोंडः । यस्तु संचारितवेन बहुप्रयोगोऽसौन्तरमार्गप्रयोजक्श्र इति विवेकः ।

-Kalā on S.R. 1,7, 101-102.

4 Br. 226B-229A is the same as Datt. 88–90A. Two letters are missing at end of line in Br. 226B.

Part of Br. 228 is a prose passage obviously not part of the verse-text but that of the prose

Vṛtti ; but it has been mistakenly taken to be part of Br. verses.

In the kaiśikī [jāti, the only] non-aṃśa is ṛṣabha. The two two-śruti notes (namely, gāndhāra and niṣāda) are said to be the nyāsas ; ṛṣabha and dhaivata should be dropped in due order [to render the jāti ṣāḍava and auḍuvita].

When niṣāda or dhaivata becomes the aṃśa, pañcama too can be nyāsa. Rṣabha too, like the aṃśas, according to some, can become an apanyāsa. In this [jāti] pañcama and niṣāda are strong [notes].

NOTE:

In this, as in another jāti, the ṣadjamadhyamā, there were more than one nyāsas. This perhaps implied that any of the notes mentioned as the nyāsas could be taken as the nyāsa proper for a particular rendering, since it is not possible to have more than one note as the last note of a gīta. Nī and ga were the two specified nyāsas, but it has also been enjoined by Dattila that when either dhaivata or niṣāda became aṃśas then pañcama, too, was added to the choice of nyāsa notes (pañcamo'pi bhavennyāso).

However, Bharata and other authorities have laid down that in the case of dhaivata or niṣāda being the aṃśa, pañcama was not just a possible nyāsa but the only nyāsa.1 Bharata says: “dhaivate'mśe niṣāde ca nyāsalḥ pañcama iṣyate” (N.S. 28, 138).

Mataṅga agrees: “nidhāvāṃśau yadā tadā/nyāsalḥ pañcama eva syādyanadā dviśrutī matau” (Bṛ 262).

Dattila has stated that, like the aṃśa notes, ṛṣabha, too, can, according to some, become an apanyāsa.

This is another way-a shorter way-of saying that the aṃśas in this jāti were in any case also apanyāsas-a fact which Bharata mentions more explicitly: “aṃśāḥ sarve caivarṣabhaṃ vinā, eta eva hyapanyāsāḥ”.

(N.S. 28, 137). Bharata also adds ri as a possible apanyāsa: “apanyāsalḥ kadācittu ṛṣabLo'pi vidhīyate” (N.S. 21, 140).

Abhinava comments that ṛṣabha became an apanyāsa only when the jāti was pūrṇa (pūrṇadasāyām).2 Ri was otherwise a weak note in this jāti ; it was dropped both for making the jāti ṣāḍava and auḍuvita. It was weak even when not dropped.

Bharata expressedly says: “daurbalyaṃṛṣabhasyaatra laṅghanaṃ ca viśeṣataḥ : ṛṣabha is here a weak note and laṅghana should particularly apply to it” (N.S. 28, 140).

Another characteristic was that although being aṃśas pañcama and niṣāda were implicitly strong, yet they have been, in addition, also explicity called strong. Bharata on this point makes a statement similar to Dattila : “balinau cāntyapancamau : the final note (i.e., nī) and pañcama are strong notes” (N. S. 28, 139).

Abhinava clarifies by saying that pa and nī are stronger in this jāti in relation to the other aṃśa notes : “pañcamanīṣādayoḥ paryāyāṃśananyāpekṣaṃ balavattaratvamiti yāvat” (A. B. on N. S. 28, 137-140).

Kallinātha has made a similar remark : “itarāṃśāpekṣayātibalulyam vidhīyate” (Kalā on S.R. 1, 7, 95-98).

1 Abhinava remarks : पयवण ध्वनतनिषादौ यदांशौ तदा पञ्चमो न्यासः। —A.B. on N.S. 28, 137-140.

2 कदाचित्तु ऋषभोऽपि इति । लोगवशाद्रायां तु नासौ । तथा पूणताय शायां तु वा भवतीति । =A. B. on N. S. 28, 137-140 ,

Bharata has mentioned a sañcāra here. He says that sañcāra should be as in ṣadjamadhyamā. In that jāti all seven notes were paryāyāṃśas and thus strong ; sañcāra was consequently enjoined to be more or less free to roam over all the notes (N. S. 28, 112 ; Datt. 73). Kaiśikī had six amśas and so here, too, sañcāra could be relatively free. Abhinava qualifies the sañcāra by saying that this jāti had six amśas (not seven as in ṣadjamadhyamā) and hence its sañcāra was slightly curtailed : “sadamśā hyeseti kiñcinnyūnasañcāreyamiti” (A. B. on N.S. 28, 137-140). Abhinava makes another remark in this connection which, however, is somewhat puzzling. He says that the sañcāra in this jāti could be yathesṭam, free as in ṣadjamadhyamā, when dhaivata was the amśa : “ṣadjamadhyamāyāṃ ca yathā cāha. dhaivatasvarāṃśāyāṃ yathesṭam sañcārastadvadiha” (A. B., ibid).

The Aggregate of Amśas

90B. iti triṣaṣṭir amśā ye teṣāmekai keṣo'mśatāṃ 91A. prakalpyāpodyate prāptam śāḍabauḍubitaṃ kvacit1

In this way there are sixty-three amśas. Among them taking individually each amśa as the amśa [proper], the process of ṣāḍava and auḍuvita in some [exceptional] cases shall [now] be laid down.

NOTE :

The statement “iti triṣaṣṭir amśāḥ” may sound puzzling. However, it points to a very simple matter. If one puts together the number of possible amśas in all the 18 jā tis the total comes to 63 :

1. ṣādjī
2. ārṣabhī
3. dhaivatī
4. niṣādavatī
5. ṣadjakaiśikī
6. ṣadjodiṣyavā
7. ṣadjamadhyamā
8. gāndhārī
9. raktagāndhārī
10. gāndhārodīṣyavā
11. madhyamā
12. madhyamodiṣyavā
13. pāñcamī
14. gāndhārapañcamī
15. andhrī
16. nandayantī
17. kārmāravī
18. kaiśikī

5 3 2 3 3 4 7 5 5 2 5 1 2 1 4 1 4 6

63

1 Br. 229B-230A is same as Datt. 90B-91A.

Bharata in a long passage (N.S. 28, 79-91) enumerates jāti-wise the number of amśas (and grahas) in each jāti, remarking at the end that thus the number of amśas in all the jātis together amounts to 63 : “ete triṣaṣṭirvijñeyāḥ sarvāśvamśāstu jātiṣu” (N. S. 28, 91). Abhinava, elucidating, sums up : “forgetting the grāma division among jātis, because the total aggregate is being reckoned, Bharata now gives the number of amśas in each. He begins with one-amśa jātis and then successively in due order recounts the jātis with all seven svaras as amśas. In this context we see that three jātis have one amśa each; three have two amśas each ; three have three amśas each ; another three have four amśas each ; four jātis have five amśas each ; one jati has six amśas and finally another has seven amśas—total being 63.”1

Enumerations of this nature were a favourite with ancient theoreticians. Bharata has also enumerated the aggregate of possible nyāsas and apanyāsas in all the jātis put together. The purpose of such aggregation is not clear. It could not, evidently, serve a musical purpose. There is, however, one likely explanation. Jātis were bound by very strict rules and bye-rules; amśa, nyāsa, apanyāsa and all other notes were strictly fixed and numbered. Making an aggregate summed up their exact total and thus the student could in a nutshell keep in mind the limited possibilities and their extent.

The grahas add up to the same number and also constitute the same svaras as the amśas with a single exception in the case of naudayantī where the graha is gāndhāra and the amśa is pañcama (Datt. 85).

In ancient texts, graha and amśa are often equated. Bharata says that in all the jātis, grahas are enumerated in the same manner as the amśas: “grahāstu sarvajātīnāmamśavatparikīrtitāḥ” (N.S. 28, 67). Its implication, as Abhinava makes clear, is that both the grahas and amśas have the same total number, i.e., both aggregate 63 (militvā amśavadbhavantīti triṣaṣṭirityarthāḥ, A.B. N.S. 28, 67). Any further equation, particularly as to 'function' of the two concepts, seems to be far-fetched if not totally unwarranted. The Vrtti on Brhaddeśī, indeed, while equating graha and amśa in as much as their aggregate is concerned, takes care to distinguish the two functionally. The Vṛttikāra says that graha is the commencing note from which the application of a jāti begins (tatrādau jātyādiprayogo gṛhyate yenāsau grahaḥ, Vrtti on Br. 197), whereas amśa was a larger concept. It was the note which determined the structure of a rāga (also a jāti by implication) and was more pervasive and thereby had a greater predominance (rāgajanakatvād vyāpakatvāccāmśasasyāya prādhānyanv- cyate, Vrtti on Br, ibid.).

1 तत्र षड्जग्रामनिरोग्राद ग्रामविभागक्रममनुद्यैवे कतिचिद्यादियोगां सप्रांशपग्रहान्तमाह । ‘मध्यमोदीच्यवायास्वादिमादि’ । तत्र तिसृणामेकांशा: । तिसृणां द्वौ (द्वौ दो) । तिसृणां त्रय: । तिसृणां चत्वार: । चतसृणां पंच । एकस्या: सप्त (षट्) (एतस्या:) सम्प्रति द्वयष्टि: । —A.B. on N.S, 28, 79-91.

Nijenhuis in Dattilam : A Compendium of Ancient Indian Music has recounted the number note-wise and not jāti-wise (pp. 298-99). Though in effect the aggregate is the same, this is not how the ancients made the calculation. The jāti here was the chief factor not the svaras.

296

A Study of Dattilam

Amśas Which Did Not Permit Sāḍavita And Auḍuvita

91B. aśāḍabā niṣāde'mse sati syāt ṣadjamadhyamā

92A. gāndhāre ca yato lopo nāṃśasamvādinarmatah

B. kaiśikī raktagāndhārī gāndhārī caiva pañcamame

93A. tathā sāḍjī tu gāndhāre dhaivate tadudīcyavā

B. gāndhārīraktagandhāryoh ṣadjamadhyamapañcamāḥ

94A. saniṣāḍah smṛtā amśā anauḍubitabhāginaḥ

B. ṣadjamadhyamajātiātau tugāndhāro'tha niṣādavān

95A. kaiśikyāmatha pañcamyāṃ kramaśo dhaivatarṣabhau

B. yojyam saptādhikeṣvevam catvāriṃśatsu śāḍabe

96A. apavādinirmuktam trimśadauduḍubitaṃ bhavet1

In case of [either] niṣāda or gāndhāra being the amśa, [the jāti] ṣadjamadhyamā cannot be rendered śāḍava, for the maxim is that the samvādi of the amśa cannot be dropped.

[Similarly] if pañcama [be the amśa in] kaiśikī, raktagāndhārī and gāndhārī ; [if] gāndhāra [be the amśa in] sāḍjī, and dhaivata in ṣadjodīcyavā [then these jātis cannot be rendered śāḍava].

It is [also] said that if in gāndhārī and raktagāndhārī the amśas are ṣadja, madhyama and pañcama along with niṣāda ; if in ṣāḍjamadhyama [the amśas] are gāndhāra and niṣāda; and if in kaiśikī and pañcamī [the amśas are] dhaivata and ṛṣabha respectively, [then these jātis] cannot be rendered auḍuvita.

Thus, with the forty-seven [amśas that permit] śāḍava, should be counted the thirty which, barring exceptions, [render the jātis] auḍuvita.

NOTE :

After noting the aggregate of amśas, Dattila lays down certain conditions for rendering jātis śāḍava and auḍuvita. In gāndharva, we have seen, only certain notes could be dropped in rendering melodic structures śāḍava and auḍuvita ; single notes and pairs of notes that could be omitted have been enumerated for each grāma separately (Datt. 32-35). Besides, in studying individual jāti-structures, we observed rules regarding particular notes which could be dropped in specific jātis. We also noted that many jātis could not at all be rendered hīna—they had to be always septatonic ; others could not be rendered auḍuvita. Now we come to injunctions which lay down further jāti-wise limtiations.

Many jātis had more than one amśa. Only when certain amśas were ruling could these jātis be rendered śāḍava and auḍuvita. Dattila (from verses 91B to 96A) enumerates amśas which did not permit either the śāḍava or the auḍuvita.

1 Br. 243B-250—excepting verse 249—are the same Datt. 91B-96A. They are interspersed with the prose Vṛtti. Br. 243B corresponding with Datt. 95B reads : योक्त॑ध्यक्ष॒रच॑वै॒ चत॑वारि॒मशा॑ट् पाडव॑

Br. 250B corresponding with Datt. 96A reads : तथापपावादिनि॒मुक्त॑ त॒क्लि॑ षा॒डो॑डु॒वि॑ते भवेत्॥

All the notes in the ṣadjamadhyamā were paryāyāṃśas. The jāti was rendered ṣāḍava by dropping ni and auḍuvita by dropping ni and ga. But in case ni was taken as the amśa, the jāti could not naturally be rendered ṣāḍava, for the same note cannot be the amśa as well as the one that can be dropped. Kallinātha says pointedly : “there is a self-contradiction in assigning the role of an amśa to a note while dropping it at the same time” (amśatvaṃ lopyatvamiti ca dharmayorvirodhāt, Kalā on S.R. 1, 7, 85-87). This non-permissibility-of-ṣāḍava rule applied also when gāndhāra was taken as the amśa. For nisāda could not then be dropped, being the samvādī of the amśa. The non-omission of the samvādī of the amśa was evidently a general maxim.

Regarding ṣadjamadhyamā, Bharata makes a similar statement in this context :

ṣaṭsvarāḥ saptame hyamśe neṣyate ṣadjamadhyamā samvadyalopādgāndhāre tadvadeva hi neṣyate

(N.S. 28, 58-59)

“When the seventh note (i.e., ni) is the amśa, ṣadjamadhyamā cannot be rendered ṣāḍava ; similarly, with gāndhāra as the amśa it can again not be rendered ṣāḍava because the samvādī [of the amśa] is not to be omitted.”1

We can come to understand in many cases why certain notes were not dropped by applying the maxim of ‘samvādyalopa’ (the non-omission of the samvādī). The three jātis, kaiśikī, raktagāndhārī and gāndhārī, belonged to the madhyama-grāma where a samvāda existed between pa and ri. In these jātis ṣāḍava was effected by dropping the ri. Consequently, when pa was the amśa, ri could not be dropped because of the samvāda and ṣāḍava was not permissible. In ṣāḍjī, ṣāḍava was effected by dropping ni which was a samvādī of ga; thus when ga was the amśa, ni was not dropped. In the jāti ṣadjoḍīcayavā which belonged to the ṣadja-grāma, ṣāḍava was effected by dropping ri which could not be dropped when, among its four paryāyāṃśas, dhaivata became the amśa, for a samvāda existed between dha and ri.

The ‘samvādyalopa’ maxim extended to the process of auḍuvita also. In the jāti ṣadjamadhyamā, auḍuvita could be effected by dropping ni and ga : when one of these was the amśa it could not be dropped and the other being its samvādī, also could not be dropped.

In kaiśikī, auḍuvita was effected by dropping ri and dha-notes which had a mutual samvāda ; therefore, when dha was the amśa, auḍuvita could not take place.

1 Abhinava comments : निषादे गान्धारे चांशौ तद्वति (?) घटस्वरा नेप्यत इत्यर्थः। अल हेतु:- संवादिनो लोपाभावान्निषादे न स्वार्त पातः। स च गान्धारस्य संवादी । —A B. on N.S. 28, 57-63.

Natyāśastra 28, 58B-59A is, as we see, parallel to Datt. 92B-93A, containig the same matter in a different language. Furlher verses are similary parallel to Datt. above. See N.S. 28, 59-64A.

However, regarding the jātis gāndhārī, rakta­gāndhārī and pañcamī, the above maxim (the detailed application of which is found in Abhinava)1, did not apply. According to Abhinava, in these cases injunction alone set the rule ('tathānirūpaṇameva hetuḥ', A.B. on N.S. 28, 57-63).

After recounting aṃśas which did not permit ṣāḍava and auḍuvita, Dattila again gives us an aggregate which may sound puzzling but it is again a matter of simple calculation, evidently, without any musical significance. The Vrtti on Brhaddeśī has detailed the calculations in a straight-forward language as follows :

"among the jātis there are four which are always pūrṇa, i.e., they cannot be rendered either ṣāḍava or auḍuvita ; these four jātis contain a total of nine aṃśas. In seven jātis, namely, ṣadjamadhyamā, ṣādjī, ṣadjodīcyavā, kaiśikī, rakta­gāndhārī and gāndhārī, seven aṃśas have been declared as exceptions to the process of ṣāḍava. Thus subtracting these sixteen out of sixty-three, we get an aggregate of forty-seven aṃśas which when used permit the jātis to be rendered ṣāḍava". (Br̥. Vrtti on 248). A similar calculation will result in thirty aṃśas for the auḍuvita (Br̥. Vrtti on 250).

Bharata does not give aggregates of the number of aṃśas permitting ṣāḍava and auḍuva though he gives indications that he too had these aggregates in mind.2 Abhinava has given the process of arriving at the numbers in almost the same terms as the Vrttikāra.3

1 'aṃśavikala­panamiti' aṃśas­vara­vikal­pan­y­ate yen. aṃśān jāta­vāy­am. aṃśa: 'pā­ṭa­ro­hita (? pā­ṭa­da­ro­hita)' iti mūla­gran­th­en vi­cā­ry­ante. tathā vakyam: 'a­ṛ­tuk­ala­panamiti' pāṭhe 'aṃśasya pā­ṭa­v­u­t­vi­ka­l­pa­s­ay vika­l­pa­n­am kathanam. asm­in­nan­de tat­nā­s­tit­i. svap­nam iti. ni­ṣā­de 'gān­dhā­re vā­ci (?) tathā­tti (?) pa­ṭh­varā ne­ṣyat iti h­ya­rya: 'a­v­at­te:'—saṃ­sā­di­no­lopa­bhā­v­a­n­i­ṣa­di­nāṃ syāt 'pā­ṭa­ḍvaḥ; sa ca gāndhā­ra­sya saṃ­vā­dī. gāndhā­rv­ān­āṃ ti­ru­ṣa­ṇa­mu­kha­meṇa pā­ṭa­ḍva­tv­am­uk­tam. sa ca madhyam­grā­me pa­ñcama­sya saṃ­vā­dī­t­i. ta­taś­ca pā­ṭa­ḍva­bhā­ve 'c­hāy­ā­de­r­ā­m pā­ṭa­jo­dī­cya­de­r­āṃ vā pā­ṭa­ḍvaḥ. ṛ­tu­ṣa­ṇa­meṇa hi tat tu. sa ca ḍva­vat­s­ya saṃ­vā­dī pā­ṭa­ḍva­grā­me. eka­lū­tyo­pa­saṃ­har­ti. ni ga pa pa sa ma ga ī­p­ye­ta­vā; a­pa­ḍi­kṣa: kāmeṇa pā­ṭa­jo­madhyama­diṣṭu pā­ṭa­jo­dī­cya­de­r­au (?) śru­ti­v­i­ci­tra­mu­khy­ā­n­a­m­śa­nā­maḥ. 'gāndhā­rī­r­ak­ta­gāndhā­ry­ā­mi­r­ya­d­i'. 'svap­nam iti'. ni­ṣā­d­aḥ pā­ṭa­ja­madhyamāyāṃ yo ḍva­v­śruti­r­nā­yā­ra­ni­ṣa­d­āu tā­va­n­n­o­ḍu­vi­ko. u­pa­saṃ­ha­re 'e­vam tu ḍva­d­rav­meṇa'iti. gāndhā­rī­r­ak­ta­gāndhā­ryo­r­a­ṣṭa, pā­ṭa­jo­madhyamāyāṃ madhyamāyāṃ hī 'pa­ñcama­sya' kāś­ci­kā­nyo 'ḍva­r­vi­ti. a­n­na kar­v­ci­tra­de­r­sa­va­lopo yu­kti­r­kṣa­ṇy­ate. ta­da­labhe tu t­u­ṅ­thā­ni­r­u­paṇa­mev a­h­e­tuḥ.

2 saṃvād­alopa­t saṃ­p­itā; pā­ṭa­s­va­reṇa vi­va­r­j­itāḥ.

—N.S. 28, 61.

again evaṃ tu ḍva­d­ra­v­me­ṇa h­e­tuḥ pa­ñca­s­va­raḥ sadā.

—Ibid., 28, 63.

3 tat­va­t­su­r­ṇāṃ jā­tī­nāṃ nitya­pūrṇa­n­āṃ na­v­aṃ­śāḥ: a­pā­ḍi­kṣa iti catuḥ pa­ñc­a­śat. te­ṣa­m­o­pi sa­pta pā­ṭa­ḍva......sat (?) iti saṃ­st­h­ā­tvā­r­i­ṣa­t pā­ṭa­ḍvaḥ: (1) ca­traḥ: pūrṇā e­v­eti, ca­tu­r­ḍu­śa­na­ma­ny­ā­nāṃ mad­hy­e ca­tra­gra­ho 'n­o­ḍu­vi­kā iti tā­sāṃ saṃ­b­a­n­dh­i­no ḍva­d­śr­u­ti­r­va­k­y­ate, ś­i­ṣ­ṭā­nāṃ ca­t­vā­r­i­ṇa­t, tat­o­pi ḍva­da­śa 'o­u­ḍu­vi­ka­t­a­m­pa­d­y­ante iti vi­śo­ḍu­vi­citāḥ.

—A.B. en N.S. 28, 64-65.

Jātis : Parents of All Melodies

Jāti 299

96B. sañkare rūpabāhulyāt jātinirdeśa iṣyate

97A. tasmād yad gīyate kiñcit tat sarvaṁ jātiṣu sthitam1

In a mixed or composite form (saṅkara), the predominant structure should be considered as the basis of indicating the [parent] jāti. Thus all the prevalent forms of singing (yadgīyate kiñcit) have their source in jātis.

NOTE :

We have had a great deal to say regarding this matter in earlier chapters.

1 Datt. T.ed. reads : ‘सङ्करे रूपबाहुल्यम्’. We have adopted Abhinava's reading who quotes part of Datt. 96B with the words :

तदेवभूतं यन्मिश्रगोयं तत्र या जातिरभिमता यन्नियम्यते मतज्ञैर्बन्धनकल्पयपयाश्चिकादिभिस्तदृधीकृतैन रूपबाहुल्येन । तत् ‘सङ्करे रूपबाहुल्याज्जातिनिर्देश’ इति दत्तिलाचार्योप्याह ।

—A.B; on N.S. 28, 141.

B.B. (I) ch. VI quotes Datt. 96B-97A along with-Datt. 18A with the words : दत्तिलोप्याह. He reads the verse as ; तस्माज्जातिप्रधानन्तु गानकवन्धं व्यवस्थितम् । यत्किञ्चिद्गीयते लोके तत्सर्वं जातिषु स्थितम् ।

TOPIC 11

VARNA

97B. atha varṇāstu catvāro jñeyā anvarthasamjñayā

98A. sthāyisañjñācirṇau caiva tathārohiavarohiṇau

B. ekasvarā pade gītiḥ sthāyivarṇo' bhidhīyate

99A. sañcārī svarasañcārad yathārthau cottarāvapi1

Now, the varṇas : [they are] indeed four. [Their nature] may be known by their] names which are self-explanatory (anvarthasamjñā). They are : sthāyī, sañcārī, ārohī and avarohī.

While [singing] a pāda if the song (gītiḥ) stays on a single note, it is known as sthāyivarṇa, [and it] becomes sañcārī by movement over notes. The latter two (viz., ārohī and avarohī) are also explained by their names (yathārthau).

NOTE :

Varṇa was the general term used to indicate musical or melodic movement over notes. The Vṛtti on Bhaṭṭadeśī calls it ‘song itself’ (varṇābdena gānamabhidhīyate, Bṛ. Vṛtti on 119-120). It was a term which encompassed all possible melodic movements. These were classified into four broad categories : sthāyī, sañcārī, ārohī and avarohī.

The movement which rendered any single note as constant, usually by repetition, was termed sthāyī varṇa. Defining sthāyī, Bharata says : “sthirāḥ svarāḥ samā yatra sthāyī varṇaḥ sa ucyate : where the same notes are constant, that is called sthyāyī varṇa” (N.S. 28, 16). Abhinava comments that in rendering this varṇa the same note, whether in the tāra, madhya or mandra octave, should be repeated at short intervals. It should not be stretched over long periods without an accent or break, as happens in the long-drawn sound of a bell. He also suggests that a note when stressed by repetition did not lose its status of a sthāyī even if one or two other

1 S.S.S. 1, 42-43, bears a remarkable correspondence with Datt. 97B-99A :

आलप्तिसंझका वर्णाश्चत्वारोरोध्वनिसंझकाः। स्वायिसञ्चारिणी चैव तथारोहावरोहिणो ।।

एकस्वरपदे गीति: स्थायिवर्णोऽभिधीयते । सञ्चारी स्वरसंचाराद्(व?)यथाव्यतिकरावपि ।।

Pārśvadeva the author S.S.S. has indeed acknowlledged Dattila as one of the early authorities (especially on tāla) to whom he was indebted. In S.S.S. 9, 1-2 he states :

तथा सलक्षणं वद्ये पूर्वशास्त्रानुसारत: । श्रीसोमेश्वरवदत्प्रभृतिविमतिस्थलस्वरं पुरा ।।

प्रोक्त' सर्वंगडिताय.....

It is thus not surprising that in S.S.S. there are quite a few verses borrowed from the Dattilam (without acknowledgment).

notes, either higher or lower, were also rendered with it.1 The Vṛtti on Bṛhaddeśī expounds the sthāyī in similar terms (Bṛ. Vṛtti on 120A). An example given by the Vṛttikāra from the ṣāḍji-jāti shows that sthāyī could employ more than one svara : sa ri sa, sa ri sa, provided that the main sthāyī note created an impression of constancy (or, as the Vṛttikāra says, it remained ‘anupahata’).2

Śārṅgadeva says that it is the holding of a single note for prolonged periods that characterises the sthāyī varṇa. Kallinātha adds that notes separately dwelt over in the same movement also constitute sthāyī such as sa s1 sa, ri ri ri (see S.R. 1, 6, 2 and Kalā).

Ārohī was the varṇa predominated by a melodic movement ascending over notes from low to high. Avarohī was the converse of this. Sañcārī had both ascending and descending movements and was characterised by an effect of wandering over notes.

In current theory also all four varṇa terms are used, but with a widely different connotation. Ārohī and avarohī are now the terms denoting structural ascent and descent characteristic of individual rāgas. They indicate formal peculiarities rather than movements. Sthāyī (with its counterpart antarā) is a basic aspect of the plan of melodic elaboration as well as of model compositions (bandish), particularly in the khayāl style of Hindustani music. In drupada compositions, two further parts are conceived of in the structural plan of melodic development : sañcārī and ābhoga. This present-day sañcārī has little to do with the ancient term.

The ancient varṇas were conceived as inseparably connected with padas (words) in a song. While a jāti, as Abhinava suggests, could be purely a note structure (jātirnāma svarasamūhamātram, A.B. on N.S. 29, 14), without necessarily implying words (as in instrumental renderings), the notion of varṇa was so inherently connected with pada that it could not be defined without it (padasambandhādhīnāstu vastvarṇavyavahāraḥ; also : varṇo nāma padasambandhamantareṇa lakṣayitumeva na śakyate, A.B. on N.S. 29, 14). The melodic movement to which a single word was sung was the unit of a varṇa, so that every word in a song demarcated one varṇa (ekapadaniṣṭhatva, A.B., ibid.). This precise demarcation of varṇas by words was imperative to avoid certain confusions. For example, an ārohī movement followed by an avarohī could be confused as a sañcārī if the two movements were not clearly demarcated by two distinct padas (words). In fact, without the pada-demarcation, the total effect could almost always have been that of sañcārī.

Dattila’s definition of sthāyī evidently suggests that he, too, held that the demarcating factor as well as the basis of a varṇa unit was the pada or a single word

1 ‘sama’ iti tulya-jatīyāḥ | tatra madhyataramadhyamūpatatayā | tasyāṃ svarasya prayogaḥ | sthāyīye varṇe: vicchidya vicchidya punaḥ prayogaḥ kartavyaḥ | tu dīrghaṇasvaranibaddhita dṛśyati tu ‘varṇa’ iti bahuvacanaprayogaḥ | svarā upasvarāṃśceṣit sarvakośa iti kecit | एतत् न साधारण्युक्तसवरधयमप्यारोहणावरोहणहस्वित्व (तद्) व्यतिरेकेṇa kaṭhaṃ pratīocyam् ।

-A.B. on N S. 29, 16.

We must, however, remember that the example itself might not be quite correct. The use of two notes might be a whim of a non-musical scribe rather than a deliberation of the Vṛttikāra. The other example of sthāyī from the jāti madhyamā is seen to be a repitition of the same note: म धयमां तु मा मा मा इति ।

• Br. Vṛtti, ibid.

302 A Study of Dattilam

in a song. Bharata gives a similar definition of varṇa and its relation to pada :

evaṃ padaṃ lakṣaṇasamyuktam yadā varṇo'ṅkarṣati

tadā varṇasya niṣpattirjñeyā svarasamudbhavā

(N.S. 29, 18)

"In this way (i.e., through sthāyī, sañcāri etc.) when a pada, which has the right attributes (i.e., which is grammatically correct) is stretched in rendering a varṇa (i.e., becomes long-drawn in singing) then arises varṇa and it is known to consist of svaras."

Abhinava concludes : "thus when a single word is stretched by a varṇa, or in other words, becomes long-drawn in singing a melodic line we have one single varṇa : tena padamekamaṃ yadā yato varṇagītikriyānukarṣati dīrghakālam karoti, tadā tato varṇastasyaikasyā sañkirṇasya niṣpattiḥ" (A.B. on N.S. 29, 18). Commenting on the expression 'lakṣaṇasamyuktam' in Bharata's definition, Abhinava (besides explaining the epithet as standing for a grammatically correct word) remarks that "this epithet implies that even in long samāsas consisting of a conjunction of many padas, the rule is that each pada be individually considered a demarcating limit (for a varṇa):" viva-kṣitasamjñā nirdeśo 'lakṣaṇasamyukta' vacanena dīrghe'pi samāsapade bhāgaśaḥ padaniyamamāśrītyāvadhinīyama iti darśayati" (A.B., ibid.).

Pada was thus inherently connected with the concept of varṇa and played a significant structural role in gāndharva. The unit of an alaṅkāra, too, was pada.

Varṇas, then, were basic melodic movements analysed into four possible types : static, ascending, descending and wandering. Surely these movements were intimately linked with the melodic structure of jātiṣ. Why, then, was the topic varṇa not expounded as part of jāti by Bharata (and also Dattila) ? It is Abhinava who raises this question as a possible objection to Bharata's scheme : "nanu varṇāderjātiśarīrānupra-veśājjātilakṣaṇa eva nīrūpaṇaṃ yuktam dṛśyate", (A.B. ibid.). The objector could also argue that varṇa should have been propounded before taking up jātiṣ, because varṇa denoted the very basic movements through which jāti structures were formulated,1 (Mataṅga, indeed, describes varṇa and alaṅkāra before taking up jāti, Br̥. 119-173). One could perhaps assent to the exposition of varṇa as part of jāti (for as Abhinava says) there exist precedences for delineations of this nature but a description of varṇa after jāti would certainly strike one as illogical.

Abhinava replies to these possible objections by saying that jātiṣ were basically a group of notes and both in gāndharva and in gāna one did find musical structures being rendered without words (tarhi gīyamānapadābhyāve kaścidapi svataḥ svaropayogo'sti gāna gāndharve vā; A.B. ibid.). But varṇa was a concept which could not even be understood without referring to pada. This is why it was described separately after the svara structures of jātiṣ and their characteristics had been expounded.

1

लब्धात्मकलभय हि विनियोगो युक्तः। दशलक्षणादिषु वा पूर्वं विनियोगः वक्तव्यः।

—A.B. on N.S. 29, 14.

2

विनियुक्तस्य लक्षणमित्यपि दृश्यते । पृथक् तु तदमिधानयुक्तमेव ।

—A.B., ibid.

Varṇa was basically a concept pertaining to song. Bharata thus says that varṇas are born of śārīra-notes (sārīrasvarasambhūtāḥ, N.S. 29, 17). Varṇa could, of course, be also rendered on instruments, for all songs are capable of such renderings; but it was in the first place a vocal concept. Abhinava, commenting on Bharata's notion of varṇas as 'śārīrasvarasambhūtāh' says:"(varṇas) are basically dependent upon sung notes. But even on the vīṇā one does find melodies rendered as resembling songs and thus varṇas are present there too; it is not that alaṅkāras (which depend upon varṇa) are not played on vīṇās."1 Abhinava here quotes a kārikā to the effect that there were elements of music which were not very clearly revealed through the voice and others which were indistinct (calitā) on the vīṇā but determinable through singing.

śārīryāṃ tvasphuṭā ye tu dāravyāṃ te vyavasihitāḥ dāravyāṃ calitā ye tu śārīryāṃ te suniścalāḥ

(A.B. on N.S. 29, 17)

Śrutis, we have seen, were better demonstrable on the vīṇā; varṇa, on the other hand, could be clearly understood only in relation to song.

'śārīrasvarasamṛpta' iti । śārīrasvarेषु मुख्येऽपि श्रितेयु लक्षणमिदमुक्त्तम् । तत्र सदृशस्वरोपलम्भ एवाऽग्यन्न वीणादाविति तद्वापि पर्यवस्यतीति न तु वीणादावलंकाराभावः ।

—A.B. on N.S. 29, 17.

TOPIC 12

ALAṄKĀRA

99B. varṇarāyāstu vijñeyā alaṅkārāstrayodaśa

100A. nāmato rūpataścaiva saṅkśepeṇa bravīmi tān

B. prasannam pūrvamuccārya śanaiḥ sandīpayetsvaram

101A. prasannādirbhavedevam prasannānto vilomataḥ

B. evam prasannamadhyāśca prasannādyanta eva ca

102A. ete sthāyinyalaṅkārāścatvārāḥ parikīrtitāḥ1

B. kvacit svare ciram sthitvā sprṣṭvā tāram tato'gnivat

103A. pratyāgacchettu tattraiva bindureśo'bhidhīyate

B. syānnivṛttapravṛttābhyāstadvadvanmandram sprṣed yadi

104A. preṅkholitam dvayorvidyāt tulyakālam gatāgatam

B. krameṇa paramaṃ tāram gatvā mandram patet tataḥ

105A. tāramandrapraśaṃno'yamalaṅkāro'bhidhīyate

B. mandrādutpatitastāram krameṇaivāvarohati

106A. mandratāraprasanno'yam, sarvasamyaḥ samo bhavet

B. urahkaṇṭhaśirasthasya triśruteḥ kampanād bhavet

107A. kampitam haritamaṃ caiva recitaṃ ca yathākramam

B. eṣāṃ ca pañca bindvādyā nityaṃ sañcārisamśrayāḥ

108A. prasannādiḥ prasannānto (tathārohā) varohinau2

B. śeṣā apyathāyogaṃ sarvavarnaśrayāḥ smr̥tāḥ

The alaṅkāras should be known as based upon the varṇas; [they] are thirteen. I shall now, in brief, relate them by their names and their forms.

Having first sung a low note (prasanna), notes should be raised gradually; thus arises [the alaṅkāra called] prasannādi. By reversing [the process] arises [the alaṅkara called] prasannānta. Similarly arise [the alaṅkaras] prasannamadhyā and prasannādyanta. These four have been described as the alaṅkāras of the sthāyī [varna].

Hold any note for a long period; then touching a higher note (tāra) like lightening (agnivat) return to the same note : this [alaṅkāra] is called bindu. If likewise a lower note (mandra) is touched, then the [alaṅkāra] is called nivṛtta-pravṛtta.

1 S.S.S. 1, 44–46 is the same as Datt. 99B–102A. The only difference is that S.S.S. 44A has हलकूज्या instead of the text above in 99B. The Datt. passage is incorporated by Pārśvadeva without acknowledgment of the source.

2 There is a lacuna of four letters at the end of Datt. 108A. The T.ed. text reads : प्रसन्नान्तो++++वरोहिणी. Our suggested reconstruction is based upon a contextual analysis of the matter and upon similar statements in other texts.

Know preñkholita to be [the alaṅkāra in which] the to-and-fro movement (gatāgatam) between two [notes] is of even duration.

Successively rising to the extreme tāra, then, should one drop to the mandra, the alaṅkāra is called tāra-mandra-prasanna. Leaping from the mandra to the tāra [and then] descending successively, is [the alaṅkāra] mandra-tāra-prassanna.

If all [the movements] are uniform, then is [the alaṅkāra] sama.

By a quiver of three śrutis in the chest (i.e., mandra), throat (i.e., madhya) and head (i.e., tāra) arise [the alaṅkāras] kampita, harita and recita respectively.

Of these [alaṅkāras], the five beginning with bindu are invariably based on the sañcārī [varṇa].

Prasannādi and prasannānta are [the alaṅkāras] of the ārohī and the avarohī [respectively]. The rest [of the alaṅkāras] are known as being duly based upon all the varṇas.

NOTE:

Alaṅkāra is defined by the Vṛtti on Bṛhaddeśī as a decorative adornment (maṇḍana) which creates a pleasing effect in a song (Bṛ. Vṛtti on 120). The form of an alaṅkāra was described and classified on the basis of varṇas.

Dattila calls the alaṅkāras ‘varṇāśrayāḥ’, i.e., ‘based on’ or ‘dependent on’ varṇas. The four varṇas, we have seen, denoted all possible melodic movements; alaṅkāras were melodic flourishes which lent colour and charm to these melodic movements. The alaṅkāras themselves, naturally enough, consisted of melodic movement (though of a specifically subtle and sophisticated nature) and hence were in this sense ‘varṇāśryaḥ.’ Bharata, too, like Dattila, has described alaṅkāras as dependent upon varṇas:

ete varnāstu vijñeyāsucatvaro gītayojakāḥ

etān samāśritān samyagalankārān nibodhata

(N.S. 29, 19)

The Vṛtti on Brhaddesī quotes an ancient kārikā which likewise says: “amī varṇāstu vijñeyā alaṅkārādisiddhaye” (Bṛ. Vṛtti on 121A).

A pada was the unit of a single varṇa. It was also the unit of an alaṅkāra.

In Sanskrit poetics, alaṅkāras such as the upamā (simile), anuprāsa (alliteration) etc. have been described through the analogy of ornaments like kaṭaka (a bracelet) keyūra (a bracelet of the upper arm) which men and women wore in order to beautify their person.1 A similar analogy has been used by the Vṛttikāra in describing melodic alaṅkāras: “tatrālaṅkāraśabdena kim ucyate ? alaṅkāraśabdena maṇḍanamucyate. yathā kaṭakakeyūrādinālankāreṇa nārī puruṣo va maṇḍitaḥ śobhamāvahet, tathā etairalaṅkāraiḥ prasannādibhiralaṅkṛtā varṇāśrayā gītirgātrṣrotrṇām sukhāvahā bhavatīti” (Bṛ. Vṛtti on 121A).

306

A Study of Dattilam

A word about the terms ‘prasanna’ and ‘dīpta’ in this context. The word prasanna in musical phraseology was synonymous with mandra or low (mandra was, in later texts, shown with a dot above a written note, e.g. śa), and dīpta with tāra or high (tāra was indicated with a vertical line above a written note e.g., śa), ; we find that Dattila uses the term sandīpana to convey a rise in pitch.

The term prasanna, mandra etc. were, in this context, often used in a broad relative sense: prasanna was a low note in relation to the dīpta or high note, whatever the octave in which they occurred. Any note which had a lower position was prasanna while a higher note was tāra (pūrvasthānāstho svaro mandraḥ parasthā-nasthāra iti, Sudhā on S.R. 1,6, 6-7). Abhinava also makes the same suggestion in saying: “dīpanam tāratā prasādo mandratā dīptirnāma śrāvakaśrotṛniṣṭho viśeṣo dharma ityanuktasamam svarasādhāraṇam” (A.B. on N.S. 29, 33-37).

The Sthāyī Alaṅkāras

Dattila enumerates and expounds four alaṅkāras as pertaining to the sthāyī: (1) prasannādi, (2) prasannānta, (3) prasannamadhya and (4) prasannādyanta. Dattila describes them in general terms as : (1) prasannādi is the alaṅkāra with a low note in the beginning and a gradually rising movement; (2) prasannānta has a low note at the end; (3) prasannamadhya is that where a low note occurs in the middle, and (4) prasan-nādyanta is one where a low note occurs both at the beginning and at the end. Bharata gives a similar description (N.S. 29, 33-35).

Regarding their actual structures, we find two divergent descriptions. The Vṛtti on Brhaddeśī illustrates these alaṅkāras as :

1. prasannādi: sa ri ga ma pa dha ni śa

2. prasannānta: śa ni dha pa ma ga ri sa

3. prasannādyanta: sa ri ga ma pa dha ni śa, śa ni dha pa ma ga ri sa

4. prasannamadhya: śa ni dha pa ma ga ri sa, sa ri ga ma pa dha ni śa

(Br. Vṛtti on 120)

Saṅgītaratnākara gives quite a different structure:

1. prasannādi: śa śa śa

2. prasannānta: śa śa śa

3. prasannādyanta: śa śa śa

4. prasannamadhya: śa śa śa

(the dot on top stands for a mandra note and the vertical line for a tāra note)

(S.R. 1,6, 3-48 and Sudbā)

1 प्रसन्नत्वं मन्द्रभाव: तत्र विन्दुलोकेनुबिभ्ज्ञानम् । तारत्वे रेखोद्वयं शिरोगता । सा (?सं) साधिति ।

—A.B. on N.S. 29, 20.

The printed A.B. text has the dot below the written note. Abhinava has not specified its exact location ; but Śārṅgadeva clearly says : मन्द्री विन्दुजिह्वा भवेत् (S.R. 1, 6, 8). Regarding the term dipta, Abhinava says : दोप्तं तारता प्रसादो मन्द्रता (A. B. on N.S. 29, 33-37). Śārṅgadeva also says: तारस्तु दीप्ति: (S.R. 1, 6, 8).

On the face of it, it would appear that the Saṅgītaratnākara is more dependable, as the sthāyī varṇa is said to remain on the same note. But the Vṛttikāra's description is earlier and also tallies with those of Dattila and Bharata who have described the first alaṅkāra as consisting of a gradual rising and the second of a gradual lowering. Abhinava explains the seeming discrepancy between the alaṅkāras and the varṇa to which they belong by observing that in the alaṅkāras of this varṇa, the insertion of different notes did not destroy the sthāyī effect (svarāntaropanipātikastailabinduḥ; A.B. on N.S. 29, 30). These alaṅkāras were graces that adorned a varṇa and the manner in which they were rendered must have played a large part in their structure and in judging their varṇa. Thus we find that the four sthāyī alaṅkāras could (as we will see) also become alaṅkāras of the ārohī or the avarohī varṇa: the manner of their rendering in such cases must have undergone subtle but sure changes.

The four sthāyī-varṇa alaṅkāras are followed by (1) bindu and (2) nivṛttapravṛtta. Bindu according to the Vṛtti on Brhaddeśī, was rendered by holding a note and then suddenly, like lightning, touching the same note in the higher octave and returning to the original place. Nivṛttapravṛtta was the converse of this and was rendered by speedily touching the identical note in the lower octave.1

Abhinava has a different description. He states that just any another higher note had to be touched in rendering bindu and not necessarily the octave of the basic note (cirakālamavasthāne svarāntaroparañjakastailabinduḥ prasṛtaḥ; A.B. on N.S. 29, 21-22). In sannivṛttapravṛtta, says Abhinava, a similar movement was made, but here it was a descending movement (avaruhya yadārohaṇaṃ tadā sannivṛttapravṛttah; A.B. ibid.). In describing bindu, Abhinava gives the analogy of a 'spreading drop of oil'; the idea is not very clear. Not knowing the exact musical nature of the alaṅkāra, we cannot grasp the purport behind the analogy used by Abhinava. Alaṅkāras like bindu and sannivṛttapravṛtta were, says Abhinava, of a subtle nature and could be understood only as part of the song they embellished. They could never be properly described—their description was only a pointer which could perhaps help one recognise them when heard in a musical context.2

Dattila says that in these two alaṅkāras, the higher (or the lower note) should be touched like lightning. Bharata, apparently, gives the exact duration of the alaṅkāra in stating “bindurekākalo jñeyaḥ” (N.S. 29, 38).

Śārṅgadeva, in describing bindu, borrows Dattila's expressive phrase 'sprstvā agnivat', but the alaṅkāra itself by his time seems to have undergone modifications

1 चिरमेकस्मिन् स्वरे वडजादिरूपे स्थित्वा तदियातरमिवनतु (पु? स्प) बस्त्रा कलामे (स्य? ) का च सिथला यद् पुनरपि समा सा मन्त्रध्वा गम्यते स बिन्दुः एककालान्तरं स्पष्टीभाव बिन्दोरेक विपर्योच्चारणात्प्रवृत्तो विनिवर्ततेवाद् निवृत्ताख्यः यथा सो सा सो निवृत्तप्रवृत्तः —Vṛtti on Br. 120.

2 यथापि चैव तैलद्रौत्‌रः यत् विशिष्टस्तवैन भान्ति तथापि प्रतीतस्य गीतसमयप्रविष्टस्याभिज्ञानादश् लिख्यते: —A.B. on N.S. 29, 21-22.

and had become more like modern ‘palṭās’.1 A similar transformation had occurred in sannivṛttapravṛtta.2 After bindu and sannivṛttapravṛtta, Dattila describes preñkholita.

The alaṅkāra preñkholita has been aptly described by Abhinava as “preñ-khito dolābandhavat””. (A.B. on N.S. 29, 21-22). Its form resembled the movement of a swing. It apparently consisted of a swinging movement between two notes – what we would call an āndolana. In fact, Abhinava, while explaining Bharata’s description of preñkholita (Bharata describes the alaṅkāra as: “gatagatapravṛtto yaḥ sa preñkholita iṣyate”, N.S. 29, 38), uses the very term āndolana to bring out its characteristic feature: “gatāgatapravṛtto yaḥ sa preñkholita iti bindukampitāndolanādityarthāḥ”

(A.B. on N.S. 29, 38). The nature of this āndolana evidently depended upon two other ancient alaṅkāras, the bindu and the kampita. Śārṅgadeva describes this alaṅkāra also as a kind of palṭā with the notation: sa ri ri sa, ri ga ga ri, ga ma ma ga, ma pa pa ma, pa dha dha pa,… (S.R. 1, 6, 41-42).

Preñkholita is followed by tāra-mandra-prasanna.

The expression “extreme tāra” (paramaṃ tāram) used in expounding this alaṅkāra is cryptic. Bharata is more definite when he says: “one should ascend gradually to the fifth or the sixth (note) in this alaṅkāra” (N.S. 29, 40). The Saṅgītarat-nākara describes this alaṅkāra as a gradual rise from a low note to its octave and then a sudden drop to the initial note (śa ri ga ma pa dha ni śa śa) (S.R. 1, 6, 57-58).

Dattila’s description seems to correspond with that of the Saṅgītaratnākara

Mandra-tāra-prasanna—the alaṅkāra next described—was very much like tāra-mandra-prasanna. In it, the first movement was a sudden leap from a lower to a higher note and then a gradual descent back to the original low note. Bharata’s description of it corresponds with that of Dattila (N.S. 29, 41). Śārṅgadeva describes the alaṅkāra on the same lines as Dāttila and notates it as : śa śa ni dha pa ma ga ri śa (S.R. 1, 6, 57, 58).

According to the Vṛtti on Brhaddeśī, however, both these above alaṅkāras had two alternate renderings: one could either ascend to the fifth or sixth note and then descend to the basic note or, alternatively, one could go upto the higher octave of the initial note (which in Bṛ. is said to be the aṃśa) and then suddenly descend (Vṛtti on Bṛ. 120).

The alaṅkāra sama probably consisted of an ascending movement which complimented a similar and equal descending movement without any sudden leaps or falls. Bharata like Dattila describes it as: “sarvasāmnyāt samo jñeyaḥ” (N.S. 29, 35).

The Vṛtti on Brhaddeśī describes sama somewhat differently. It says that in this alaṅ-kāra the same range of the seven svaras is sounded in all three octaves: “sthānatraye’-pi sadrśadhvanih saptasvaroccāraṇaḥ samaḥ” (Bṛ. Vṛtti on 120). The idea is not

1 अथ बिन्दुः सप्नूतादिः स्वरमं कृत्वाड्मनवपरं स्पृष्ट्वाडऽद्र्ध; स्पर्शोऽनाखिलः कलाः; सडेरिस रिङ्गारिर गेडमग मऱपम परेघप घरेनिध

—S.R. 1, 6, 37-38.

2 सप्तमगिरि रिखपरमग वानिधपम ..

—S.R. 1, 6, 48.

clear and the accompanying notations are obviously corrupt. Śārṅgadeva interprets sama as a kind of palṭa.1

The three alaṅkāras, kampita, harita and recita, were obviously similar except for a difference in the octave (sthāna) where they were rendered. The expres-

sion “a quiver of three śrutis” used by Dattila in describing them is far from clear. Bharata, describing kampita, says that it consisted of “a quiver lasting three kalās”

(kampitaśca kalātrayam, N.S. 29, 43). Abhinava speaks with disapproval concerning the interpretation of kalā as śruti by some authorities and says that kalā here meant

a unit of time (kalātra na śrutih api tu kāla-kalāḥ, A.B. on N.S, 29, 43). Elsewhere (A.B. on N S. 29, 21), Abhinava says that kampita, harita and recita resulted from a

certain subtle quivering of the three-śruti note (triśruteḥ svarasya) and adds that though these three alaṅkāras could be recognised in a melody, yet they could not at

all be indicated in words and were thus more subtle than bindu and preṅkholita etc. which could at least be indicated.2 Abhinava (and also Dattila) seems to imply that

these alaṅkāras were used only on the notes ri and dha (and pa in madhyamagrāma), the three-śruti notes.

Dattila enumerates thirteen alaṅkāras in all. Five, he says, belong to the sañcāri varṇa, one each to the ārohī and avarohī varṇas and four to the sthāyī varṇa. Other

alaṅkāras belong to varṇas appropriate to them (Datt. 108). It should be noted that the alaṅkāras prasannādi and prasannānta are in his description common to two varṇas.

The former is both a sthāyī and an ārohī alaṅkāra and the latter is both a sthāyī and an avarohī alaṅkāra. In the classification of Bharata and others also, we come across

alaṅkāras common to two or more varṇas.

While Dattila enumerates only thirteen alaṅkāras, Bharata and Mataṅga have accepted thirty-three alaṅkāras.

Mataṅga and the Vṛttikāra say that there are thirty-three well-known alaṅkāras. They obviously had the Nāṭyaśāstra in mind when naming this figure, for the

Brhaddeśī puts great reliance upon Bharata in describing the alaṅkāras.3 Abhinava refers to the ‘Ṭīkāāra’ (or Ṭīkākārās, for he says ‘ṭīkākrdbhih’) in this context who had

spoken of thirty-three alaṅkāras on the basis of works by Sadāśiva and others: “ṭīkākṛdbhistu sadāśivamataṃdigranthāntaralikhitam ‘trayastriṃśadimeproktā alaṅkārāḥ’

(A.B. on N.S. 29, 43).

1 सरिगममरिस रिगमपपमगतिरि गमपधपमगतिनिरनिदमम ।

—S.R. 1, 6, 40.

2 ‘द्रुतिकम्पितकुरुरास्तु’ शिरोऽवष्टभ्य कण्ठनिनिघृष्टस्य त्रिश्रुतेः स्वरस्य कम्परूपा इत्यभियार्तिनामपि दर्शयितुमशक्या इति न विलिखिता इति ।

—A.B. on N.S. 29, 21-22.

3 Mataṅga introduces alaṅkāras with the words : यस्मिन् वर्णा स्थिता ये च अलङ्काराः मनोहराः । तानिदानीनि प्रवध्यामि भरतोक्तविधानतः ॥

—Br. 125

and ends the subject with: अलङ्कारास्वसिद्धदेवमते मयोदिता:

—Br. 170.

The Vṛttikāra says: इदानीन्तुरसिद्धदास्ववस्तिसंभदलङ्कारापादेन (?) ततः प्रयोगतश्च कथमन्ते

—Vṛtti on Br. 120.

Datilla, we have said, has listed prasannādi as a sthāyī alaṅkāra as well as an ārohī alaṅkāra (although there is a lacuna here in the text, the purport appears to be clear). Bharata includes this alaṅkāra in the sañcārī list also (N.S. 29, 21-22). The inclusion of this alaṅkāra among ārohī alaṅkāras seems logical because the alaṅkāra was formulated through a movement of ascent, but the reason for its inclusion among sañcārī alaṅkāras is not clear. Again, Dattila lists the alaṅkāra prasannānta as sthāyī as well as avarohī; its avarohī form is clear, for it consisted of a movement of descent. Bharata has termed this alaṅkāra an ārohī alaṅkāra (N.S. 29, 24) but we do not know why. Preṅkholita, the alaṅkāra which involved a constant swing-like āndolana between two notes, has been aptly listed as a sañcārī alaṅkāra by Dattila, whose dictum "nityam sañcārī-samśrayalḥ" (Datt. 108) obviously implies that this alaṅkāra along with the other four of its group-bindu, nivṛtta-pravṛtta, tāra-mandra-prasanna and mandra-tāra-prasanna—belonged to sañcārī alone. Bharata lists preṅkholita as an ārohī-alaṅkāra (N.S. 29, 23-24); the basis of his classification is not clear.

Mataṅga also gives a varṇa-wise classification of alaṅkāras, avowedly on the basis of Bharata (bharatoktavidhānataḥ, Br. 125). His list, however, does not entirely agree with Bharata's list as accepted by Abhinava.

The Vāyupurāṇa also describes thirty three alaṅkāras (chapter 87). On the nature of the alaṅkāras, the Purāṇa says: "just as a woman is adorned by various ornaments, so is a varṇa adorned by alaṅkāras arising from the varṇas themselves" (Vāyu, 87, 25). Later theorists like Śārṅgadeva enumerate as many as seventy alaṅkāras (S.R. 1-6). One reason why the number of alaṅkāras described by Dattila is so small, might be his attempt at brevity. It is likely that he considered only thirteen alaṅkāras to be the important ones and others as included within these. Another cogent reason, as we have argued earlier, could be that Dattila represented an earlier, somewhat less florid tradition of gāndharva where fewer alaṅkāras were used.

Of the thirty three possible alaṅkāras, Bharata considered only about seventeen as valid in dhruvās (N.S. 29, 30-32). His list of alaṅkāras suitable for dhruvās contains some names not found in the saptarūpa list, such as ūrmi and apāṅga.

Conclusion of Svarā Topics

109A. iti ślokāśatenedamuktam svaragatam sphuṭam

Thus, in a hundered verses, the topics relating to svara (svaragatam) have been clearly expounded.

NOTE :

Dattila began the exposition of svara topics after enumerating svaroddeśa with the 8th verse and concludes the section with the 108th verse. He thus takes not hundred but (to be exact) a hundred and one verses for describing the subject.

Section II

ON TĀLA

A SHORT PREAMBLE

109B. atha tālaṃ pravakṣyāmi (yathāśāstram pramāṇataḥ)

110A. tālāt sāmyam bhavet sāmyādiha siddhiḥ paratra ca1

Now I shall expound tāla (based upon authentic accounts in the śāstras). Through tāla is achieved sāmya (equipoise), [and] sāmya is the source of fulfilment here (in this world) and hereafter.

NOTE :

This verse forms a short introductory preamble concerning the nature of tāla. Tāla (from √ tāl, ‘to establish’)2 has been defined by Abhinava as the action which reveals definite demarcations in musical forms on the basis of time units (cf. tālaḥ paricchityātmakakalakhaṇḍe kriyārūpo dravyātmā sa eva gītakriyā pramāṇaparicchedopāyaḥ, A.B. on N.S. 31, 1). Tāla is a large term, for it denotes not only rhythm, tempo, measure of time, etc. but also the process and actions through which these are achieved and demarcated.

In modern music, drums are utilised to keep tāla ; distinctive beats on drums arranged in specific rhythms and groupings effect different tālas. In ancient gāndharva music, the instrument utilised for keeping tāla, as Bharata indicates, was basically ghana (N.S. 31, 1) ; sound was produced by striking on metal (for details, see ch. V).

The function of tāla, according to Dattila, was, as we have seen, to measure svara. When the measure (tāla) was in perfect accordance with the measured (svara) the result was sāmya, a pivotal term which, though not found in Bharata. has been often used by Abhinava in the context of tāla and its function in gāndharva. The notion of sāmya occurred also in the work on gāndharva by Viśākila—as a remark by Abhinava suggests.3 In verse 110A Dattila explicitly hints at the adrṣṭa-end of gāndharva in saying that “sāmyādiha siddhiḥ paratra ca” (see ch. V for a discussion of sāmya).

1 T. ed. 109B reads प्रवद्यामि+++++++++

2 This is an oft-repeated etymology in saṅgīta texts. Abhinava says : तल्लु प्रतिष्ठाकरणं इति तालः’ —A.B. on N.S. 28, 11-12. Sārṅgadeva says : तालस्थतमप्रविष्ठायामिति धातोर्धृजिस्मृतः। गीतं वाद्यं तथा नृतं यतस्ताले प्रतिष्ठिततम् ॥ —S.R. 5, 2.

3 विशिष्टाचार्याः: साम्यादिहु सिद्धिं परलेति वदन् प्रादीदूषुः। —A.B. on N.S. 33, 1. Mark that Viśākḥila’s reported words are very close to those of Dattila.

THE TĀLODDEŚA

110B. tatra jñeyāḥ kalāḥ pātāḥ pādabhāgāstathāiva ca

111A. mātrā ca parivartaśca vastu caiva viśeṣataḥ

B. vidyāryaṅgalayānvitāḥ pāṇiryatịḥ praka (raṇam) tathā

112A. + + + + +vajrā ? ca gītirmārgāśca śāstrataḥ

B. ityuddeśaḥ padārthānāṃ jñeyastālagatō budhaiḥ

113A. saṅkṣepenātha vakṣyāmi sūkṣamasameṣāṃ vinirnayaḥ

[In tāla] one should know specifically the kalās, the pātās, the pādabhāgas, the mātrā, the parivarta, and also the vastu, the vidārī, the aṅga, the [different] layas, the pāṇi, the yati and the prakaraṇa, the avayava, (?) the gīti and also the mārga. These, according to the śāstra, are the topics pertaining to tāla the wise should endeavour to understand.

1 shall now, in brief, decree the precise nature of these.

NOTE :

Dattila has listed fifteen topics under tāla, whereas Bharata has twenty-one. Thus Bharata does not list kalā, but enumerates the various kalās as separate topics. The discrepancy between Bharata's and Dattila's tālodeśas, though of import as indicating theoretical differences, is not as basic as it might appear (see ch. I).

Dattila expounds the tāla-topics in the same methodical way as svara-topics, and his exposition strictly follows the listed order.

There exists a lacuna in the Dattilam text where the topics are enumerated. Line 111B reads 'prakaka + + + +'. The intended word is obviously 'prakaraṇam', which indeed is expounded as the 12th topic following 'yati'.2

Again in line 112A, five letters are missing, after which the word 'vajrā' occurs. 'Vajrā' is a term that does not fit into the picture.3 It is not included by Bharata

1 Kṣirasvāmin in the nātyavarga of his țikā, Amarakośodghatāṇa, on the Amarakośa quotes Datt. 110B-111A with the words यद् दत्तिलः. The reading is the same as above.

2 Nijenhuis thinks that prakaka...is a wrong reading for the term prastāraka, which has been used by Nārada in the Saṅgīta Makaranda in connection with tāla (see Dattilam : A compendium of Ancient Indian Music, p. 320). Nijenhuis's conjecture is evidently misconstrued. Firstly, the Saṅgīta Makaranda is not an early work, and to ascribe a concept to the ancient Nārada on its basis is not satisfactory. Secondly, prakaraṇa occurs as a topic in Bharata, too, who, we have seen, has the entire subject-matter of gāndharva in parallel with Dattila. Prakaraṇa was indeed the most important topic in gāndharva-tāla, for it denoted the seven main tāla structures used in rendering gāndharva. Thirdly, as we have remarked, prakaraṇa is indeed the name of the 12th topic in the lakṣaṇa portion of the Dattilam itself where the topics are expounded. After having dealt with yati, the 11th topic in the uddeśa, Dattila in his lakṣaṇa introduces prakaraṇa with :

अथ प्रकरणं नाम पदकाविधीयते

—Datt. 155B.

3 Nijenhuis has, however, accepted the term 'vajrā' at its face value (see Dattilam : A Compendium of Ancient Indian Music, p. 37).

among the topics in tāla, and is, in fact, a minor variety of a specific tāla structure used with others in a more complex tāla structure called ovenaka. Its enumeration among the general topics to be discussed under tāla is clearly due to a scribal error. The word that should have occurred is ‘avayava’, a topic found in Bharata but erased in the manuscript from the Dattilam list. While taking up the topics enumerated step by step in a logical order, Dattila does not deal with vajrā though avayava does apparently occur, and it immediately follows prakarana as in the listed order.

The word ‘avayava’ in the Dattilam text was either qualified by an epithet or linked with the rest of the sentence by a preposition, as the number of missing letters indicates. These elude reconstruction. As in the section on svara, so in this section, Dattila promises to give a brief viniṛnaya—authoritative exposition—of the topics after enumerating them. In this case, he adds that his exposition will be ‘sūkṣma’, i.e., ‘subtle’, ‘intricate’, ‘all-pervading’—besides being brief. Dattila gives greater space to tāla and indeed his exposition concerning this aspect of gāndharva seems to have commended greater attention from subsequent theorists than his exposition of svara.

TOPIC 1

KALĀ

113B. tatrāvāpo’thā niṣkrāmo (vikṣepaśca praveśanam)

114A. (śa) myā tālaśca vil(iñe)yāḥ sannipātaśca samyamaḥ

B. kaiścinniṣeṣakāle(na)1 kalā saptavidhoditā

115A. layasthitiyā ca mārgeṣu kalpitam taiḥ kalāntaram

B. sarvā ceyam kalā yeṣāṃ ghaṇṭā nādavatī (sthitā)2

116A. (te)ṣām (citrādi)3 mārgeṣu viśeṣo’yam kalāśrayah

B. dvimātrā syāt kalā citre caturmātrā tu vārtike

117A. aṣṭamātrā tu vidvadbhirdakṣiṇe samudāhrtā

B. mātrā tu laukikī neha kintu paiñcaguṇā tataḥ

118A. tayākāryam kalāmānam prayogo’tha niyamyate4

Kalā should be known as sevenfold : āvāpa, niṣkrāma, vikṣepa and praveśa, śamyā, tāla and the seventh sannipāta.

Some believe that [these] arise through [the unit of time called] nimeṣa.

They believe that the state of the tempo (laya) differentiates kalās in the [three] mārgas.

[But] those who hold that [the nature] of these kalās is like a resounding bell (ghaṇṭā nādavatī) make following distinctions in the kalās depending upon the mārgas : in citra (mārga) the kalā is of two mātrās, in vārtika of four, and in dakṣiṇa the wise describe it as of eight mātrās.

[The term] mātrā in this context is not that of common usage, but is five times that [duration]. By this [mātrā] should kalā be measured. Now the prescribed application.

NOTE :

Kalā in Dattila is a complex term.

Dattila defines kalā first as a group of actions like āvāpa, niṣkrāma, etc. and then also as a unit of time—a unit which, according to some theorists, was based on nimeṣa. Kalā was thus a unified notion both indicative of the actions which manifested or formed the time measure in tāla and the units forming the time measure itself. These two aspects of forming tāla are in fact inseparable.

Abhinava argues that since the units of time in tāla are unthinkable without the

1 T.ed. reads nimeṣakāle+

2 T.ed. reads घण्टानादवता++

3 T.ed. reads तेषां म++

S. S. S. 6B-7A is same as Datt. 113B-114A, except that S. S. S. reads विशेषेऽपि. S.S.S. 7, 14B-15A is same as Dattila 116B-117A. The debt to Dattila is not acknowledged.

actions that demarcate them, the one notion is inherently connected with the other (cf. na kālaḥ kriyāvyatirekaḥ, api tu sarveṣāṃ paricchedahetuḥ kāla ityucyate. parichedaśca kriyaiiva, A.B. on N.S. 31, 1).

Two differing opinions are mentioned by Dattila regarding kalā as time-unit. The exact difference between the two, however, is not very clear.

The first opinion, according to Dattila, apparently held that kalā was constituted of nimeṣas and was, in various mārgas, distinguished by the state of tempo. Dattila expresses the second opinion through an analogy: that of a resounding bell. Its bearing on the matter remains obscure. However, theorists holding this view did measure kalā in terms of mātrās—on this point the meaning is clear. The number of mātrās in a kalā, according to them, differed in various mārgas.

In the last analysis, however, mātrā too was measured in terms of nimeṣas (N.S. 31, 4) and there seems to emerge no basic difference between the two differing opinions mentioned by Dattila.

Kalā was measured in nimeṣas. Nimeṣa was, according to one Indian tradition, the smallest unit of time. It was the time taken to utter a short Sanskrit vowel (such as a, o, i, u,) which metrically consisted of one mātrā.1 Mātrā in ordinary or customary measure was thus equal to one nimeṣa.

In gāndharva this common (laukika) measure was somewhat modified or rather multiplied. Bharata says that the standards of measure current in ordinary practice—such as kalā, kāṣṭhā, nimeṣa—did not apply to tāla in this context, because the mātrā here was equal to five nimeṣas (N.S. 31, 2-4). Dattila expresses the same idea in saying that “mātrā in this context is not that of common usage, but is five-times that” (Datt. 117B).

Mātrā was ordinarily measured as the time taken to utter one short or laghu vowel. But in this context, as Abhinava explains, “the standard time through which one unit of measure in singing is to be demarcated from another is equal to five nimeṣas, that is, five short aksaras : nimeṣāḥ pañca iti gītakālasya paricchedakālānta-ramaṃ pramāṇam; pañca nimeṣāḥ hrasvākṣarapañcakam yāvat” (A.B. on N.S. 31, 4).

1 Thus the encyclopaedic Viṣnudharmottarapurāṇa, in enumerating standard units of time and their measure, records :

लघ्वक्षरमात्रा निमेषः परिकीर्तितः । ब्रह्मन्; सूक्ष्मतरः कालो नोपलभ्यो भूतलम् ॥

—Viṣnudharmottara, 1, 73, 1.

According to this tradition, 2 nimeṣas constituted a truṭi, 10 truṭis equalled a prāṇa, 6 prāṇas equalled 1 vināḍikā, 60 vināḍikās made 1 nāḍikā and 60 nāḍikās made a full 24 hour cycle of day and night (Viṣnudharmottara, 1, 73, 1-3). Calculation shows that nimeṣa in this measure was of 0.2 seconds.

Kauṭilya (in the second chapter of the Arthaśāstra) records another tradition with a somewhat different standard for measuring time. According to this tradition, truṭi was the smallest unit of time : 2 truṭis equalled 1 lava, 2 lavas made 1 nimeṣa, 5 nimeṣas made a kāṣṭhā, 30 kāṣṭhās made a kalā, 40 kalās made a nāḷikā, 2 nāḷikās made a muḥūrta, 15 muḥūrtas made a day or a night, i.e., 12 hours. However, the value of the nimeṣa was approximately the same according to this tradition, too ; for calculation shows nimeṣa as equal to 0.24 seconds. Bharata seems to have had this Kuṭilyan measure in mind, for he says : या लौकिको काला काष्ठा निमेषच्च स्मृता बुद्धैः (N. S. 31, 2). Kalā and kāṣṭhā as units of time are not found in the Viṣnudharmottara but they occur in the Arthaśāstra.

318 A Study of Dattilam

This five-nimeṣa mātrā was the standard for measuring kalā as a time-unit in gāndharva. Kalā is said to be not a uniform unit but differing in duration according to the mārgas. It is stated that in citra mārga the kalā was of two mātrās, in the vārtika of four and in the dakṣiṇa of eight mātrās.

The term mārga literally means ‘way’. In music, especially in gāndharva, it signified broadly a mode of rendering tāla. The chief feature of mārga, as such, was laya or tempo. Citra mārga was distinguished by a fast tempo (druta laya) and in it a kalā or each unit of time, says Dattila, was constituted of two mātrās. Vārtika mārga, on the other hand, was distinguished by a medium tempo (madhya laya) and in it a kalā contained four mātrās and was twice in measure compared to citra. Dakṣiṇa mārga was rendered in a slow tempo (vilambita laya) and had a kalā of eight mātrās.

This notion of the kalā measure as increasing and decreasing with mārga was, as we shall see, to be understood in a rather special sense. Ordinarily, as a standard, kalā in gāndharva was understood as an inflexible two-mātric or ten-nimeṣa time-unit.

The Actions Which Demarcated Kalā

118 B. āvāpasaujñakam jñeyamuttānāngulikūkīcanam

119 A. adhastalena hastena niṣkrāmākhyam prasāraṇam

B. tasyā dakṣiṇataḥ kṣepo vikṣepākhyaḥ paribhāṣitaḥ

120 A. atha cākuñcanaṃ jñeyaṃ praveśākhyamadhastālam

B. śamyā dakṣiṇapātastu tālo vāmastu kīrtitaḥ

121 A. ubhayohastayoh pātah sannipāta iti smrtaḥ1

The [kalā] called āvāpa should be known [through] folding the fingers with [palm] upwards. That called niṣkrāma is [formed] with palm downwards and [fingers] extended. Casting of the [hand] to the right is defined as vikṣepa. And folding it with palm downwards should be known as [the kalā] called praveśa. The beat formed with the right hand is śamyā [and] that with the left is known as tāla. The beat [formed with] both hands is known as sannipāta.

NOTE:

After describing the measure of kalā as a time unit, Dattila now describes its other aspect: kalā as a set of actions that demarcated the time units. The first four actions forming kalā in this sense—āvāpa, niṣkrāma, vikṣepa and praveśa—were formulated as unsounded beats, whereas the last three—śamyā, tāla and sannipāta—were sounded. In tāla, all beats are, as a rule, formulated by an action usually of the hand.

1 S.S.S 7, 7B-10A is same as Datt. 118B-121A. The passage is incorporated without acknowledging Dattila as the source. The passage occurs in conjunction with Datt. 113B-114A which forms S.S.S. 6B-7A as already noted. In the S.S.S. version, Datt. 119A reads अतस्तलेन; 119B reads परिभाष्यते and 120A reads प्रवेशाख्यमवस्तलम्

hands and palms, but in case of gāndharva, certain characteristic gestural patterns had been prescribed with these movements, thus giving them a ritualistic significance (see ch. V). It was believed that only by keeping strictly to the prescribed movements, could transcendental merit be acquired.

Bharata has recounted an additional sounded beat called dhruva which Dattila does not include in his list.1 For our discussion on this point, see ch. I.

शाम्या तालो ध्रुवश्चैव सन्निपातस्तथा परः । इति शाब्देन संयुक्तो विनेयोऽपि चतुर्विंशः ॥

—N.S. 31, 31.

TOPIC 2

PĀTA

121 B. evamete trayaḥ pātāḥ kalā + + + + + + +

122 A. (eteṣāṃ śa)ptavidhye pi trikalāyāṃ prayojayet

Among the kalās, these [last] three are known as the pātas. Though they are seven, they should be used in the three [modes] of kalā.

NOTE :

These seven kalās stood for the seven beats in gāndharva tāla. Among them only three were sounded or struck beats. Pāta literally means a struck beat. Therefore, strictly speaking, only the three sounded beats among the kālas were pātas. Kallinātha says : “sounded beats are suggested by two terms, pāta and kalā, but beats without sound are signified by only one term : ‘kalā’.” Vemabhūpāla makes a similar remark in Saṅgīta Cintāmani. Dattila too has evidently used the word kalā to denote the sounded ones alone. Pāta was in effect a part of the larger concept of kalā. Bharata does not include pāta as a topic in his list, though he does distinguish the sounded beats as pātas.

A further clarification of the concept of kalā may be worthwhile before we proceed to describe the three modes of kalā mentioned here by Dattila. In the Dattilam (and also the Nātyaśāstra), the term kalā has been used with somewhat varying connotations (besides its dual meaning in Dattila, already indicated), and unless the different senses or aspects of the concept are kept distinct, confusion might result. Basically, kalā was a unit of time measured on the basis of two mātrās or ten nimeṣas. Kalā also denoted the beats, sounded or unsounded, that demarcated time. But, as we have seen, the measure of kalā was not uniform, but changed according to mārga. Dattila observes : “kalās in different mārgas have different number of mātrās; in citra two, in vārtika four, and in dakṣiṇa mārga eight.” This difference in the measure of kalā itself should be understood in terms of what might be called the different modes of kalā-application. The mode of kalā-application in the citra mārga was ekakala; it was consi-

1 T. ed. reads : + + + + + + स्तविच्छेदे पि. Our suggested reconstruction of the missing words is based upon implications of the description of the matter contained in Sārṅgadeva, Kallinātha and Vemabhūpāla.

2 पातः कला तु सा जेयति सा सङ्घवदक्रिया पात कलेति सङ्गादवेनोच्यते । नि:शब्ददक्रिया तु कलासङ्ख्ययवोच्यत इति तु शब्दस्यार्थः ।

-Kalā on S.R. 5, 6.

3 आवापश्चाथ निःकामो विशेषपक्षक प्रवेशकः । एवमेषा तु निषण्णा कलेरेय प्रकल्पिता ॥

सङ्ख्याव्यापि चतुर्योङ्कता ध्रुवसाम्या तदैव च । तालमच्च सन्निपातमच्च सा स्यातं पातः कलेति च ॥

-S.C., ch. I, section on vādya.

4 एषामनःरपातास्तु पातसङ्ख्या: प्रकीर्तिता: । शम्या तालस्तु विद्वेय: सन्निपातरस्त्रिधा च ॥

सङ्ख्याहस्तनिपातः स्याच्छब्दस्य तालस्य वामतः । हत्स्थयोरन्तु समः पातः सन्निपात इति स्मृतः ॥

-N.S. 31, 36-37.

dered the basic mode. In vārtika (also called vṛtti mārga), the mode of kalā application was dvikala. In dakṣiṇa mārga, kalā measured 4 times the basic citra-

mārga unit of kalā, and the mode of application was known as catuṣkala. When Dattila says that kalā has different measures in different mārgas it is evidently to

the ekakala, dvikala and caṣṭuskala modes that he is referring. Thus kalā itself remained the same (= 10 niśas) but the unit of time measure in different mārgas

changed in terms of the number of kalās: citra had a 1-kalā unit, vārtika a 2-kalā unit and dakṣiṇa had a 4-kalā unit. However, the notion of kalā seems to have had still

more complexities (see note on ‘yati’ and ‘mārga’).

The ekakala mode was also known as yathākṣara and was considered the prototype mode. It was called yathākṣara for a reason. A number of kalās (both as

time-units and actions forming sounded and unsounded beats) distributed into groups, formed tālā. There were two such basic group formations: those based on a

duple arrangement (yugma) and those based on a triple arrangement (ayugma). These arrangements were expressed in terms of certain word-formulas, to the metric

syllabic structure of which the ekakala mode conformed (and was hence known as yathākṣara). These yathākṣara structures Dattila describes next.

122B. kalānāṃ samudāyo'tha yugmo'yugmo'thavā bhavet

123A. yugmaśaccatputasthastah syādayukcācapuṭāsrayāḥ

B. tālākṣarāṇāmetesāṃ samsthāpya gu(rulāghavam)

124A. yugmāntyaṃ plutavat kṛtvā kalāyogaṃ prakalpayet

A group of kalās can either be yugma or ayugma. Yugama is based upon caccatputa(h) and ayugma depends upon cācapuṭah.

In these letters of [yathākṣara] tālas, having established the laghu and the guru and rendering the last [syllable] in yugma as pluta, one should duly group the

kalās.

NOTE:

After a short aside on the pāta, Dattila comes back to the concept of kalā; for though Dattila has given to pāta an independent status as a topic (unlike Bharata)

yet pāta was a notion subsidiary to that of kalā.

The basic structures of tālas had been classified into two broad categories : the yugma and the ayugma. The prototype of all yugma (lit. ‘even’) rhythms was

called the tāla caccatputaḥ which was conceived of as a unit comprising four kalās (thus also called caturasra). The prototype of ayugma (lit. ‘odd’) rhythms was conceived

to be cācapuṭaḥ—conceived as a unit consisting of three kalās (also called tryasra).

1 The Hindustani kaharvā has a basic yugma arrangement whereas dādrā has an ayugma mode. In Karnatic music tālas are classified into tryasra (ayugma) and caturasra (yugma) types—a very methodical classification. In ancient times, too, tryasra was a synonym for the ayugma and caturasra for the yugma. We shall often come across these terms.

The terms cācapuṭah and cacatpuṭah are themselves meaningless, but they constituted word-formulas which when analysed letterwise (yathākṣara) and metrically on the basis of mātrās contained in each syllable, yielded the basic prototype of the yugma and ayugma modes of grouping kalās. Dattila does not analyse the words cācapuṭah and cacatpuṭah in detail, but simply gives general instructions in a verse (123B-124A) regarding the arrangement of mātrās in these tālas assuming a syllabic analysis as known to his readers.

In Sanskrit prosody, metre is measured in terms of mātrās (i.e., the time taken in uttering a short vowel). A laghu is a syllable containing one mātrā. Guru contains two mātrās and pluta three.

Bharata gives the full metrical analysis of the term cācapuṭah: Cā + ca + pu + ṭah. It contains a guru syllable followed by two laghus and then another guru. The total number of mātrās are six and we thus have a basic triple time structure :

traysraścācapuṭah prokto gurulaghvākṣarānvitaḥ ādau gurvakṣaraṃ jñeyaṃ laghunī guru caiva hī tryasraḥ sa khalu vijñeyaḥ tālāścācapuṭo budhaiḥ

(N.S. 31, 9)

In the yathākṣara yugma, i.e., cacatpuṭah—cac + cat + pu + ṭah—the distribution of laghu and guru js : guru + guru + laghu + guru. The total number of mātrās is thus seven; one short of the required number of mātrās that could form the four kalās (a kalā=2 mātrās) of which this tāla-prototype consisted. Hence Dattila has enjoined that the last syllable in the cacatpuṭah should be considered as pluta, i.e., containing three mātrās. This gives the required number of mātrās, viz. eight. Bharata in analysing cacatpuṭah on the ‘akṣara’ or syllabic basis does not say that the last syllable should be considered as pluta ; he simply enjoins it to be pluta :

ādau dve guruṇī yatra laghu ca plutam eva ca sa vijñeyaḥ prayogajñāistālāścaccatpuṭāśrayaḥ

(N.S. 29, 10)

"That tāla is known as based on cacatpuṭah by experts which has two gurus in the beginning followed by a laghu and a pluta." Śārṅgadeva says that the last syllable in this yathākṣara should be known as pluta (though is actually not so) : ayamekakalaścaccatpute tvantyaṃ plutam viduḥ (S.R. 5, 19). He notates its metric arrangement thus : SSIS (the last ‘S’ standing for a pluta).1

1 प्रत्येक तो नामतींगर्स्स्तन यथाक्षरम् । अयमेककलश्चच्चपुटेऽन्त्यं प्लुतं विदुः ॥ SSI इति यथाक्षरम्च्चपुटः ॥ SISI इति यथाक्षरम्च्चपुटः : ॥ —S.R. 5, 19.

Caccatpuṭah and cācapuṭah were thus the two basic tāla-prototypes or yathākṣaras. Bharata calls them the yonidvaya (N.S. 31, 7), for with them as base were constructed other tāla or kalā modes.

Abhinava reports that Āgamas and Purāṇas had traced these yathākṣara word-formulas to Śiva himself.1 He reads two verses in this context (perhaps as part of the Nāṭyaśāstra for he attaches them to N.S. 31, 8-11) which state that the syllables ‘caccatpuṭah’ were uttered by Lord Śiva in four breaths, after which he stopped. He did not utter the last syllable as pluta because it is not deemed fit that great personages should end theis utterances with a pluta. The second syllable was also not uttered as pluta for such a practice is followed only by the lowly. The pluta, in other words, was left as understood.

Abhinava quotes :

"tadāha caccatpuṭastviti devaścaturbhirniśvāsairakṣarakṣarāṇāṃ catuṣṭayam udīrya tasyātīte tu viśrānto girijāpatih plutāntanyāsato nāyamuttamānāṃ vidhīyate dvitiyo na tatheā tena hyadhamānāṃ pratīyate" (A.B. on N.S. 31, 8-10)

Though Śiva had not uttered the last syllable of caccatpuṭah as pluta yet obviously it was meant to be pluta and considered so in tradition.

Abhinava stresses that these prototype (yonirūpa) word-formulas had a special adṛṣṭa significance attached to them : “yonirūpamadrṣṭaviśeṣasya hetubhūtam kīrtya-mānamāgame prasiddham” (A.B., Ibid.).

The Nāṭyaśāstra speaks also of basic tāla units comprising five, seven, nine, ten and eleven kalās; these Bharata calls sañkīrṇa.2 But these units, Bharata says, were not employed in the gāndharva gītakas (saptarūpe) or in the dhruvā-songs of ancient dramas (na hyeṣā mupayego'sti saptarūpe dhruvāsu vā ; N.S. 31, 25).

Gāndharva, as we have observed, was a zealously guarded form of music which limited itself to only certain enjoined and approved forms. It is thus small wonder that in respect to tāla, too, gāndharva revealed a strict selectivity. But as to why the complex rhythmic forms based on units of five, seven, etc. were not employed in dhruvā (which contained a very pliant form of music allowing all kinds of modes) is

1 Again we notice the connection of gāndharva with Śiva. The four letters, says Abhinava, of the yathākṣara formulas were pronounced by Śiva Himself and were thus charged with spiritual significance and their utterance resulted in transcendental merit :

पुराणे वाग्मादौ भगवती महेश्वरस्य वक्तचतुष्टोमूत्कृष्टोभूतमेतदर्शचतुष्टकर्मेतद्वतामनःसंस्मृत्याद्यैर्युक्तम्

—A.B. on N.S. 31, 8-10.

The traditional basic syllabic formulas of Sanskrit grammar such as अइउण् ऋलक्.......... etc. were also said to have been devised by Śiva and were thus called Māheśvara sūtras.

कलाः पञ्च तथा सप्त पुनर्नव च कीर्तिताः । दशैकादश चैवते संकोर्णाः समुदाहृताः ॥

न ह्येषामुपयोगोऽस्ति सप्तरूपे ध्रुवासु वा ।

—N.S. 31, 24-25.

324

A Study of Dattilam

somewhat difficult to understand. Perhaps it was the very oddity of these uneven rythmic structures which made them unsuitable for dhruvā, for the complex structure of such rhythms may have hampered with the direct communication of poetic meanings. These rhythms were certainly used in non-dramatic popular forms.

Measure of Kalā

124B. kalāṃ guruṇi yuñjīta laghunyadhakalāṃ tathā

125A. plute sārdhakalāmevaṃ bhavedekakalo vidhiḥ

NOTE:

In these lines, Dattila lays down the measure of kalā, as well as its arrangement in the yathākṣara yugma caccatpuṭah. One mātrā in ordinary usage—which amounted to the time taken in uttering a short vowel—was in gāndharva considered as a duration of time taken in uttering five short vowels (Datt., 117). Two mātras, we learn here, formed one kalā. Kalā thus was a unit of time equal to ten common metric mātrās. Or, to avoid confusion between the two different measures accorded to mātrā, we can say that one kalā was equal to ten nimeṣas, since a nimeṣa was the time taken for uttering one short vowel.

In yathākṣara or ekakala modes of tālas, a guru equalling two mātrās was rendered as one kalā. Consequently a laghu was half a kalā and pluta one and a half kalās.

Syllables and Kalā

125B. yathākṣareṣu yoktavyāḥ pātās ete yathākramam

126A. sannipātaśca śamyā ca tālaḥ śamyā tathaiva ca

B. eṣa pātavidhirjñeyo yugme cāyugma eva ca

127A. śamyātal(au ca)1 yugme dvistālaḥ śamyamathāpi vā

NOTE :

The yathākṣara patterns described above were to be applied in the ekakala mode. The yathākṣara syllabic arrangement in yugma was : guru, guru, laghu, pluta.

1 T. ed. शम्याताल+ युग्मे

Our suggested reconstruction is based upon the description of the beat formations found in other authoritative texts, chiefly the N.S. Bharata descrbes the structure in this context in words similar to Dattila :

शम्यातालौ द्विरम्यस्तो तालः शाम्या तथापि वा

– N.S. 31, 12A.

Kalā

325

A two-mātrā-unit was equal to a kalā unit of time. Thus the pattern of tāla in terms of kalās was in caccatputaḥ : one kalā + one kalā + half a kalā + one and a half kalā. In rendering the ekakala mode of tāla these kalās were formulated through pātas alone, i.e., only through sounded beats. The beats enjoined by Dattila for these kalās have three alternatives :

1. sannipāta, śamyā, tāla, śamyā.

2. śamyā, tāla, śamyā, tāla

3. tāla, śamyā, tāla, śamyā.

The yathākṣara metric pattern in cācapuṭaḥ was : guru, laghu, guru, laghu ; or in terms of kalās : one + half + one + half. The beats enjoined were the same as in the first alternative (given above) for rendering caccatputaḥ. Thus in the ekakala mode, though yathākṣara caccatputaḥ had four kalās and cācapuṭaḥ had three, the number of beats (pāta) in both cases was four. This was because in the yathākṣara mode the syllabic structure of the word formulas comprising the tālas was adhered to. Both cācapuṭaḥ and caccatputaḥ had four syllables and consequently both were formed with four beats. Bharata says : “yathākṣarakṛtaiḥ pātaistālo jñeyo yathākṣaraḥ” (N.S. 31, 39).1 The number of kalās forming a duple and a triple grouping respectively, distinguished the two tālas.

Bharata has given the same three alternative beat arrangements for the caccat-putaḥ as has Dattila (N.S. 31, 11-12). The first alternative which begins with a sannipāta, he calls the ‘śuddha’ alternative :

sannipātastatata śamyā tālaḥ śamyā tathaiva ca

evamekakalaḥ śuddho yojyaśaccatputo budhaiḥ

(N.S. 31, 117)

He adds that the second formation beginning with śamyā was used in āsāritas (these were structures in the gītaka vardhamāna) and the third in the gītaka pāṇikā (N.S. 31, 13-14). These two gītakas were not within the saptarūpa fold, though they were gāndharva forms.

Dattila states that the latter two of the three alternative beat-arrangements were renderable only in caccatputaḥ (yugma) and not in cācapuṭaḥ (ayugma). Bharata, however, considered that all three alternatives were permissible in the cācapuṭaḥ also, but the first alternative—beginning with sannipāta—was the strong one (balavān).2 Abhinava comments that in cācapuṭaḥ the beat-formation beginning with sannipāta was the one mostly used: “cācapuṭaḥ sannipātādireva bhedaḥ pracuraprayogaḥ” (A.B. on N.S. 31, 15). It was, of the three, the only beat-formation for cācapuṭaḥ which was

1 Mark Abhinava’s comments on this line :

यथाक्षरक्तः: पार्वर्तिस्थिति यावन्ट्यक्षराणि तावत्संख्याेकस्थाक्षरसंख्याकलोपैरीप्ति

2 च्चत्पुटस्य ये भावा: सन्निपातादयक्षन ये । त एव भेदा विविधेया बुदैप्रेच्छाचपुटे पृथक् ॥ सन्निपातादिकस्तवस्य बलवानीतरौ तथ्यो । पट्कालो पट्कलश्चैव तालो हस्तामत: प्रवर्तते ॥ -N.S. 31, 14-15.

further expanded with the addition of other beats to form dvikala and catuṣkala modes. This was the reason why, Abhinava says, theorists like Viśākhila and others had considered this formation to be the only valid one in cācapuṭaḥ.1 Dattila, on this point, was evidently in agreement with Viśākhila. Vemabhūpāla in his Sangītacintāmani has indeed noted that Dattila did not consider the two latter alternatives as valid in cācapuṭaḥ but Bharata did. He says, “The beats beginning with sannipāta have been unequivocally laid down for cācapuṭaḥ. But the other two alternatives beginning with śamyā and tāla (respectively) have been decreed by Bharata (alone); Dattila and others do not mention them.”2

Later theorists like Śārṅgadeva, too, have described the beat formations of the two yathākṣara tālas. They have also notated the structures, using diminutives for the name of the beats and have listed the alternatives for caccatpuṭaḥ as follows :

1. S S I S

saṃ śa tā śa

2. S S I S

śa tā śa tā

3. S S I S

tā śa tā śa3

(S stands, as in Sanskrit metrics, for a guru mātrā; I for a laghu. Further, sam = sannipāta; śa=śamyā, tā=tāla. Similar short forms were current for the other beats too. We do not know when these short forms became accepted in practice. They are to be found in Abhinava but not in Bharata; nor in Dattila, in spite of the fact that he was zealously after brevity).

Śārṅgadeva notes that the two latter possibilities were, according to Bharata, valid in cācapuṭaḥ also : “anyadbhedādvyam cācapuṭepyasti munermatam” (S.R. 5, 29), implying that other ācāryas held a different view.

एवं भेदयसम्भवेन चापुटस्येतरेगा सास्यमू-वादयो विविधं दर्शयति। सन्निपातादिविशेषैरग्रहद्विकलचतुष्कलसिद्धिः। तथा चाचार्यविराधिलादिषु स एवं भेदो दर्शितः॥

-A.B. on N.S. 31, 15.

भेदस्तु सन्निपातादिर्यथाचापुटे स्फुटम्। भेदो शम्यादितालादयोग्यो वा चापुटे कव्वित्॥

-S.C. ch. I, section on vādya.

चत्वारस्तु इकले सङ्घतताङ्ग यथाक्रमम्

S S I S

सं श ता थ

यदवा शताश्रतता तालःशम्या वा द्विभंगेन्दिह

S S I S

श ता श ता थ

• S R. 5, 27-28,

The yathākṣara tāla ṣatpitāputrakah

127B. uttaraḥ pañcapānyākhyah ṣatpitāputrakākṣaraḥ

128A. ayugmotthah plutādyantastathā cāhātra kohalaḥ

B. dvau tu cācapuṭau kṛtvā dvitīyopānyake kramāt

129A. ādyantayorniyujyainam ṣatpitāputrakam viduḥ

B. sannipātaśca tālaśca tasyādyantau yathākramam

130A. (dvi)rmadhye tālaśamye ca (pātā)dibhyo (ka)lāvidhih

The uttara [tāla, also] called pañcapāni [is formed] with the syllables ṣatpitā-putrakah. [This tāla] arises from the ayugma. In it the first and the last [syllables] should be rendered as pluta. Regarding it, Kohala says : “having juxtaposed two cācapuṭahs, place the first syllable at the second [in cācapuṭah] and the final syllable at the last but one [in cācapuṭah] such [in cācapuṭah] is known as ṣatpitāputrakah.” In its commencement and its end, respectively, are sannipāta and tāla ; in between is a pair of tāla and śamyā formed twice. These are the beats that form the kalās [in this tāla].

NOTE :

Uttara was another yathākṣara tāla. Though not ‘yonirūpa’ or prototypal, it was still an important formation. This tāla had three names : uttara, pañcapāni and ṣatpitāputrakah (S.R. 5, 23). The first was the name proper, the second appears to be a descriptive epithet and the third represented the word-formula or syllables that formulated the tāla. It was a form of ayugma and contained twelve mātrās or six kalās.

Metrically analysing the syllables of ‘ṣatpitāputrakah’ (a strange word, indeed, which literally means ‘a son of six fathers’) into mātrās we obtain only ten mātrās : ṣaṭ (guru) + pi (laghu) + tā (guru) + put (guru) + ra (laghu) + kah (guru) = 10 mātrās. However, it had been ordained, as Dattila has stated, that the first and last guru syllables were to be considered pluta or three-mātric. Thus the tāla had 12 mātrās.

Dattila has described the metric arrangement of this tāla in a short aphoristic phrase characteristic of his laconic style: ‘plutādyantaḥ’ (Datt. 128)).2 Bharata, however, elucidates the whole metric structure syllable by syllable. “The first letter of the ṣatpitāputrakah”, he says, “is a pluta, the second laghu, third and fourth are gurus, the fifth is laghu and the last is pluta: ṣatpitāputrakṛtah pañca-pañirudāhṛtaḥ/ādyam plutam dvitīyam ca laghu yatrākṣaraṃ bhavet/trtīyam ca guruni pañcamaṃ laghu/plutāntaḥ ṣatpitāputrah” (N.S. 31, 17-19).

Dattila has given a dictum from Kohala in which the syllabic structures of two juxtaposed cācapuṭahs has been compared with that of a ṣatpitāputrakah. The purpose and nature of this comparison is not very clear. However, we attempt at a reconstruction of the idea.

1 Datt. 130A in the T. ed. reads :

विमध्ये (?) तालशम्ये च ++ पद्म्यो + लाविधि :

Our suggested reconstruction follows descriptions found in other texts.

2 Śārṅgadeva echoes this phrase from the Dattilam in describing the ṣatpitāputrakah :

घटपितापुत्रकस्योक्तयःप्‍वेः सौरिप तथा द्विधा । यथाक्षरे विशेषचिह्नं प्लुतमाधानतोमवेत् ॥

—S.R. 5, 22.

Kohala—according to Dattila—had said that having placed two cācapuṭahs together the first syllable of ṣaṭpitāputrakah should be placed where the second syllable in cācapuṭah occurs, i.e., under ‘ca’; the last syllable of ṣaṭpitāputrakah was to be placed under ‘pu’, the last but one syllable of the second cācapuṭah.

The number of syllables or letters in two juxtaposed cācapuṭahs amounts to eight, while in ṣaṭpitāputrakah there are only six syllables. The following chart makes the comparison between two juxtaposed cācapuṭahs and a ṣaṭpitāputrakah (as was perhaps intended by Kohala) clear :

Cā ca pu ṭah Cā ca pu ṭah ṣaṭ pi tā put ra kah

And in terms of metric structure :

S I I S S I I S Ś I S S S I Ś

(signs: I=laghu, S=guru, Ś=pluta)

We notice that the metric structures correspond throughout except at the beginning and end, but here, too, there is a correspondence: the pluta ‘ṣaṭ’ is equal to the three mātrās which cā +ca make and the pluta ‘kaḥ’ similarly equals three mātrās of the pu+ ṭah.

There was thus a general correspondence between two cācapuṭahs and a ṣaṭpitāputrakah and the above structural collation shows that the latter tāla was indeed an off-shoot of the former—it was ‘ayugmotthaḥ’ as Dattila describes it. Kohala’s purpose in collating the structures was perhaps just an easy practical method to demonstrate this fact, and it is, apparently, to this end that Dattila quotes him.

Though ṣaṭpitāputrakah was born of the ayugma, it was yet considered a basic tāla in itself. In spite of a correspondence with cācapuṭah it had a fundamental difference as to its beat-structure in the ekakala or yathākṣara mode. In ekakala modes, beats were formed on the basis of syllables and thus ṣaṭpitāputrakah (which had six syllables) had only six beats which formulated its six kalās (one and a half + half + one + one + half + one and a half) whereas two cācapuṭahs had to be formulated with eight beats.

Dattila has given the beat-structure of ṣaṭpitāputrakah in a short formula-like dictum (which, moreover, has some letters missing in the manuscript)—a manner of instruction which he repeats again and again as we shall further see. Bharata gives details arranged syllabically which reveal the following structure:

ṣaṭ pi tā put ra kah1 saṃ tā śa tā śa tā

1 ......पठ्यपाठः: परम् ॥ पटकारे सन्निपातः स्यात् पिकारे ताल एव च । ताकारे तु भवेच्छाया पुकारे ताल एव च ॥ तकारे तु भवेच्छाया ककारे ताल एव च । द्वितीयोऽङ्येऽयमेव स्यात् सन्निपातस्ततः परम् ॥ also : सन्निपातस्ततः तालः शा्म्यातालस्तथैव च । शा्म्या चैन हि तालश्च पट् पातास्तस्य कीर्तिता: ॥ —N.S. 31, 20.

The same structure has been given by Śārṅgadeva and others.1 Dattila, too, has, evidently, laid down this very formation. We have tried to reconstruct missing words from his injunction on the basis of these available authoritative descriptions. Dattila says that the first beat is the sannipāta which is followed by tāla+śamyā repeated twice; the final beat being tāla or in other words : sam tā śa tā śa tā, as described by Bharata.

1 Śārṅgadeva describes the beats as :

उत्तरेऽसं तत्स्ताल: भ्रातालो डिरनन्तरम्

S I S S I S

सं ता शा ता शा ता

इत्येककलपट्‌पितापितृकलाकविधिः;

—S.R. 5, 30.

TOPICS 3 AND 4

PĀDABHĀGA AND MĀTRĀ

130B. guruplutāni hitvātha dvimātrān parikalpayet¹

131A. pādabhāgaṃścaturbhistairmātreti paribhāṣyate

[Now] discarding [the divisions] of guru, pluta [etc.] render [all divisions] into [uniform units of] two mātrās.

With four of the pādabhāgas arise what is a pāribhāṣic mātrā.

NOTE:

In order to render an ekakala yathākṣara tāla into a dvikala, the number of metric yathākṣara mātrās was doubled and divided uniformly in terms of gurus instead of the divisions into pluta, laghu, etc. Each guru as before was equal to one kalā; but in dvikala two such kalās made a single unit and this is why the mode was known as dvikala. One unit of two kalās in the dvikala mode was termed the pādabhāga, and four such pādabhāgas made a larger unit of tāla called mātrā—which term in this sense had a strict technical connotation confined to tāla structures in gāndharva. It should be noted that this mātrā was quite different from the metric mātrā discussed earlier (which equalled one nimeṣa) and also the mātrā as a time unit for measuring kalā in gāndharva (which equalled five nimeṣas). Thus Dattila has called the mātrā here, pāribhāṣic mātrā.²

The pādabhāga and the pāribhāṣic mātrā were intimately connected with the dvikala (and also catuskala) modes of grouping kalās and the resultant tālas. And these two modes were, we shall see, very often used in forming gāndharva tāla-structure; thus pādabhāga and its related notion of mātrā (pāribhāṣic) was of the greatest importance in gāndharva.

Let us try with the help of an illustration to graphically understand the way in which the dvikala mode was formed from the ekakala yathākṣara and how the resultant structure was divided into pādabhāgas and mātrās.

1 A. B. on N.S. 31, 40 quotes Datt. 130B with the words : तदुक्तं दत्तिलेन. The reading is the same as in Datt.

2 This mātrā, says Śārngadeva, was used in the gītakas, the gāndharva tāla structures :

पदभागैस्तु भिस्तर्तव्यं स्यान्मद्रकादिषु

—S.R. 5, 21.

Kallinātha, evidently echoing Dattila, calls this mātrā ‘pāribhāṣiki’ :

मद्रकादिपु वद्मायणे गीतकेपु च तैर्द्विकलोकैः श्चतुष्कलोकैः चर्तुभिः पादभागोः मात्रा स्यात् । पारिभाषिकी मात्रा भवेत् ।

—Kalā on S.R. 5, 20-21,

Pādabhāga and Mātrā

331

The caccatputaḥ in its yathākṣara mode contained 8 mātrās (gāndharva time-units; each=5 nimeṣas) unevenly distributed as: guru + guru + laghu +pluta. For arriving at the dvikala form 8 mātrās were multiplied by two, thus giving 16 mātrās. These were then evenly divided into 8 gurus and the uneven distribution of the yathākṣara into guru, laghu, pluta was discarded as Dattila enjoins (Datt. 130B). The 8 guru were further divided into groups of two, each group having two kalās (one kalā = two mātrās = one guru); this gave the mode its name. Each of these two kalā groups constituted a single pādabhāga and the dvikala cacatputaḥ structure had four such pādabhāgas. This aggregate of four pādabhāgas was known as a mātrā (pāri-bhāṣic). The matter can thus be illustrated through a chart:

Mātrā Pāribhāṣic

| 1 | | 2 | | 3 | | 4 | (dvikala cacat-putaḥ)

| S S | | S S | | S S | | S S |

p ā d a b h ā g a s

It was only in the yathākṣara form that the beats followed the guru, laghu or the syllabic pattern of the tāla letters. In the dvikala and catuskala, however, the guru became the basis of a uniform division and the beats followed this guru unit.1

Dattila has given the process of dvikala formation and the related exposition of the notions of pādabhāga and mātrā (pāribhāṣic) in a very condensed, skeleton form. Details can only be filled in with the help of other texts. Bharata, too, is somewhat elliptical on this point,2 but Abhinava offers some comments which are helpful. He also quotes Datt. 130B in the course of his remarks on this point (A.B. on N.S. 31, 40-50). It is, however, in Śārṅgadeva that the actual form of the dvikala mode, metrically arranged, is depicted. He notates the dvikala cacatputaḥ arranged into pādabhāgas as: “SS SS SS SS—iti dvikala cacatputaḥ” (S.R. 5, 20). Each kalā, he says, in this mode is formed with a guru: “guruḥ kalātra dvikale” (S.R. ibid.).

The dvikala cacaputaḥ was formed on the same principle as the yugma form. Here the structure emerges as SS SS SS. One cācaputaḥ cycle had only three pādabhāgas and could not form a complete mātrā (pāribhāṣic).

1 यथाक्षरकृतः पादस्तालो जेयो यथाक्षरः: । गुरुद्दैर्घ्यरतः विश्रब्धः: स एव द्विकलो भवेत् ।। -N.S. 31, 39-40.

2 Bharata's definition of pādabhāga and mātrā, too, is as short as that of Dattila: त्रिकलश्र्चतुष्कलश्च व पादभाङ्: प्रकीर्तितः । क्त्वारः; पादभमानस्तु मानेभति परिचीयते ॥ -ibid., 31, 52.

131B. ādyam prakalpya niṣkrāmam dvitīyam tu praveśanam

132A. sarveṣām pādabhāgānām tataḥ pātān prakalpayet

B. śamyā dvitīyā kartavyā tālaśamye tu madhyame

133A. sannipato'vasāne ca dvikalo'yam vidhih smṛtah

[In dvikala caccatpuṭah] one should form niṣkrāma at the beginning and praveśa as the second in all the pādabhāgas. Then one should arrange the pāt as [thus]: śamyā should be rendered as the second tāla and samyā should occur in the middle and sannipāta as the final. Such is known the process (vidhi) [of beats] in the dvikala.

NOTE:

In his zeal for brevity, Dattila becomes so laconic in his descriptions that they often read like obscure mathematical formulae. Fortunately, Bharata and also later authors have given details enabling us to interpret Dattila meaningfully. Bharata recounts all the eight beats of dvikala caccatpuṭah in due order.

The beats are

S S S S S S S S

ni śa , ni tā, śa pra , ni sam

(abbreviations for the seven beats: ā=āvāpa,

ni=niṣkrāma, vi=vikṣepa, pra=praveśa,

śa=śamyā, tā=tāla and sam=sannipāta).

In the light of this, let us consider Dattila's description. He first makes a general rule that niṣkrāma should be formed in the beginning and praveśa as the second in all pādabhāgas. He then qualifies this general rule by particular details: śamyā, he says, should be formed as the second beat-this, in effect, removes praveśa from the

1 T. ed. reads पाद (?ता) न

2 Bharata describes the structure as follows, giving the exact finger-gesture for each of the un-sounded beats to be applied:

कनिष्ठाङ्गुलिनिष्क्रामं शम्या चैव ततः भवेत् । कनिष्ठानामिकाभ्यां तु निष्क्रामौौ विधीयते ॥

ततश्च तालः; कर्तव्यं चैव तु पंचभि: । प्रवेशो मध्यमाङ्गुष्टः कर्तव्यस्तर्जनीकृतः ॥

निष्क्रामं सन्निपातोद्भते तर्जन्यमस्तकलौ भवेत् ।

—N.S. 31, 41-43,

"First the niṣkrāma with the little finger ; then the śamyā. Again niṣkrāma, indicated with the little and ring fingers. Then should follow tāla; śamyā should be rendered the fifth beat. The sixth is the praveśa rendered with the middle finger; then occurs niṣkrāma with the index finger. Sannipāta is the last beat. Thus is rendered the form having 8 kalās (i.e. the dvikala caccatpuṭah)."

Śārṅgadeva gives the same array of beats for the dvikala caccatpuṭah:

निषो निताप्रविष्टं द्विकलके यूरुमके मतः:

S S S S S S S S

नि शा नि ता या प्र नि सं

—S.R. 5, 30,

first pādabhāga replacing it by śamyā. Then he enjoins that tāla and śamyā should occur in the middle—middle in Dattilam, evidently, stands for the two central beats. This means that praveśa is to be removed from the second pādabhāga and substituted by tāla and that niṣkrāma is removed form the third pādabhāga and replaced by śamyā. The final injunction is that sannipāta should te the last beat; consequently praveśa is removed from the last pādabhāga to be replaced by sannipāta. We thus arrive at a beat structure, same as that of Bharata.

In verse 131B-132A, Dattila in effect composes a general formula for the tāla formation of all dvikala pādabhāgas; niṣkrāma in these structures was by a general rule the first beat and praveśa the second, unless either or both were replaced by sounded beats through a specific injunction. This formula is utilised by Dattila in depicting the beat-structures of other dvikala forms, too, as we shall see in the very next verse which describes the beat structure of the dvikala pañcapāṇi or ṣaṭpitāputrakaḥ.

We find an indication in Bharata also of a suggestion that the structure ‘niṣkrāma + praveśa’ was considered the general prototype beat structure for all pādabhāgas in the dvikala mode. Bharata says: “niṣkrāmaśca praveśaśca dvikale parikīrtitau” (N.S. 31, 35). In itself the statement is ambiguous but Abhinava explains it as meaning that in the dvikala mode the first kalā should be manifested through niṣkrāma and the second through praveśa (in every pādabhāga) as a general rule: “prathamakalākalaṅakena niṣkrāmeṇa dvitīyakalākālakaḥ praveśena vyañjaya iti sāmyāvalakṣaṇam”.

However, line 2 of the verse from Bharata quoted above reads: “āvāpaniṣkrāmakṛto dvikalo yoga isyate” (N.S. 31, 35). This line apparently gives quite a different general structure for the dvikala than the preceding line. Abhinava suggests that the two kalās indicated here (in the second line) were the first two kalās of the catuṣkala form of the pādabhāga—these as we shall see were generally formed with āvāpa and niṣkrāma. Abhinava’s words are: “when extension of the kalās is indicated at one end (i.e., when kalās are increased to four, in the catuṣkala pādabhāga) then the two end-kalās are indicated with āvāpa and nikṣepa (=niṣkrāma?)—tathābhūtapāryantikavitātībhāvasūcite cāvāpanikṣepakale’tra pāryantikama-vadhikam sūcayataḥ” (A.B. on N.S. 31, 35). Some theorists, says Abhinava, gave a different interpretation and considered that āvāpa and niṣkrāma could be the two generally ordained kalās of the dvikala mode also. Abhinava vehemently criticises such theorists. He say that these people are contradicted by the living traditions of the art: the authoritative evidence of texts other than the Nāṭyaśāstra as well as by guru-paramparā : forms and techniques as taught and explained by teachers; they are also, he adds, contradicted by the evidence of actual practice. He concludes that no attention should be paid to such erroneous views.1

1 अन्ये त्वेककले आखुते यथा निष्कामस्तथा द्विकलनायामपि भवत्यावाप इति मन्यमाना यथाश्रुतमेवावापनिष्क्रामातमकद्विकलं मन्यन्ते । तेऽपि पूर्वपराशास्त्रतत्स्वरूपपरालक्ष्यविरोधादूनेक्ष्या एव ।

—A.B. on N.S. 31, 35.

334 A Study of Dattilam

Beats of the dvikala ṣaṭpitāputrakah

133B. pañcapāneścaturthī tu śamyā kāryā tathāṣṭamī

134A. tṛtīyaṣaṣṭhanavamā(s)tālāḥ śeṣaṃ yathocitam

Whereas, in pañcapāni, the fourth and eighth [beats] should be formed with śamyā, the third, sixth and ninth with tāla. The rest being according to the general rule (yathocitam).

NOTE :

In the dvikala pañcapāni or ṣaṭpitāputrakah there were twelve kalās distributed in six pādabhāgas, and the tāla was rendered with twelve beats. According to the general formula posited by Dattila in the preceding description, each pādabhāga was to be formed with niṣkrāma as the first and praveśa as the second beat. In the case of pañcapāni, Dattila qualifies this general structure by stating that śamyā should replace the fourth and the eight beats and tāla should occur at the third, sixth and ninth beats. We thus have :

The Pattern according to general rule :

ni pra, ni pra, ni pra, ni pra, ni pra, ni pra,

Qualified pattern:

ni pra, tā śa, ni tā, ni śa, tā pra, ni pra,

The pattern given by Nāṭyaśāstra is the same as the qualified pattern here, except that sannipāta is stated to form the final beat (N.S. 31, 47 to 50). Although Dattila does not state so explicitly, it seems to have been another general rule that sannipāta formed the final beat in all dvikala and caṭuṣkala forms (see Datt. 160 A). This can be inferred from other structures given by both Dattila and Bharata. Perhapss Dattila here does point to sannipāta as the final beat in saying, ‘śeṣaṃ yathocitam’. This phrase follows after the naming of each particular beat which were to replace the beats in the general formula and clearly implies that some feature had still remained unsaid. The unsaid feature itself supposedly followed as a general (yathocita) dictum. This seems to have been the final sannipāta.

Before coming to the caṭuṣkala mode, let us take up some other matters of interest in connection with yathākṣara and dvikala tālas.

Variant yathākṣaras :

Dattila has expounded only three yathakṣara tālas, namely, caccatputah (yugma), cācapuṭah (ayugma) and ṣaṭpitāputrakah (another mode of the ayugma). Bharata has described two more yathākṣara structures : udghaṭṭaḥ and sampakveṣṭākaḥ (N.S. 31, 21-22).1

1 The N.S. 31, 21 reads सम्पकवेष्टक संजितः but this spelling is erroneous; the fourth syllable should be द्र because Bharata himself says that this yathākṣara consisted of five guru syllables (गुरुपञ्चाक्षर). Abhinava reads the name as सम्पकवेष्टाक इतिन his commentary and Sārṅgadeva also has the same spelling.

(1) Udghaṭṭaḥ, as its syllabic structure indicates, was a ayugma structure consisting of three gurus : SSS. The beats given by Bharata are ni śa śa : ‘niṣkrāma followed by two śamyās.1

(2) Sampakveṣṭākaḥ was also another yathākṣara ayugma. In its syllabic structure it contained five gurus which make only 10 mātrās. However, as Bharata has stated, the first and last syllables were ordained to be pluta2. The metric structure thus was : Ś S S S Ś.

Bharata has decribed the beat-structures of the sampakveṣṭākaḥ—which contained five beats as it had five syllables—immediately after that of ṣaṭpitāputrakaḥ. Bharata says that (unlike ṣaṭpitāputrakaḥ), sampakveṣṭākaḥ began with tāla. This was meant to imply that the rest of the beat structure was common to the two yathākṣaras. Abhinava comments that in sampakveṣṭākaḥ the sannipāta was to be left out from the ṣaṭpitāputrakaḥ structure (thus making it ‘tālādi’) for it had five beats. He gives the resulting structure as : tā śa tā śa tā.3 The ṣaṭpitāputrakaḥ structure was : saṃ tā śa tā śa tā.

Udghaṭṭaḥ, Abhinava says, was just a mode of the cācapuṭaḥ itself and was only sometimes used in practice; but it was rendered with a form somewhat different (from the cācapuṭaḥ) : “udghaṭṭopi yathākṣarasya cācapuṭasyai va rūpaṃ kvacit prayogād rūpabhedenoktaḥ” (A.B. on N.S. 31, 53). Similarly, he says, the yathākṣara structure called ‘sampakveṣṭākaḥ’ was a variant of the pañcapāṇi or ṣaṭpitāputrakaḥ : “evaṃ sampakveṣṭāko bhedaḥ pañcapāṇīereva” (A.B., ibid.).

These two forms though different in their yathākṣara forms, retained no differences with cācapuṭaḥ and ṣaṭpitāputrakaḥ in their dvikala (and also caṭuṣkala) modes. Abhinava reports that “neither in lakṣaṇa (i.e., śāstric exposition) nor in lakṣya (musical practice), does one observe a separate dvikala structure for these two: “etyevaṃ ca tayor na pṛthagdvikalābhyo lakṣaṇe lakṣye ca dṛśyate” (A.B., ibid.).

Śārṅgadeva follows Bharata in describing sampakveṣṭākaḥ. He says that in sampakveṣṭākaḥ as in ṣaṭpitāputrakaḥ (of which the former was a variant), the first and last syllables were to be considered as plutas. The resultant metric structure being : Ś S S S Ś:

sampakveṣṭā’ko’pi bhedaḥ ṣaṭpitāputrakasya saḥ tadvadyathākṣaraḥ kāryaḥ plutamādyaṃtayorbhavet S S S S S iti yathākṣaraḥ sampakveṣṭākaḥ

(S.R. 5, 24)

1 द्वयथ्रं सर्वगुरुं कृत्वा निष्क्रामं स्वनन योजयेत् । शाम्याद्वयं ततस्त्वेष उद्‌घट्टः कथितो बुधैः ॥ -N.S. 31, 22.

2 तालादिस्थ यथाङ्गेsड्यः सम्पक्वेष्टकसंज्ञितः । गुरुपञ्चाक्षरावृत्याsन्त्यमातामसमन्वितः ॥ -ibid., 31, 21.

3 cf. गुरुपञ्चग्रहणान्न षट्चपाणिप्रभेदनेनित तावता माङ्गाभिधाने sसन्निपात नियमो लब्धः: तेन तासं तासं ता तालस्येव भेदः:. A.B. reads तार्सं तासं ता but सं here is obviously a scribal error for या. There was no yathākṣara structure which contained so many sannipātas. If it did, Bharata would have especially said so.

336 A Study of Dattilam

The pāribhāṣic mātrā and the dvikala :

The two entities, mātrā and pādabhāga, occured only in the dvikala and caṭuṣkala modes. Bharata thus says:

dvikalaścaṭuṣkalaśca pādabhāgaḥ prakīrtitaḥ

catvāraḥ pādabhāgāstu mātreti paricodyate

(N.S. 31, 52)

In the cācapuṭaḥ dvikala ayugma structure there occured only three pādbhāgas—or a pādbhāga less than the mātrā (pāribhāṣic). In dvikala ṣaṭpitāputrak aḥ the structure had six pādbhāgas and thus one and a half mātrās. This was certainly an odd arrangement. In the caṭuṣkala mode, too, though the size of the pādbhāga became doubled to four kalās, mātrā (pāribhāṣic) remained equal to four pādbhāgas and thus the basic ayugma structures still contained mātrās in fractional number. We do not know why no separate pāribhāṣic mātrās was conceived for the ayugma, which being based on the triple-time rhythm, was certainly structurally quite distinct from the yugma and hence deserved a different mātrā (pāribhāṣic) unit based on the triple form.

There is a dictum of Bharata which, as explained by Abhinava, affords some clue as to the reason for this dominance of the duple. Regarding the two differing tāla structures, yugma and ayugma, Bharata says that their basic nature (prakṛti) was one:

tryasraśca caturasraśca sa tālo dvividhaḥ smṛtaḥ

dvividhas yāpi tālasya tvekā prakṛtirisyate

(N.S. 31, 7)

"Tāla is of two modes: tryasra (ayugma) and caturasra (yugma). Both these modes of tāla have one basic nature."

Abhinava comments: "in both yugma and ayugma structures the guru mātrā (metric) forms the basic unit (after the laghu pluta division is discarded)— thus both the ayugma and the yugma forms are constructed with guru units, which is their basic nature (prakṛti)."1

Keeping these remarks in mind, we, indeed, notice that both the yugma and ayugma structures in their dvikala and caṭuṣkala modes were based upon a duple unit, the guru which constituted kalā, the basic time measure. There was no kalā with a triple time-base for the ayugma. Again these guru units were combined into pādbhāgas, and here, too, we notice that a yugma-combination formed the structural basis both for yugma and ayugma tālas: the dvikala pādbhāga was formed with two gurus and the caṭuṣkala with four gurus (gurudvayaṃ gurucaṭuṣkaṃ cātra mūlaprakṛtiḥ; A.B. on N.S. 31, 7). In fact, the entire structure of all dvikala and caṭuṣkala tālas was built up of units based on multiples of two. This evinces that in gāndharva there was

'एका प्रकृतिरिति' । गुरुरपो यो मेन्द: सोऽत्रैकस्वभावे लघुप्लुतादिच्छेदविभागोऽनूद्य: सोऽत्र प्रकृतिमूलभूतः । तदपेक्षया च तस्यैतद्विभागम् ।

—A.B. on N.S. 31, 7.

a pronounced leaning towards the yugma unit. This was perhaps the reason that led the pāribhāṣic mātrā, too, to be constructed uniformly of four pādabhāgas both for yugma and ayugma structures.

Abhinava, explaining Bharata's dictum regarding mātrā (viz. 'catvāraḥ pāda-bhāgāstu mātreti paricodyate', N.S. 31, 52), remarks that in dvikala modes eight kalās made a mātrā and in catuṣkala modes (where each pādabhāga contained four kalās) sixteen kalās made a mātrā. Even for ayugma tālas, he says, the same rule applied. He gives an actual example: that of the gītaka prakarī where sixteen kalās of the ayugma tāla made a mātrā (pāribhāṣic).1

Mātrās (pāribhāṣic) of the ayugma were evidently formed by repeating the basic tāla structure. Whole number of mātrās could only have been obtained by multiplying the basic ayugma structures in such appropriate numbers of cycles that the resulting kalās formed a multiple of eight in the dvikala and sixteen in the catuṣkala2. Bharata, describing the number of kalās contained in the various ayugmas (N.S. 31, 28-29), implies that this is what, indeed, occurred. Ayugma tālas, he says, have six classes (ṣadbhedaḥ) in terms of the number of kalās contained in them, namely, those with (1) three kalās (e.g., ekakala cācaputaḥ), (2) with six kalās (e.g., ekakala ṣaṭpitāputrakaḥ), (3) with twelve kalās (e.g., dvikala ṣaṭpitāputrakaḥ), (5) with forty-eight kalās and (6) with ninety-six kalās.3 The last two forms, namely, those having 48 and 96 kalās respectively could only have been formed through repetition because the catuṣkala mode which was the largest accepted mode could not have had more than 24 kalās in any of the basic ayugma structures. We notice that the repetition was such that the number of resultant kalās was a multiple of 16 (and 8) in both cases. Abhinava explains that a 48 kalā rendering of ayugma was to be found in the gītaka called madraka (aṣṭacat-vārimśattālā iti, tālastu yo madrakaśya prastutaḥ prayogaḥ A.B. on N.S. 31, 28-29). After 48 kalā in this gītaka, he says, occurred a vidārī (a sectional pause or division).4 A similar unit of 96 kalās was found in the gītaka prakarī.5

This further reveals that in gāndharva forms it was the duple time rhythm which was more predominent and even triple time forms were grouped in terms of the

1 मालासङ्ख्यं चत्वारः पादभागास्तु मातृगेति । तेनाष्टौ द्विकलेऽचतुःकुले च षोडशकला इति माता । तस्या अवयवो युग्मो भागः । न तु यथेष्टप्त च प्रकृत्यादि मातृ षोडशकालैव भवतीति ।

व यथास्थितालस्तु पदप्रमेयस्तिकालः । कला द्वादश चैव स्यात्तु चतुर्थः षटिरेव च ॥

-N.S. 31, 28-29.

4 cf. विदार्यों अवश्यमेव विधेया इति तस्य समुदायस्य भागतया कृत्वाऽऽड्तचतुर्वारिंषट्कलसङ्ख्यादियुक्तम्

-A.B. on N.S. 31, 28-29.

5 एवं षण्णवतिकले प्रकृत्या: प्रयोगे

-A.B. on N.S. 31, 28-29.

338

A Study of Dattilam

duple or yugma. Abhinava consequently remarks that : "in this way, through

pādabhāgas and mātrās, all tāla varieties terminated ultimately in the caturasra (or

yugma) and thus caturasra was the predominating structure."1

Another point worthy of attention here is that Bharata has recounted groups

of kalās only upto 96. Larger groups could be built by further multiplications of two

but these were evidently not permissible in gāndharva. "Numbers in themselves",

comments Abhinava, "have no limiting rule, but gāndharva is like the Vedas; it exists

from time immemorial (vedavadanādi); usages in it are strictly determined (niyata)

and in gāndharva a grouping of 96 kalās (is the largest per missible) cluster.2"

The Catuṣkala Mode

134B. vinyasya madhye vikṣepamādāvāpameva ca3

135A. sarveṣām pādabhāgānām prayuñjīta catuṣkalam

The application [of tāla] in catuṣkala should be [accomplished] by placing

āvāpa in the beginning and vikṣepa in the middle of all [the dvikala] pādabhāgas.

NOTE:

Dattila now comes to the catuṣkala pādabhāga. He elucidates its beat

formation again with a characteristic short formula. He says that two extra beats

were to be added to the pādabhāga previously described (i.e., the dvikala) in order to

obtain the catuṣkala.

The dvikala pādabhāga had two kalās. The catuṣkala was double the dvikala :

"dvirbhāvād dvikalasyaapi vijñeyo'tha catuṣkalaḥ", (N.S. 31, 40). It had four kalās.

Consequently, (the total number of kalās in a catuṣkala caccatuputah was 16 (4 pāda-

bhāgas of 4 kalās each= 16 kalas); in cācapuṭa the number of kalās was 12 (3 pādabhā-

gas of 4 kalās) and in ṣaṭpitāputrakah 24 kālas (6 pādabhāgas of 4 kalās). Dattila

has not enumerated the number of kalās that obtained in a catuṣkala pādabhāga in so

many words but in laying down two extra beats (both unsounded) for it, he implies that

the catuṣkala pādabhāga had four kalās.

The catuṣkala pādabhāga had two kalās more than the dvikala. Thus two more

beats had to be added to each pādabhāga of the dvikala in order to render it catuṣ-

kala. There was a simple general maxim for making this addition : āvāpa was to be

placed in the beginning and vikṣepa in between the beats of the dvikala mode in all

its pādabhāgas to render them catuṣkala. The catuṣkala forms thus were:

Caccatputah:

S S S S S S S S S S S S S S S S

ā, ni, vi, śa ā, ni, vi, tā ā, śa, vi, pra ā, ni, vi, sam

1 एवं हि पादभङ्गमालादिति चेद् गेयदस्य चतुरश्रप्रयवसानाच्चतुरस्रस्यैव प्राधान्यम्

-A.B., ibid.

2 सङ्ख्याया अनियमाद गान्धर्वस्य वेदवदानादिसिद्धत्वेन नियतप्रयोगत्वाद्वात्तु वण्ण्यते कलाावान्तर समुदाय :

-A.B. on N S. 31, 30-32.

3 A.B. on N.S. 31, 51 quotes Datt. 134B with the words : 'तथा च दत्तिलाचार्य: स्पष्टमवोचत्

The read ing Athinava gives of the line is the same as in Datt.

Pañcapāṇi:

S S S S S S S S ā, ni, vi, pra, ā, tā, vi, śa, ā, ni, vi, tā, S S S S S S S S ā, ni, vi, śa, ā, tā, vi, pra, ā, ni, vi, saṃ

Cācapuṭaḥ:

S S S S S S S S ā ni vi śa ā tā vi śa ā ni vi saṃ

The maxim given by Dattila for the formation of caṭuṣkala pādabhāga has a parallel in Bharata:

eṣa tvāpavikṣepavyavadhānena paṇḍitaiḥ vidhiścaṭuṣkalo jñeyaḥ prāguktāṅgulībhiḥ kṛtaḥ

(N.S. 31, 51)

We observe that Bharata has not described the exact placing of the two extra beats to be introduced into the caṭuṣkala pādabhāga as Dattila has. But Abhinava in his commentary discusses the location to be accorded to āvāpa and vikṣepa. Taking ‘vyavadhāna’ in this context to be the key-word in Bharata’s kārikā (quoted above), he concludes that the norm for the dvikala pādabhāga which was enjoined as generally consisting of the beat structure ‘niṣkrāma+ praveśa’ (see Datt. 131B) had to be enlarged for the caṭuṣkala in such a way that an obstruction—vyavadhāna—was created between the two beats of this structure. This was done firstly by displacing niṣkrāma from the first position it occupied and relegating it to the second place (which, comments Abhinava, was what ‘vyavadhāna’ meant in this context). Similarly, praveśa had to be displaced from the position it occupied in the dvikala form and vikṣepa introduced in its place. In this way āvāpa and vikṣepa created a ‘vyavadhāna’ between niṣkrāma and praveśa (and resulted in the caṭuṣkala pādabhāga). Abhinava adds that this matter has been clearly expounded by Dattila and quotes Datt. 134B.2

Śārṅgadeva has also spoken of the rule by which the caṭuskala beat-structure was to be arrived at by enlarging the dvikala. He gives the exact placing of the two

1 This conforms with the beat structures as notated by Śārṅgadeva and others (see S.R. 5, 31-32).

2 यदावापविक्षेपाद्‌वाघातौ स्थनादपसारयेत् तथा पणिदते: पूर्वानुसंधानकृतलेशतुष्टिकलविघातय: । ततस्तु श्चायापेनानास्था-नमाक्राम्यता निष्‌क्रमस्थाने हितोर्ध्यायघानमार्जवम् । प्रयोगस्थ च व्यवधानं विस्तारोकरुणामिति । यदा (वा)पयोगेन निष्‍क्‌रामस्य च विक्षेपयोगेन प्रवेशस्य च प्रयोको व्यवस्थित: कारय: । आवापविक्षेपाभ्यां तयोरेवच व्यवधानमिति ध्यायाताम् । सर्वथा नावपिक्‍षेपो निरस्तरो निष्‍क्‍रामप्रवेशयोमध्ये (अन्ते आदी वा श्रृङ्‌यो) इति मतस्‍थ्यम् । तथा च इतिक्‍लाचाय: ‘विनियत मध्य वि क्ष पमा दधा (वा) परमेव च’ । -A.B. on N.S. 31, 51.

new beats that were to be introduced and was perhaps influenced in this matter by Dattila's description. He also notates the actual beat structure of caṭuṣkala forms.1

We notice that in caṭuṣkala pādabhāgas the first two kalās are almost invariably formed with āvāpa and niṣkrāma (see beat-structures). This is perhaps what Bharata had indicated in saying: 'āvāpaniṣkrāmakṛto dvikalo yoga iṣyate' (N.S. 31, 35). We have already argued that Abhinava appears to have been in favour of such an interpretation (see note on Datt. 131B-133A).

Finger Mudrās For Making Unsounded Beats

135B ādyadiyīyamādyāntāt pādabhāgād viduh kramād

136A. kaniṣṭhānāmikā yuktā madhyamād deśinīgatān

B. yugmaṁ madhyahīnāṁ syādādyaṁtādika uttaraḥ2

137A. niṣkrāmaśca praveśaśca dvikale parikīrtitau

B. evam kalādi yoktavyam viśeṣo yatra nocityate

In dvikala [tālas] niṣkrāma and praveśa are said to be [indicated as follows] :

[In yugma] the first two, the middle and the last pādabhāgas are known in due order as indicated by the little, the ring, the middle and index fingers.

In ayugma the middle finger is omitted.

In uttara, the first and the last [finger indications] are added. In absence of any specific injunction, kalās should thus be rendered (yoktavyam).

1. आदावधिक आवापे विश्रवेप्सितनिस्तरा । द्विकले पादभागः स्यात्पादभागकृतभूषिते ॥

SSSS SSSS SSSS SSSS

आनिविष्ट आनिवित आभिमुख्र आनिविसं

इति चतुष्कलचञ्चुपुटकलाविधिः:

SSSS SSSS SSSS

आताविष्ट आनिविसं

इति चतुष्कलचञ्चुपुटकलाविधिः:

SSSS SSSS SSSS SSSS SSSS

आताविप्र आनिविता आनिविषा आनिविसं

इति चतुष्कलषट्पितापुत्रकलाविधिः

—S.R. 5, 31-32.

2 Abhinava quotes Datt. 135B-136 twice. When commenting on N.S. 4, 315. he quotes the lines with the words: 'सङ्क्षिप्य कथ्यतेऽत्र तद्', indicating that Dattila had expressed in brief a matter which in the N.S. is expounded at length. This matter says Abhinava (when commenting on ch. IV) is described by Bharata with the words "कनिष्ठानामिकानिर्गतम्" "वर्जनीकृतोऽनिकृताम्" etc. in the tālādhyāya (i.e., N.S, ch. 31). Abhinava was here clearly pointing at N.S. 31, 41-49 where Bharata gives detailed instructions for the finger-mudrās for niṣkrāma and praveśa in various pādabhāga units. A.B. reads Datt. 135B as 'आचादितो मध्यमात्ततः' and Datt. 136 as : कनिष्ठानामिकामिका युक्तोमध्यमादर्शिनो गतौ:. Commenting upon N.S. 31, 40-40, Abhinava again quotes Datt. 135B-135 with the words: 'तदुक्तं दत्तिलाचार्येण. Here he adds this short tippani to these lines: "विदुरिति क्रामिति स्वार्थितः पदैरेकं ज्ञातमित्य सूचितम् । यत्र तु पादोपपादकैर्न भवति तत्र सार्धं लिख्यते'.

In this instance Abhinava's own reading of the Datt. passage differs somewhat from his previous reading: Datt 135B reads as in T.ed. But Datt. 136B has 'अयं मध्यमहीनः'. T.ed. here has many lacunae. But Abhinava gives a full reading.

NOTE:

The ekakala modes of tālas were rendered with sounded beats or pātas alone but the dvikala and catuṣkala modes contained unsounded beats also. The gestures used for indicating praveśa and niṣkrāma have already been described in verses 119-120. Here we find more details specifying particular finger indications which qualified the main gestures. Bharata gives a beat by beat detail regarding these additional finger indications while describing the beats in the pādabhāgas of each of the three basic dvikala modes, caccatputah, cācapuṭah and ṣaṭpitāputrakah (N.S. 31, 41 to 50). Dattila, characteristically, formulates a simple general rule or maxim from which the details follow. Abhinava, indeed, quotes the above verses from Dattilam with the words : “dattilācāryeṇa sāṃkṣipyoktametat”, (A.B. on N.S. 4, 315).

In the light of details given by Bharata, Abhinava and of Śārṅgadeva, the rule formulated by Dattila works out to be as follows. In the first pādabhāga of the yugma dvikala tāla (i.e., caccatputah), niṣkrāma or praveśa—whichver of these silent beats occurred—was to be indicated with the little finger in addition to the larger gestures already ordained for these beats. In the second pādabhāga, niṣkrāma and vikṣepa were to be indicated by both the little and the ring fingers. In the third pādabhāga, the middle finger, too, was to be added. Finally, in the fourth pādabhāga the, index finger was also to be included (N.S. 31, 41-43).1

In ayugma, says Dattila, the middle finger indication was to be omitted. By ayugma he means the cācapuṭah. Ayugma had only three pādabhāgas. The first two pādabhāgas retained the same finger movements as the yugma, but for the third pādabhāga it was not the middle finger that was used—as should have been done if the due succession was to be maintained—but the index finger. The gesture with the middle finger was omitted (ojasya pādabhāgeṣu kalā madhyāṅgulim vinā, S.R. 5, 34). The reason was ritualistic. Abhinava says that this rule depends upon adṛṣṭa alone: “adṛṣṭasāmarthyāt tarjanyeva gṛhyate madhyamā mu tyajyate” (A.B. on N.S. 31, 40-50). In uttara (i.e., pañcapāṇi or ṣaṭpitāputrakah) in the first four pādabhāgas, niṣkrāma and praveśa were indicated as in cactapuṭah but for the two extra pādabhāgas, the fifth and sixth, Dattila enjoins the first and the last gestures successively (ādyantādhika uttaraḥ, Datt. 136B). That is to say—as Bharata clearly discribes—the tenth kalā (in the fifth pādabhāga) which was rendered with a praveśa was to have the additional finger movement indicated by the little finger and the eleventh kalā (in the sixth pādabhāga) which was a niṣkrāma was to be indicated through the index finger.2 The ring and middle fingers, which obtained as per succession, were omitted.

We see that in yugma tālas, finger-indications began with the little finger : the first pādabhāga was invariably indicated with it and successive fingers upto the

1प्रथमे पादभागे स्यात्कलाऽनुक्ता ल्या कनिष्ठया । तथा चानुमयान्यत्त तालस्यो मध्यमया तथा ॥ -S.R., 5, 32-33.

2कनिष्ठिकाप्रवेशार्ष्तु दयामी तु कलां प्रवेत् । निष्क्रामश्वैव तर्जन्या तदनन्तरभिष्यते ॥ -N.S. 31, 49B-50A;

index were added for subsequent pādabhāgas. In ayugma structures, too, gestures were so arranged that in case of the repetition of a tāta-cycle it was always to the little finger that one returned for the first pādabhāga. The reason for beginning with the little finger was, as Abhinava says, ritualistic and followed Vedic practice. In chanting mantras of the Ṛgveda and Yajurveda the accented syllables which were indicated through gestures always began with the kaniṣṭhā (the little finger): “yajusi rci vā svaritapradarsane kaniṣṭhatā ārabhya svarānām kurvate” (A. B., ibid).

Line 137B of Dattilam also evidently concerns gestures. T.ed. reads ‘kāla’ in this line which we think is a mistaken reading for kalā, the more appropriate term in this context. Kalā here, evidently, refers to the beat, and this line seems to refer to gestures in the catusḳala mode and seems to bear the same contention as N.S. 31, 51 where Bharata says: “experts make the aforementioned indications in the catusḳala also but āvāpa and vikṣepa intervene in between.”1 The finger indications for niṣkrāma and pravesa in the various pādabhāgas of the dvikala have been described. In catus-kala, niṣkrāma and pravesa were indicated by the same movements but here āvāpa and vikṣepa intervened as decreed. No additional specific finger movements have been laid down for āvāpa and vikṣepa. They were, apparently, indicated only by the movements described in Datt. 118-19.

1 एवं त्वावापविक्षेपव्यवधानेन पण्डितैः। विधिश्र्चतुष्कलो ज्ञेयः प्रागुक्तस्त्रिकलो लिपिः कुरः॥ — N.S. 31,51,

TOPIC 5

PARIVARTA

138A. pādabhāgo'tha mātrā vā samastasatāla eva vā

B. gīyate pari(vṛtyā cetparivar)tah sa ucyate1

Repeated singing of a pādabhāga or a mātrā or a complete tāla is called pari-varta.

NOTE :

The mātrā referred to in this verse, obviously, is the pāribhāṣic mātrā of four pādabhāgas. The term parivarta literally means to revolve or turn around. Parivarta is listed as a tāla topic in Bharata, too, but he does not elucidate its nature in his exposition. He does, however, use the term in specifying tāla particulars (see also ch. 1). Abhinava also uses the term. And it is clear that by parivarta they both meant a repetition2, though the form and nature of this repetition is not clearly laid down as in the Dattilam.

A few words are missing from Datt. 138B, but the intention is clear. The Sangītacintāmaṇi of Vemaḷhūpāla has a definition of parivarta which seems to have been borrowed almost verbatim from Dattilam:

pādabhāgāsya mātrāyaḥ tālasya sakalasya vā

āvartanaṃ tu yadgīte parivartah sa ucyate

(S.C. ch I, section on vādya, verse 82).

Our suggested reconstruction of the missing letters is based mainly upon this verse. Śārṅgadeva has a definition similar in spirit but he names only the pādabhāga

1Datt. T.ed. 138B reads : गीयते परिव + + + पदन; स उच्यते ।

2Bharata describes an application, where the tāla-limb called āsārita was formed thrice. Here he uses the word parivarta:

ततश्चायार्षितं बुद्धे; १ परिबर्त्ति स्थितियौ स्तमतश्चोपोहानं भवेल् ॥

—N.S. 31, 335.

Abhinava, following Bharata, often uses ‘parivarta’ to denote a repeated formation of any element in tāla. Describing vardhamānaka. where elements termed khaṇḍikā had to be repeated in a certain order and were formed ten times, he says:

संहृत्य त्याज्योऽपेक्ष्यवृत्तितवे दशपिरवर्तंयोमे इहोच्यते and again आसारितारामत्न संयोजे योजनाविशेषाधीने प्रयोने कण्ठकानामेव सम्बन्धी ताल उत्कृष्ट दश परिवर्ताः ।

—A.B. on N.S. 31, 147.

Bharata says:

अत्र तालविधिच्छन्दतः प्रायशस्तुग्रवन्धने । आदावृत्तटतः: कार्य: परवर्तैक एव च ॥

—N.S. 31. 265.

Abhinava explains:

परिवर्तेक व्यपदेशाच्च द्विकलेडपि परिकल्प्यामातस्तवादनस्य स्तादितश्चचक्रुट्टताल: कर्तव्यसैवनेतत; ।

as a unit for repetition; mātrā and total tāla structures are not named, though they are implied through the use of an 'etc' added after pādabhāgā: “āvṛttih pādabhāgādeḥ pari-vartanamiṣyate” (S.R. 5, 43). Kallinātha comments that by 'pādabhāgādeḥ' was meant the different limbs of a tāla as well as the tāla itself.1

Later expositions of parivarta are perhaps based on Dattilam or another similar text on gāndharva because Bharata does not have it.

1 पादभागादे:; इत्यादिषुर्देनावबध्यो तालो गृह्यते । तस्य आवृत्तिर् अभ्यास: युतः पुनः करणमिति यावत् ।

TOPIC 6

VASTU

139A. yathā (krameṇa) 1ṣaṇmātrāstistro'dhyardhāpi vā bhavet

B. kecit prādhānyamātrena vastusamjnā pracakṣate

140A. apanyāso'tha vastvante syādamśo nyāsa eva vā

B. sannayāso vātha tasyātra rūpaṃ vi(dvadbirucyate)2

141A. prathamāyā vidāryāyā madhye nyāsatvabhāk svaraḥ

B. na cedamśavivādī syāt sannyāsaso'bhidhīyate

A vastu, in due order [consists of] six or of three mātrās or is further reduced by half (adhyardha). The term vastu is used by some because it predominates.

At the end of a vastu is either apanyāsa or amśa or nyāsa or sanyāsa. The form of the last ('tasya', i.e., of sanyāsa), in this context, is known by the wise [as follows]: in the first vidārī the note which is accorded the position of nyāsa (nyāsatvabhāk svaraḥ) is called sanyāsa provided it be not a vivādī of the amśa.

NOTE :

A vastu was one of the main tāla components with which the structure of gītakas was built. We have already defined a mātrā (pāribhāṣic). Many components of the gītakas were measured in terms of mātrā. Vastu was, apparently, a representative of all major tāla components. Nānyadeva, in his Bharatabhāṣya (ch. 8-2,3) thus says: "tāla predominates in all the gītakas, mātrā is termed tāla and with mātrā is formulated the vastu".3

In the section on svara, Dattila had first enumerated the basic structures and concepts with which the musical form of jātiṣ was constructed and understood. Similarly, in the section on tāla, he first describes and defines the basic concepts of tāla and its groundwork of structures with which the major tāla-forms called the gītakas were built. The gītakas were complex tāla-structures utilising simpler elements that have been described earlier.

The steps Dattila had followed in describing tāla may here be summed up. Beginning with the basic unit of time (kalā) in tāla, he defines its measure in terms of

1T.ed. reads यथा++++. Our suggested reconstruction is based upon descriptions in later saṅgīta texts, especially the S. Raj. Kumbha almost reflects Dattila's words in saying:

प्रकृत्या मत्त्रके वापि तथा च्चेदापरत्वके । अर्थ(द्रै) होन क्रमा व स्तु षणमात्रादूतरोरितरमु ॥ मा पष्ठयोमादिके वस्तु प्राधान्यात कुलचिन्मतम् । —S. Raj 2,4,1. 19-20.

2T.ed. reads:

तस्याद्र रूप वि� (?) + + + + :

3सर्व्वस्यामेव गीतानां तालस्यैव हि मुख्यता । ताला तु गदिता मात्ता मानातो वस्तु कल्पना ॥ —B.B.(I) ch. VIII.

mātrā (metric) the duration of which in gāndharva was defined as equal to five ordinary mātrās equal to five nimeṣas. Then he describes the various beats with which time was demarcated (these, too, were denoted through the term kalā). Subsequently he takes up the basic tālas, namely, yugma (caccatputah) ayugma (cācaputah) and uttara (ṣatpitāputrakah). He describes the syllabic structure through which these tālas were represented, sketches their form and the number of kalās contained in them. He also recounts the pātas or beats which demarcated their structure. Next he takes up the notion of pādabhāga which formed the basis of the two other modes of tāla formation, namely, dvikala and catuṣkala. He delineates the structure of tāla and the kalā unit in dvikala and catuṣkala and introduces the concept of the pāribhāṣic mātrā. He then sums up the general rules for the arrangement of beats in the pādabhāgas of the dvikala and catuṣkala modes, giving detailed beats for the caccatputah and uttara tālas.

Now, with Datt. 139 he takes up vastu. And with this he enters into the actual form of the gītakas.

Dattila enumerates three types of vastus: those formed with six mātrās (pāri-bhāṣic), those with three and those with one-and-a-half mātrās. It appears from descriptions given by later theorists that in the gītaka called madraka the vastu was built of three mātrās. Bharata thus characterises prakarī as having an expanded form: “prakaryāstu prakīrṇatvāt” (N.S. 31, 203). Abhinaya elucidates: “a vastu should here be expanded upto the limit of six mātrās.”1 In aparāntaka the vastu was of one-and-a-half mātrās.

The second line of Datt. 139 is not very clear, though Dattila’s contention appears to be that vastu was the model or representative form of all tāla-components (tālāngas) in gītakas. This, he adds, was according to some the reason why the name vastu was accorded to it. There is no clue as to the theorists who held this view.

In verses 140 and 141, the nature of the association that vastu has with its accompanying melody is expounded and some rules governing structural fusion are specified.

The notions of apanyāsa, amśa and nyāsa have already been explained (in the svara section). The connection of apanyāsa with vidārī has also been noted. Sanyāsa was a more restricted concept, a corollary to apanyāsa. Apanyāsa had been defined as the note accorded the position of nyāsa in a vidārī; sanyāsa was the note accorded the position of nyāsa only in the first vidārī of a composition.

Dattila defines sanyāsa in dealing with tāla because it has an innate connection with the nature of tāla formation. Bharata defines sanyāsa as part of jāti-lakṣaṇas. But the two definitions are quite parallel. Bharata says: tatra prathamavidārīmadkye nyāsasvaraprayuktastu vivadanaśilam muktvā sannyāsah so’ bhidhātavyah (N.S. 28, 73) “That note is called sanayāsa which acts as the nyāsa in the first vidārī and is not a vivādī.”

1 'प्रकीर्णत्वाद् विशिष्टतया च विस्तारस्तः वसु तनु पञ्चमातमेव कियत इति -A.B. on N.S. 31, 203,

Vastu

347

Bharata says that the sanyāsa note should not be a vivādī but he does not specify the exact nature of vivāda as Dattila does in saying ‘na cedamśavivādī syāt’. Abhinava, like Dattila, remarks : “amśasya vivādī yo na bhavati” (A.B. on N.S . 28, 73).1 We have had occasion to remark that vidārī was a pause, a sectional ending within the body of musical structures. Dattila’s remarks make it clear that vidārī scanned not only tāla but also the accompanying svara-structures. His observations reveal that tāla and svara (with its accompanying pada) were inseparably and intimately interwoven into one unitary structure and imply that each portion and aspect of this whole was pre-determined according to well-set rules of vidhi and niṣedha, do-s and dont-s.

The Nāṭyaśāstra contains no distinct and specific exposition of the term vastu, though being a key concept the term naturally recurs often in its descriptions.

1 The Vrttikāra on Br̥haddeśī makes a similar qualification : हृदानीं न्यासोभधीयते । तत्र प्रयमाविदारी मधये न्यासस्वरा: (?रः) प्रयुक्तः (?तः), सन्यास: सौधम्भधीयते । ग्रंथास्य द्विवादी यथा न भवति प्रयमविदार्य्यन्ते यदि प्रवृत्तो यदा भवति तदासौ सन्यास इत्यर्थ: —Br̥. Vr̥tti on 197A.

Sārṅgadeva, too, says : अंशाविवादी गीतस्यार्थाविचारีसमाप्तिकृत् सन्यास; —S.R. 1, 7, 47-48.

TOPIC 7

VIDĀRI

142A. padagītisamāptau1 yā prthagītiriva sthitā

B. gītapeśim vidārīm tāmāhurgītavidāṛāṇām

143A. pramānāvarjam tasyāstu rūpam vastuvadisyate

A portion of a song (gitapeśi) which at the completion of a pada or gīti stands as if it were a separate gīti, is called vidārī.

The measure of its form, if not indicated, should correspond with the vastu.

NOTE :

The purpose of vidārī, as the word itsef suggests (from √dṛ =‘tear asunder’ or ‘split’), was to divide or scan a song into parts. Bharata has defined vidārī as the consummation of pada or varṇa.2 Abhinava elucidates : “the end of an intermediary sentence (avāntaravākya) or the consummation of a musical varṇa (such as sthāyī, etc.) on the nyāsa or the apanyāsa is called vidārī.”3 He analyses the word vidārī as ‘vidāraṇaṃ vidārī’ and calls the term ‘anvartha—self-explanatory’ (A.B. on N.S. 32, 17).

Dattila’s definition is akin to that of Bharata, though vidārī, according to Dattila, divides not pada or varṇa but pada or gīti. However, it appears likely that gīti in this context had the connotation of varṇa. The varṇas encompassed all musical movements possible in song—gīti—and Dattila may have had this in mind.

Vidārī was an important notion in ancient Indian music. Abhinava says that “melody or song rendered without the charm of diversity imparted to it by vidārīs, is (monotonous) and sounds as if a book being read : vidārīvaicitryaśūnye hi gītavyavahāre svarālāpaḥ pustakavācanādiva” (A.B. on N.S. 31, 200). Abhinava again remarks elsewhere that “the formation of a vidārī imparts great charm to music : vidārī niveśyamānā śobhātiśayaṃ puṣṇāti” (A.B. on N.S. 32, 17). Indeed, pauses within a musical structure are like sentences and paragraphs in prose and stanzas in verse and are just as essential.

Vidārī has been defined in terms of varṇa, that is, the pada units in a song (padavarṇasamāptistu vidārī, N.S. 32, 17). But its measure has been given in terms

1 T. ed. reads : padagītiḥ; samāpyo

2 पदवर्णसमाप्तिस्थु विदारीत्यभिसंजिता ।

—N.S. 32, 17.

3 पदवर्णप्रकाशितुविदारीति । अवान्तरवाक्यस्य समाप्तो स्थाय्यादिवर्णस्य च न्यासेन वा समाप्निविदारी ।

विदारणं विदारी ।

—A.B. on N.S. 32: 17.

of a tāla-component (Datt. 143A). Dattila says that the extent of a vastu normally

comprises one vidārī division. Here we certainly have another indication that a

pre-planned correlation existed between the sung melody and tāla.

Vidārī evidently depended to a great extent on punctuations existing in the

words of a song. Abhinava suggests this in remarking that vidārīs occurred at the end

of clauses (vidārī cāvantarapadasamūhavicchedair ūpā, A.B on N.S. 29, 14).

Dattila has explained vidārī as a gītapesi. Abhinava, with the same con-

tention, says : “gītakhanda vidārī” (A.B. on N.S. 32, 17). ‘Peśī’ literally means a

carved piece (of meat etc.) and ‘peśikṛ’ means ‘to carve into pieces’. The Vṛtti

on Bṛhaddeśī in expounding vidārī uses both Dattila’s and Abhinava’s expressions :

he equates them, implying that vidārī was based on the act of splitting or dividing

a song and thus consisted of a portion of a song : “gītapesi gītakhaṇdamiti yāvat”

(Bṛ. Vṛtti on 197A).1

Abhinava at one point somewhat cryptically remarks : “a gītaka is rendered

with songs formed with varṇas like sthāyī etc.; divisions or sections in the song (gīta)

being sung should be made as per the gītaka structure (tadanuvartī). A section or

division has to be carved out of the body of the object being divided (which in this

case is the gītaka ); this act of craving is effected through the vidārī which thus

imparts sāmyā to the (musical) structure. Therefore vidārīs are enjoyed as being

constructed through the tāla gītas (i.e., the gītaka tāla structures—because the main

body which they render in to sections in that of the gītakas).”2 Abhinava’s purport

appears to be that though vidārī divides the whole melody into sections yet the act

of division is made on the basis of tāla.

Dattila says that when the measure of a vidārī has not been specified, it must

be taken to equal a vastu. This seems to imply that the vastu was the largest

possible unit of vidārī. The vastu itself, we have seen, could be of differing length

in different gītaka tāla-structures. This would, consequently, have shortened or

lengthened the vidārī section according to the gītaka in which it was formed. Dattila

here has not laid down the standard measure or length of a vidārī in gītakas not

formed with vastu. However, he seems to have taken vastu as the representative

form and consequently the major tāla-components of the four other gītakas were,

apparently, in this matter to be taken as the largest measure of a vidārī.

Vidārīs could also be smaller in measure than a vastu. They could even be

formed on the basis of single padas and varṇas or small clusters of these contained

1. A full treatment of vidārī is not available in the Br. However, in defining apanyāsa and

sanyāsa, some explanation of vidārī was necessary. The Vṛttikāra to that end says :

ननु विदारीखण्डन किमुच्यते । पदनां (पदानां) स्तादिस्थियादि विदारणा खण्डनमिति यावत् । गीतकेशो गीतखण्डनमिति

यावत् ।

• Vṛtti on Br. 197A.

'गीतकेपी' here is obviously an erroneous reading for 'गीतपेशी' ।

तथा हि । स्थाय्यादयो ये वणीरताः-नानाप्रचालनमान गीत गीतव मिति तदनुवर्तीं विच्छेदः क्रियते । परिच्छेदस्य परिच्छेद

(या) तनुत्वादिविशिष्ट साम्यास्ति परिच्छेदेनश परत्वम् । श्रुत एवैयं तालगीतोनिर्दिष्टा ।

-A.B. on N.S. 31, 190-91.

within a vastu. Later authorities, including Abhinava, have given the name of mahāvidārī to a vidārī which equalled a vastu, and avāntaravidārī to the sub-sections of mahāvidārī (see A.B. on N.S. 31, 197-98).1

Bharata has laid down a limit for the number of vidārīs which could be formed in any gītaka. "The vidārīs," he says, "should not be less than three and never exceed eleven in number. But, he adds, "the maximum number of vidārīs that can be formed is twenty-four."2 Abhinava explains the first statement as giving the possible numbers of mahāvidārīs and the second as giving the avāntara vidārīs in a gītaka (A.B. ibid.). Dattila, as we see in our text, makes no such injunction. Later authorities were aware of Dattila's position and Śārṅgadeva clearly states: "teachers like Dattila have not laid any rule as to the number of vidārīs" (S.R. 5, 74; see also ch. I).

The mahāvidārīs, according to Abhinava, equalled in measure a vastu: "atra vastunyekā mahāvidārī" (A.B. on N.S. 31, 197). Thus in the two gītakas madraka and prakarī-each of which was formed with three vastus (also, alternately, with four vastus)-there were at least three mahāvidārīs: "madrakasya prakaryā vā tāvattatra tisro mahāvidāryā ityanuyogah" (A.B. ibid.); this according to Bharata was the minimum number of vidārīs obligatory in a gītaka. The gītaka aparāntaka could be formed with seven vastus or alternately with five or six. In it, consequently, one could have upto seven mahāvidārīs. Hence Abhinava remarks : "one can form vastus more (than three)-i.e., four etc.-upto the number seven, as in aparāntaka : bhūyo-vastutvam tvartha (?) caturthavastutvaprabhṛti yāvat saptavastutvam yadāparantake" (A.B., ibid.).

The maximum permissible number of mahāvidārīs is stated to be eleven. This number, according to Abhinava, is obtained in the gītaka aparāntaka in the following manner : seven vastus in aparāntaka collectively constituted a large tāla-component called śākhā, another such tāla-component complementing the śākhā (and formed with half the vastu), was the pratiśākhā which consisted of four vastus. One could have four vidārīs in the pratiśākhā (in addition to the seven in the śākhā) and consequently a total of eleven in aparāntaka-"tatra (aparāntake) prativastu vidārī sapta vidāryā iti śākhā. ardhacaturtheṣu vastuṣu pratiśākheti. tadgatāścatastro vidāryā ity ekādaśeti prayogah" (A.B. ibid.)

Abhinava also gives some interesting details regarding the shorter or avāntara vidārīs. Bharata, we have seen, also speaks of a maximum of twenty-four vidārīs (N.S. 31, 198). This according to Abhinava was the total possible number of avāntara vidārīs in a gītaka. He takes the example of madraka. Madraka could be composed of four vastus, each consisting of three mātrās (pāribhāṣic); short vidārī sections were to be made at each half mātrā-thus each vastu had six avāntara vidārīs, the

1 cf. विदारी तु गीतखण्ड द्रिधा च सा । महत्यवान्तरा चेति महती व्याप्तवस्तुका इ । समाप्या पदवर्णौत्कृष्टा न्तरा स्वन्तर मता । —S.R. 5, 72-73.

2 न.यत्रैकादशावर (कादशावर) विदार्यः परिकीर्तिता। चतुरुच्वशमितरेतासां प्रमाणं परमं स्मृतम् ।। —N.S. 31, 197-98.

total in the gītaka being twenty-four: "caturviṃśati etāṣāṃ pramāṇamiti. anenedamucyate, trimātre vastuni mātrārdhe yāvad vidārī kartavyā. yena caturvastuke madrake piṇḍīkṛtya caturviṃśatiḥ. prativastu ṣaḍiti gaṇanayā" (A.B. on N.S. 31, 198-99).1 Abhinava then takes another example. When the vastu consisted of six mātrās and three vastus had to be formed—as in the gītaka prakarī—then short vidārīs could not be formed at every half-mātrā unit (as in madraka); for, this would have resulted in 36 vidārīs and with more vastus (for prakarī could be formed also with four vastus) further avāntara vidārīs would have resulted. Thus vidārīs in order to conform to the rule limiting their number were here to be formed at every mātrā and not half a mātrā: "ṣaṇmātre tu prakṛte vastuni nārddhamātrāyāṃ vidārī kāryā; evaṃ hi trivastuke ṣaṭtrimśad2vidārikādhikātve'dhikavidārikāpi prakarī syāt. etasmānniyamāttu na bhavatīti jñātamātrāyāṃ tara vidārītyuktam bhavati" (A.B., ibid.).

Some theorists had interpreted the avāntara-vidārī rule—where the maximum possible number is said to be twenty-four—in quite a different manner. From Abhinava's suggestions the interpretation appears to be as follows. The structure of any tāla-formation could be taken to form a single unit divisible through vidārīs. Take for example the ekakala caccatpuṭaḥ. Taking it as a single unit, one vidārī was obtained. If, however, each kalā was to be made into a single vidārī section, four vidārīs were obtained. And on further divisions of half-kalās and quarter-kalās, one could obtain eight and sixteen vidārīs respectively. The same analysis could be applied to the ayugma cācapuṭaḥ and the ṣaṭpitāputrakaḥ, which could yield vidārīs upto a maximum of twenty-four (i.e., 6 kalās × 4 quarter divisions = 24). Thus, according to this view, the rule regarding the number of permissible divisions (viccheda) referred not to gītaka structures but to time-units of tāla-measure in general (sāmānyakālābhiprāyeṇa).3 In this view the number of divisions permitted in each kalā were four. Abhinava has illustrated the nature of these divisions with the example of two actually sung syllables, namely. 'ru dra'. But the reading as available in the Abhinava Bhāratī does not reveal the illustration clearly: "yathā ru..... dra ...... ityekasyāmapi kalāyāṃ catvāro vicchedāḥ" (A.B., ibid.).

Abhinava introduces his comments on N.S. 198A with the words 'matāntaramāha' which seems to imply that the statement regarding 24 as the maximum number of permitted vidārīs was, according to Abhinava, not Bharata's own view but that of another ācārya (or school of thought) whose opinion, however, Bharata may have incorporated out of deference. We do not know which ancient ācārya or school of thought Abhinava (or Bharata) here had in mind.

2 The A.B. here has पड्‌त्रिकत् which is obviously an error, because three vastus of 6 mātrās, each mātra containing two vidārīs, would yield a total of 36 and not 26 vidārīs.

अपरेऽपि सामान्यकालाभिप्रायेण यद्वासम्भवं विच्छेदसंस्याप्रतिपादनं प्रलोकाद्‌वस्य तात्पर्यमाहुः। तथाहि । चच्चापुटे अपरेऽपि सामान्यकालाभिप्रायेण यद्वासम्भवं विच्छेदसंस्याप्रतिपादनं प्रलोकाद्‌वस्य तात्पर्यमाहुः। एककालविच्छेदे एका विदारी । प्रतिकलविच्छेदे चतस्रः। कलार्धविच्छेदेऽपि तद्‌द्वयं विच्छेदे चतुस्त्रिकलद्वयविभागोद्भवेऽष्टौ। एवं द्वयत्रे वाच्यम् । यावत् पञ्चपाणो चतुर्विंशतिः। —A.B. on N.S. 31, 198-99.

TOPIC 8

AÑGA

143B. vak śyamānam mukhādyaṅgaṃ vijñeyam ca caturvidham

144A. ekakam cāvagāḍham ca pravṛttam vividham tathā

B. nyāsāntamapi vāṃśāntam vidāryekakasaṃśritam

145A. avarohyavagāḍham tu prāyaśoṃsāntamiśyate

B. vidāryoḥ (ntarahīneṣu)¹ svaraṣvekāntareṣu vā

146A. evam nyāsāntamārohī pravṛttamabhidhīyate

B. arohaṇam tu dvividham² tathā caivarārohaṇam

147A. nyāsāpanyāsayogena bhavedantarmārgataḥ

B. śuddhaḥsvamśavaśāt tatra jñeyamantaramārgataḥ

148A. nyāsāpanyāsayogena vikṛtāsu ca jātiṣu

B. sāṃmugdho³ thārdhasāṃmugdho vivṛddhaśca yathābhidhaḥ

149A. samo madhyo'tha viṣamo vividhastrividhaḥ smrtaḥ

B. ipāthamātrā samā gitīḥ³ sāṃmugdho vividho bhavet

150A. asampūrṇavidāryādisaṃyogāṃ caiva prakalpayet

B. nyāsānto vividhaḥ sarvo muktvā dvaigeyakam bhavet

151A. nity am ca dvividārikch, pūre⁴ jñeye tu tu śaṭ pare⁵

Mukha and others, of which we shall speak subsequently, constitute aṅga; this is of four types: ekaka, avagāḍha, pravṛtta and vividha.

Ekaka is based upon [a single] vidārī [and] ends with a nyāsa or with an amśa.

Avagāḍha [contains] the avarohī (i.e., the descending varṇa) and should, in most cases, end with the amśa. Vidārīs [in it should be formed] on notes immediately following [the amśa] or [on notes] once removed from [the amśa]. Similar is the pra-vṛtta, [but formed with] the ārohī (i.e., the ascending varṇa) and it ends in nyāsa.

1 T.ed. reads : विदार्योः +++

2 T.ed. reads : द्विवियं

T.ed. reads : चाड़मात्रा स महामोति : We have accepted the reading in B.B. (I).

3 T.ed. reads : इतत्सं च द्विविधारोका पूर्‌

4 Many lines from the passage on aṅga have been quoted by later authorities :

Abhinava quotes Datt. 143B with the words : ‘तथा च दत्तिलचायोऽक्तम्’ (A.B. on N.S. 31, 190-91).

His reading agrees with the Datt.

Abhinava also quotes Datt. 147B-148. Before quoting the lines he observes : ‘अत्र कश्चिद्विशेष उत्कः ;’ then the quotation is given—which agrees with Datt.—after which Abhinava says : इति दत्तिलाचायः:

Nānyadeva in B.B. (I) ch. VIII quotes Datt. 148B-150—these five lines in the same order as in Datt. Dattilam is not mentioned as the source but the context seems to show that Nānyadeva had the Dattilam in mind. The B.B. (I) reading is quite corrupt at places. However, it reads Datt. 149B as पाठमात्रासमामोति:, a cogent reading which makes sense while the Datt. does not.

We have accepted the B.B. (I) variant.

[In this context] ārohaṇa and avarohaṇa are of two kinds: [firstly] formed by using nyāsa and apanyāsa, [and secondly] depending upon antaramārga. In the śuddha (i.e., unmodified jātis) they should be known [as formed] through antaramārga, which depends upon the aṃśa. In vikṛta (or modified) jātis [they are rendered] by utilising nyāsa and apanyāsa.

Vividha is said to be of three kinds termed sāmudga, ardha-sāmudga and vivṛddha; [these are respectively] sama, madhya, and viṣama [in nature].