The Music and Musical Instruments of Southern India and the Deccan c6: Chapter II

The Music and Musical Instruments of Southern India and the Deccan c6

Language: English

Chapter II

c6.1

Preservation of music in Southern India—Modern Karnâtik system believed to be closely akin to the ancient—Sanskrit treatises—Definition of s'ruti—System of twenty-two—Formation of scales (grâma)—Relation of major and minor modes—Deviation of intervals of system of twenty-two from those of just intonation—Comparison of ancient scale with European diatonic—Results of recent observation—System of twelve hinted at in Sanskrit—Murchanas—Râga—How differing from mode—Tâla or rhythm—Time, how estimated—Ancient varieties of rhythm—Ancient notation.

c6.2

THE theory, modes, and notation in present use throughout the whole of India are derived from that taught originally by the earlier Sanskrit musicians; but owing to the South of India having been less disturbed by internal commotions, and having been more subject to Hindu rule than either the Deccan or Northern provinces, the science of music would seem to have been maintained and cultivated long after the original art had been lost in the North.

c6.3

Hence Southern Indian music, or, as it is more usually called, Karnâtik, bears, as far as we can judge, a very close resemblance to what the Sanskrit must have been, and in many cases we can clearly trace the development and refinements introduced from time to time upon the original râgas.

c6.4

One of the principal Sanskrit works upon the theory of music is the Sangíta Ratnâkera, or “Ocean of Music,” written by Sârnga Deva at a very early date, probably about A.D. 200. This book, with the exception of Bhârata S’astra, noticed hereafter, is considered to be the oldest reliable musical work extant, and is quoted in many subsequent treatises. The Sangíta Dâmódara is also a well known and very ancient work.

c6.5

Later than this we have the Sangíta Naráyana, by Naráyanadéva, in which the Dâmódara is frequently quoted.

c6.6

There are also the Râganâva,¹ or “Sea of Passions”; the Sangíta Ratnâvali; the Sábha Vinóda,² or “Delight of Assemblies”; and the Sangíta Darpana, or “Mirror of Music,” written by Dâmódara Misra—all works of more or less value in the elucidation of the music of the ancients.

c6.7

¹ This work, together with the Sábha Vinóda and Râga Darpana, is quoted by Mirza Khan, but there seems to be no copy known at present.

c6.8

² This work is mentioned by Sir William Jones. As far as I have been able to ascertain, there is at present no known copy in existence, unless, indeed, this work is identical with one called Sangíta Vinóda, a copy of which is in the library of H.H. the Maharajah of Bikanîr.

c6.9

The Râga Vívodha, or “Doctrine of Musical Modes,” by Somanath, or Soma Raj, is one of the most valuable of the ancient treatises that have been handed down to us. This book was evidently written at a much later date than the Ratnâkera, which it quotes from several times; but it is doubtless a very ancient composition. The first, third, and fourth chapters explain the doctrine of musical sounds—their division and succession, the variation of scales by temperament, and the enumeration of modes. The second chapter contains a minute description of different vinas, with rules for playing upon them.³ The last chapter contains strains noted in letters. The whole work is singularly clear from mythological references.

c6.10

³ Sir W. Jones.

c6.11

The Persian writer, Mirza Khan, under the patronage of Aazim Shah, wrote a work entitled “Tohfuht-ul-Hind,” which contains a chapter upon music, the information for which was extracted, with the assistance of Pandits, from various Sanskrit works. Mirza Khan describes four principal systems of music—viz., those of Isvara; of Bhârata; of Hanumân, or Pavân; and of Callanath. All four are mentioned also in the Râga Vívodha.

c6.12

The main principles contained in most of these ancient works are very similar. The differences consist mostly in the names and the constitution of the different modes and râgas.

c6.13

Written at a still later date than any of the former, to judge by all appearances, is the Sangíta Pârijâta,⁴ or “Flower of Music,” by the Pandit Ahobala.

c6.14

⁴ According to the Mahabhârata, Pârijâta is the name of the celestial tree in Indraloka.

c6.15

The system of music described in the Pârijâta differs from that of the Ratnâkera, in that it admits of greater intervals than a tone or four s’rutis, and of less intervals than a semitone or two s’rutis, being, therefore, capable of forming numerous enharmonic scales. All the notes, except the first and fifth, are occasionally shifted above or below, and the fourth is never omitted in the scale.⁵

c6.16

⁵ See Preface to edition of Sangíta Pârijâta, edited by Kâlivâra Védantabágisa. Calcutta, 1879. See also the list of modern Karnâtik scales given upon pages 32—35.

c6.17

This work contains the key to the present Karnâtik system, and many of the râgas contained in it are practically the same as those now in use in Southern India.

c6.18

The ancients divided their octave into twenty-two intervals, called s'rutis.

c6.19

The names of these s'rutis are differently given by the various authors, but the following list, taken from the Sangíta Ratnâkera, is the one which, at the present day, is best known:—

c6.20

c6.21

⁶ "In the arrangement of the s'rutis, modern usage is diametrically opposite to the classical one: the latter placing them before the note to which they respectively belong, while the former gives their position after the notes. It is difficult to determine when or by whom the alteration was effected. The arrangement of the frets of the vina and other stringed instruments accords with the modern acceptation of the principle. According to the rule laid down in the classical treatises, the disposition of the notes is reversed in the case of Daravi (lit., wooden—i.e., stringed) instruments, and out of this reversed arrangement, perhaps, the modern theory about the arrangement of the position of the s'rutis has been worked."—"The Musical Scales of the Hindus." S. M. Tagore. Calcutta, 1884.

c6.22

The exact definition of what constituted a s'ruti is difficult to determine, but it is thus vaguely given by the Sangíta Ratnâvali:—"A s'ruti is formed by the smallest intervals of sound, and is perceivable by the ear; it is of twenty-two kinds" (i.e., as shown above); also "every distinct audible sound is a s'ruti; it is a s'ruti because it is to be heard by the ear."

c6.23

The scales are formed from the s'rutis, four s'rutis being allotted to a major tone, three to a minor (which would appear to have been of a pitch somewhat flatter than the tone and sharper than the semitone; doubts, however, exist as to whether the intervals of the s'rutis were equal or not), and two to a semitone.

c6.24

The s'rutis are differently arranged in grâmas, or scales, three in number—viz., Shadja-grâma, Madhyama-grâma, and Gandhâra-grâma.

c6.25

The literal meaning of grâma signifies "a stopping place" or "village." Hence the word came to be used for scale, since the s'rutis are arranged in a scale as mankind in villages.

c6.26

The Shadja-grâma consists of two tetrachords similar to each other, and separated by a major tone—nearly our diatonic major scale.

c6.27

The Madhyama-grâma is formed from the preceding by a transposition of the major tone, between Pa and Dha, and of a minor tone between Dha and Ni—precisely our diatonic major scale.

c6.28

Hence the two grâmas stand in the following relation:—

c6.29

| | Sa | Ri | Ga | Ma | Pa | Dha | Ni | Sa |
| Shadja-grâma | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 1 | 2 | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 1 | 2 |
| Madhyama-grâma | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 1 | 2 | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 1 | 2 | 3 | 4 | 1 | 2 |
| | Sa | Ri | Ga | Ma | Pa | Dha | Ni | Sa |

c6.30

The third grâma is called the Gandhâra-grâma. Its construction is not clearly laid down, and if it ever existed in practice its use has long been discontinued. According to the Dâmódara its construction is only known in Indraloka, the mythical heaven of the god Indra, thus dispensing with the difficulty conveniently.

c6.31

The Sangíta Pârijâta mentions that it merely differs from the other grâmas in that the note Ni will have four s'rutis, and that Sa will consequently have only three.⁷

c6.32

⁷ The disposition of the s'rutis in this case would be preceding their respective notes, otherwise Sa would be shifted one s'ruti, and this is not so, as has been already remarked when mentioning the Pârijâta first.

c6.33

The Sangíta Darpana points out three changes in the scales in forming the Gandhâra-grâma from the Madhyama-grâma.

c6.34

Upon this Mr. Paterson⁸ makes the following remarks:—

c6.35

⁸ "On the Grâmas or Musical Scales of the Hindus."—"Asiatic Researches." Vol. IX. Calcutta, 1807.

c6.36

1st. Gandhâra takes one s'ruti from Rishaba and becomes of three—*i.e.*, by rendering the note Ga flat, the interval between Ri and Ga is reduced a semitone, and that between Ga and Ma becomes a minor tone.

c6.37

2nd. Panchama loses one s'ruti to Gandhâra. I am at a loss to know how this can take place. I rather suspect an error in the text, and would propose to substitute Dha, the sixth note, instead of Gandhâra. The three s'rutis of Panchama make the interval between the fifth and sixth; by losing one, it is reduced to a semitone; but it cannot lose this one to Gandhâra, which is the third note. There are but two methods of reducing this interval to a semitone—one by raising the fifth note, the other by rendering the sixth flat. But here the interval between the fourth and fifth remains unaltered. It must in this case be done by making Dha, the sixth note, flat; or, in the language of Hindu music, by giving one of Panchama's s'rutis to Dhaivata.

c6.38

3rd. Suddhasvâra gives one s'ruti to Nishâda. Here Nishâda is rendered one s'ruti flat. Suddhasvâra is not the name of a note, but is explained to me to be a term applied to a note possessing its full complement of s'rutis. It may, therefore, in this case be applied to Dhaivata; for although it may give one s'ruti to Nishâda, yet it gains one from Panchama, and still retains four complete s'rutis. If these conjectures are admitted, and we compare it with the Madhyama-grâma, to which these changes evidently refer, it will stand thus:—

c6.40

That the Hindus probably by this division of the octave meant nothing more than what I have before supposed may appear from the following table, in which the intervals between each note and the note above it are taken from Mr. Malcolm's series of the octave in the two modes (as given by Mr. Chambers under the article "Scale"). This I have done in order to compare these intervals with the s'ruti of the Hindus, and to show the differences—

c6.41

c6.42

| The difference between | Proportion of the intervals between each note and the note above it. | What they ought to be if the scale was divided into 22 parts, or the whole string into 44. | What they are as stated by the Hindus. | The difference between | Proportion of the intervals between each note and the note above it. | What they ought to be if the scale was divided into 22 parts, or the whole string into 44. | What they are as stated by the Hindus. |
| I & 8/9 | 1/9 | 4 8/9 | 4 | I & 8/9 | 1/9 | 4 8/9 | 4 |
| 8/9 & ⅘ | 4/45 | 3 4/9 | 3 | 8/9 & ⅚ | 1/18 | 2 4/9 | 2 |
| ⅘ & ¾ | 1/20 | 2⅕ | 2 | ⅚ & ¾ | 1/12 | 3⅔ | 3 |
| ¾ & ⅔ | 1/12 | 3⅔ | 4 | ¾ & ⅔ | 1/12 | 3⅔ | 4 |
| ⅔ & ⅗ | 1/15 | 2 14/15 | 3 | ⅔ & ⅝ | 1/24 | 1⅚ | 2 |
| ⅗ & 8/15 | 5/75 | 2⅕ | 4 | ⅝ & 5/9 | 5/72 | 3⅛ | 4 |
| 8/15 & ½ | 1/30 | 1 7/15 | 2 | 5/9 & ½ | 1/18 | 2 4/9 | 3 |

c6.43

In a paper read to the Royal Society,⁹ in 1877, upon the Hindu division of the octave, Mr. Bosanquet shows that the fifths and thirds produced by dividing the octave into twenty-two equal intervals do not deviate very widely from the exact intervals which are the foundation of the diatonic scale, the fifth being only ·07 or very nearly 1/3 of a comma¹⁰ sharp, and the major third ·045 or nearly 1/3 of a comma flat.

c6.44

⁹ "Proceedings of Royal Society." Vol. XXVI., page 372.

c6.45

¹⁰ Comma of 81/80 = ·21506.

c6.46

He also gives the following table in order to show the deviation of the other intervals of the scale from those of just intonation:—

c6.47

SYSTEM OF TWENTY-TWO.

c6.48

| Interval. | Difference of | Units. | Interval. | Exact Interval. |
| Fourth ... ... | Fifth and Octave | 9 | 4·9091 | 4·9805 |
| Major Tone ... | Fourth and Fifth | 4 | 2·1818 | 2·0391 |
| Minor Tone ... | Third and Major Tone | 3 | 1·6363 | 1·8240 |
| Major Semitone | Third and Fourth | 2 | 1·0909 | 1·1174 |
| Minor Third ... | Fifth and Third | 6 | 3·2727 | 3·1564 |
| Minor Semitone | Major Tone and Minor Third | 1 | ·5454 | ·7067 |

c6.49

“In regarding these numbers,” he observes, “we must remember that as far as European musicians are concerned, the deviation from equal temperament is the most important thing in a melodic point of view.

c6.50

“Intervals which deviate widely from equal temperament sound out of tune to the European ear, and as harmony is not employed, the justification which derivation from perfect concords is felt to give in harmony has no opportunity of asserting itself.”

c6.51

This calculation of Mr. Bosanquet’s was made on the assumption that all the s’rutis were equal. That such could not have been in reality the case, or that the employment of the system of twenty-two never entered practically into Indian music, would seem to be from all evidence almost certain.¹¹

c6.52

¹¹ See also “The Twenty-two Musical S’rutis of the Hindus.” S. M. Tagore. Calcutta, 1886.

c6.53

This will be more evident by a reference to the following comparative diagram of the primitive Sanskrit Shadja-grâma and the European diatonic scale, as drawn for the Râjah Sir S. M. Tagore, and published in his work upon the "Musical Scales of the Hindus," from data supplied by the ancient treatises, the measurements being those of a string 90 inches long:—

c6.54

PRIMITIVE SANSKRIT SHADJA-GRÂMA. 90 inches.

c6.55

| 20 | 16 | 9 | 15 | 13⅓ | 10⅔ | 6 |
| 20 | 16 | 9 | 15 | 12 | 12 | 6 |

c6.56

EUROPEAN DIATONIC SCALE.

c6.57

The only difference, it will be seen, is in the fact that the sixth is in the European diatonic scale flatter than the ancient one; so that the ancient Sanskrit sixth had apparently the same ratio, theoretically, as the Pythagorean sixth of the Greeks.

c6.58

This seems probable, for the historian Strabo says that among the Greeks those who regard all Asia as far as India as a country sacred to Dionysius, "attribute to that country the invention of nearly all the science of music."¹²

c6.59

¹² Strabo, book X., iii.

c6.60

But as concerns string measurements by the monochord, the late Mr. A. J. Ellis, F.R.S., in a most exhaustive paper read before the Society of Arts,¹³ notices the above table of the Râjah's and remarks: "These divisions are made on the supposition that the vibrations are inversely as the length of the strings, which all my observations and experiments show is not the case on any practical instrument."

c6.61

¹³ On "The Musical Scales of Various Nations," by A. J. Ellis, F.R.S., F.S.A., &c.—See "Journal of Society of Arts," 27th March, 1885, No. 1,688, Vol. XXXIII.

c6.62

Mr. T. M. Venkatas'esha S'astri, a well-known authority upon theoretical music in Southern India, in a letter to the author, says that the word "s'ruti" appears to have undergone a great change in its meaning, and he inclines to the belief—but on what grounds it is difficult to say—that the suddha septa svâras, as understood at present in Southern India, meaning, as already explained, the seven notes of the scale containing their full number of s'rutis, are as follows:—

c6.64

This tends to account in some measure for the preference given to the scale Mâyamâlavagaula, described in another chapter.

c6.65

Mr. A. J. Ellis has given a table in the above-mentioned paper¹⁴ of the comparative differences in the chromatic scales formed by the division of a string into parts, according to both the ancient directions and to what the Râjah Sir S. M. Tagore states¹⁵ to be the modern Bengali use. The figures are in cents—that is to say, the hundredth parts of an equal semitone. He says:—

c6.66

¹⁴ See previous note.

c6.67

¹⁵ “The Musical Scales of the Hindus.” Calcutta, 1884.

c6.68

I give the number of degrees (s'rutis) and the calculation of their value on both plans, old and new, with the names of the nineteen Indian notes, assuming that the pitch varies inversely as the length of the string, as shown by the position of F and the octave, and that any errors thus arising have been corrected by ear.

c6.69

INDIAN CHROMATIC SCALES.

c6.70

| Degrees | I | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 |
| Notes | C | D♭♭ | D♭ | D | E♭♭ | E♭ | E | E♯ | F | — | — |
| Old | 0 | 51 | 102 | 153 | 204 | 264⅔ | 325⅓ | 386 | 442 | 498 | 549 |
| New | 0 | 49 | 99 | 151 | 204 | 259 | 316 | 374 | 435 | 498 | 543 |
| Degrees | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 |
| Notes | F♯ | F♯♯ | G | A♭♭ | A♭ | — | A | B♭♭ | B♭ | B | B♯ |
| Old | 600 | 651 | 702 | 753 | 804 | 855 | 906 | 966⅔ | 1027⅓ | 1088 | 1144 |
| New | 589 | 637 | 685 | 736 | 787 | 841 | 896 | 952 | 1011 | 1070 | 1135 |

c6.71

The only values agreeing in each are C, D, F, while new E♭ is the just minor third—a mere accident. The nine degrees from C to F vary from 49 to 63 cents, and then there is a sudden break, after which the thirteen degrees from F to the octave vary from 45 to 65 cents. This is the first intelligible presentment of the Indian scale which I have been able to effect. It will be seen that C, D♭, D, E♭, E, F, F♯, G, A♭, A, B♭, B are represented pretty well by our equally tempered notes; but that the seven intermediate notes—D♭♭, E♭♭, E♯, F♯♯, A♭♭, B♭♭, B♯—could only be tempered in the quarter-tone system used in Syria. Hence in the usual transcription these seven notes are identified with some of the others. . . . These comparisons necessarily injure the original character of the music and give it a harmonisable appearance which is entirely foreign to Indian music.

c6.72

Mr. Ellis also examined a vina from Southern India, now in the South Kensington Museum, with the following result:—

c6.73

| Cents | 0 | 97 | 195 | 312 | 397 | 515 | 596 | 692 | 782 | 883 | 997 | 1092 | 1207 |
| Notes | G | A♭ | A | B♭ | B | c | d♭ | d | e♭♭ | e | f | f♯ | g |

c6.74

He remarks: “This is very close indeed to our scale of twelve semitones, and may be taken for it”; thus proving scientifically what has been found by experiment to be invariably the case with vinas used by Karnâtik musicians (vide page 31—footnote).

c6.75

In a communication received from Mr. A. J. Hipkins, it is not acknowledged that there is any connection between the old Indian grâmas of 22 quarter-tones, or s'rutis, and the modern European scales, with major and minor tones and semitones, are founded upon the knowledge and practice of harmony. The Indian scale intervals ought to be understood as they are explained by native writers—namely, as a tone,¹⁶ a ¾-tone, and a ½-tone, composed of 4, 3, and 2 s'rutis respectively. With this conception of intervals, and it must be borne in mind the ¾-tone is still approved of in the East, a division of the octave into 24 equal quarter-tones becomes impossible. For as it was essential to secure an approximately perfect fourth with 9 s'rutis, and a fifth with 13, the division of the octave by 22 was the only one available. The error in the fourth of 9 equal s'rutis of a 22 division is no more than ⅕-comma, in melody scarcely noticeable, but the error in a 21 or in a 23 division could not have been easily tolerated. The s'rutis thus being a little wider than exactly equal quarter-tones, 54 6/11 cents instead of 50, the Indian grâmas in most intervals come near to those of our just intonation scales, but this resemblance is accidental, as the foundation is different. It must, however, not be forgotten that this scale of probably equal s'rutis was theoretical, and has long since been superseded by another and more practical system, and that equal measurements of a string will not represent accurately this old Hindu conception.

c6.76

¹⁶ Mr. Hipkins states: "The Pythagorean tone—i.e., the distance by which a perfect fifth overlaps a perfect fourth—which is here meant, is 2·04 equal semitones, and the greater tone that from the harmonic seventh completes the octave is 2·31. The four s'rutis amount to 2·18."

c6.77

The comparison of the 22 s'ruti scale with the European one of just intonation is as follows. The figures are equal semitones to two places of decimals (or the late Mr. A. J. Ellis's cents).

c6.78

| Indian | 2·18 | 3·82 | 4·91 | 7·09 | 9·27 | 10·91 | 12·00 |
| European | 2·04 | 3·86 | 4·98 | 7·02 | 8·84 (or 9·06, with comma added to make 5th to D.) | 10·88 | 12·00 |

c6.79

The following table, kindly sent me by the late Mr. Ellis, shows the results obtained from a most minute and careful examination made by him and by Mr. A. J. Hipkins of a beautiful old vina, in perfect condition, now in my possession. This instrument is between two and three hundred years old, and is from the collection in the Tanjore Palace. The results, as will be seen, tend to prove that the frets were purposely arranged for something like equal temperament. We see, therefore, that in India much the same results have been independently arrived at by native musicians as have been attained by subsequent science in Europe.

c6.80

| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
| Frets. | Millimetres from nut. | Sounding lengths of string. | Pitch or numbers of double vibrations. Nov. 18, 1887. | Cents in the interval from the lowest note. | Intervals in cents from note to note. | Intervals in cents calculated from column 3. | Cents in the interval from the lowest note from column 3. |
| Nut | 0 | 555 | 210·7 | 0 | 0 | 0 | 0 |
| 1 | 31 | 524 | 222·8 | 97 | 97 | 99 | 99 |
| 2 | 59 | 496 | 236·6 | 201 | 104 | 96 | 195 |
| 3 | 85 | 470 | 246·2 | 270 | 69 | 93 | 288 |
| 4 | 110 | 445 | 260·9 | 370 | 100 | 94 | 382 |
| 5 | 134 | 421 | 275·0 | 461 | 91 | 96 | 478 |
| 6 | 156 | 399 | 289·2 | 548 | 87 | 93 | 571 |
| 7 | 179 | 376 | 310·5 | 671 | 123 | 104 | 675 |
| 8 | 200 | 355 | 325·3 | 752 | 81 | 99 | 774 |
| 9 | 219 | 336 | 346·3 | 860 | 108 | 95 | 869 |
| 10 | 236 | 319 | 365·6 | 954 | 94 | 90 | 959 |
| 11 | 253 | 302 | 385·8 | 1047 | 93 | 95 | 1054 |
| 12 | 269 | 286 | 421·5 | 1200 | 153 | 94 | 1148 |
| 13 | 286 | 269 | 447·2 | 1303 | 103 | 106 | 1254 |
| 14 | 301 | 254 | 472·5 | 1398 | 95 | 99 | 1353 |
| 15 | 314 | 241 | 494·1 | 1476 | 77 | 91 | 1444 |
| 16 | 327 | 228 | 523·6 | 1576 | 100 | 99 | 1548 |
| 17 | 341 | 214 | 553·3 | 1672 | 96 | 107 | 1650 |
| 18 | 352 | 203 | 583·4 | 1763 | 91 | 91 | 1741 |
| 19 | 363 | 192 | | | | 97 | 1838 |
| 20 | 373 | 182 | The rest of the notes were too high for my forks. | Auxiliary Table. | The great uncertainty in these intervals renders any hypothesis as to the intended nature of the scale impossible. But col.7, in conjunction with the auxiliary table in col. 5, shows that the mechanical rule of making the successive lengths as 18:19, or 17:18, or 16:17 hence that the maker intended the frets to be arranged for equal temperament. | 102 | 1990 |
| 21 | 383 | 172 | | | | 98 | 3028 |
| 22 | 392 | 163 | | Cents. | Intervals. | 93 | 2121 |
| 23 | 401 | 154 | | 92 | 128:135 | 99 | 2220 |
| 24 | 410 | 145 | The height of the string above the frets varied from 1 to 5 millimetres. | or larger limma | | 104 | 2324 |
| bridge | 553 | 0 | | 94 | 18:19 | | |
| | | | | 99 | 17:18 | | |
| | | | | 100 | 84:89 | | |
| | | | | 105 | 16:17 | | |
| | | | | 112 | 15:16 | | |
| | | | | 151 | 11:12 | | |
| | | | | 155 | 32:35 | | |
| | | | | | Applied in column 7. | | |
| 100 times these numbers give the cents in equal temperament. | to the beginning of the bridge, but the real sounding length is as 555 millimetres. | From these lengths are calculated columns 7 and 8 on the hypothesis that pitch varies inversely as the sounding lengths | A cent is the hundredth part of an equal semitone. | | | | Compared with column 5, this shows the erroneous nature of the hypothesis that pitch varies as the sounding length on fretted instruments. |

c6.81

The system by which the octave is divided into twelve semitones is clearly hinted at in the Sangíta Darpana, which states that there are seven pure tones (Suddha or Prakrita), which appear to refer to the intervals composing the diatonic major scale; and twelve impure tones (Vikrita), by which we may conclude that the chromatic scale is implied.¹⁷

c6.82

¹⁷ For other information about Prakrita and Vikrita notes, see "The Musical Scales of the Hindus." S. M. Tagore. Calcutta, 1884.

c6.83

The s'rutis are arranged in their different svâras, or intervals of the scale, according to the “murchanas.” Of what these murchanas really consisted is very doubtful.

c6.84

Sir William Jones states that the murchanas, of which there are seven in each grâma, appear to be no more than “seven pieces of diapason multiplied by three, according to the difference of pitch in the compass of the three octaves.”

c6.85

This view seems to be that taken by Kóhala, an ancient musician, from whose pen fragments of a treatise in Sanskrit are still remaining.¹⁸

c6.86

¹⁸ The Sangíta Darpana gives a totally different meaning to murchana, describing the murchanas as the permutations produced by a method somewhat like change ringing; the number of murchanas being the continued product of the number of notes employed. Hence, from a grâma of seven tones we get 5,040 different murchanas. The method of producing these permutations is called “ Kundameru” by native musicians.

c6.87

Mr. Paterson, on the other hand, conjectures that they are the intervals of each grâma, and arranges them in the following classification:—

c6.88

Shadja-grâma .......... Sa to Ri 1st 2nd ,, ,, Ga ,, 3rd 1st Tetrachord. ,, ,, Ma ,, 4th Pa ,, Dha ,, 2nd ,, ,, Ni ,, 3rd 2nd Tetrachord. ,, ,, Sa ,, 4th Sa ,, Sa ,, Octave Madhyama-grâma .... Sa ,, Ri 2nd ,, ,, Ga Greater third ,, ,, Ma 4th ,, ,, Pa 5th ,, ,, Dha Greater sixth ,, ,, Ni 7th ,, ,, Sa Octave Gandhâra-grâma .... Sa to Ri 2nd ,, ,, Ga Minor third ,, ,, Ma 4th ,, ,, Pa 5th ,, ,, Dha Minor sixth ,, ,, Ni 7th¹⁹ ,, ,, Sa 8th

c6.89

¹⁹ Whether major or minor is not stated by Mr. Paterson.

c6.90

The seven intervals of each scale are arranged in what are called râgas.

c6.91

Sir W. Jones employs the term râga as synonymous with mode.

c6.92

Mode and râga are, however, perfectly distinct from each other—Mode being termed *thât*, and not râga or râgini. Now *thât* consists in determining the relative intervals between several sounds, which constitute an octave with respect to each other. A râga is formed from these, in its composition employing the whole or less number of the intervals of the *thât*, and with a peculiar melodic style of its own; in fact, a melody type formed upon a mode. This, however, is more fully explained in another chapter.

c6.93

In no two Sanskrit works do we find that the râgas agree either as to their names or their notation; the modes or scales of these different râgas are not given, and, in most cases, it is therefore only a matter of conjecture as to how they were performed. In almost these works a somewhat similar classification of the râgas and râginis has been adopted. There are six principal râgas personified as demigods, each of which has a certain number of râginis (personified as the wives of the râgas)—sometimes five and sometimes six—appended to it.

c6.94

The following classification is that of Hanumân²⁰:—

c6.95

²⁰ See "Sangíta Sâra Sángrahâ," a collection of various Sanskrit authorities, edited by S. M. Tagore. Calcutta, 1875.

c6.96

I.—Bhairavâ . . . . . Dha, ni, sa, ga, ma, dha. Madhyamadi ....Ma, pa, dha, ni, sa, ri, ga, ma. Bhâiravi ....(1.) Ma, pa, dha, ni, sa, ri, ga, ma. (1.) Ascending. (2.) Dha, ni, sa, ga, ma, pa. (2.) Descending. Vangâli . . . . . Sa, ga, pa, ni, sa, Ma, pa, dha, ni, sa, ri, ga, ma. Varâti . . . . . Sa, ri, ga, ma, pa, dha, ni, sa. Syindavi . . . . . Sa, ri, ga, ma, pa, dha, ni, sa, Sa, ga, ma, pa, dha, ni, sa.

c6.97

II.—Mâlavakûsika.....Sa, ri, ga, ma, pa, dha, ni, sa. Tódi . . . . . Ma, pa, dha, ni, sa, ri, ga, ma. Sa, ri, ga, ma, pa, dha, ni, sa. Kambâvati ........Dha, ni, sa, ri, ga, ma, dha. Gauri . . . . . Sa, ga, ma, dha, ni, sa. Gunâkeri . . . . . Dha, ni, sa, ga, ma, pa, ni. Sa, ga, ma, pa, ni, sa. Kakobhâ . . . . . Dha, ni, sa, ri, na, ma, pa, dha.

c6.98

III.—Hindola . . . . . Sa, ga, ma, pa, ni, sa, ni, pa, ma, ga, sa. Velâveli . . . . . Dha, ni, sa, ri, ga, ma, pa, dha. Râmakeri . . . . . Sa, ri, ga, ma, pa, dha, ni, sa. Sa, ni, dha, ma, ga, ri, sa. Déshaks'ya ........Ga, ma, pa, dha, ni, sa, ga. Palamângeri ....Pa, dha, ni, sa, ri, ga, ma, pa.

c6.99

Lalíta . . . . . Sa, ri, ga, ma, pa, dha, ni, sa, sa. } . . . . . Dha, ni, sa, ga, ma, dha, dha. IV.—Dîpaka ........... Sa, ri, ga, ma, pa, dha, ni, sa. Kadâra . . . . . Ni, sa, ga, ma, pa, ni, ni. Kanada . . . . . Ni, sa, ri, ga, ma, pa, dha, ni, ni. Deshi . . . . . Ri, ga, ma, dha, ni, sa, ri. Kaumodi . . . . . Dha, ni, sa, ri, ga, ma, pa, dha. Nâtika . . . . . Sa, ri, ga, ma, pa, dha, ni, sa. V.—S'ri-Râga ......... Ni, ga, ma, pa, dha, ni, sa, ri. Vasantha . . . . . Sa, ri, ga, ma, pa, dha, ni, sa. Mâlava . . . . . Ni, sa, ga, ma, dha, ni. Mâlava-s'ri ......... Sa, ri, ga, ma, pa, dha, ni, sa. Dunâsri . . . . . Sa, ga, ma, pa, dha, ni, sa. Asâveri . . . . . Dha, ni, sa, ma, pa, dha. } . . . . . Ma, dha, ni, sa, ri, ga, ma. VI.—Mégharâga ..... Dha, ni, sa, ri, ga, ma, pa, dha. Gauri . . . . . Dha, ni, ri, ga, ma, dha. Déshakâri ........... Sa, ri, ga, ma, pa, dha, ni, sa. Bhupâli . . . . . Sa, ri, ga, ma, pa, dha, ni, sa. } . . . . . Sa, ga, ma, pa, sa. } Gaurjeri . . . . . Ri, ga, ma, pa, dha, ni, sa, ri. Dakhâ . . . . . Sa, ri, ga, ma, pa, dha, ni, sa, sa.

c6.100

The Sangíta Naráyana shows that thirty-six “modes” or râgas are in general use, and the rest very rarely applied to practice. These modes are shown by Sir William Jones in his essay on the musical modes of the Hindus, and will be found in many respects similar to those described in the Sangíta Darpana, and shown above. Thirty-six modes from the Râga Vívodha, and thirty-six from the work of Mirza Khan, have also been described by the same eminent scholar, and need not, therefore, be reproduced here.

c6.101

The rhythm of the early music seems to have been very complicated, and the most exact directions as to the value of notes and the division into “tâlas” or rhythmical periods are given. These, again, vary in different authors.

c6.102

For example, in order to estimate the relative time value of successive notes, the sage Anginayya gives the following poetical directions:—

c6.103

Take one hundred petals of the lotus flower, place them then one upon the other; and when pierced with a needle, the time in which the point passes through a single petal is called one second; eight such seconds are called one lavâ; eight lavâs one koshtâ; eight koshtâs one nimishâ; eight nimishâs one kalâ; four kalâs one anudruthâ; two anudruthâs one druthâ; two druthâs one lâgu; two lâgus one guru; three lâgus one plutha; four lâgus one kakupathâ.

c6.104

Of these, later on in his work, he employs the following, calling druthâ a half matra, and this he takes as the limit.

c6.105

Hence we can deduce the following table:—

c6.106

c6.107

The common and triple time here implied is striking, and in some measure tends to prove that the employment of triple time is not of such comparatively recent introduction as some writers endeavour to show.

c6.108

Formed upon the above basis, several hundreds of different tâlas or measures—many of them extremely complicated—are given.²¹

c6.109

²¹ Extracts from Anginayya's work have been given by the Râjah Sir S. M. Tagore, in "Sangíta Sâra Sángrahâ," which see for further information upon this subject, p. 207 and following pages.

c6.110

The rhythm of some of these ancient tâlas is still employed in practice, although the complicated system of signature is no longer in use. The following table, taken from the work of Anginayya, comprises some of these ancient tâlas, together with their value in modern notation.

c6.111

The sign of C or rest, although its value is not definitely stated by Anginayya, can be—judging from the performances of modern Mridang or Tabla players—correctly taken to be of the same value as the note immediately preceding; or else it may be employed as a "dot," when placed after a note, to lengthen its value by one-half.

c6.112

Captain Willard, in his "Treatise upon the Music of Hindustan," has given another and very complete list of these ancient tâlas, differing in many ways from the following; but, unfortunately, he does not state from what authority he gathered them.

c6.113

To judge by the very complicated nature of many of these tâlas, and the fact that they vary widely in almost all the authorities, it seems hardly likely that they were ever in very common use; but they are, nevertheless, interesting as showing the great variety of rhythm that can be produced by such simple means as beating the two hands together—the earliest kind of rhythmical accompaniment; and some slight idea of the peculiarities of modern Indian drum playing can also be gleaned from this table:—

c6.114

There exist, as will be explained later on, at present in Karnâtik music seventy-two modes or scales, all formed from seven of the twelve semitones in the octave differently disposed upon a tonic of equal pitch. If we are to understand that the râgas described by Sir W. Jones were simply such scales or modes, the only way in which it is possible to reconcile the theory to the present system is to imagine that for ascending the scale they employed a perfect fourth (or, as a Karnâtik musician would say, for Arohâna Suddha-madhyama), and for descending the augmented fourth (Práti-madhyama), or possibly the reverse, like the Chinese practice at the present day.

c6.115

The râgas are, however, differently distributed in the different works, and the modern custom differs widely from the directions of the Sanskrit.

c6.116

The notation given in all the ancient treatises is very similar to that at present in use, letters only being employed to express the notes. The following, a fac-simile of the most ancient form of notation, is from the work of Sóma, and has been thus rendered into the European notation by Sir William Jones, who remarks:—

c6.117

I have noted Sóma's air in the major mode²² of A, or Sa, which, from its gaiety and brilliancy, well expresses the general hilarity of the song; but the sentiment, often under pain even in a season of delights, from the remembrance of pleasures no longer attainable, would require in our music a change to the minor mode; and the air might be disposed of in the form of a rondo ending with the second line, or even with the third, where the sense is equally full, if it should be thought proper to express by another modulation that imitative melody which the poet has manifestly attempted: the measure is very rapid, and the air should be gay or quick in exact proportion to it:—

c6.118

²² It is rather difficult to understand why Sir William Jones employs this key instead of that of C, which seems to have been the most natural that would have suggested itself.

c6.119

यमनः ॥ ॥ म०रि ग० म० ध म०धप० म ग म०स० म ध म ध नि प प दि म से॥ ध मूं गं मं रि मं नि सॅ० वि प० म ध ॥ म ध ग वि प० प दि म मं गं नि प ध० म प ॥ म ध स० म ग ग रि स ध स नि प ध प म ध० म गं म० ध प० म प० म गं म० गं म० ग ग रि म ध

c6.120

TRANSLATION OF THE ABOVE.²³

c6.121

²³ This translation must, of course, be more or less hypothetical; and as it is so entirely different in character and style to all modern Indian music, and airs heard now in India which are said to be very ancient, its correctness appears to be very doubtful. A comparison with the examples quoted later will show how widely it differs.

Linked texts (6)

Saṅgītaratnākara — Śārṅgadeva, Sanskrit
Nāṭyaśāstra — Bharata, Sanskrit
Mahābhārata — Vyāsa, Sanskrit
Saṅgītanārāyaṇa — Gajapati Vīrasrī Nārāyaṇadeva 1675, Sanskrit
Saṅgītapārijāta — Ahobilasūri 1600, Sanskrit
Rāgavibodha — Somanātha 1584, Sanskrit