The Music of Hindostan c17: Chapter XII. Melody
The Music of Hindostan c17
- Language
- English
Chapter XII. Melody
c17.2
He will put a girdle round the earth in forty minutes and forget to mention the fact, but he may weep at the breaking of a pencil point. And as his poems come he knows not whence, so he is content to let them fly he knows not whither. You ask for yesterday's masterpiece, and it is not to be found. ‘Did I write that,’ he asks with Shakespearian negligence, ‘and was it really good? Never mind, I'll do you another.’
c17.3
*The Perse Playbooks.*
c17.4
AND now if some one should assert, and we were inclined to humour him, that the Hindus, having worked at the laws of melody for at least 3,000 years, must have something to teach us, what could we say in his support?
c17.5
1. First that melody, both time and tune, is in a mode; that is, that it must be homogeneous, and that nothing can be allowed in it which would destroy that homogeneity. We also recognize this in our tunes, though it would not in every case be easy to say how or why. We realize that a melody is good in proportion as everything in it is in keeping; we at least feel that something is wrong if we patch two melodies together, thus:
c17.6
BEETHOVEN. Ex. 443.
c17.7
MOZART.
c17.8
Here the second half seems to defeat rather than to supplement the first, because the descending thirds, though they balance, do not develop the ascending, and the repeated quaver motion is monotonous. Or, if we reverse the process, the one note we do not desire after the first section is the high E♭; but we get it three times, and the climax consequently drags. Also the sixth bar going over exactly the same ground as the third at half the pace forms an anticlimax.
c17.9
MOZART. Ex. 444. BEETHOVEN.
c17.10
The homogeneity of a melody is based differently with them and with us. The musical sentence is in both cases made up of salient and abeyant notes, just as the verbal sentence is made up of nouns and verbs in prominence, and the other parts of speech in subordination. But with them the salient notes are fixed by long association; with us they are made such by the momentary impulse of the harmony or counterpoint. The distinction is not unlike that between verse and prose as the chosen vehicle of a people’s literature; poetry fills a form which is brought to it, and the words assert their value by fighting against it, though they yield to it; prose creates its own form in the process of uttering itself, and this form emerges from the conflict and balance of phrases rather than of words. In the same sort of way individual notes have in modal music their value as filling out with detail a form already supplied, while for harmonized music points of structure, such as melodic figures, have value as creating and articulating form as the music proceeds. In the first it is the note, in the second the cluster of notes that is to be true to ‘mode’.
c17.11
Still, phrases after all are made of notes; and just as the notes of Hindu melodies must be ‘in’ an assigned mode (of tune or time), so must the notes of ours, if they are to be strong, have a firm diatonic and rhythmic basis. Chromatic notes easily become sentimental—become ‘weak forms of strong things’—when they are unduly dwelt upon and distract attention from the big words of the sentence (as in the march of the Persian soldiers in the *Fall of Babylon*), not when they are there only to make intelligible the big words which without them would be too startling (as in ‘O du mein holder Abendstern’).¹ Diatonic is, of course, an expansive word. The diatonic scale is not the same to one generation as to the next, any more than modes are stereotyped for every time and place. But inside the accepted diatonic the notes must move as a compact whole, or the melody become heterogeneous.
c17.12
¹ But we may fancy how, when later on these startling words had become commonplaces to him, Wagner may have smiled contemptuously at the greasy effect of the chromatics, as perhaps the mature Browning may have regretted the false note struck by the first word in the pretty jingle of ‘Menace our heart ere we master his own’.
c17.13
Again, they must have a firm tread along some high road of rhythm whatever flowers they may stay to gather by the wayside or whatever excursions they may make to points of interest. Rhythmic strength lies in subordinating the attractive figures to the steady march of the whole, not in pursuing them for their own sweet sake (as in some of the more dreamy episodes of Mendelssohn’s Fugues), nor in hunting them to death (in the various forms of Rosalia).
c17.14
These principles, which we recognize as the foundations of our melody, we find more naïvely formulated in the systems of *Rāg* and *Tāl*, much as we read in Plato simple statements of the eternal problems of ethics and politics. These systems lay it down that chromatic alterations can only take place on the weak (*alpa*) notes of the scale, that is, in unessential places; and that, though any amount of cross-rhythm is permissible, it must justify itself by resolution at a particular point (the *sam*).
c17.15
2. Melody proceeds by step (conjunct) or by leap (disjunct). In Indian melodies the disjunct intervals amount to 10 per cent. and the conjunct to 90 per cent. We will deal first with the 10 per cent.
c17.16
The surprising thing in Indian melodies is their frequent use of intervals which appear difficult to us. We consider, for instance, such intervals as:
c17.17
Augmented Second. Diminished Third. Augmented Fourth. Ex. 445.
c17.18
to be unvocal, but they employ them quite commonly. What makes intervals vocal is the fact that the note to which the leap is made is familiar. For us, it is made familiar chiefly by its forming some intelligible harmony with the note which is quitted. We can realize this in the phrase at the end of Brahms’s *Mainacht*:
c17.19
Ex. 446. Und die ein - sa - me Thrä - - - ne bebt mir heis - ser, heis - ser die Wang’ . . . . her - ab
c17.20
The leap ‘die Wang’’ is peculiarly dissonant; it consists of two diatonic semitones, which are greater than a whole tone by 20 cents; but it is made intelligible for us because it is a passage from the Neapolitan Sixth to the dominant chord, which is a perfectly familiar sequence. The other leaps are all justified to our ear as harmony notes.
c17.21
But if we look at an Indian melody, one in *Bhairavi*, for instance:
c17.22
Ex. 447. Bhairavi.
c17.23
of which the sixth of the scale (A♭) is *aṃśa*, with E♭ for its consonant, we see that the leaps are mainly to those notes, or round them, but not anywhere else. The tune swings between these two, generally in conjunct motion but with an occasional leap to one or other of them, and comes to rest between whiles on the tonic. Intervals as constituent parts of chords hardly ever occur, because chords form, for an Indian, no intelligible relationship.
c17.24
The explanation of the augmented Second and diminished Third is that they are felt as conjunct intervals, not disjunct. The augmented Second is familiar to us in this sense in the ‘harmonic’ minor scale:
c17.25
Ex. 448. Neh-met hin mei-nen Leib, Neh-met hin mein Blut . . . auf dass ihr mein . . . . . . . . ge-denkt.
c17.26
The B♮ has here a melodic appropriateness after the A♭ which the harmony immediately seizes upon and justifies; the harmony by itself (on double basses and drum) forms but a weak link. In Indian melodies the augmented second occurs also when the note between is transilient, that is to say, omitted not merely on that occasion but as a regular thing. The following instance in Sohani:
c17.27
Sohani.
c17.28
Ex. 449.
c17.29
consists almost entirely of conjunct motion. The G is transilient, and the F♯ optionally F♮. (This is the Poona form; elsewhere the A♭ is also A♮. Cp. Ex. 264.) It contains therefore, as conjunct intervals, one diminished Third (F♯–A♭) and two augmented Seconds (D♭–E and A♭–B); there is also a disjunct interval, the diminished Third, B–D♭. The *amśa* is E, the consonant B, and the melody swings between these two, making a pause on or otherwise emphasizing the tonic, C, on occasion. Hindus have said to me more than once that they like Wagner best of our music, and one sees perhaps in these two melodies a point of contact.
c17.30
The augmented Fourth and diminished Fifth, which we avoid, are common with them. Although any sequence may be justified by its context, yet we consider on the whole the first interval slightly vulgar, and the second a little mawkish as melodic progressions, perhaps because they emphasize unduly the notes of a rather trite harmony. Where harmony is non-existent there can be no such feeling, and they are freely used in the following melody between the *amśa*, D♭, and the Fifth of the scale, G:
c17.31
Gaula.
c17.32
Ex. 450.
c17.34
No less surprising than the presence of intervals which seem to us unvocal is the comparative absence of what we should call vocal intervals—the Thirds, Perfect Fourth and Fifth, and the Sixths. All these occur, but again are limited, as a rule, to leaps which approach a salient note. Such a passage as:
c17.35
SCHUBERT.
c17.36
Ex. 451. Das Wan-dern ist des Mül - lers Lust, das Wan - dern.
c17.37
would be very unlikely to occur in their music. There would be no way of hitting the diminished Fifth (‘ist des’) unless the D were the *aṃśa*; and if it were, it would be impossible that so important a note should occur only once, and at such an insignificant point of the phrase. The intervals at ‘Müllers Lust’ would be practicable if C# were *aṃśa*, but they would not be at all typical, as the Third which immediately follows (‘Wandern’). A Third, in fact, is habitually taken as a changing note:
c17.38
Ex. 452.
c17.39
or a grace-note:
c17.40
Ex. 453.
c17.41
or, especially in ascent, in a sort of tentative way:
c17.42
Ex. 454.
c17.43
indeed, in any way but as a harmonic interval:
c17.44
Ex. 455.
c17.45
To continue Schubert's tune :
c17.46
Ex. 456.
c17.47
Here there is nothing unvocal from an Indian point of view, but the phrase repeated *totidem verbis* at different levels is not in their manner. A reason for this may be that their melodies swing from salient note to salient note, and that if the salient notes were in the important places in the first phrase they would not be so in the second, and vice versa, so that the point of the contrast would be lost; it would be like reading poetry thus :
c17.48
The young men march before him in all their strength and pride,
The tender little infants they totter by his side.
c17.49
It is conjunct motion that Hindu practice, rather than theory, accounts the stronger. It has been maintained ¹ that melody tends to be developed along the line of the common chord. This appears to be the case in some parts of the world, among the Chippewas, for instance :
c17.50
¹ See Oskar Fleischer, *Ein Kapitel der vergleichenden Musikwissenschaft*, in the first quarterly volume of the International Music Society, p. 17.
c17.51
Ex. 457.
c17.52
but it is emphatically not the case in India. The reason may be that there is more real contrast to be got out of step than leap; because the step must often be to a consonant interval, and in that case the second note repeats to some extent what has already been said by the first. It has often been noticed that Beethoven's melodies are full of conjunct motion—the great tune of the Choral Symphony, for instance, which matured during half a century with Goethe's *Faust*; and perhaps this motion was what he admired in our national anthem when he undertook to 'show the English' what a fine tune they had—though they have hardly been as grateful as they should have been; perhaps, too, the injudicious use of disjunct motion accounts for some of the insipidity of the tunes of Shield, Henry Bishop, and T. H. Bayly, and the judicious admixture of both kinds of motion for the virility of Arne and Purcell.
c17.53
Conjunct motion, too, is eminently suited to the speaking voice, which makes small changes in pitch; moreover small intervals seem to be more effective than large against a drone, as may be seen in Schubert’s *Leiermann*. The song has contained plenty of disjunct motion, but in the last line:
c17.54
Ex. 458.
c17.55
where the maximum of effect is to be produced, it becomes suddenly conjunct.
c17.56
Before we leave this point there is one thing that their songs have in common with our Folk-songs which seems worth mention. When a rising or falling Fourth is bridged by a passage note, this is usually nearer to the note which is being quitted than to that which is being approached. Thus:
c17.57
Ex. 459.
c17.58
are commoner than
c17.59
Ex. 460.
c17.60
The fact that the former kind is preferred suggests that such motion is really the stronger melodically of the two; and this seems to be borne out also in our music. We realize how much is lost by the not infrequent substitution of an A for the second note (G) in ‘The bonny banks of Loch Lomond’.
c17.61
Ex. 461.
c17.62
We feel also that Beethoven’s
c17.63
[Figure: Ex. 462.] &c.
c17.64
is better suited to a reflective *Andante con moto*, but Brahms’s
c17.65
Ex. 463.
c17.66
to an austere ‘first movement’, and Mozart’s
c17.67
Ex. 464.
c17.68
to a robust *Allegro molto*.
c17.69
3. This leads to another point. The typical forms of ascent and of descent are apt to differ in one and the same *Rāg*, which is then called ‘crooked’ (*vakra*). The Hindus are most strict about the observance of the type. Its origin probably lies in this tendency which we have been discussing—the tendency to choose, between two consonant notes, the note of passage next to that which is being quitted, thus:
c17.70
Ex. 465.
c17.71
Its effect certainly is the avoidance of tautology and anticlimax. There are plenty of instances in our melody where this principle of contrast is not employed, because the harmony can supply all the contrast which is required:
c17.72
*Ein' feste Burg.*
c17.73
Ex. 466.
c17.74
Upper tetrachord of tonic. Lower tetrachord of dominant.
c17.75
BEETHOVEN. Quartet, Op. 18, No. 5.
c17.76
Ex. 467.
c17.77
To dominant. To tonic.
c17.78
But there are others in which a different form is adopted for ascent and descent, and consequently the contrast which harmony offers can be dispensed with:
c17.79
SULLIVAN.
c17.80
Ex. 468. Take a pair of spark-ling eyes
c17.81
The whole of this departs but little from tonic harmony, and the main contrast lies in the scale:
c17.82
Ex. 469.
c17.83
But a large number of these ‘crooked’ forms of Hindu *Rāgs* consisted in creating gaps in ascent to be filled up in descent, or vice versa. This not only avoids anticlimax, but induces climax. To pass over a note immediately creates a desire for it, and it then becomes a fit note to bear the climax:
c17.84
BACH.
c17.85
Ex. 470. *
c17.86
BACH.
c17.87
Ex. 471. *
c17.88
BACH.
c17.89
Ex. 472. * * *
c17.90
BACH.
c17.91
Ex. 473. * * * *
c17.92
In the first three cases the notes marked (*) have only been heard as passing notes, and now they are made substantive; in the last case three of them have not yet been heard at all. There are special reasons, no doubt, for constructing the theme of a fugue in this way, but incidentally it has great melodic point. Two notes are similarly kept in reserve in the tune of Sullivan just given, and they (F and A♭) are the making of it. Similarly also in the ‘Swanee River’ the 7th and 4th of the scale are kept in reserve, in ‘Come lasses and lads’ the 4th, in ‘Farewell Manchester’ the 7th; whereas the ‘Vicar of Bray’ and the ‘Bay of Biscay’ put all their goods in the shop window. The last line of Schubert’s ‘Wandern’, again,
c17.93
Ex. 474. das Wan - dern, das Wan - dern.
c17.94
where the D, hardly heard previously, now comes by its full rights, is a peculiarly grateful instance of this, and gains by the repetition of the phrase.
c17.95
4. What gives its peculiar flavour to a melody, what defines more than anything else the character of a particular *Rāg*, is the position of the *aṃśa*. Round it, and round its consonant or correlative (*saṃvādi*)—a note at the distance of a Fourth or a Fifth from it—are grouped the phrases of the song. They derive their meaning, their point, from their relation to it. It is easy to see that the character of a phrase would completely alter according as the *aṃśa*, the salient note of the scale, were made the salient note of the phrase, or not. The *aṃśa* may be at any place in the scale, not excluding the keynote or drone note. The first section (astai) of the tune usually circles round the *aṃśa* and the second (antara) round its consonant. The aesthetic effect of a high and a low *aṃśa* is quite different, and in this respect the *aṃśa* and all that it implies may be compared with what we know as tessitura, though that is rather a ‘range’ of notes, whereas the *aṃśa* is a single note. Over against the *aṃśa* stands the drone note, or keynote (*kharaj*). These two notes divide between them the functions of our tonic. Hindostani music is apt to begin on the *aṃśa*, and to end, if not there too, on whatever was the first note, or the end of the first phrase, of the song. This leaves the melody, to our ears, in the air. In the Carnatic music it is the rule, rather than the exception, for the melody to end on the drone note, or its Fifth. Yet, even so, the drone note has not the finality of our tonic. When the *aṃśa* is at F, for instance, any meaning that we could attach to a tonic is utterly driven out of C. The purpose of the drone is rather to maintain a steady level by comparison with which the *aṃśa* appears as high (and the tune spirited) or low (and the tune reflective).
c17.96
This is a conception which is and must remain foreign to our music as a whole; it is impossible for us to put any note besides the tonic in this position of ‘central’ except the Third or the Fifth; and to build tunes only on that principle would limit the possibilities enormously. Still we may realize the idea of this centre in those of our tunes in which the Third or the Fifth is so treated. Many of Mozart's circle round the Fifth and gain their sprightly character from that fact:
c17.97
Ex. 475. Andante. Sonata.
c17.98
and
c17.99
Ex. 476. Andante. Eb Symphony.
c17.100
Ex. 477. Allegro. Quartet.
c17.101
All our composers have been fond of the Third as 'centre', in slow movements especially; it will be enough to look at the melody of the Finale of the Choral Symphony:
c17.102
Ex. 478. Allegro assai. BEETHOVEN.
c17.103
The 'central' note is F#; and there is a skill in the way in which it is made central without being monotonous. It is central both because of the frequency of its use, and because, though the tune lies mainly above them, this is balanced by the dramatic drop to A at the climax. But it is not monotonous because (leaving aside repetitions of phrase, which are not strictly repetitions of note) it is never treated twice exactly in the same way although it occurs eight times; either the note is on a different beat of the bar, or of a different length, or it is an ascending instead of a descending phrase. And these are exactly the principles—position of central note, variety of metre, and contrast of ascent and descent—which India says are at the bottom of good melody.
c17.104
The *amśa* need not, and frequently does not, occur in a melody any oftener than other notes; nor indeed need any of the other salient notes. But their presence is felt in the appoggiatura, which they render possible. It is difficult to imagine any system of music which did not give an important place to this the most emotional element of tune; the Chinese tunes seem, indeed, to show small trace of it; but we have not yet, perhaps, very trustworthy records of them. The Hindu tune to be quoted presently shows half a dozen instances of appoggiatura, and the poignancy of
c17.105
Adagio molto. BEETHOVEN, Rasoumoffsky Quartet. Ex. 479.
c17.106
the tender grace of
c17.107
BRAHMS, Aeolsharfe. Ex. 480. Ihr kom - met, Win - de, fern her - ü - ber, ach, von des Kna-ben, der mir so lieb war, frisch grü - nen-dem Hü - gel.
c17.108
and the impetus of
c17.109
BRAHMS, Magelone Lied. Ex. 481. Wie soll ich die Freu - de, die Won - ne denn tra - gen?
c17.110
are closely bound up with this device.
c17.111
We may learn of this possibility in an Indian tune too, but we cannot instinctively feel it, because the *Rāg* is not ringing in our ears. We do not know, we have to be told, that in Behāg, for instance, B and E are strong and A and D weak notes, before we feel the force of the appoggiatura in:
c17.112
Ex. 482.
c17.113
But if we cannot select any other than these two notes (the Third and the Fifth) for ‘central’ note, because to do so would upset our tonality, and if to confine it to these two notes and insist upon it in every tune would cramp our harmony without securing for our melody any compensating variety, yet, on the other hand, harmony can establish a ‘centre’ anywhere and everywhere in a tune at will, and some of the most exciting tunes are built entirely on this change of centre (not of tonic); as, for instance, Brünnhilde’s apostrophe to Grane at the end of the Götterdämmerung,
c17.114
Ex. 483. Espressivo. Götterdämmerung.
c17.115
where, though the tonic, E, is clearly felt throughout, the ‘centre’ moves upwards to a climax,
c17.116
Ex. 484.
c17.117
and in so doing forms a broad ‘inner’ melody of its own, upon which an identical figure is carried up, and by its identity gives to the tune its unity.
c17.118
This cannot be done in Indian melodies because each note has a fixed meaning, not, as with us, one assignable by the harmony at will. In them an identical figure, when it is transferred to another part of the scale, takes on a new meaning, because its relationship to the *amśa* has been altered. Consequently a principle of structure is then lost to Indian music. The Hindu melody has to get what symmetry of structure it can—and it will be seen presently that this may be a good deal—in spite of the fixity of the note values. As a result their melodies have, apart from the fact that they are of an extempore nature, less cohesion than ours.
c17.119
5. With regard to rhythm Hindu melodies tell us two things. First, that variety of metre is more important than variety of accent, and secondly that cross-metre greatly enhances the interest.
c17.120
With us, too, variety of metre is a sure mark of a strong tune. There is the often cited violoncello sonata of Beethoven:
c17.121
Allegro ma non tanto. BEETHOVEN, Op. 69. Ex. 485. p
c17.122
in which no two bars have the same metre. An even better instance perhaps is:
c17.123
SCHUBERT, Wanderers Nachtlied. Ex. 486. Ü - ber al - len Gip-feln ist Ruh.
c17.124
Other themes attain the variety by syncopation:
c17.125
BACH, Wohltemperirtes Clavier, No. 43. Ex. 487.
c17.126
But the truth is that we have not attached much importance to the breaking up of metre. One reason for this may be that, both for the sake of clinching the tonality and also because of the variety which is thereby obtained, we love to contrast different levels of pitch; and the best way of making this difference of level felt is to keep the melodic figure constant. This may easily degenerate into a mannerism.¹
c17.127
¹ The greatest sinner in this respect is Tschaikowsky who, in examples too numerous to need quotation, works some poor little rhythmical figure literally to death; and the same is also true of a large number of Franz’s songs, of *Für Musik*, for instance :
c17.128
But the most infallible and effective way of lending interest to a theme is by cross-metre.
c17.129
PURCELL, ‘I attempt from love’s sickness to fly.’
c17.130
Ex. 488.
c17.131
The pairs of phrases overlap. In each case the *ictus* of the phrase is shifted, thus:
c17.132
Ex. 489.
c17.133
At (*a*) and (*c*) the reply arrives a beat late, at (*b*) half a beat early, and at (*d*) it is prolonged; and the result of it all is a flowing five-bar rhythm, a wonderful instance of art concealing art.
c17.134
FRANZ, Op. 10, No. 1.
c17.136
Nun die Schat-ten dun - kel, Stern an Stern er - wacht:
c17.138
Welch ein Hauch der Sehn - sucht flu - tet durch die Nacht!
c17.139
In the Andante of Brahms's 3rd Symphony the whole theme is built on a subtle cross-rhythm:
c17.140
Ex. 490. Andante. BRAHMS. 3rd Symphony. (Expanded.)
c17.141
(Expanded.)
c17.142
(Contracted.) (Expanded.)
c17.143
(Still more expanded.)
c17.144
The slow movement of Beethoven's seventh violin sonata may be simplified thus:
c17.145
Ex. 491. Adagio cantabile. a b c d
c17.146
The four phrases marked *a*, *b*, *c*, *d* have this in common, that they all make the high or the low F the climax of the rhythmical figure (it might almost be called the *aṃśa*). They also increase in sweep and complexity. But the point here is that they are ‘strettoed’— that the F in *c* and *d* increasingly antedates its arrival, and that the cross-metre which this involves does much to heighten the interest.
c17.147
We will now put side by side a good specimen of each system, for comparison of method, not of merit. For it must be remembered that the Indians do not make ‘tunes’ in our sense; all they do is to display *Rāg* and *Tāl*. Still, most of those given in the last chapter are tunes with a beginning, middle, and end, though it may be doubted whether a Hindu would consider them to be typical of his music at its best. Florestan's air from *Fidelio* compresses the maximum of point and balance into four bars, and a singularly beautiful tune in *Rāg Kānaḍā* attains with curious felicity a more ornate symmetry in sixteen:
c17.148
Ex. 492. BEETHOVEN.
c17.149
Following the skeleton given below the tune, the four sections are of equal length (practically three crotchets) but iii begins one quaver late, and iii and iv overlap. The A in iv is an appoggiatura for G, and is to be considered as a long G. The semiquaver in III corresponds to the acciaccatura in II. Then, ii repeats i in double time, at a higher level, preceded by a link note; iii reverses ii; iv repeats ii at the level of i, doubling the time of ii (quadrupling the time of i) and lengthening out the link note. Again, I is unadorned, II adds a semiquaver (acciaccatura), in III this becomes two semiquavers, and in IV three semiquavers. There is not a single note which is not doing duty. What makes this closeness of structure possible is that harmony allows of any note being taken at any interval; in an Indian tune, for instance, if the leap in II were possible that in IV would be less possible, and vice versa, and that in III would be impossible in any case; or again, if the downward progression in II were possible the upward in III would very likely not be so in that particular *Rāg*.
c17.150
Now we will take this tune in *Rāg Kānaḍā*, which is obviously well balanced; but before it is dissected and its beauty explained away the reader is recommended to turn back to p. 310, Ex. 416, and enjoy it. It is set to a delicately constructed *Omvī* (a Maratha form of verse), of which the English is unfortunately not forthcoming. The words are:
c17.151
Māgō sān|gītālēn|lākṣaṇēn
Mūrkhā āṇ|gīn cātū|ryā bānē
Ātēn ē|kā sāhā|nē ǎsō-
nā mūrkhā
c17.152
and the scansion : Ma Ra Ra Ma Ta Ya Ma Ra Ra [Ma Ta] Ya (see p. 196.)
c17.153
The bars in the music follow this scansion roughly, the first and second lines being farsed with a ‘jubilation’ (see p. 255); but beyond this the music pays, as far as its structure is concerned, no attention, as indeed our best melodies do not, to the poetical form; it only sees that a long syllable is provided with a long note as ours see that a stressed syllable comes on the musical accent.
c17.154
The tune has 16 bars; 5+5 in the astai, and 6 in the antara. There is thus a suggestion of cross-metre (4 time and 5 time) in the general plan. The astai ascends to a climax for three bars and descends for two (twice over); the antara does the same for three and three bars, enlarging the intervals for its downward sweep from the second (higher) climax. The amśa of Kānaḍā is E♭, with consonant note B♭ above and below; and at these places the most conspicuous figure (marked ‿‿) is introduced. The astai is concerned only with the lower tetrachord, the antara (until the final cadence) only with the upper, according to rule. Kānaḍā is one of the Rāgs with a queer intonation (see App. II); but though that makes it more difficult (and more beautiful) it does not affect the question of structure:
c17.155
Ex. 493. Astai. Mā - go . . . saṇ - gī - ta - līṇ la - kśa - neṇ . . . .
c17.156
Mūr - khā . . . aṇ - gīṇ cā - tū - rya bā -
c17.157
Antara. - ne . . . . . . . . Ā - tāṇ e - kā . . . śa -
c17.159
The main interest here centres in the metrical plan. This is based on a contrast of what we may irreverently call pot-hooks and hangers. They are marked *a* throughout, placed for the former below the stave and for the latter above. A succession of these makes up a section which is understood to be *A* in contrast to *B*. *B* is itself an expansion of *a*; it is a trisyllable as against a disyllable, an opposition which we saw to be important in verse. But for fear the *a*’s and *B*’s should become monotonous they vary in length; the *a*’s are of 2, 3, or 4 quavers, the *B*’s of 6, 7, or 8. The ‘pot-hooks’ have on the whole the best of it in the *astai*, but the ‘hangers’ drive them off the field in the *antara*; *B* acts as an impartial umpire, giving decisions in favour now of one side now of the other. The ingenious way in which in the first five bars the last note of the ascending metre is made the first note of the descending melody is worth notice. It will be seen here too that, though on quite different principles, not a note is idle.
c17.160
It is true no doubt that these tunes could be analysed in quite other ways, and equally true that their respective composers never thought for a moment of this or any other analysis. But it is only by analysis that those of us who are not composers can realize such a complex thing as the act of composition. The point is that a tune is not a tune at all until it has something that we must call closeness of structure. It might seem as if nothing could be less fettered or more impulsive than melody; beauty is there—wherever it is that tunes come from—and the composer has only to stretch out his hand and seize it. Man’s will is free, no doubt; but he is bound by his own past—in a tune, as in anything else; his tune must be a related whole, just as his life must be a consistent one. That is what is meant by structure. The closer the structure, the more lasting the pleasure which the tune gives; and to make such tunes is art, through whatever conventions.
c17.161
These two tunes epitomize for us the music of West and East. On the one side a repression of what is petty, a rejection of what is transient, a soberness in gaiety, an endurance in grief. On the other a vivid insight, the eager quest after wayside beauty, the dexterous touch that turns it to account. The one seems to say—Life is puzzling, its claims are many, its enthusiasms hardly come by; but we will hammer out a solution not by turning away from ugliness but by compelling it to serve the ends of beauty. The other—Life is simple, and beauty close at hand at every moment whenever we look or listen or wherever we go; the mistake is in ourselves if we do not train our eyes and ears and hearts to find it.
c17.162
Who would wish to decide which way was the best? Both are human. There is no need to decry one in favour of, or to exalt one at the expense of the other. Are those Europeans who smile at ‘tomtoms’ sure that they understand the grounds of their faith in their own music? And are those Indians who scoff at Equal Temperament, at the dullness of European song and the screaming tones of the European voice, able to reach the governing principles of their own art through the mass of tradition and imagery in which it has become involved?
c17.163
Art expresses. It finds words or tones for what was hitherto unnamed; it actually calls into being an experience which has not previously existed; it does not communicate to us, or acquaint us with, an antecedent experience. And what it expresses is the fact of emotion, not particular feelings. It is not that words or tones cannot indicate these, but that directly they do so the result ceases in so far to be a work of art. That music can, as most would agree, express emotion more immediately than other arts, depends on the fact that ‘in its ideal, consummate moments, the end is not distinct from the means, the form from the matter, the subject from the expression; they inhere in and completely saturate each other; and to it, therefore, to the condition of its perfect moments, all the arts may be supposed constantly to tend and aspire’.¹
c17.164
¹ Pater, The Renaissance, 1904, p. 139.
c17.165
But if music at its best obliterates those distinctions, it will be at something less than its best when it maintains them. Thus both the ‘popular’ music which exalts matter at the expense of form, and extreme ‘absolutism’ in which the form dwarfs the matter, depart, to some extent, from the ideal. Indian music knows perhaps less than ours of this falsehood of extremes. It seldom descends to the merely popular because it refuses to compress itself into a square tune, something that you can carry away with you like an umbrella, and, as Corney Grain used to add, ‘it doesn’t much matter if you never bring it back’. Nor has it the deadly insistence of the cook’s song in the *We’re Here*, which was ‘like something ever so bad but sure to happen whatever you did’, because it is never sung twice in the same way. Nor is the tune a mere adjunct to the words, as rhyme or alliteration are to an advertisement, for the words are set to the tune, not the tune to the words.
c17.166
On the other hand, it seldom runs riot in absolutism. Since it is vocal rather than instrumental, melodic and not harmonic, it seldom travels far from those broad human requirements which gave, in song, the first impulse to musical expression, and which, through whatever subsequent developments, still form its ultimate justification.¹ Since concerted song is rare the personal element is not put in commission, and the appeal of the music is therefore more immediate. At the same time it never ceases to be a purely musical appeal. For whereas in our songs development has proceeded increasingly on other lines than those of pure melody,² in India the singer’s tones can still carry all the artistry which his mind can conceive. And while in England, especially, concerted music has always been highly prized, and rightly so, for its social element,³ it is apt to fall short of the highest ideals, since it is never so easy to find an artistic crowd as an artistic individual.
c17.167
¹ A string quartet, for instance, which may be taken as our highest achievement in absolute music, still finds its final justification in the fact that the instruments individually or collectively ‘sing’, that is, are employed upon phrases which are thoroughly ‘vocal’ in all but the grace of words. Modern composers seem increasingly to ignore this fundamental of their art and to forget to ‘sing’. If they will not take warning by the fate of Spohr’s saccharine harmonies and of Berlioz’s evasion of the main issue, they can expect no immortality.
c17.168
² Since Schubert the centre of interest has been more and more transferred from the voice part to the accompaniment. To hear the accompaniment alone would give a better idea of the composer’s thought than to hear the voice part alone. The latter would, in fact, be wholly unintelligible in a large number of modern songs, especially in those which aim at creating atmosphere or at elaborating niceties of declamation.
c17.169
³ And there is something, therefore, to be said even for the Crystal Palace and other monster concerts.
c17.170
Again, the fact that in India the composer and executant are one and the same person, brings it about that he sets, as composer, no problems which he cannot, as executant, himself solve, and that mere academicism is non-existent and that his songs ‘come off’.
c17.171
When all composition is improvisational, and when notation is rarely employed for recording music and never for performing it, there can be no such thing as ‘visual’ music—music which makes a pretty pattern to the eye but disappoints the ear; and of this we have a surprising number of examples in the madrigal period, and it is by no means unknown in later times.¹ The Indian practice, described on page 85, of setting the syllables of the words to notes which bear the same Sol-fa syllable, is quite as childish; but it appeals, at any rate, to the ear, not the eye.
c17.172
¹ See an article by Alfred Einstein, *Monthly Journal of the International Musical Society* for October, 1912.
c17.173
Are there not singers amongst us who have felt a desire to break loose, if it were possible, from the trammels of our tonality—from its ‘closes’ and ‘half closes’, its ‘conceded modulations’, its ‘unity of design’, its rhythmical rigidities and its fussy logic—and to let the melody bear them along on light wings of fancy; to find, in fact, a music which is free like that of the woods in spring-time, where, without rule, the uncouthest tones like the crudest colours all harmonize, where, unguided and unthwarted, sound and silence answer one another, where the inchoate and incomplete are made good by the motherly bounties of Nature, and ‘unbroken perfection is over all’? Something of this is in the careless profusion and the unstudied rapture of Indian song:
c17.174
Singing hymns unbidden,
Till the world is wrought
To sympathy with hopes and fears it heeded not.