The Music of Hindostan c11: Chapter VI. Rāga

The Music of Hindostan c11

Language: English

Chapter VI. Rāga

c11.1

I'll take my pipe and try
The Phrygian melody;
Which he that hears
Lets through his ears
A madness to distemper all the brain;
Then I another pipe will take
And Doric music make
To civilize with graver notes our wits again.
RANDOLPH.

c11.2

The *Rāgs* of Hindostan vary¹ so much with the locality (those of the Carnatic less so) that there seemed to be no better way of showing them than to place in parallel columns the *Rāg* as given by different authorities (see the Table of *Rāgs*, opposite). In the first column appear sixty-three *Rāgs* as given me by Ashreka Ganga Rām, whose teacher was Balwant Rao of Gwālior, a state which has a high reputation for music. Ashreka seemed to be particularly sure of his knowledge; and from no one else was I fortunate enough to get any certain information about the *Rāgs* which employ quarter-tones (see Nos. 2, 26, 33, 35, 36, 40, 41, 49 in App. II). The second column is taken from the *Bālsangīt-bodh* (Juvenile Singing Tutor) of T. B. Sahasrabuddhe, of the Poona Gayan Samāj (singing school). The third column is extracted from a list of eighty-three *Rāgs* in B. A. Pingle's *Indian Music* (Bombay, 1898);

c11.3

¹ All the same, a *Rāg* is one thing, not several. It is, as will be seen, so sharply characterized that although in this or that respect one *Rāg* trenches on the domain of another, yet it never does so in all respects, and therefore a change of one of its constituent notes does not obscure its individuality; just as a man through all sorts of not easily definable varieties of vocabulary or pronunciation, or even grammar, may yet be speaking pure English. This is what an Indian believes, although he will tell you in the same breath that the *Rāgs* are to be found in their purity only in the school of music to which he belongs. The true *Rāg*, of which all those with which the traveller meets are manifestations, is stored up in the heavens—or in the bosom of the most educated musician, whoever he may be.

c11.4

the mordents show where he places the *mūrchhaṇas* (grace notes), and the *mūrchhaṇa* comes as a rule on the *aṃśa*; as these *mūrchhaṇas* are generally a Fourth or a Fifth apart the list confirms the view that the *aṃśa* was placed at the limits of the tetrachords. The list in the fourth column was given me, in response to a request for the thirty commonest and most popular *Rāgs*, by Upendrakisor Ray of Calcutta; and of these he named Nos. 2, 10, 11, 14, 32, and 60 as quite the commonest of all. ‘This list’, he writes, ‘was obtained by analysing a collection of more than one thousand songs. But its value is somewhat impaired by the fact that most of these songs were popular, not classical. Writers of popular songs are not always good musicians, so their opinion as to the *rāgas* has to be accepted with some reserve; the “popular” Behāg, for instance, is C D E F G A B♭ C, popular Lalit is C D♭ E F G A♭ B C, popular Multān is C D E♭ E♮ F G A♭ A♮ B♭ B♮ C’ (as compared with Nos. 60, 42, and 41). With regard to the descriptive epithets, which are interesting to compare, and which on the whole agree with those of the second column, he says—‘I give them with a good deal of diffidence; in fact I am not very sure about the majority of them. The quality of restraint is a predominating feature of Indian music, and this precludes any free exhibition of the emotions. Our gaiety and sadness often merge into each other: our most impassioned passages fail to stir anybody up. There is any amount of enjoyment of the deepest kind, but it is undemonstrative and reposeful; it leads not to action but to abstraction.’

c11.5

In these lists the notes shown by a crotchet head are those which are comparatively lightly touched; which are (1) used in ascent only or descent only, as the B♮ and B♭ in No. 11, or (2) are more or less optional, as the F and F♯ in No. 63. In this *Rāg* the usage is E F E and G F♯ G, never, as a rule, E F♯ E or G F G; while in the passage from E to G, or the reverse, either an F or an F♯ may be used, but F♯ more commonly is used. This is very much the same as the mediaeval *Musica ficta*. And (3) some *Rāgs* are on their way from being transilient to being complete; though all the notes are potentially there, some of them are as a rule jumped. This is the case in No. 60, where the D and the A are occasionally heard, but never dwelt upon. It did not seem worth while to go into the details of each *Rāg* in this matter, even if sufficient and accurate knowledge had been forthcoming; but these three principles throw a good deal of light on the conception of *Rāg*.

c11.6

The ♭ or ♯ over a note means the lowering or raising by some kind of *śruti*, see p. 117.

c11.7

The mark [=] under a note shows the Predominant (*amśa*).

c11.8

*Graha* [Gr] and *Nyāsa* [Ny] are respectively the notes on which a melody is supposed to begin or end; this rule seems not always to be obeyed.¹ The final note has nothing like the importance it has in harmonized music.

c11.9

¹ And this makes it probable that it is not the true explanation. There is another, quite different. The *Graha* is the note which an individual singer chooses for his *Sa*. Thus if a bass voice chooses C, a tenor would take F, and a soprano B♭. The *Graha* is therefore the ‘clef’. And this seems to be in complete agreement with the statement in Bharata that the *Graha* is identical with the *amśa*. It seems probable that *graha* originally meant the upper and *nyāsa* the lower extremity of the tetrachord (*aṅga*), and that these meanings have been forgotten.

c11.10

*Śuddha* [Shud] means ‘pure’; *Chāyālaga* [Chāyā] means ‘transitional’ (from *Chāyā* shadow); *Sankirṇa* [Sank] means ‘mixed’; and *vakra* [vakr] is literally ‘crooked’ and means that the notes are taken not in direct order from *Sa* to *Sa*, but that the *Rāg* has certain melodic figures as characteristics.

c11.11

The twenty-four hours are divided into eight watches, beginning at 6 a.m. The Hindus attach much importance to, and are wonderfully unanimous as to, the ascription of a *Rāg* to a particular hour of the day or season of the year,² although ‘advanced’ musicians may be found among them who say it has no meaning. A musical distinction, however, may be extracted from the *Rāgs* themselves, independently of the melodies sung in them or of the words to which these are set. Taking column 4, an analysis shows that if we may divide the *Rāgs* into ‘sad’ and ‘merry’, the ‘sad’ have an average of three flats to an average of two flats in those which are ‘merry’. Again, if we divide them into morning and evening *Rāgs*, the morning have an *amśa* ranging about G, the evening about E. And the same thing, though in a less marked degree, may be deduced independently from the second column.

c11.12

² The first thing, sometimes the only thing, that an Indian who is not very familiar with the science of music can tell you about a particular *Rāg* which is being performed is that it is a morning, or that it is an evening *Rāg*—though this perhaps does not prove more than that the *Rāgs* are as a matter of fact confined in their performance to the appropriate time of day.

c11.13

So that there is a real musical basis for this ascription of time of day and of *ēthos*. The ascription of time of year is possibly more fanciful.

c11.14

But the truer explanation of this deep-seated feeling is more likely to lie in the history of *Rāga*, as far as we may hope to know it. What is antecedently probable, and, indeed, is evident from the data in the first two chapters of this book, is that songs were sung long before *Rāga* as such was formulated. The Hindu account is that there were four sources of *Rāg*: (1) local tribal song, (2) poetical creations, (3) the devotional songs of the mystics, and (4) the labours of the scientific musicians. The principal ground for this belief is, no doubt, that all four causes are in full operation in India to-day, and are evolving the *Rāgs* of the next thousand years; a secondary reason for thinking it to be the case is to be found in the names of *Rāgs* such as (1) *Kānaḍā*, the Carnatic; (2) *Hindol*, a swing; (3) *Jogi*, a mystic; (4) *Sārang*, from Sāraṅgadeva, a musician of the thirteenth century. The most important of these sources, because the most constant and widespread and unconscious, is the first—tribal song. Unfortunately, we do not yet possess nearly sufficient data to enable us to trace the actual formation of *Rāga* from this source, for India; but we may get an idea of it from our own Folk-song.

c11.15

The fourteen versions of a song variously called ‘The seeds of love’, ‘The sprig of thyme’, &c., given in the annexed example, have been taken from the volumes of the *English Folk-song Society's Journal*, and one of them from *Sharp's Folk-songs of Somerset*. When the English counties have been thoroughly explored, there will prove to be not fourteen, but very likely forty or four hundred of such versions. The first seven of these are, however, enough to show how a melody may take various shapes and yet centre in a common mode; and also how by a slight shift of the centre of gravity, brought about by the battle of its several phrases for supremacy, the melody slips from one mode to another. At the end of each melody is set the scale in which it appears to lie, with the strong (*balin*) and weak (*alpa*) notes marked, and after these the name of the *Rāg* to which in principle, though not, of course, in detail, this scale corresponds.

c11.16

Supposing a variety of scales to have been brought into use by tribal song from various sources, next comes the poet and weaves

c11.17

FOLKSONG VARIANTS

c11.18

Border Melody

c11.19

ii. 288 (2) Deskar. 3.

c11.20

The garden well perfumed

c11.21

i. 86 Hamirkaliān 59

c11.22

I sowed my seeds of love

c11.23

i. 88

c11.24

I sowed the seeds of love

c11.25

ii. 24 (3) Kamodkaliān 55

c11.26

I sowed the seeds of love

c11.27

ii. 23

c11.28

Sovay Sovay

c11.29

F.S.S. Kamodkaliān

c11.30

So the soldier was seen alone

c11.31

ii. 24 (1) Pancham 8

c11.32

I once had a sprig of thyme

c11.33

ii. 288 (1)

c11.34

The seeds of love

c11.35

i. 209 (3)

c11.36

The seeds of love

c11.37

i. 209 (2)

c11.38

The bailiff's daughter of Islington

c11.39

i. 209 (1) Plagal major mode

c11.40

Come you lads and lasses gay

c11.41

i. 210 (1)

c11.42

My old thyme it is all gone

c11.43

i. 210 (2)

c11.44

I sowed the seeds of love

c11.45

ii. 24 (2)

c11.46

N.B "Sovay Sovay", the sixth of these variants, has in the "Folksongs of Somerset" a piquant mixture of two-time and three-time, which is neglected here for the purpose of comparison (cp. p 134)

c11.47

his mythology round them. He speaks of six original Rāgs which were *śuddha* (pure). The six are variously given. The six of the *Rāgavibodha*, as authoritative as any, are, though under other names, Nos. 13, 26, 37, 47, 49, 51. Of these, modes were taken, sometimes called *mūrchana* (rise and fall), sometimes *jāti* (species). The next step was to cross the *jātis* and produce hybrids ¹ (*Chāyālaga*, *Saṅkīrṇa*, *Sālanka*, *Miśra*, &c.). The poet speaks also of the six original Rāgs having wives (*bhārya*) and sons (*putra*), and it seems hardly worth while to translate these into the plain prose of ‘mode’ and ‘hybrid’. The exact pedigree of the *putras* seems to be irrecoverable: their names are all there (though no two lists give them alike), but not a trace of form or feature.

c11.48

¹ I have been unable, however, to find any explanation of the way in which this was done.

c11.49

Meanwhile the philosopher has been at work upon the psychology of the music, and, without any special theories as to the emotional value or the ‘mental effect’ of this or that note or combination, determines on general grounds the *ēthos* of the *Rāg*, its appropriateness to occasion or to circumstance, or its effect upon the body, or its affinities with the various affections of the spirit or aspirations of the soul. These views linger in some of the names of the *Rāgs*: *Dīn-ka-(pūria)*, ‘of the day’, *Shām-(kaliān)*, ‘evening’, *Basant*, ‘spring’, *(Gaurī-)mano-hāri*, ‘pleasing to the mind’; or in the legends about them, as of *Dīpak* (= *Māravā*, No. 49), which enflamed not only the mind, but the body of the singer so that even the Jumna could not put out the fire. Descriptive names of this kind were commoner and more elaborate in old days. The famous *Mahā-vyut-patti* (Sanskrit-Tibetan vocabulary, seventh century or later) has a list of sixty *Rāgs*, the translations of a few of which are—‘with a voice like the thunder-cloud’, ‘like the trumpeting of the elephant’, ‘like a sparrow’, ‘like the serpent king’, ‘like Indra’, or again, ‘smooth’, ‘refined’, ‘making fully to comprehend’, ‘not puffed up’, ‘delighting every organ’, ‘having the chief voice in all sorts of sounds’, and so forth.

c11.50

Last of all, when by these three processes the number of *Rāgs* had reached hundreds or thousands, came the mathematically-minded musician who tabulated the scales as deviations from a normal scale (*grāma*), calculated the departures in quarter-tones (*śruti*), and rung the 13678 changes (*prastāra*) on the seven notes (svara); reduced the 108 traditional ways of counting time by syllable (akṣara), i. e. a development from the longs and shorts of poetry, to thirty-five ways of counting it by time-unit (mātrā,¹ lit. syllable) which are rhythms (tāla); and provided both Rāg and Tāl with an intricate memoria technica to which not every one possesses the key.² This kind of musician is commoner in the Carnatic than in Hindostan, where theory has reared a less imposing edifice.

c11.51

¹ It is noteworthy that both mātrā and akṣara are terms of prosody. Cp. the early reckoning of musical time in Europe, Oxford History of Music, vol. i, p. 168 seq.

c11.52

² See Chinnaswami Mudaliar, Oriental Music, pp. 16 and 25.

c11.53

The aesthetic effect of the amśa is important. Whatever may have been its origin—and there is every probability that it was the tonic in the same sense as the Greek mesē—its present function is to define the tessitura. With harmonized melody we are not accustomed to think much about this—until perhaps we have to consider what sort of song we shall be able to sing when we are tired.³ But in unharmonized music very much, more than anything else, depends on the general level in which the song lies. A song with a high tessitura has quite a different character from a song with a low one. It was in this very point that in Greece the ‘plaintive’ modes (e.g. the Mixolydic) were distinguished from the ‘soft and convivial’ (e.g. the Lydic), according to Dr. Macran (Harmonics of Aristoxenus, 1902, pp. 39–81). An example of the plaintive kind of Greek song with a high tessitura (and the Mixolydic fifth) may be seen in Pachtikos (Aismata Hellēnica, Athens, 1905, No. 133), a funeral dirge circling about B, B♭, and A, and descending to a tonic E. There is no doubt also of the low position of the amśa in the convivial Kaliāns (Nos. 50–59), or of its high position in the serious Bhairavī (32) and Bhairav (37). Similarly, the theme of Mozart’s Clavier Sonata in A major owes perhaps some of its alertness to its high, and that of the Choral Symphony something of its restraint to its middle tessitura; but we can hardly be said to have erected this into a principle of our art.

c11.54

³ Aristotle spoke of modes with a low mesē as ‘old men’s scales’.

c11.55

An English-speaking Indian habitually chooses the word ‘tune’ to translate Rāg. And the fact is he does not attach importance to the tune of a song. A song may be sung in many ways—the more the better, as a rule—so long as it is within the welldefined limits of the *Rāg*; on the other hand, the least departure from the prescribed notes surprises us, not always pleasurably. This has one advantage, at any rate, that it is unnecessary to record Indian tunes in notation. The names of the *Rāgs* to which the *Gita-Govinda* (eleventh century) was sung have been preserved, and even in the absence of a notation it is open to any musician of to-day of sufficient skill to sing them as expressively as Jayadeva himself.

c11.56

On the other hand, the absence of definite melody precludes concerted music. Several rehearsals are necessary in order to get a dozen people to sing a hymn in unison, and to overcome the lust for ‘grace’ and *contrappunto alla mente*. When some years ago *Bande Mātaram* was being sung there were probably six different versions at half a dozen different street corners of Calcutta, of which the following, taken down from the lips of a blind man in Dalhousie Square, and sung at the top of his voice without causing a single passer by even to turn his head—so little did the mere tune convey anything—is one:

c11.57

Ex. 250. Maestoso. Ban-de mā-ta-ram su-ja-lām su-fa-lām Ma la - ya-ja-shi-ta- - lam Sho-shya shyā-mala khe - trām Ban - de mā-ta-ram Ko - ti ko - ti kan-the Ka-la ka-la ni - nā - de Di sap - ta ko - ti Bhu-je shob-he ka-ra - bā - le

c11.58

I bow before thee, O mother, who art washed clean by rivers, by raindrops, by oceans. By the help of thy life-giving waters and the free breezes born in the Malaya mountains thy crops yield and thy fruit-trees bear. A thousand thousand nurtured on thy lap are making prayer to the Almighty Father and to thee. O mother Durga, in your ten hands you have as many weapons always ready to protect your children. But because we have lost our devotion therefore you have no sympathy for us.

c11.59

We have seen nothing in the facts of *Rāga* so far with which to meet a possible suggestion of the frivolous, that if you were to put the phonogram of an Indian melody on to the cylinder fore part behind and hear the tune backwards it would sound just as well. One answer to this would be that in that case all the ascending passages would come out as descending and vice versa, and that many (not all) *Rāgs* have a distinct form for ascent and descent. And this in two ways, or at least as two manifestations of the same instinct. The pentatonic (*oḍava*) and hexatonic (*ṣāḍava*) modes are often pure (as in Nos. 1 and 24), but sometimes mixed (as in Nos. 4 and 11): in the pure mode the ‘leap’ is taken absolutely; in the mixed mode it is taken with a subsidiary grace-note, or in a sort of tentative way. In descent the ‘leap’ is bridged by a passing note:

c11.60

Ex. 251.

c11.61

in ascent, if it is not taken absolutely, the voice climbs up to it, as it were

c11.62

Ex. 252.

c11.63

If we refer back to Ex. 226 (*Śardula-vikriḍita*) for a moment we shall find an example of this. The scale is

c11.64

Ex. 253.

c11.65

and the weak note is B; the gap A-C has to be negotiated, and this is done thus:

c11.66

Ex. 254.

c11.67

*Behāg* (No. 60), for instance, is divided into tetrachord (B C-E) and pentachord (E F G-B) with the gaps occasionally filled. We will take the tetrachord

c11.68

Ex. 255.

c11.69

The B usually, but sometimes the E, is given as *aṃśa*; they are at any rate about equally prominent notes. Next in importance is the C, and least important the D. In effecting the passage from B to E such phrases will occur as while the descent from E to B may be by leap, but will generally be by step:

c11.70

Ex. 256.

c11.71

Ex. 257.

c11.72

In most Rāgs, even if heptatonic (sampūrṇa), there is still a feeling that some notes are strong (balin) and others weak (alpa), and these weak spots tend to be negotiated in this ‘crooked’ (vakra) way. And if the Rāgs are actually transilient (varja) they are apt to be vakra in ascent, and to have the gap filled in descent—rather than the other way. So that to turn ascent (ārohaṇa) into descent (avārohaṇa), or vice versa, would be completely to falsify the Rāg. Other answers to the ‘frivolous’ person could be made on the grounds of Time and of Form (see Chapters VIII and XI).

c11.73

Similarly other Rāgs have typical phrases:

c11.74

58. Kaliān.

c11.75

Ex. 258.

c11.76

54. Yaman-kaliān.

c11.77

Ex. 259.

c11.78

55. Kamod-kaliān.

c11.79

Ex. 260.

c11.80

59. Hamir-kaliān.

c11.81

Ex. 261.

c11.82

57. Shām-kaliān is not vakra.

c11.83

But it must not be thought that every melody necessarily contains the typical phrase of the Rāg. In India, as in Europe, the rules are ‘the obedient humble servants’ of the composer; he obeys them, of course, but in the spirit, not the letter, just as the best sculptor’s work approximates most nearly to, but never exactly coincides with, the nine-unit or ten-unit or other system which the case may require; or, as we ourselves should maintain, while a given chord has its typical resolution, yet that music is chiefly to be prized which in obeying this ‘rule’ manages still to say something new that, when closely examined, really disobeys it.

c11.84

What surprises us most in this music is its predominantly conjunct motion.¹ Leaps of all sorts are taken, including those which do not appeal so much to us—the tritone and the augmented second. But if small they still have the character of being a, if not the, next door note; if large, they seem to be merely a means of getting quickly to some other contrasting register. They do not seem to be used for the pure pleasure of the interval itself, as in many examples we might quote from our music:

c11.85

¹ So little is there any idea of the possibility of disjunct motion that there seems to be no distinctive name for it. The leap of a Third is called Abhyucchraya, ‘pressed together’; of a Fourth, Prastāra, ‘extension’; and of a Fifth, hiṃkāra (making the sound Hiṃ or Om, the syllable of the sacred name; the application is obscure). A singing exercise in mixed leaps is also called vārida (?). The old words nigraha, ‘subduing’ (sc. suppressing the intermediate note), and praveśa, ‘binding together’ (sc. the neighbouring notes)—disjunct and conjunct motion—seem now to be forgotten.

c11.86

Ex. 262. Handel. Beethoven. Wagner. R. Strauss. &c.

c11.87

Contrast with these such a melody as the following:

c11.88

Brahma Samāj hymns. Dhrūpad. Jhil² Surphakt.

c11.89

² A Jhil is a Rāg without grace-notes.

c11.90

Ex. 263. Pra-ti-di-na ta-ba gā-thā gā-ba ā-mi shū-ma-dhur Tu-mi de-ha mo-re ka-thā tu-mi de-ha mo-re sūr Tu-mi ja-di thā-ka ma-ne bi-ka-cha ka-ma-lā-sa-ne

c11.98

Upendrakisor Ray’s translation.—Day by day shall I sing thy sweet song. Give thou me the words, give thou me the tune. If thou abidest in my heart in the full-blown lotus seed,¹ if thou makest my soul overflow with thy love, if thou hearest my song, seated in my heart, if thy noble and beautiful eyes lend me the nectar of their glance, if thou on my sorrows placest thy hand pitifully, if thou from my happiness chasest away pride—give thou me the words, give thou me the tune.

c11.99

¹ God abides in the innermost recesses of the heart in the form of a lotus seed. The lotus is the type of purity; it may be offered, even when not fresh (as other flowers may not) unless previously offered to another. Hence ‘lotus lips’, ‘lotus feet’, &c.

c11.100

It is true this is a Folk-song, to which words have been added by Rabindranath Tagore, and a *Jhil* (i. e. without grace-notes); but it will do as well as one in a formal *Rāg* to show the general preference for step over leap.

c11.101

Another thing which completely baffles our ears is the way even the conjunct motion is taken. The following hymn of the Brahma Samāj:

c11.102

Brahma Samāj hymns. Rāg Sohani; Tāl Surphakta. ‘Prathama āditāla.’ ♩ = 130. Music and words by RABINDRANATH TAGORE.

c11.103

Astaï

c11.104

Ex. 264. (4 + 2 + 4)

c11.105

Antara.

c11.107

Sanchari.

c11.109

Ābhog.

c11.111

N.B.—Crotchets or minims would have been more appropriate; quavers are used in order to suggest the rhythm better.

c11.112

Upendrakisor Ray’s translation.—Thy power is from all time; from all time is thy supreme radiance in the skies. Thine is the first word. Thy joy lives in each new year afresh in the heart. In the firmament of thy mind glisten the sun and the moon and the stars. The wave of life vibrates in the atmosphere. Thou art the first poet; the master of poems art thou. Thy deep-voiced utterances find voice in praise and prayer which ascends from all the world.

c11.113

sung as I heard it on their anniversary (January 25) with a choir of twelve voices supported, in unison, by a small organ and two violins, with drum, was extremely impressive. In particular the phrase marked (a) at the end of each section seemed to give it that sort of sublimity which we recognize in those chords that smote on Samuel Butler’s ear on his entrance into Erewhon.

c11.114

Both these peculiarities, a preference for conjunct over disjunct motion, and unusual sequences in conjunct motion, are made possible and almost demanded by the absence of harmony. They are demanded as a means of contrast. Motion through the notes of a chord, when there is no idea of forming chords at all, does not give enough contrast; E and G say more or less the same thing as C. And again, when the ear has become accustomed to the sound of F♯ G A♭—a note with an upward and downward leading note—to pass direct from F♯ to A♭ is a very striking divagation from the normal, and gets a high value from the consciousness of the norm behind it.

c11.115

The claim is occasionally made that this music has harmony, or a ‘sort of’ harmony, in it. Such feeling for harmony as is involved in the construction of a scale at all, especially of such an elaborate and balanced scale, or in the appreciation of the greater sonority of C G C over the simple C in the drone, it has. But of harmony, in the sense either (1) of the sequent intervals in melody being chosen for their consonance, or (2) of two notes being sounded simultaneously for the sake of the consonance, there is none; there is less even than among the North American Indians whose music,¹ as far as at present investigated, is at a very early stage. And it seems impossible that there should be any. Hindu music, though it has enormously refined the detail, has not advanced beyond the principles of Greek music. This may not seem an inducement to deny them harmony to those who still think that the Greeks had harmony. But at least this is true that harmony in the sense of the Organum and of Dufay did not become possible till the Greek modes had long lost any feeling of the *mesē* (*aṃśa*) and of the functions (*dunameis*) of the other notes.² And the first thing that harmony would do, if now applied even tentatively to Indian music, as some advocate, would be to get rid of that feeling and those functions, and with them of the grace-notes and all that makes *Rāg* worth having. As the *Rāg* now is, its notes are like the pieces on the chess-board; harmony, by investing them all equally with powers of its own, would make them like the pawns. Hence the serious menace to Indian music of the harmonium, which has penetrated already to the remotest parts of India.³ It dominates the theatre, and desolates the hearth; and before long it will, if it does not already, desecrate the temple. Besides its deadening effect on a living art, it falsifies it by being out of tune with itself. This is a grave defect, though its gravity can be exaggerated; it could also be lessened by a revised tuning. A worse fault is that it is a borrowed instrument, constructed originally to minister to the less noble kind of music of other lands. It has taken a century to invent and perfect the pianoforte; if she must have the fatal facility of a keyed instrument, India could well spare a century or two for inventing something that should do justice to her music.

c11.116

¹ There is one curious point of contact, the more curious that any direct connexion is unthinkable: the Chippewas draw a picture of a tune, by the help of which another man can sing it, and the Hindus set store by their *Ragmālas*, sets of paintings of gods or men affected by the passion which the particular *Rāg* expresses. See also Ex. 457.

c11.117

² That in virtue of which they took their part in the mode as *vādi*, *samvādi*, &c.

c11.118

³ I was present for an hour at a concert in Trivandrum at which this appalling instrument never ceased, and I found in the Salt Range (Northern Panjab) a ‘Teacher of Song and Harmonium’.

c11.119

The question as to what foreigners are to make of this music is still more difficult. To add harmony to it is to kill it; and yet it is impossible for Europe to understand melody except in terms of harmony. Harmony has been successfully added to songs whose musical thought is nearer our own in Stanford's *Irish Songs* or Bourgault Ducoudray's *Chansons de la Basse Bretagne*. But are the *Chansons de la Grèce et de l'Orient* a complete success? And is the problem quite so simple as Herr Polak¹ supposes; though he exhibits the curiosities of Indian music, does he convey the art? It is quite possible to select, as Lady Wilson² and others have done, those songs which have definite cadences—and there are a few—or even to invent cadences for those which have none, to give them a frankly European dress, and to suggest the kind of mood in which a foreigner listens to Indian music. That is a definite thing well worth doing; but it does not bring one any nearer to an understanding of the indigenous art. The best authorities on Bengālī say that it is untranslatable, and one has heard the same of Sanskrit; and there seems to be no way of understanding the music except to set to work and learn its language.

c11.120

¹ *Die Harmonisirung indischer, türkischer und japanischer Melodien*, Leipzig, 1905.

c11.121

² *Five Indian Songs*, Paterson & Sons.

c11.122

We now examine a *Rāg* in practice, and for this purpose take one of the commonest—*Yamankaliān*. Of the three instances here given:

c11.123

*Yamankaliān.* (1)

c11.124

Ex. 265.

c11.129

the first is a *Svarāvarta*, an exercise sung to the *Svaras* (Sa, Ri, Ga, &c.) used as a teacher's model of the *Rāg*; the other two are simple songs for children. We learn that the E is made prominent by the melody circling about it, or stopping there; that the passage E–G or G–E may be taken as a leap, but that if the movement is slow an F♯ is inserted as a grace-note (not an F♮); on the other hand, an F between two E's is always made natural. Of the following songs, ‘Dukhera beshe’ is a hymn of the Brahma Samāj; I do not know why the mode is called *mishra* (mixed). ‘He mora debatā’ is a poem. Both are *Khyāls*, and by Rabindranath Tagore. ‘Hādi e illah’ is a *Thumri* from Benares. It was sung to me by a Mohammedan in Bombay with very genuine feeling. The words are Hindostani. The *Rāg Kalyānī* does not differ much in this particular melody from *Imankaliān*, except that occasion might arise, but does not, for F♯'s.

c11.130

Brahma Samāj. V. 40. Mishra Imankaliān. Jhampa. Tonic D. (3+2.) ♩= 125.

c11.131

Ex. 266.

c11.132

Du-khe-ra be-she e-se-chha ba-le to-mā-re nā-hi da-ri-ba he.

c11.134

Ye-khā-ne bya-kha to-mā-re se-khā ni-bi-ra ka-ri kha-ri-ba he Ān-dhā-re mūkh ḍhā-ki - le swā-mī to-mā-re ta-bu chi-ni-ba ā-nu

c11.137

Ma-ra-ṇa rū-pe ā-si-le pra-bhū cha-ra-ṇa kha-ri ma-ri-ba he

c11.139

Ye man ka-re dāo-na de-khā to-mā-re nā-hi da-ri-ba he

c11.141

Na-ya-ne ā-ji jha-bi-ccha ja-la ja-ru-ka ja-la na-ya-ne he

c11.143

ba-ji-ccha bu-ke bā-jūk-ta-ba ka-thi-na bal-ri bau-kha-ne he

c11.145

Tu-mi se ā-chha ba-kshe kha-re ba-da-na tā-hā jā-nāk mo-re

c11.147

Chā-ba-ṇā ki-cchā ka-ba-na ka-khā cha-hi-yā ra-ba ba-da-ne he

c11.149

Na-ya-ne tā-ji jha-ri-cche ja-la jha-ru-ka ja-la na-ya-ne he.

c11.150

(The translation of ‘Dukhera beshe’, by Upendrakisor Ray.)—Though thou comest in the garb of war, yet shall I not fear thee; rather shall I hold thee tight to the place where it hurts. Though thou hidest thy face in darkness, yet shall I know thee. If thou comest as Death I shall clasp thy feet and die. In whatever manner thou showest thyself to me, I shall not fear thee. Tears fall from my eyes to-day: let the tears fall from my eyes. My hurt aches in thy hard embrace: let it ache. Let that pain tell me thou hast me held to thy breast. Nothing shall I want, nothing shall I say, but simply keep my gaze fixed on thy face. Tears fall from my eyes to-day: let the tears fall from my eyes.

c11.151

Imankaliān (Tonic F). Ektāla. ♩ = 70.

c11.152

Astaī.

c11.154

Ex. 267. He mo-r(a) de - ba-tā . . . bha-ry-a e de-ha prān Ki am-ri-ta tu-mi . . . chā-ha ka-ri-bā-re pān . .

c11.156

Antara.

c11.158

he mo-ra de-ba-tā . . . Ā-mā-ra na-ya-ne

c11.160

to-mā-ra bī-sa chhā-bi De-khy-a la-i-te . . . .

c11.162

sadh jai ta-ba ka-bi Ā-mā-ra mu-gdha . . .

c11.164

sra-va-ne ni-ra-be ra-hi Su-ni-ya la-i-te chā-ha

c11.166

ā-pa-nā-ri gān . . He mo-r(a) de-ba-tā . . .

c11.167

Sanchari.

c11.169

Ā-mā-r(a) chi-tte to-mā-ra sri-shti khā-ni

c11.171

Ra-chy-a tu-li-chhe bi-chi-tra e-ka ba-ni

c11.172

Abhog.

c11.174

Tā-ri sā-the pra-bhu mi-li-ya to-mār prī-ti

c11.176

Jā-gā-ye tu-li-chhe . . . ā-mār sa-kal-(a) gī-ti

c11.178

Ā-pa-nā-re tu-mi de-khi-ccha ma-dhu-ra ra-se Ā - mā-ra mā - jhā - re . . . . ni - je - re ka - ry - ā dān . . .

c11.181

He mo-r(a) de - ba - tā. . .

c11.182

(The poet's, and composer's, own translation.) — What divine drink wouldst thou have, my God, from this overflowing cup of my life! My poet, is it thy delight to see thy creation through my eyes, and to stand at the portals of my enchanted ears silently to listen to thine own eternal harmony? Thy world is weaving words in my mind, and thy joy is adding music to them. Thou givest thyself to me in love, and then feelest thine own entire sweetness in me. (*Gītāñjali.* No. 65.)

c11.183

Kalyānī. d=40. *p Molto adagio.*

c11.184

EX. 268.

c11.185

Hā - - - - di e il - lah Hā - - - - di e il - lah Sa - - hi-ba Sa - - - hin se ta - - - ra-ka Ri - ma ra-hī - - - ma a - - kī ma hā - li-man hā - - - - li-man Pā - ka bī - - - ne a - je pā - ka la - ti - - - fa pā - - - ka - jā - - *pp* *Sotto voce.* *D.C.* ta par - de po - - - sa dā . . . na - bī - na.

c11.186

Mighty is God the Lord. O Lord God, from thee cometh salvation. Mighty, Almighty, All-knowing, All-loving! Day by day thy holy presence is nearer to us: in the darkness thou art near us.

c11.187

The melodies ‘Dukhera beshe’ and ‘Hādi e illah’ do not lose so much as some others in notation: if they are sung with a little faith and goodwill some impression at least will be conveyed of their sincerity and quiet fervour. ‘He mora debatā’ is perhaps more difficult; the melody will strike a European at first as running up hill only in order to run down again. There is a great temptation to regard the close as in A minor and to supply a G harmony to the fourth bar; and these temptations must be firmly resisted. The fourth bar rises to the tonic (not the seventh of G) through the ‘strong’ notes A and E (not C); and an F pedal must be held on in order to get the plaintive character of the close

c11.188

Ex. 269. - tā - -

c11.189

(not at all unlike that of the Sixth Ecclesiastical Tone) which is on the ‘strong’ note and therefore in a sense a ‘full’ close, but which is not the tonic, and therefore points away, as it were, to something beyond it. Also the rhythm of the ninth bar is difficult. The eighth and ninth bars are taken as one sweep of melody and in strict time, but with the prosody of the words crossing this time in the ninth bar, as if it were

c11.190

Ex. 270. Sadh jai ta-ba ka . bi Ā-mā-ra &c.

c11.191

It offers just that sort of pleasure which we get in lines of blank verse

c11.192

—Dúncan is in his gráve;
After lífe’s fítful féver he sleéps wéll;
Treáson has done his wórst ; nor steél, nor poíson,
Málice doméstic, fóreign lévy, nóthing
Can toúch him fúrther.—

c11.193

where the assertiveness of the individual words fights against and yields to the compelling calm of the metre.

c11.194

The next thing is to consider the list of *Rāgs* as a whole. We notice that they are re-entrant: that is, they return upon themselves after making the full circle as, by an accepted adjustment, our keys do, but as our modes do not.¹ The order here adopted is to show this; it is not the order of the native treatises. Any point might have been, and the simplest form of our major scale, is taken as the point of departure. Nos. 1–5 are transilient, and modes of each other. Nos. 6–8 transitional. No. 9, the major scale. Nos. 10–14 substitute B♭ for B♮. Nos. 15–21 flatten the E as well. Nos. 22–5, the B♭ and E♭ are fully established. No. 26, anomalous. Nos. 27, 28, A flattened. Nos. 29–31, though they seem to show a good deal of licence, are really transitional between the two-flat and the four-flat scale.¹ Nos. 32, 33, D♭ (of some sort) established. At this point the European circle of modes stops, because the next modal shift would flatten the G. This mode with a flat Fifth will be discussed later under *Sāman* chant. The *Rāgs* proceed, after reaching four flats, by sharpening the leading note to the tonic and then the leading note to the Fourth as well.² No. 34, B♮. Nos. 35–8, B♮, E♮. No. 39, a tentative sharpening of F, as the leading note to the Fifth. Nos. 40, 41, the F sharpened instead of the E♭. Nos. 42–8, all three leading notes sharpened. Nos. 49–52, after a little remaining hesitation over the A♭ and D♭ we get back in Nos. 52–63, through the *Kaliāns*, to the place we started from. These sixty-three are only the best known out of some hundreds of *Rāgs*; but to have shown them all would only have taken the principle of re-entrance to another place of decimals, as it were.

c11.195

¹ These refuse to admit the Locrian, so that there is a gap between the Lydian at one end and the Phrygian at the other.

c11.196

¹ There is the same sort of hesitation about our Aeolian mode which frequently shows a Dorian sixth and a Phrygian second.

c11.197

² What has hitherto been described as successive flattening might, of course, have been taken in reverse order and described as sharpening. The two are the same thing looked at from different points of view.

c11.198

The *Rāgs* have been arranged in this way not merely to obtain a symmetrical system, but also to bring out an essential point. They fall, in the first instance, into nine or ten groups, of which the centres are perhaps:

c11.199

1. *Bhūpaliān.*
9. *Shankarābharaṇa.*
14. *Jhinjoti.*
23. *Kāfi.*
28. *Āsāvarī.*

c11.200

32. *Bhairavi.*
37. *Bhairav.*
44. *Basant.*
50. *Pūriakaliān.*
53. *Kaliān.*

c11.201

But secondly, each group consists of an average of seven *Rāgs* (out of a total of sixty-three; out of a total of 500, an average of fifty or sixty). And a *Rāg* gets its special flavour not so much from its being just what it is, as from its not being something else, closely allied to it, which is present all the time in the musician's consciousness. Any one can hear the difference between, say, *Jhinjoti* and *Bhairav*; but the more of a musician a man is the more he can hear and enjoy the subtler distinction between, for instance, *Jhinjoti* and *Desh*, or the still subtler between *Jhinjoti* and *Tilang*. It might appear that as a given *Rāg* is taken so differently (see Nos. 27–9, 46) in different parts of the country these distinctions do not much matter; but it must be remembered that the distinctions are accurately adhered to in any one district, or by any particular school; and that it is of no use to test the songs of Bengal, for instance, by the system of Gujarāt. And some of these characteristics are so strikingly beautiful as not easily to be missed, or to be forgotten, even by a foreigner: the strong legato from F♯ to E in *Imankaliān*, for instance, or the pentatonic suggestion of *Behāg*, or the plaintive alternation of the two B’s in *Khamāj*, or the of the two E’s in *Kāfi*, or the old-world sound of the scale of *Kānaḍā*.

c11.202

In No. 59, *Hamīrkaliān*, the sharp Sixth (A♯) appears. It is sounded very likely in others, but I could get no certain information. Similarly, the eight *Rāgs* with exceptional intonation given in Appendix II is probably not a complete list; they are merely those of which the true intonation has been preserved in one part of the country, Gwālior; at Calcutta, for instance, I heard of *Āsāvarī* (No. 28) as being of exceptional tuning, but of none of these; and Mr. Clements gives other instances from the Deccan.

c11.203

Express mention is seldom if ever made in the books of the Drone. It is taken for granted of course; though, to those who have never heard it, it is not at all an obvious concomitant of melody.

c11.204

In fact, at first it is a little confusing, even irritating; but one gets to discount it as readily as one does the hammering of the pianoforte keys or the scraping of the violin bow. It is there not only because without it, especially with quarter-tone scales, the singer would feel, as a Benares musician said, ‘like a ship without a rudder,’ but in order to consolidate the melody. Not that a song cannot be sung without it; a ryot returning from his work, or a mother nursing her baby, does not have first to go and fetch a tambura; but, as the boy said when asked why he had godfathers and godmother at his baptism—‘because, I suppose, it makes it a sweller thing’. The drone may be a drum, carefully tuned, or two drums, in unison or at the octave, or the drone strings of the *vīṇā*; but is as a rule, and with the best singers, a *tambura*, a large *vīṇā*-shaped instrument, with one gourd (sound-board) instead of two, and with open strings—two or three Sa's and one Pa. The drone strings of the *vīṇā* and *satār* or of the *surbahar*, a large *satār* common in Bengal, often have amongst them one tuned to Ma, and there is, curiously, no objection to the Pa and the Ma sounding together. The effect of this may be heard in the ‘wilderness’ part of Granville Bantock’s *Omar Khayyám*.

c11.205

Turkomani melody. Chorus. Tenors and Basses. GRANVILLE BANTOCK.

c11.206

Ex. 271.

c11.207

But they avoid putting either of these notes on the drone if the *Rāg* which is being sung omits it. Thus, for *Bibhās* they would tune the Ma to Ga, in spite of the ‘thick’ effect of a major Third so low in the compass; for *Malkos*, in which there is no Pa and the Ga is an E, and neither this nor E would be available on the drone (which must be as consonant as possible), the Pa string would be tuned in unison with the Ma string; and it is only for the European that the re-inforced Fourth of the scale upsets the tonality. The strings of the tambura are long and sonorous, and the Fifth overtone (Ga) comes out very clearly from the fundamental whatever the other strings may be; and this still further adds to the dissonance of the Fourth. It is curious, too, that the B♮ and D♭ of *Bhairav*, and the B♭ and D♭ of *Toḍi* should be sung, and evidently with special pleasure, against a C drone¹; it recalls the cadence of the Organum:

c11.208

¹ Cp. the last chord of the *Matthew Passion*.

c11.209

Ex. 272.

c11.210

(except that it is the C which is stationary, and the D comes down to it); and the dissonance upon a strong beat of Descant:

c11.211

Ex. 273.

c11.212

(except that it is much more common than that apparently was). Thus the drone is of considerable use too in forcing a strong appoggiatura, which is the peculiar, perhaps the only, melodic point of the snake-charmer’s pipe:

c11.213

Ex. 274.

c11.214

About the simultaneous presence of Ma and Pa in the drone a word more may be said. It is interesting that the ‘consonance’ of the whole tone which appeals to the Hindus was one which the Greeks recognized too,² and we see that it is the logical outcome of quintal harmony. That the appreciation of the whole tone is historically prior to that of the major Third may also be gathered from the form of the pentatonic scale which may be pronounced typical. This is not *Bhūpaliān* [Figure: Ex. 275.]

c11.215

² See Gevaert, *La Musique de l’antiquité*, vol. i, p. 98, and Westphal’s edition of Plutarch’s *De Musica*, p. 46.

c11.216

but *Sārang* [Figure: Ex. 276.]

c11.217

or *Arabhī* [Figure: Ex. 277.]

c11.218

and these entail no harmony except that of the Fourth and its derivatives, among which the whole Tone. Indeed the construction of the *Sa-grāma* and *Ma-grāma* (see pp. 109 foll.) point to a stage of Indian music when the Pythagorean Third (derived from the Fifth) was being exchanged for the harmonic Third (heard directly) which has left its impress in this and other ways on the national art. There is of course no hesitation about this interval (the harmonic major Third) nowadays. A singer when preluding often touches the common chord of the key note, and will do this even when the Third of the *Rāg* he is about to sing is minor.

c11.219

What, then, are the main differences, if we consider melody alone without rhythm for the moment, between this music and ours? First of all the song is cast in one definite mood throughout. This mood calls up associations, unnameable, but yet distinct, of similar song heard under similar circumstance; so much so that to tell an Indian what *Rāg* a singer chose on such and such an occasion is to tell him a good deal about the song. We have no similar classification. We have plenty of ‘mood’ in our music, but we use it to articulate the balance of the song, not to differentiate one song from another.

c11.220

Secondly, in the chosen ‘mood’ (*Rāg*) the notes stand out from each other as clearly as the faces of our friends do to the mind’s eye. One of our systems describes the Third of the scale as the ‘restful’ note, the Fourth as ‘awe-inspiring’, another note as ‘desolate’, another as ‘enterprising’. But in this music any given

c11.221

¹ Miss Gilchrist also (*Folk-song Society Journal* for 1911) makes out the original Scotch pentatonic to be [Figure:]

c11.222

On the other hand, Frances Densmore (*Chippewa Music*, Bureau of American Ethnology, 1910, p. 9) shows the typical North American Indian scale to be what is here called Bhūpaliān, and next to that Miss Gilchrist’s form. (These two are Helmholtz’s ‘Fourth’ and ‘Second’ form of the pentatonic.)

c11.223

note of the scale may be any or all of these according to the ‘mood’ in which it occurs. Its character depends not on its pitch relation to a tonic only, though it does so depend partly, and on its relation to the tessitura of the song, which we also feel to be part of its essence (though the, to us, more important distinctions of harmony largely cut across this), but also on the extent to which it is *ex officio* a substantive or a passing note. The whole scale is made up of a hierarchy of notes; so that the passage from G to A♭, let us say, is not merely an exaltation of pitch, nor merely an exchange of the perfect consonance C–G for the imperfect C–A♭, but a passage from a stable to an unstable note of the scale (for A♭ may under certain circumstances become A♮) and from a permanent to a temporary (for the A♭ may be altogether omitted, or omitted in certain conditions), or from a note of universal application to one which requires special treatment (for the A♭ may be used in descent, but not in ascent).

c11.224

An instance, *Rāg Kedāra*,

c11.225

Ex. 278.

c11.226

as used in the following melodies, slightly different from the version given in the list of *Rāgs*. The A is lightly touched, the E still more so. The F is predominant (*aṃśa*), the F♯ is used in *Musica ficta*, i. e. between two G’s. In the first melody the B♭, in the second the B♮, is adopted. In both the E has a special treatment; it is used only as a pendant to the F. The scale is seen to be a filled up *Sārang*,

Linked texts (2)

Gītagovinda — Jayadeva, Sanskrit
Rāgavibodha — Somanātha 1584, Sanskrit