The Music of Hindostan c15: Chapter X. The Sāman Chant

The Music of Hindostan c15

Language: English

c15.s1 Authorities for the Sāman Scale — 178 parts

Chapter X. The Sāman Chant

c15.2

It is a bemoanable pity to consider how few there are who know, but fewer who consider, what wonderful-powerful-efficacious Virtues and Operations Musick has upon the Souls and Spirits of Men Divinely-bent.
THOMAS MACE.

c15.3

An inquiry into the ethos and structure of some of the oldest music of which there is any record is not a mere piece of antiquarianism; it throws that sort of light on the music of modern India which a knowledge of the geology of a country throws upon its scenery.

c15.4

Before coming to the Sāmaveda itself it will be necessary to glance at the recitation of the other Vedas, especially the Rgveda. For the Sāmaveda has borrowed the large majority of its text from the Rgveda, only altering or expanding the actual words so as to make them suitable for chanting; and much of the grammatical and prosodical determinations of the Rgveda has been given in the Sāmaveda a musical meaning.

c15.5

The Rgveda is recited now, as it has always been, to three tones; for the accent was originally a mark of musical pitch, and became a mark of stress only after the beginning of our era. The ‘raised’ (udātta) and ‘not raised’ (anudātta) sounds represent the two main pitches of the speaking voice; this is the prose accent of the Brāhmaṇas; and there is nothing to show whether these two prose accents had any musical relationship. In addition to these two there was, for the reciting voice, the ‘sounded’ tone (svarita), which is shown on philological grounds to have been originally between the two others in pitch, but which is in the Rgveda above the udātta. No explanation is given as to what is exactly meant by ‘sounded’,¹ although it seems clear from the treatment of svarita that it means ‘graced’. The svarita is in effect ‘a falling accent of a dependent nature, marking the transition from an accented to a toneless syllable. It regularly follows the udātta, to the rise of which its fall corresponds in pitch.¹ But also, ‘the first part of the svarita sounds higher than the udātta’ (ibid.), so that the svarita is practically an ornamented udātta falling to an indefinite pitch below it. This indefinite pitch is called pracaya (‘throng’), in allusion to the number of unaccented syllables which occur there in succession; the pracaya is toneless, as distinguished from the udātta and anudātta, which are toned.² Before the rise from the pracaya to the next udātta or svarita, the voice descends to a ‘lower’ (sannatara) or ‘more lowered’ (anudāttatara) note, which is below the anudātta.

c15.6

¹ Haug (Ueber das Wesen, &c., p. 74) appears to accept the view that svarita from Svara (like getönt from Ton) means ‘accentuated’. But (1) this looks like a false analogy from the double meaning of Ton, and (2) svarita must surely be connected with svāra, which meant, technically, a downward slide through the svaras, i.e. the svarita was that note to which this slide was attached.

c15.7

¹ A. A. Macdonell, Vedic Grammar, p. 77.

c15.8

² It is tempting to see in the pracaya the counterpart of the pyknon. The pyknon was the cluster of three notes at intervals of a diesis (quarter-tone) which occurred in the Enharmonic genus (or of a semitone in the Chromatic). Thence there was a leap to the mesē just as there was a leap from the anudātta, or pracaya, to the udātta. But the pyknon was not an amorphous cluster; one of its three notes stood out from the other two, was in fact of the nature of a tonic. This is closely analogous to the toneless pracaya in which the toned anudātta stood out. And this ‘toneless tone’ may, further, be in principle the forerunner of the drone, frequently relegated to the drum, which might be called from one point of view toneless, from another, toned. The formation of the svarita tells us that the interval which was taken as a leap in ascent was bridged in descent; and this is borne out by the practice of modern Rāga, which prefers on the whole disjunct motion in ascent and conjunct in descent.

c15.9

That is the original theory of the recitation; it will be seen, however, that it is modified in practice. This example, from Haug, p. 52, shows the svarita marked with a perpendicular line over the syllable, the anudātta with a horizontal beneath³; the udātta and pracaya are unmarked:

c15.10

³ Devanāgari letters are surmounted by a continuous horizontal line. A perpendicular line stands out well against this, and a horizontal line against the tails of the letters below.

c15.11

Ex. 340.

c15.12

This is the Poona form of recitation taken down by Haug from the lips of Maratha Brahmans.⁴ Only one form of the svarita, the appoggiatura F E, is here employed. But the noticeable thing is that the *pracaya* (marked P) is the same note as the *udātta*, not as the *anudātta*; if in a chant of only three notes it is to be distinguished from the *anudātta*, it could not well have any other place; still, this change appears to await explanation.

c15.13

⁴ Haug (loc. cit.) gives four more of these stanzas.

c15.14

In the following specimens¹ from Madras the *svarita* appears as a single note (F) and as an upward appoggiatura, E F.

c15.15

¹ See M. Śeṣagiri Śāstrī, *Descriptive Catalogue of the Government Oriental MSS. in the Madras Library*, vol. i, pt. 1, pp. 3, 4.

c15.16

Ex. 341.

c15.17

Ex. 342.

c15.18

In the second of these examples the syllables *Co-da* obey a rule that ‘when more than one *anudātta* precedes an *udātta*, all but the last are sounded lower’. Dr. Erwin Felber’s book, *Die indische Musik der vedischen und der klassischen Zeit*, contains (p. 106, No. 433) an example of recitation of the *Ṛgveda* ‘in the style of the old *Ṛṣis*’, which is more florid and of larger compass than any of these.

c15.19

The *Yajurveda* is also recited on the notes D, E, F; although in the half dozen times I heard it D, E, F♯² was quite as common. The following is from a Mysore singer:

c15.20

² Noted here for convenience as C, D, E; I made no note of the absolute pitch.

c15.21

Ex. 343.

c15.22

but the passage is too short to show the compass. At any rate, Dr. Felber¹ gives two specimens of *Yajurveda* on pp. 108, 116, which show a compass of a Fourth and a Fifth respectively; his example (p. 106) of the *Ṛgveda* reaches a Sixth. This is perhaps only one more instance of what is so common in music, the ‘official’ rules lagging decades or centuries behind the practice. His three examples (of both Vedas) show a foundation scale D, E, F, though one of them begins in E, F♯, G, and continues in E, F, G.¹

c15.23

¹ The names of the notes throughout this chapter indicate relative pitch only. They are chosen at any absolute pitch which is suitable, as fragments of the Sāman scale, which is explained, or at any rate discussed, later.

c15.24

The *Atharvaveda* does not appear to be recited, at least not in accordance with rules. But all reading of the *śāstras* is recitation. Students read, i. e. recite, in the same room. Children are taught to read so. A boy read me some of the *Ṛgveda*, employing the accents correctly, but stopped when he reached a certain page ‘because he had not learned any further’.

c15.25

It will be noticed in these examples that with the exception of an occasional grace-note or appoggiatura—and a great many more grace-notes are put in than appears in the text, for seldom is a note ever sung, as we should say, clean—the syllables are allotted one note each; there are no melodic figures like the ligatures of Gregorians. For these we turn to the

c15.26

*Sāmaveda*. The symbol round which the elaborate ritual of the *Sāmaveda* gathers is that sacrifice of which the drinking of the juice of the *Soma* plant was the central point. The virtues of this juice are recapitulated in the ninth book of the *Ṛgveda*, from which mainly the words for the *Sāman* chants are taken.² *Soma* is translated ‘moon-plant’; and the *Sāmaveda* is specially connected with the worship of ancestors, whose abode was the moon. Great care was taken not to deviate from the original melody-types and rhythms, and the religious efficacy of the hymns was held to depend largely on the right application of directions contained in the Brahmanic explanations (*Brāhmaṇa*, not later than the sixth century B. C.) of the Vedic text (*Saṃhitā*). The expense of the full ceremonial was not small; the Soma sacrifice involved days in performance and months in preparation. A full description of its elaborate and gorgeous ritual is to be found in the *Aitareyabrāhmaṇa* of the *Ṛgveda*, translated into English by Martin Haug, 1863, and its close connexion with the fire-worship of the Zoroastrians is there detailed. Other creeds have now swept across India and buried that of the Vedas in oblivion. The memory of the words and chants, and possibly of parts of the ritual, lingers chiefly in the South. Even there the true *Sāma*-singers are few, and impostors are beginning to flourish.

c15.27

² There is nothing to show that the chants are later than the words; in fact, since *Sāmans* are often mentioned in the *Ṛgveda* there is a probability, beyond the intrinsic likelihood, that they are older.

c15.28

The body of hymns is divided into two sections, which may be centuries apart—the early, according to the latest authority,¹ the *Uttarārcika*, and the late, the *Pūrvārcika*. The early hymns are divided into (1) the ‘investigated’ (*Ūha*), and (2) those which are ‘to be investigated’ (*Ūhya*) or which are ‘secret’ (*Rahasya*). The later hymns are in three divisions, addressed to Fire (*Agni*), the Rain-god (*Indra*), and Soma *Pavamāna*.² These are arranged internally according to their metres, and according to the ‘jubilations’ (*stobha*) and the doxologies (the parts of the hymn which are genuine *Sāman*). Only seventy-five verses, out of many hundreds, are not borrowed from the *Ṛgveda* or other source; such borrowings are known technically as *yoni* (womb). A distinction is also made between those verses of the *Ṛgveda* which have one melody (*eka-sāmi*) or many (*bahusāmi*), showing that *Sāman* is, properly speaking, music, not music and words, nor words for music. In fact the conception is the reverse of ours. We speak of setting words, that is, putting music to them; but the *Sāman* is a melody for which words were found (chiefly in the *Ṛgveda*). This is the case too with modern poets; Rabindranath Tagore says that when the general idea of the poem comes into his mind, he first thinks of the melody (generally his own) and then writes words to it. We are reminded of Burns’s two manful attempts to find words for that ‘crinkum-crankum tune, Robin Adair’, with its ‘cursed, cramp, out of the way measure’, and of his better success with Duncan Gray, ‘in Scots verse’.

c15.29

¹ Caland, *Jaiminīya-saṃhitā*, p. 4.

c15.30

² *Paramāna*, clarifying.

c15.31

The rhythm is, as in Plain-song, determined by the words, which decide where it will be appropriate to take breath. The two versions of the first hymn of the *Sāmaveda* which follow are from A. C. Burnell (*Ārṣeyabrāhmaṇa*, Introd., p. xlv) and Śeṣagiri Śāstrī (loc. cit.); and the notation is in accordance with the directions of the Gānas (the Sāmaveda text as actually sung) as they understood them, checked in the first case by the verification of a musician, and in the second by the usage as the writer knew it: Burnell, Ārṣeyabrāhmaṇa, p. xlv.

c15.32

Ex. 344.

c15.33

O - gnā - i ā - yā hi vo - i - to - yā - i gṛ-ṇā-no ha -

c15.34

Śeṣagiri Śāstrī, Cat. Sanskr. MSS., Madras.

c15.35

O - gnā - i ā yā - hi ī ve - i - to yā ā i gṛ-ṇā-no ha-

c15.36

- vya-dā - to - yā - i to - yā - i nā - i ho - tā sā - - - -

c15.37

- vya-dā to yā ā i to - yā - ā - i nā - i ho - tā sā ā ǎ

c15.38

- tsā - - i bā - - - - - - au ho bā - rhī - - - - shī

c15.39

tsā - ā - ī bā - ā ā ā au ho bā - rhī - i - ī - ī - shī

c15.40

Here, and elsewhere, a long mark over a vowel sometimes represents musical, not prosodic, length.

c15.41

The actual words of the hymn are: Agna āyāhi vītaye gṛṇāno havyadātaye | Ni hotā satsi barhiṣi || ¹

c15.42

¹ Translation: ‘Come, Agni, praised with song, to feast and sacrificial offering; be seated as Hotar on the holy Kuśa grass.’

c15.43

The bar-lines (given by Burnell only) are breath marks. The last note of each ‘bar’ (parvan) is accented (vṛddha).²

c15.44

² Lit. augmented. It appears to be occasionally augmented in time-length also, counting 3 to the 2 of the long (dīrgha) and the 1 of the short (hrasva). Later theory (see p. 256) has added lengths between these. We are reminded of the Greek time-unit (chronos prōtos), ‘two-time long’ (disēmos) and ‘three-time long’ (trisēmos).

c15.45

The next two examples have no such authority. They are merely the versions of two gurus, whom I heard in Mysore and Bangalore, of an identical verse, and represent possibly two different sects (śākhā). The tonic is D in both cases; that of Mysore ‘closes’ by means of an unusual leap (C, E, C, D), and that of Bangalore by the device of an ‘Alleluia’ (stobha):

c15.46

[Figure: Mysore. Ex. 345. Ka - yā na-ś ca - - i tra - - ā bhūd ū Bangalore. Ka - yā na-ś ca - ay - tra - bhūvad ū - - -
- tī - sa-dā - vṛ-dhaḥ sa-khā au - u - ho - o - hai ka-yā - -
- tī - sa-dā - vṛ-dha sa-khā - au - ho - hai ka-yā - - śa - ci - ṣṭha - yo ha śa - ci - ṣṭha-yo . . ha him 1 - ā - vṛ - to . . hai]

c15.47

There was, not unnaturally, great difficulty in inducing priests to sing these chants, and one man, at Bangalore, offered in lieu thereof ‘Sanskrit in the style of Sāman’. But the unusual content of the words shows its influence in the more or less rhythmical phrases induced by the prosody, which take the place of those vocal phrases and melodic figures which flow naturally from the liturgical use of the words in the Sāmaveda:

c15.48

[Figure: Bangalore. Sanskrit in the style of Sāman. Ex. 346. He jih-ve - ra-sa-sā - ra-jñe - sar - va - dā -
- ma-dhu-ra-pri - ye - nā - rā - ya-ṇā - - khya pī - yū ṣam - pi - ba jih-ve - - - - ni-ra-nta - ram - - -]

c15.49

On the other hand, at Tanjore a man got up of his own motion and sang *Sāman* before a roomful of people, and, what was more surprising, before a European in their presence. It appeared that he had a theory of his own as to how *Sāman* should be sung, and was seeking converts. He has evidently taken the scale of the *Nāradaśikṣā* (see below, p. 259) as if it were in the modern notation, F–G, instead of the ancient, G–A, and the mode is therefore changed. Further, the essence of the *Sāman* chant is that it circles about one note, F (or E, or D), and only leaves that to form a downward cadence. His version draws attention away from the E to the A below, which in the true *Sāman* is quite a subsidiary note, and it begins, therefore, to be in a regular *Rāg*, which *Sāmagās* assert positively that their chants are not. Moreover, he has forgotten after the first line to put in those characteristic prolongations of notes by vocalizing syllables which add so much to the solemnity of the chant:

c15.50

Tanjore.

c15.51

Ex. 347.

c15.52

The words are spelled phonetically; his original text is not available.

c15.53

*Form.* The *Sāman* is divided into five parts:
(i) *Prastāva*, introductory ascription of praise, preceded by the syllable *hum*, sung by the *prastotṛ*, or assistant priest.
(ii) *Udgītha*, preceded by *om*, sung by the *udgātṛ*, or priest.

c15.54

(iii) *Pratihāra*, the joining in, preceded by *hum*, sung by the *pratihartṛ*, who ‘joined in’ at the last syllable of the *udgītha*. Sometimes the *pratihāra* is divided into two, in which case there is—
(iv) *Upadrava*, the recession, formed out of the last few syllables of the *pratihāra*, sung by the *udgātṛ*.
(v) *Nidhana*, the coda, consisting of one or at the most two syllables, sung by all three priests, followed in some cases by *om*.¹ The following are the first two sections of the first hymn of the *Sāmaveda*:

c15.55

¹ These five *bhakti* (see p. 255) are the subject of Dr. Simon’s latest work, *Das Pañcavidhasūtra*, Breslau, 1913.

c15.56

Burnell, *Ārṣeyabrāhmaṇa*, p. xlv.

c15.57

I. 1. 1. Prastāva. Udgītha. Ex. 348. O - gnā - ī ā yā - hi . . vo - i - to - yā - i

c15.58

I. 1. 2. A-gna ā yā - hī vī - ta - yā - i gṛ - ṇā - no ha-vya - dā - gr - ṇā - no ha-vya-dā - to - yā - i to - yā - i - tō - - - yā - i

c15.59

Pratihāra. Upadrava. na - ī ho - tā sa - - - - - tsā - - - i ni ho - tā sa - tsi bā - rhā - - - - i - ṣi bā - rhā - - i

c15.60

Nidhana. vā - - - - - au ho bā - rhī - - - - - ṣī ṣi . . . . . au ho va bā - rhī - ṣi . . . . . .

c15.61

The verse, *āgna āyāhi* (*Ṛgveda* VI. 16. 10), is sung three times over to form a *stoma* (group); Burnell only gives two of the three, and they are arranged here under one another for comparison. Of the remaining nine verses of this particular hymn, eight are from the *Ṛgveda*, and the last, which forms a doxology, belongs exclusively to the *Sāmaveda*. Besides the first verse, the fourth and fifth also form *stomas*; the sixth, seventh, and eighth are sung twice, and the other four only once¹; in all, nineteen sections (*bhakti*). The usual number was fifteen. Mention is also made of nine and twenty-one *bhakti*; so that a basis of three or five or seven *stomas* is implied, varied probably in any particular case.

c15.62

¹ These are called *stotras*.

c15.63

The practical directions for the singing of these hymns are given in the *Gānas*. These give the text of the *Sāmaveda* (which has been culled, with modification, from the *Ṛgveda* chiefly) with the syllables vocalized (i. e. with prolongation and modification of vowels) and farsed (i. e. with insertion of vocal syllables between those of the text). These insertions (*stobha*, praises) are the exact counterpart of the *jubila* interpolated in Plain-song in the ninth and tenth centuries, out of which ‘tropes’ were developed in the eleventh,² as the *bhakti* of the *Sāmaveda* are also of the ‘strophes’ of Plain-song, which, like them, were of indefinite number; the idea, too, of singing the hymn in *stotras* (stanzas) and *stomas* (groups) is reproduced in the sometimes single, sometimes repeated performance of the strophe.

c15.64

² See W. H. Frere, Winchester Troper, Introduction, p. xxxii seq.

c15.65

*Rhythm*. The metres of the hymns are six (later eight) chief ones, in which the governing factor is the number of syllables, varying from twenty-four to forty-four for the stanza. Quantity of vowel is taken no account of except at the cadence. These need not detain us now, as the rhythm of the chants ignores the poetical metre entirely. In the true *Sāman* there are only three time-units: a short (*hrasva*) note for a quantitatively short syllable, a long (*dīrgha*) for a quantitatively long; syllables made long ‘by position’ count as short. An occasional important syllable still further lengthened or, especially at the end of a phrase (*parvan*), stressed, or both lengthened and stressed, is called augmented (*vṛddhi*) or prolate (*pluta*). This also finds its analogue in Plain-song.

c15.66

The system of the *Mātrālakṣaṇa* (date?) is an advance upon this, and is placed here for comparison beside the modern South Indian scheme of time-values:

c15.67

c15.68

¹ Virāma appears to be used indifferently for a rest or a dot with the time-value of an *anudruta*. The word means in general ‘stop’ (of punctuation); and is used in particular to show that a consonant is not followed by the *ă* which is inherent in it.

c15.69

² Cp. German *anderthalb* = 1½, *dritthalb* = 2½, &c.

c15.70

³ The sign in one modern notation for the longest note is x.

c15.71

These interminable theorizers also subdivided the *anudruta*. The *anudruta* is the equivalent of:

c15.72

| 2 × | 2 = | 4 Kalā | syllabic instant.⁴ |
| 4 × | 8 = | 32 Nimeṣa | moment. |
| 8 × | 32 = | 256 Kāṣṭhā | 2½ of an hour. |
| 16 × | 108 = | 1728 Lava | 2000 of an hour. |
| 32 × | 512 = | 16384 Kaṇa | atom. |

c15.73

⁴ The syllable *ka* has two ‘instants’—*k* and *a*.

c15.74

The use of astronomical subdivisions is interesting, though hardly illuminating (there are other explanations of the *kāṣṭhā* and *lava*). The only one of these terms in practical use is the *kalā*, which properly means 1/16. The *kalā* is 1/16 of the *Laghu* (their unit) as the semiquaver is 1/16 of the semibreve (our unit). Cp. the German *Sechzehntel* and the American ‘sixteenth note’.

c15.75

Scale. In the accompanying list of authorities for the *Sāman* scale the first six are ancient, the last four modern; the two sets are separated by something like twenty centuries. The list is probably not complete; there is no indication of place of origin. They are collated on two assumptions: (1) that the words *kruṣṭa*, *prathama*, &c., name the same note wherever they occur, that is, that *kruṣṭa*, for instance, was not at different periods applied to two different notes, and (2) that the two identifications of the *Sāman* names with the notes of the secular scale by authorities III, VII, VIII, and X are to be trusted. If either of these assumptions should be shown to be wrong the evidence would have to be reconsidered.

c15.76

Names of the *Sāman* notes. As extensions of a nucleus—*prathama*, *dvitīya*, *tṛtīya*, *caturtha* (first, second, third, fourth)—occur three others—*kruṣṭa*, *mandra*, *atisvārya*.

c15.77

*Kruṣṭa*. In more than one of these authorities the order in the original text is *prathama*, *dvitīya*, *tṛtīya*, *caturtha*, *mandra*, *kruṣṭa*, *atisvārya*. *Kruṣṭa* is used of the highest note; and as the sixth in a downward series this word has given trouble. Burnell proposed to read *kṛṣṭa*, ‘that to which *karṣaṇa* has been applied.’ He did not, however, explain why *karṣaṇa* should be applied in the sixth place especially, and it appears now that it is typically applied to the highest note. But there is nothing to show that the order of the names is the melodic order; and it is at least possible that in the order given above the first five name the ordinary notes of the scale (see p. 258, l. 12) and the last two the extraordinary ones (which happen to be at either end of the five). Meanwhile, if order does matter, that of No. III, in which *kṛṣṭa* actually occurs, is the order of the original. Moreover, No. V alludes to the seven *svaras* as *kruṣṭādi*, ‘beginning on *kruṣṭa*’; so there is little doubt that the *kruṣṭa* is above the *prathama*, and that another statement of Burnell’s that *kruṣṭa* and *prathama* are the same note is not universally true.

c15.78

There is another interest attaching to the word *kruṣṭa*, ‘highest.’ It shows us, as do also the ordinals *prathama*, *caturtha*, that the Hindus regarded the treble as high and the bass as low; as we do, and as the Greeks, who named the lowest note *hypatē* (highest), did not. This reminds us that high and low in music are merely a metaphor, though a very natural one. In a somewhat similar way Europeans regard the harmony of a song as ‘underneath’ the

Authorities for the Sāman Scale

c15.s1.1

| I Ṛkprātiśākhya. At least fourth cent. B. C. | kruṣṭa head | prathama palate | dvitīya brow | tṛtīya ear | caturtha nose | mandra | atisvārya (formed by karṣaṇa from the mandra) | 1 2 3 4 |
| II Bṛhaddevatā. At least fourth cent. | kṛṣṭa | prathama udātta Ga prathama | dvitīya svarita Ri dvitīya dvitīya | tṛtīya pracaya Sa • tṛtīya | caturtha anudātta Ni • caturtha | mandra • mandra | atisvārya (Ahvārakas) (Taittirīyas) | 5 6 7 8 9 |
| III Taittirīya- Prātiśākhya. At least fourth cent. | kruṣṭa | prathama | dvitīya | tṛtīya | caturtha | mandra pañcama | atisvārya atisvāra anusvāra ṣaṣṭha antya | 10 |
| IV Sāmavidhāna- brāhmaṇa. Later than I, II, III. | | | | | | | Alludes to the whole series as ‘kruṣṭādi’, that is, beginning with kruṣṭa. | 11 |
| V Puṣpasūtra. Fifth to second cent. | | | | | | | | |
| VI Sāmatantra. Prob. not modern. | | | | gi | di | bi | | 12 |
| VII Sāmaparibhāṣā. ‘Quite modern.’ (R. Simon.) | kruṣṭa Ni {C (G) | prathama Dha {B (F#) | dvitīya Pa {A (E) | tṛtīya Ma {G (D) | caturtha Ga {F (C) | mandra Ri {E (B) | atisvārya Sa {D (A) | 13 14 |

c15.s1.2

c15.s1.4

Relative pitch of the notes.

c15.s1.5

melody; whereas the Maoris think of the song as ‘here, down below’, and the harmony as going up to the roof under which they are singing and ‘coming down’ to them.

c15.s1.6

Mandra, low. This is the usual name for the sixth note; it is once (No. IV) called pañcama, in continuation of the ordinals. That it was a substantive note of the scale we see perhaps in No. I, where it appears even though the tṛtīya is omitted. But that it was on a different footing from the five notes above it is indicated in line 4, where they are all head-notes as opposed to the mandra, a chest-note. It gave its name also to the lower octave in the secular scale as opposed to the middle octave (madhya, or madhyama) and the upper (tāra, or uttama).

c15.s1.7

Atisvārya, extremity of the cadence. This is a curious name for the lowest note of the full scale, to which we should naturally have attached some idea of a tonic. But it is obvious that it is not a substantive note of the scale at all. In line 4 it only exists by permission of the mandra; in line 10 there is an uncertainty about the pronunciation of its name (atisvārya, atisvāra, anusvāra), and alternatives are suggested—ṣaṣṭha (sixth) and antya (last); in lines 17, 19, and 25 the ‘musical hand’ does not provide for it, and line 4 does not appropriate to it an organ of the body. It is clearly an ‘extra’ note; and it may be observed that the Greeks similarly regarded their scale as stopping normally at the B, while they called the note A below it ‘extra’ (proslambanomenos).

c15.s1.8

Tetrachords. There are indications that the nucleus of the scale was, successively perhaps, the tetrachords:

c15.s1.9

EX. 349. and

c15.s1.10

The tetrachord E–B. In line 9 this is the usus of the Taittirīyas, confirmed by line 29. In line 16 these same notes are appropriated to the four Ṛgveda accents. Lines 1 and 2 are very difficult because we do not know how to place the three notes of line 2. Aṣṭama (eighth) in a scale of seven notes is surprising. But it will be seen at the foot of the diagram that the scale actually contains eight notes, and it is just conceivable that aṣṭama, which is obviously extra to the accepted seven, names the F♯, which is alternative to the F♮. This leaves sama (level) and śukra (bright) to fill three places; and all that can be said is that atisvārya, the ‘additional’

c15.s1.11

note, is the most likely of the three to have been left unfilled. In line 1, where we can place the notes, it is remarkable that *tṛtīya* should be omitted in the series of four; and this may be an indication that the D was not on the same footing with the others, and that in this treatise, at any rate, the real nucleus of the scale was held to be F E–C B, i. e. composed of the Fourths F–C and E–B.

c15.s1.12

The tetrachord F–C. On the other hand, the fact that the four ordinals (first to fourth) are assigned in every treatise to the notes F, E, D, C, and that in line 6 the *Ṛgveda* accents are appropriated to these notes, makes it clear that the tetrachord F–C has an equal claim with that of E–B to be considered the foundation of the scale.

c15.s1.13

The hesitation between these tetrachords is reflected in the ‘musical hand’ of line 17, where F and E are both assigned to the first finger, and C and B both to the fourth; and the hesitation remains, in respect of C and B, in lines 19 and 25.

c15.s1.14

It is clear also that the scale consisted normally of only five notes for a considerable time. In the *Puṣpasūtra* (9. 26) we find that ‘the Kauthumas sing the majority of their chants to five tones, a few to six, and two of them to seven tones’. In line 12 we find Sol-fa names for five notes only. That no more than five were contemplated may be seen from the choice of consonants. They are the medials of the five classes of consonant (guttural, palatal, cerebral, dental, and labial) proceeding in regular order from throat to lips. Lines 17 and 25 are a further indication that the normal scale consisted of five notes. All this lends some support to the interpretation given above of line 1.

c15.s1.15

The scale was increased to six by the addition of *kruṣṭa*. That this was an afterthought is seen in the fact that it is independent of the nomenclature by ordinals, in the hesitation as to *kruṣṭa* or *kṛṣṭa*, and in the confusion introduced into the musical hand (lines 19 and 25). Finally, a seventh note was tentatively admitted in the *atisvārya*.

c15.s1.16

*Consonance* (*saṃvāditva*). The consonance which was suggested by two tetrachords, F–C and E–B, is carried out further in line 23, where the seven notes are divided into two similar tetrachords. This consonance holds whatever the intervals may be between the notes within the tetrachord. For the intonation of these we depend on the identifications in lines 14 and 21 of the *Sāman* notes with those of the secular scale, which, as we saw (p. 114), was at least as old as the oldest of these treatises.

c15.s1.17

In lines 6, 15, and 21 Sa is identified with tṛtīya, and this gives the Sa-grāma in a plagal form from N(i). In line 14 Sa is identified with atisvārya, and this gives the Ma-grāma in a plagal form from N(i), or the Sa-grāma in an authentic form from M(a). But S(a) and P(a) are saṃvādī; therefore the notes are D and A (not D and A). The two scales are therefore

c15.s1.18

Sa-grāma. EX. 350.

c15.s1.19

and

c15.s1.20

Ma-grāma. EX. 351.

c15.s1.21

We saw in Chapter IV that the Gandhāra-grāma was, according to one explanation,

c15.s1.22

EX. 352.

c15.s1.23

and it is possible that the Sāman scale may throw some light upon it. That scale is reckoned downwards, and therefore the note E is called R(i); but the grāmas are reckoned upwards, and therefore E is G(a). If the Ga-grāma begins on E, it has for its upward tetrachord E, F, G, A, that is, a semitone and two major tones. These make more than a Fourth, therefore either the semitone is the Pythagorean comma (256/243, 90 cents), or one of the tones is a minor tone, as in the lower tetrachord, E, D, C, B. The upper tetrachord would in the first case be the same as the early Greek Doric, and would contain the ditone, instead of the major Third. It seems likely that in some way of which we cannot now trace the history, the gāndhāra-grāma is a relic of this old tuning. The Comma of Didymus (81/80) has to be negotiated somewhere in the scale. We, in Europe, put it between the A and the D (major tone, semitone, major tone), the Indians between the E and the A (semitone, major tone, major tone). In so placing it, it is possible that they reduced an original E 256/243, F 9/8, G 9/8, A to a later E 16/15, F 9/8, G 9/8, A.

c15.s1.24

We have other hints which point to a correspondence between the Greek and Indian scale. The Aryans came into India from the North. Gāndhāra is the same word as Kandahar, which as late as Alexander's time (fourth century B. c.) included Afghanistan and the Panjab, with a capital near Peshawar. The Gāndhāra-grāma is the oldest of the three grāmas, and may preserve an early Aryan tradition. The Aryans spoke Sanskrit, which is closely allied to Greek. Evidence of the correspondence of the Greek and Indian systems has been accumulating in these pages; and that the fundamental scale of Greece, the Doric tetrachord, should be identical in form with the oldest form of Indian tetrachord would be more striking evidence than any. But it is unfortunately not more than circumstantial.

c15.s1.25

The Gāndhāra scale gave its name no doubt to the Gāndhāra note. The names Madhyama and Pañcama have been discussed in Chapter V. Of Ṛṣabha, Dhaivata, and Niṣāda nothing is known; they may be place-names, like Gāndhāra, or fancy names, like the majority of the śrutis. Of Ṣadja there is something more to be said. Referring once more to the list above we find in line 4 the seventh note left without an assigned organ of the body. In line 14 that same note is called Sa. We remember also that the atisvārya is not a regular note of the scale, but is introduced into it as an extension of the other six. Elsewhere, in the Sangīta-sāra-sangraha, for instance, ṣadja, which undoubtedly means ‘born of six’, is said to be produced from six organs of the body, viz. nose, throat, chest, palate, tongue, and teeth. Putting these together we see that the name ṣadja is merely the mythological expression of the musical fact that the atisvārya was an ‘extra’ note on a different footing from the others.

c15.s1.26

There is, however, another tradition as to the intonation of the Sāman scale. M. Śeṣagiri Śāstrī (Descriptive Catalogue, pp. 3, 4) gives the scale as Ma, Ga, Ri, Sa, Dha, and says that the sound is that of the Rāg Abhogi. That Rāg is:

c15.s1.28

Drone D¹ (which does not matter here, the drone being a later invention), *aṃśa* E, omitted notes C and A. This is an old tradition, as may be seen from three passages: (1) in an early book of the *Mahābhārata* the notes of the secular scale are given as Sa, Ri, Ga, Ma, Dha, Pa, Ni; (2) in the *Nāradaśikṣā* (I. 2. 16) the order is Ma, Ga, Ri, Sa, Dha, Ni, Pa; and (3) in I. 7. 4 it is Ma, Ga, Ri, Sa, Dha, Ni (Pa omitted). In all three passages the first five notes are the same, and the last two are out of order; this looks as if the first five were substantive and the last two optional, thus:

c15.s1.29

¹ The European note (not the svara).

c15.s1.30

Ex. 354.

c15.s1.31

and this is the scale of Śeṣagiri Śāstrī and of the Rāg *Abhogi*; and since the *aṃśa* of *Abhogi* is at E, it divides thus:

c15.s1.32

Ex. 355.

c15.s1.33

and its lower tetrachord is similar to the *usus* of the *Āhvārakas* given in line 26 of the list. We must accept this transilient scale, therefore, as a collateral form of the *Sāman*.

c15.s1.34

*The musical hand.* The basis of numeration is in some countries five, in others a dozen (England) or a score (France); in India it is four. Rupees, for instance, are frequently counted by placing one finger on each and sweeping them into the hand. Bearing this in mind, it may be possible to introduce some method into these discrepant accounts of the musical hand. The root idea of this ‘cheironomy’ is that the thumb touches successively the tips of the four fingers. This is seen in the *svāra*, or cadence, where the hand is held ‘in the shape of a cow’s ear’ (see the second of the two illustrations opposite), and the thumb passes over the four fingers in succession. Line 17 may accordingly represent two systems of counting:

c15.s1.35

| | Ga | Ri | Sa | Ni | Dha |
| (*a*) | | 1st | [2nd] | [3rd] | 4th |
| (*b*) | 1st | [2nd] | 3rd | 4th | |
| (line 17) | 1st | 1st | 3rd | 4th | 4th |

c15.s1.36

Similarly line 19:

c15.s1.37

| | Ma | Ga | Ri | Sa | Ni | |
| (*b*) | | [1st] | 2nd | 3rd | 4th | |
| (*c*) | thumb | [1st] | 2nd | 3rd | 4th | (pentachord) |
| (line 19) | thumb | thumb | 2nd | 3rd | 4th | |

c15.s1.38

Sāmagah, Madras.

c15.s1.39

Sāmagah, Bangalore.

c15.s1.40

Lines 24 and 25 combine these. We have therefore:

c15.s1.41

| | Ma | Ga | Ri | Sa | Ni | Dha | |
| (a) | | | E | D | C | B | (tetrachord) |
| (b) | | F | E | D | C | | (tetrachord) |
| (c) | G | F | (F♯) | E | D | C | | (pentachord) |

c15.s1.42

as bases, perhaps successive bases, of the Sāman scale.

c15.s1.43

*The accents.*

c15.s1.44

| | Ma | Ga | Ri | Sa | Ni | Dha | Pa |
| (i) line 6 | | U | S | P | A | | |
| (ii) ,, 16 | | | S | A | U | P | |
| (iii) ,, 23 | S | U | A | S | U | A | S |

c15.s1.45

There was a doubt, as we saw, about the relative position of udātta and svarita: in the Rgveda the svarita is above, but there is reason to think it may originally have been below the udātta. This doubt, reflected in (i) and (iii) above, is natural if we consider the nature of the svarita, which started on the udātta, rose a little above it, and fell much below it. It was natural that in recitation, where the question of exact pitch did not come up, the fall should be the conspicuous thing, and in its general effect the svarita come to be considered below the udātta; but that, when the pitch came to be defined in chanting, the rise should rather attract attention, since a fall or cadential phrase took place for other reasons than the fact that the syllable had a svarita accent, whereas the rise was peculiar to the svarita note, and therefore in its particular effect the svarita was above the udātta.

c15.s1.46

That in (i) the pracaya is above the anudātta and in (ii) below it need cause no trouble, since, as the pracaya is ‘toneless’, there is really no sense in assigning it to any particular tone.

c15.s1.47

The position of the anudātta above the udātta in (ii) is more difficult. Perhaps we are justified in calling Ga (in ii) udātta on the analogy of (i) and (iii). This would fulfil all that is required of the pracaya and anudātta, that they should be somewhere below the svarita (and the upper udātta). The lower udātta then appears as the other terminal of the tetrachord, with the svarita above it.

c15.s1.48

In (iii) the accents appear to lose their grammatical meaning entirely and to be given a musical meaning. The udātta is now the terminal of the tetrachord, at Ga and Ni, with the svarita above it and the anudātta below it. The udātta is ‘raised’ above the others not in pitch, but in musical eminence. The svarita loses its sense of ‘graced’ note, and becomes a note of certain relative pitch. The *anudātta* is no longer of indefinite ‘low’ pitch, forced under certain conditions ‘lower’ (*anudāttatara*), but becomes definite.

c15.s1.49

The theory of consonance thus determined has its counterpart in Greek music:

c15.s1.50

c15.s1.51

These are quite different from Hucbald’s *excellentes*, *superiores*, *finales*, *graves*, with which Burnell compares them, and which involved a series of disjunct, not conjunct, tetrachords.

c15.s1.52

An example of the chant as sung by a *Sāmagaḥ* in Madras shows the general musical effect. The breve at the beginning shows the actual pitch of the first note (a Fifth higher); the chant itself is given at the pitch which suits the scale just discussed. The fragmentary syllables are merely phonetic spellings of the apparent sounds of the phonograph, with gaps where the words are inaudible. It was extremely difficult to get even this short specimen.

c15.s1.53

Sāman chant. Madras. (Phonogram.)

c15.s1.54

Ex. 356.

c15.s1.55

Herr Felix Exner was far more successful in 1904. His results have now been worked out by Dr. Erwin Felber, provided with text and translation by Herr Bernhard Geiger, and were published in 1913 under the title *Die indische Musik der vedischen und der klassischen Zeit* as a contribution to the history of recitation. Dr. Felber has very kindly given permission for their transcription here. The actual pitch is given, as before, by the breve at the beginning of each example. The *Sāmaveda* notation with which each of his examples is supplied is omitted here, because it is probably the notation of a different sect from that to which the singer belonged, who has, at any rate, taken no account of it except in a very general way. Felber's numbers are given for reference. He discusses the melodic structure of these chants in his third chapter (pp. 38–43).

c15.s1.56

No. 427. Sung by a Brahman priest in Calcutta, forty-two years old, born in Jodhpur. Original text:

c15.s1.57

Mahi trīṇām avar astu dyukṣaṃ mitrasyāryamṇaḥ | durādharṣaṃ varu[ṇasya] |

c15.s1.58

Translation:

c15.s1.59

May the mighty aid of the three gods be ours—the help of Mitra and Aryaman and the invincible Varuṇa.

c15.s1.60

Felber. No. 427. ♩=79. *Sempre glissando, piano.*

c15.s1.61

Ex. 357.

c15.s1.62

Molto allegro. hā - - - i ma-hi t-rī - ṇām a - va - ra - stū -

c15.s1.63

e dyuk - ṣam-mi-tra - syā - rya - m - ṇāḥ

c15.s1.64

du-r - ā - dhā - - - - - - - r - ṣā . . m va-ra-u

c15.s1.65

This example (No. 427), and three others (Nos. 428–30) which it has seemed unnecessary to transcribe, shows, like the Madras example just given, little more than the scale of the Ṛgveda (see line 28 of the diagram on p. 259):

c15.s1.66

Ex. 358.

c15.s1.67

the F being, on the whole, the most prominent note.

c15.s1.68

Felber 443. Sung by two Brahmans of Tanjore aged forty and seventeen—a priest and a student in Madras.

c15.s1.69

Original text:

c15.s1.70

punānaḥ soma dhārayāpo vasāno arṣasi |
ā ratnadhā yonim ṛtasya sīdasy utso devo hiraṇyayaḥ ||

c15.s1.71

Translation:

c15.s1.72

Purified, O Soma, by drops, thou movest concealed in the waters; rich in treasure thou residest in the womb of the sacrifice, a golden fount divine.

c15.s1.73

Felber. No. 443. = 106. Agitato. Ex. 359. f pu-nā . . . . nā - ḥ so . . ma dhā - rā . . yā

c15.s1.74

ā - po vasā - - - na-ā . . . . sā . . . sī

c15.s1.75

ā - rā - t - na - dhā - ḥ yo ni mṛ-tā . . .

c15.s1.76

Musical notation for "sya sā . . . . dā . . sī ut-sā - ḥ dā-i-vo . . ."

c15.s1.77

Musical notation for "hi - rā . . . . nyā . . yā - ḥ."

c15.s1.78

Felber, No. 444. Another version of 443, sung by the Madras priest alone.

c15.s1.79

Original text of the second verse:

c15.s1.80

duhāna ūdhar divyam madhu priyam pratnaṁ sadhastham āsadat | āpṛchyam dharuṇaṁ vājy arṣasi nṛbhir dhauto vicakṣaṇaḥ ||

c15.s1.81

Translation:

c15.s1.82

Milking the honey he loves from heavenly udders, Soma took his seat on his ancient throne. Mightily thy stream empties itself into a vessel of honour, and purified from the touch of men thine eyes behold what is far away.

c15.s1.83

Felber. No. 444. d=116-149. Agitato.

c15.s1.84

Musical notation for "Ex. 360. pu-nā-naḥ so . . mā - dhā - rā . . . . . . yā . ."

c15.s1.85

Musical notation for "ā . po - sā - no a - r - ṣa - syā ra-tna-dhā yo"

c15.s1.86

Musical notation for "ni - mṛ ta - sya sā . . . . . . i - da - sā - i"

c15.s1.87

Musical notation for "o - hā - u - vā ut-so de - vo hi-rā . . . hā - i"

c15.s1.88

Musical notation for "o - hā - u - vā ṇya - yā . . . au - ho - vā"

c15.s1.89

Musical notation for "ut-so de - vo hā - i - ra ṇyā - - - - yā - ḥ"

c15.s1.98

The scale of these two (Nos. 443 and 444) is the same:

c15.s1.99

Ex. 361.

c15.s1.100

the melodies pivot on F and D. The chromatic beginning of No. 444 is quite common in secular singing, from which it seems to have been imported here, in order to establish a tonic.

c15.s1.101

The two which follow show a different scale:

c15.s1.102

No. 426.

c15.s1.103

Ex. 362.

c15.s1.104

No. 425.

c15.s1.105

Ex. 363.

c15.s1.106

The F♯ is essential in No. 425 and accidental in No. 426; in both it is used chromatically, which is on the whole unlike secular song. I heard something like it in Madras, but was unable to note it. Two chants were sung: one distinctly chromatic, described as *ucca* (high); the other as distinctly diatonic, described as *nīca* (low). I was more than once asked by the singer whether he should sing in the *ucca* or the *nīca* voice; and it seems possible that he may have been referring to a difference of scale rather than of pitch—*ucca* G F♯ F♯ E, &c., *nīca* (G) F E D, &c.

c15.s1.107

No. 426. Sung by a boy of fifteen in the Sanskrit College, Calcutta.

c15.s1.108

The text is not from the *Ṛgveda*, but a *stobha*, i.e. composed not of verses but of jubilations consisting of sentences and alleluias. (*ho-i*, &c.). In the Pada form:

c15.s1.109

ut nayāmi ādityam ādityam prāñcam yantam ut nayāmi ahorātrāṇi ahaḥ rātrāṇi aritrāṇi dyauḥ nauḥ tasyām asau ādityaḥ ādityaḥ īyate tasmin vayam īyamāne īyāmahe [. . .].

c15.s1.110

Translation:

c15.s1.111

I bear aloft the advancing sun, I bear it aloft; days and nights are the oars, the sky is the ship. On this fares forth the sun; and with it we are borne to that well-loved spot whose name consists of three letters.

c15.s1.112

*Note.* That which is named in the adoration of *om* (spelled a + u + m), which is an image of the world of Brahma.

c15.s1.113

Felber. No. 426.

c15.s1.114

♩=88. Sempre accel. Ex. 364. un - na-yā-mi ho - i un - na-yā-mi ho - i un - na-yā-mi ho - i ā - di-tyam prāñ-caṃ tam un - na - yā-mi ho - i ā - di-tyam-prāñ-caṃ tam un - na-ya-mi ho - i a - ho - rā - trā - trā - ṇya-ri ho - i a - ho - rā - trā - trā - ṇya-ri ho - i

Linked texts (15)

Nāradīyāśikṣā — Nārada, Sanskrit
Ṛgvedasaṃhitā — Sanskrit
Bṛhaddevatā — Śaunaka, Sanskrit
Taittirīyaprātiśākhya — Sanskrit
Puṣpasūtra — Gobhila & Vararuci, Sanskrit
Aitareyabrāhmaṇa — Mahidāsa Aitareya, Sanskrit
Māṇḍūkīśikṣā — Māṇḍūka, Sanskrit
Mahābhārata — Vyāsa, Sanskrit
Sāmavedasaṃhitā — Sanskrit
Atharvavedasaṃhitā — Atharvan & Aṅgiras, Sanskrit
Sāmatantra — Sanskrit
Saṅgītasamayasāra — Pārśvadeva, Sanskrit
Pañcavidhasūtra — Sanskrit
Ārṣeyabrāhmaṇa — Sanskrit
Sāmavidhānabrāhmaṇa — Sanskrit

The Music of Hindostan c15: Chapter X. The Sāman Chant

Contents

Chapter X. The Sāman Chant

Authorities for the Sāman Scale