Society for Theory

Variation and Process in South Indian Music: Some "" and Their "Sangatis" Author(s): Robert Morris Source: Spectrum, Vol. 23, No. 1 (Spring, 2001), pp. 74-89 Published by: University of California Press on behalf of the Society for Music Theory Stable URL: http://www.jstor.org/stable/746059 Accessed: 14/10/2008 19:11

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=ucal.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].

University of California Press and Society for Music Theory are collaborating with JSTOR to digitize, preserve and extend access to Music Theory Spectrum.

http://www.jstor.org Variation and Process in South Indian Music: Some Kritis and their Sangatis

Robert Morris

While Caratic music, the classical music of South India,' has There are many different musical forms used in Carnatic com- its roots in the religions and cultures of the Indian subcontinent, it positions. The most prevalent are the , kirtana, , also has some surprising similarities with the common-practice art , and javali.4 In this paper I focus on perhaps the most im- music of the West. For example, concerts of Camatic music feature portant of these, the kriti. I identify a number of melodic and performances of completely composed devotional , most rhythmic procedures that project long-range continuity and pro- created by great musicians of the past.2To be sure, these composi- mote a sense of autogenic process in two contrasting kritis, tions are interspersed with highly structured types and forms of "Karpagame"and "Mokshamugaladha."The musical analyses also improvisation, but there is a large concert repertoire of pieces that address issues of compositional identity and change in a tradition are performed essentially identically-within a variety of distinct where fixed performance and improvisation exist side by side. traditions and styles-from one concert to another.3Most of these Before I describe the basic musical concepts and practices of vocal compositions are not simple monodies, but, like their , I need to point to three features that make its Western counterparts, are complex and multifaceted composi- analysis difficult for Westerners. First, there is the problem of no- tions, often lasting eight to ten minutes. tation. Contraryto common knowledge, Indians do have an effec- tive notation for their music, but it demands intimate knowledge of and structuraldetail. It cannot be read 'Caratic music is to be distinguishedfrom North Indian, or Hindustani, performance practice by music. While both are based on the same melodic and rhythmicprinciples, their a musical outsider any more than a lead sheet, fake book, or specific , talas, and performancepractices are different.In addition,North arrangementcan be read by a musician versed only in the perfor- Indianmusic has been influencedby Islam and the musical culturesof Pakistan mance of Western classical music. A number of scholars and mu- and Persia. sicians, both Western and Indian, have transcribedCarnatic music 2This concert format is relatively recent, introduced by the renowned Carnaticcomposer/singer Ariyakudi Ramanuja Iyengar in the 1930s. 3North Indian music has a different performance practice and setting. constraints,sometimes with and sometimes without a rhythmicframework. A Ratherthan being mainly a recital of composed songs, a North Indian concert similar item on a Carnaticconcert is the ragam-tanam-,which would is comprisedof only a few but lengthy concert items, each involving improvi- occupy perhapsa thirdof the concert. sation on a single based on short phrases from old compositions often 4Tamiland Sanskritwords are renderedin transliterationaccording to Pan- only known to professionalmusicians. Much emphasis is placed on the artist's Indian conventions, without diacriticalmarks, and with pluralsformed by the ability to develop the characterof the raga successfully within improvisational additionof "s." Variation and Process in South Indian Music 75

into Western music notation, but our notation is not suitably struc- There are thousands of ragas, but only about one hundred are tured to convey essential features of Indian music.5 ubiquitous. A given composition or improvisation is usually based A second problem is the ubiquity of melodic ornamentation, on only one raga; however, a sequence of different ragas, called a much of which involves continuous modulation of pitch; straight raga-malica, is permitted in certain musical genres.6 tones without ornamentationare, in fact, deviations from the norm. As mentioned above, talas are cycles of beats (aksharas) that Indeed, while the term is correctly translated as "scale de- are continuously repeated. They provide a rhythmic framework gree," it also denotes ornamentation. In this way a svara might be for composition and improvisation. As with ragas, a given concert better defined as a precise but fluid melodic shape performed on item will use only one tala throughout.7Each tala cycle (avarta) and in precise vicinities of fixed, finely tuned scale degrees. is divided into subdivisions called angas.8 The most common A third difficulty for the Western analyst is the of talas are: Adi of 8 beats; Rupakam of 6; Mishra Chapu of 7; and Carnatic music, which is based on additive cycles of beats called Khanda Chapu of 5. talas. The music coordinated by the beat-structure of a tala need A kriti is a three-section, vocal composition that sets a text in not project the tala in any way; in fact, the music can employ all Sanskrit or a South Indian language, usually Telugu, Tamil, or kinds of deviations and that become sensible only Kannada. The text often alludes to Hindu myths and religious lore against an internalized grid of tala beats. Despite these problems, and exhibits much wordplay and wit. Quite a few kritis are devo- a novice can learn to appreciate Carnatic music if issues of tional, asking for succor and spiritual support from a personal notation, ornamentation, and rhythm are specifically addressed in god. The kriti was developed into its present form by the com- analytic discourse. poser in the early nineteenth century, who introduced Indian music is based on sophisticated rules and practices, but into the kriti extensive melodic/rhythmic variations called sanga- none of these pertain to what Westerners would define as tis. Before Tyagaraja, kritis were for the most part usually com- or counterpoint. These rules and practices are codified into two posed of only a few phrases, which would be repeated with slight structural concepts: raga () and tala (rhythm). Ragas are variation; but with his use of sangatis, the kriti was expanded into partially ordered sets of pitch-classes, each with its own set of lak- a long, dynamic composition, often containing a climax replete shanas, or structural features. Ragas occupy a place somewhere with intricate motivic development. I should point out that some between a tune and a scale. Raga lakshanas specify a raga's reper- sangatis in many kritis performed today were not only created by toire of pitches (which can vary from four to twelve), its ascend- the kriti's composer, but were added by more recent musicians in ing and descending melodic patterns, characteristic motives and the composer's lineage, who usually attribute their additions to phrases (called sancharis), beginning, cadential, and stressed the original composer. Furthermore,master musicians may intro- (jiva) pitches, melodic ornamentation, and the like. Because many duce improvised sangatis in performance that range from slight ragas are associated with a time of day or season, social function, any one or more of nine basic affects (rasas), and, in some cases, 6Forlists of South Indianragas with theirproperties and features,see Bhag- particular gods and goddesses, a raga can set an emotional mood yalakshmy1990, SubbaRao 1996, and Kaufmann1976. My descriptionsof the and Saramatiare drawn from as well as become a complete and identifiable musical world. ragasMadhyamavati primarily Bhagyalakshmy. 7However,there are performancegenres that are not coordinatedby talas, such as alapana and tanam. Fortranscriptions of Carnaticmusic in Westernnotation, see Catlin 1980, 8Forinstance, the three angas of the seven-beat tala Misra Chapu are of 3 Kumarand Stackhouse1987, Powers 1958, Wade 1988, and Visvanathan1974. beats, 2 beats, and 2 beats. 76 Music Theory Spectrum

variations on the sangatis already in the composition to novel "Karpagame," the first of the two kritis I will discuss, was variations that may be remembered and retained in future perfor- composed by the twentieth-century composer Papanasam Sivan." mances. In this way, some kritis have come to have multiple ver- Example 1 provides basic information about "Karpagame."The sions, not all of which are accepted by every Camatic lineage. composition is set in the raga , a pentatonic raga. The three sections of a kriti are called pallavi, anupallavi, and The tonic is notated as C,12 and the most important cadential . The pallavi and anupallavi are set melismatically while melodic tones are the second, fifth, and higher octave. The stressed the charanam is usually more syllabic. The text setting must con- jiva are the second, fourth, and scale degrees.3 form to the rhythmic structureof the tala, which may set the text's Ragas often prescribe ornaments: in Madhyamavati these include meter in a predictable mapping of feet to beats. The first line of periodic vibrato undulations on the second degree touching a pitch the pallavi text is similar to a refrain; it is repeated at the end of about a sixth-tone higher; other ornaments used in this raga are the pallavi, the anupallavi, and sometimes the charanam. This rep- shown in the example. Some of these ornaments many confuse the etition is reflected in the music so that the first musical phrase of Western ear. For example, the ornament on the fourth degree, F, the pallavi (henceforth, the pallavi phrase) finishes the anupallavi, (in the fifth "measure"of the first line of the example) slides from sometimes preceded by a transition I call a return phrase.9 The G to E or vice-versa, which suggests that the note E is perhaps in charanam may or may not return to the pallavi phrase, and there the raga. But this is not the case. may be more than one charanam in a kriti. The pallavi usually is The tala used in "Karpagame"is Adi, the rhythmic cycle of composed around the tonic note in the middle register, or madhya eight beats mentioned above, partitioned into angas of four, two, sthai, while the anupallavi is in the higher range, or tara sthai, and two beats, as shown in the example. To the lower right of the usually around the tonic note an octave higher. ? example the form of this kriti is outlined. There are four main While the kriti is a vocal form, it is often adapted to instrumen- sections to "Karpagame":the pallavi, the main section in the mid- tal performance, a natural development since the Camatic vocalist is almost always accompanied by an instrument performing in "The performancetranscribed in Examples 3, 4, and 5 was played on the vina (a lute) the celebrated virtuoso S. Balachander and was unison. When kritis are played as instrumental solos, however, plucked by recordedin the late 1960s. The omits the (drum) im- be to make them more effective transcription mrdangam they may idiomatically adapted provised .The choice of Balachander'sperformances to intro- concert pieces. This practice, of course, is familiar to us in the duce niceties of Caratic music to Westernmusic theoristsmay surprisethose West, where popular vocal music is adapted for performance on alreadyknowledgeable in Indian music, since Balachanderwas either reviled saxophone, trumpet, and other instruments in jazz sets. But many or praisedby critics for often introducinghis own sangatis and variationsinto of the traditionalliterature. Nevertheless, I have Bala- Carnatic instrumentalists take great pride in playing kritis in an performances compared chander's of (NonesuchHS-72003) and "Mokshamuga- idealized vocal that the characterof Carnatic recording "Karpagame" style suggests typical ladha"(Odeon 1044) with other,more traditionalvocal performancesby K. V. vocal performance practice. Narayanaswamyand Alleppey Venketesan(both disciples of AriyakudiRama- nujaIyengar) and found Balachander'srecordings comparable. '2Notatingmodal compositions in C is an ethnomusicologicalconvention that facilitates comparativeanalysis. The tonic pitch of Balachander'svina is 9Thereseems to be no correspondingCaratic termfor "returnphrase." approximatelyEb. '?Thegeneral range of a kriti may cover three octaves. Carnatictheory uses '3Eventhough Madhyamavati is a pentatonicraga using five svaras,I num- the termsthai to denote an octave rangefrom tonic to tonic; the lowest octave is ber scale degrees here based on the universeof seven svarasavailable in Indian called mandrasthai; the middle,madhya sthai; and the highest, tara sthai. music. Variation and Process in South Indian Music 77

Example 1. Overview of "Karpagame"by PapanasamSivan Composition:"Karpagame" by PapanasamSivan Pallavi: Raga Madhyamavati eduppu arudi scale: ornaments: A J? A >> > > --L ~--C ~or for or for r* * X - I -!. m J I ui-ci vt o/ T) P r tonic (oscillation) (Whole notes indicatte (oscillation) Anupallavi: cadentialpitches; accented notes indic;ate (jiva) stressedpitche s.) TalaAdi 8 beats = 4 + 2 + 2 Charanam1: 1 2 3 4 5 6 7 8 Charanam1: r r r r clap 5th 4th 3rd clap wavewave clap wavewave finger tap finger tap finger tap die and lower part of the tessitura of the composition; the anu- Charanam2: pallavi, the second section in a higher range, centering on the high tonic;14 and two , or verses, followed by the return 4 ara.U 1 phrase and the pallavi phrase. The pallavi and anupallavi are filled out by sangati phrases. The example shows the first phrase of each of the four sections. Looking at the pallavi phrase, we see that it begins on the last quarter of the first beat of the tala. This starting Kriti form: place is called the eduppu, and is usually not found on the first Pallavi (+ sangati-s) beat of the tala in Tyagaraja's music. As we will see, all melodic Anupallavi(+sangati-s) Returnphrase (3x) phrases that return to the pallavi phrase must progress to the Another in the is the eduppu. important point pallavi phrase arudi, Charanam1 (+repeats) which is on beat 5. The arudi (also called padagarbham) is the di- Charanam2 (+repeats) viding point of the pallavi phrase and functions as the consequent Returnphrase (3x) of the music beginning with the eduppu.'5 Pallavi Phrase Listening to the performance of "Karpagame"that I transcribed discloses that the notation in Example 1 presents a simplified '4Theanupallavi is followed by a returnphrase that provides a brilliantre- transitionto the opening phraseof the pallavi. '5Fora definitionof arudi,see Sambamurthy1964, 23. 78 Music Theory Spectrum

Example 2. Notations of Carnaticmusic while destabilizing the opening pallavi phrase, prefixing and shift- ing its pitch motives in various In 3(a), the Detailednotation ways. Example pallavi phrase is followed by eight sangati phrases, numbered 2 through n 9, that are aligned beat-for-beat with each other and with the SJ ? ai5or, 5 Pallavi phrase. We can discern a slow but inexorable process of rhythmic saturation over the course of the section. Only the Dt on the third beat appears essentially unmodified by rhythmic elabora- Simplifiednotation tion, but even this is not the case in the actual performance, for a detailed transcription would show that the vibrato evolves toward 41 r 9 Y I n i I I greater complexity over time. Nevertheless, beat 3 and, to a lesser extent, beat 7 are the pallavi's stable beats. Note how beats 3, 7, and 5-the arudi-support one of the main tones of the raga. The Carnaticnotation position of the eduppu presents a mild rhythmic challenge and functions as an "upbeat"to the D of beat 3. Lines 2 and 3 place a ,,,s ,rs, r , rs r , m, Ir , s mp C at the very beginning of the tala, thereby stabilizing the rhythm. However, the addition of the tie from the end of line 3 to 4, which image of the music. On the top of Example 2, a much more accu- serves to introduce the low B, (performed with a sliding orna- rate notation is given for the first phrase of the pallavi. There we ment), produces a greater complexity to the onset of the eduppu. can appreciate the nuances of rhythm and ornamentation that give But this new Bb also creates a new pattern: the third beat is now life and character to the performance. A simplified notation medi- prefixed by the pattern of two sixteenths and a sixteenth rest ates between the rigors of detailed transcription and a particular played twice, which recontextualizes the function of the eduppu dialect of Indian notation shown below the simplified transcrip- by extending the beginning of the phrase back one sixteenth. The tion. This dialect shows rests or ties by commas, pitches by abbre- end of line 4 pushes the D and C (previously on the last two beats viated Indian solfege,'6 and rhythm by letting each symbol stand of the tala) earlier, so that a new prefix of sixteenths can be placed for a half-beat or, if underlined, a quarterbeat.17 In order to use the before the Bl in line 5. The end of line 6 contains another prefix. Indian notation one has to know the intricacies of the raga and tala These rhythmic, melodic, and other modifications complicate the systems as well as remember the proper ornamentations of the profile of the music so that each time a new sangati is played, the music. The simpler notation will serve well enough for most of relative positions of rhythmic stability and repose within the tala my analysis, but I will use more detailed transcriptions when move and eventually exchange. These transformations are more needed. complex than those found in similarly structured passages in be- Now let us examine the transcriptions in some detail. We will bop jazz and in the music of Steve Reich. see that the sangati phrases expand the pitch range of the music In a second pass through the nine lines of Example 3(a), we will follow the upper contour of the music to see how the highest 16Theseven scale (swaras)are named,from low to Sa, Ri, Ga, degrees high, of each are treated over the The note G Ma, Pa, Dha, Ni. See Wade 1988, 37. pitches sangati pallavi. '7Forinstance, some other forms of Indian notation use capital letters for remains the highest until the Bl on the last beat of line 6, after notes twice as long as notes writtenin lower case. which the range expands one tone higher to C in line 7. Mean- Variation and Process in South Indian Music 79

Example 3(a). Pallavi of "Karpagame"with sangatis while, the lower side of the music extends from C to low G. The three of the lines 1-5 always appear after rather than PallaviPhrase: high pitches at the beginning of the tala's anga and only within rather than on beats. Furthermore, these high notes are in the middle of orna- 1. ments until the end of line 4, where the G begins to introduce the prefix of the eduppu mentioned above. The first appearance of a high note on a beat occurs in line 6, but its articulation is some- 2. what compromised by the ornamental at that point. On line 7, beat 6, the most stable high pitch yet, a C surrounded by even sixteenths from D to D, suggests a motivic transformation(by oc- or 3. ri I~ ~ 3 tave displacement) of the local D-C-D motion of the eduppu arudi. Only the eighth line has a high note sustained for more than a quarter of a beat, but it is reached via a glide on the second half of beat 6 and dies away, only to be abruptly brought down in 4. thirty-second notes to the tonic of the raga at beat one of line 9. U. w This gradual strengthening of the metric position of high notes brings them into relief, preparing for the register of anupallavi, 5. which stresses the high C throughout. Focusing on the durations of groups of notes provides another view of the pallavi. On the left of Example 3(b), the music from 6. the first two beats of lines 2 and 3 is grouped into a 3-3-2 pattern, R~~~j according to placement of the tonic note that begins each division. In contrast to this grouping, the next line of the pallavi, line 4, begins without a pitch on beat 1. Instead, a C of six sixteenths' 7. duration passes over the beginning of beat 1 to reach the B on the second half of the beat. This extension induces a 6-3-3 pattern that moves the tonic note to the second place within the 3s of the pat- 8. tern, confounding the place of the eduppu, as noted earlier. In ad- dition, it demonstrates a principle of placing the same notes in dif- ~~b . nJ, d# JW;_01 ferent additive patterns simply by deleting and adding single 9. notes: in this case, the pair of 3s in the beginning of 3-3-2 is moved to the end of 6-3-3. While this observation may seem an overly detailed way of analyzing this admittedly minute fragment of the pallavi, it is a simple example of a particularly classic Indian form of transformation called prastara in ancient treatises on rhythm. Prastaraliterally means "spreading out," but it actually 80 Music Theory Spectrum

Example 3(b). Rhythmicdetails in line 2 of pallavi of "Karpagame" Considering all of the transformationsand relations introduced the I have cited-and there are more-it is remarkable line 3 line 3 line 4 by sangatis A =M that the core phrase from the eduppu to the arudi, except for the 7 7 low Gs at the on lines 6 and remains 9) 4 v9U eduppu 8, completely pre- * 1 I I ,3 , 3 2LL ,6 ,,3 ,,3 served and embedded in all sangatis. Like the pallavi, the anupallavi follows a course of increasing but one that contrasts with that taken the Example 3(c). Rhythmicdetails in line 3 of pallavi of "Karpagame" rhythmic complexity, by pallavi. Example 4(a) displays the eight lines of the anupallavi. line 4 line 5 Line la is not a sangati or even part of the composition, but a pas- sage filling in the time of one tala cycle prolonging the high C. ,3 ,3 ,4 , 15 15 The upward gesture on the last beats of the line is improvised by Balachander.This filling passage is a typical performance practice in anupallavis and is sometimes extended to two or three tala denotes a technique by which a series of durations is consistently cycles during which the performer may freely improvise. permuted to form rhythmic variety.18 The anupallavi contrasts with the pallavi in that the stability of The pallavi also illustrates another example of patterns of the arudi on the fifth beat of the pallavi theme is now undercut durations-those that grow or contract by single units. Example first by rests in line 1, beat 5, and then in later lines by a complex 3(c) shows the music from line 4, beat 6, to line 5, beat 3. I have ornament starting on the last half of beat 4 and lasting five six- bracketed the music here into the expanding pattern 3-3-4-5-5, a teenths. As we can see, the last four beats of the sangati phrase typically South Indian construction. Note how the two 5s, which gradually grow in rhythmic complexity, line by line. There are form a double prefix to the eduppu, are now shown to be logically two places of rhythmic stability in each sangati. One is on beat 3, connected to the preceding tala cycle. (Incidentally, such rhythmic which is preceded by the same 2-1-2 pattern from the eduppu patterns are routinely improvised by performers.) Native musi- found in the pallavi. The other occurs on beat 7 of each line ex- cians would say (using technical Indian music terms and con- cept lines 5 and 6, where the high tonic C is delayed until the last cepts) that this 3-3-4-5-5 pattern is easily constructed by taking sixteenth of the beat. These two points of stability cut across the twenty sixteenths of 5 beats grouped 4-4-4-4-4, and then taking 4-2-2 beat-pattern of the tala, since the middle of the tala's first one unit from each of the first two 4s and appending a unit to each division of four beats is made stable and the arudi on beat 5 is of the last two 4s.19We shall encounter a similar additive structure passed over. Heightening this against the tala's beat in the returnphrase. structure is a parallelism between the two stable beats. Both beats are preceded by the same 2-1-2 pattern, with the first 2 always ar- '8Prastarais the tenth category of the Tala Dasa Pranas. Sambamurthy ticulated by a long and percussive slide from G to C. This slide 1964, 182 defines prastaraas follows: "Prastaraof a tala anga means splitting promotes considerable rhythmic tension since it slides right over up the anga into its possible component angas and presenting them with all the first sixteenth of a beat. varieties in a tabularmanner." He shows the "8 Prastaras"of possible possible Examples 4(b) and 4(c) show how these parallel 2-1-2 antici- an of four beats as: 3 2 2 1 1 1 2 1, 1 2, 1 1 1 1. anga 1, 2, 1, 3, of beats 3 and 7 are even more articulated. 19Documentationof this kind of talk among Indianmusicians (using indige- pations prominently musical nous terms and concepts) can be found in the ten extensive interviews in Each 2-1-2 pattern lasts five sixteenths and is preceded by Nelson 1991. events that can be grouped into six followed by five sixteenths, Variation and Process in South Indian Music 81

Example 4(a). Anupallaviof "Karpagame"with sangatis AnupallaviPhrase:

A 1.

(fill) la.

2. L; r_ 'rr r r_r' ?r - " 3. ORSv02r rQSJ or I r

4. 1 4 7 -- _

5.

6.

5 1 3o 7. : ' , ~~~I+IES creating an underlying pattern of 6-5-5 leading to each stable beat. vides a basic outline of the music, which cleverly progresses back Once again, Indian performers are well aware of this rhythmic to the opening of the pallavi phrase, but stops short by sustaining partition: 5 times 3 is 15, 1 sixteenth less than 16 sixteenths, half the second degree with a microtonal vibrato. The more accurate of the tala cycle. transcription at the bottom of the example reveals that each six- The increasing rhythmic tension which builds up and through teenth note is either made up of a slide of two thirty-second notes the last sangatis of the anupallavi finds release in the return or left unornamented. Taking the unornamented notes as the last phrase. The transcription of the return phrase in Example 5 pro- event in a note sequence, we see that the end of the return phrase 82 Music Theory Spectrum

Example4(b). Rhythmicdetails in line 2 of anupallaviof "Karpagame" Charanam 1 oscillates about G, while charanam 2 starts and ends on C; both charanams are associated this line 2 1 high by high-low-high (2 2) contour. L , I- r Ir r : r In charanam 1, a fresh melodic idea occurs on beat 3 involving ,5 1 ,5 , 5 the notes F, D, and Bb--the jiva svaras of the raga. This collection or of notes is often reserved for passages near the end of improvisa- ,6 tions, so it is appropriatethat they appear in the charanam rather than earlier in the composition. As usual, the 7th beat is a place of detailsin line 3 of of Example4(c). Rhythmic anupallavi "Karpagame" repose, and a new rhythmic parallelism is found between the first line 2 line 3 and fourth beat. This parallelism and the stable 7th beat produce a 1 2) 3-3-2 beat which once cuts across the tala's 4-2-2 r-_ pattern, again design and recalls the 3-3-2 pattern found on the sixteenth-note level at the of lines 2 and 3 of the The second ,6 , ,5 , L5 , ,4 beginning pallavi. charanam conforms to the tala by having a rhythmic structure of 2-2-1-1-1-1 produced by cutting each anga of the tala in half: is based on the kinds of additive structures we have seen before. 4-2-2 becomes 2+2, 1+1, and 1+1. The ascending melodic se- Here, the pattern is so common that it has a name: gopuccha yati quence of beats 5 through 7 aptly sets some of the raga's charac- is the Carnatic term for systematic contraction in the length of teristic sanchari phrases, making this last charanam the simplest rhythmic patterns.20In this case, the process does not exactly land and most straightforwardsection of the piece. at the eduppu of the pallavi phrase, but rather overlaps it, ending Now I turn to a rather different Carnatic kriti, "Mokshamu- just before the second note of the pallavi. Alternatively, one might galadha," composed by Tyagarajaand set in the raga Saramati and say that the grouping pattern of 6-5-4-3 sixteenths ends on the Adi tala. Example 6(a) contains the first half of a transcriptionof a start of the prefix to the eduppu in the sangati on line 4 of performance by Balachander that was originally issued in 1965 on Example 3(a). Odeon Records.21The right side of the example shows the scale I shall not dwell on many features of the two charanam phrases of the raga. Its notes comprise our Aeolian mode, which is called found in Example 1. One can see that they are much more rhyth- in Caratic music. The raga is heptatonic in ascent mically stable than either the pallavi or anupallavi. Perhaps the and pentatonic in descent, an of notes that is some- lack of tied notes (except in one case) and the anapest rhythmic what rare among ragas in general; most ragas have the same or motives account for the charanams' direct rhythmic appeal, which fewer notes in ascent than in descent. Since the two notes used only comes in part from the practice of setting the text of a charanam in ascent, the second and fifth, are melodically vital in Saramati,the syllable for note-something one can only surmise from this raga has two faces to the knowledgeable Indian listener: one is as- performance on the vina. The overall melodic contour of both sociated with ascent, and stresses the fifth; the other is associated charanams also contrasts with contours in the earlier sections. with descent and has no fifth or second. The descent alludes to a well-known pentatonic raga called , which consists of 20Gopucchayati literally means "cow's tail," which tapers toward the end. The term denotes any transformationfrom large to small, fast to slow, many to few, in the structureof music or text compositions.There are six types of yati, 21Therecording was recently released on the compact disk S. Balachander which is the ninthcategory of the TalaDasa Pranas(see note 18). (Carnatic- Veena),EMI 147746. Variation and Process in South Indian Music 83

Example 5. Returnphrase of "Karpagame"

ReturnPhrase of "Karpagame"

bll 11 a r- e o I I I ,,,, ,,, l I I r I 3 I I I I - p I v ~ ~ ~ccc 2 1I 42 m - ^ C J J jJ

6 ?h^F n etc. (on to Charanams)

Phonetictranscription

6 5 4 3 r ^B 'Lirj.rjFtr'r ^ i t14r r r &r f '7fi ru ff ^S ^rjrj

,^^^^ f-

2 zzzzEzzI '7 7 etc. * 4'

eduppu only the descending notes of Saramati.22The third, sixth, and sev- The composition conforms to the kriti form but omits a distinct enth degrees of Saramati take sliding ornaments (just as they do return phrase (which is not an essential part of a kriti). As ex- in Hindolam), but the third also takes an oscillation ornament that pected, the pallavi phrase is situated in the lower octave, followed can make it sound as if it were a slow, measured between by the higher tessitura of the anupallavi. Nevertheless, the pallavi the second and fourth degrees. does have sangati phrases that ascend into anupallavi territory,as 22Thevina player Geetha Ramanathanmany years ago mentioned to me in line 8 of Example 6(a). The charanam, beginning on line 21, this connectionof Saramatito Hindolam. seems not to contrast much with the other sections because it is I_

84 MusicTheory Spectrum

Example 6(a). Transcriptionof "Mokshamugaladha"with sangatis Raga Saramati

1. -bl ji IJ J -I: 11 a ornaments J- J_ I s r >~' for ,I.t iF' b"-- b - 3. I2J. I J or for

5 4:6.~j~ , I3 -q ~C r 19<-r- for :8.4 77 r ..... :11 i2jj iy- I_ I _ 11

I _ = ;1 I I-IT ~~ - - - - I 4

12. _J a 4d

I J iJ R 14. 'bl l__F~F iO 1A - .,

s ~TI rf q J: -I . A Y I1 Variation and Process in South Indian Music 85

Example 6(a) [continued]

15.-r 1

16. r r

17. L1

18.

19. . .L ,L -l 1 1

20. j 11

C i i 5

7 22.A J7 J j /

23.- j = I1 J

24.- . J 256.-~K i7 ~~ 71

25. '2 J- ' --~.^ > ...... J

26., J Y, ?JJ | 86 Music Theory Spectrum

not set syllabically, although it is definitely less melismatic than via two sangatis to AS, and then further to F just before the begin- the pallavi or anupallavi, having a longer text that is set by only a ning of line 17. Finally, it descends low enough to launch the few pitches per syllable. However, because we are studying an in- pallavi phrase at line 19. The anupallavi thus exhibits an inverse strumental performance in the present analysis, the difference be- pitch contour in relation to the pallavi and complements it in an- tween the text setting of charanam and the other parts is lost. Even other way: the amount of ornamentation in the anupallavi is of a so, "Mokshamugaladha"has a much more prominent non-standard far greater density than that in the pallavi, taxing the ability of feature; while its pallavi phrase (line 1) is only one tala cycle Western notation to present it accurately. long, one of the sangati phrases (lines 8 and 9), as well as the anu- Rather than continue to the charanam, I want to return to the pallavi and charanam phrases, is two cycles long. This has led opening of the piece in order to examine the presentation of the some musicians and scholars to double the length of each tala beat pallavi phrase. The first three lines of the piece are almost identi- so that all phrases last one cycle except for the pallavi phrase, cal, starting on beat 2, which is the eduppu. The rise to the fourth which always has to be performed twice in a row in order to fill line, completed on beat 7, is accomplished by an oscillation that out one cycle. alternates D and F from beat 5 through 6. This alternation is actu- Considering the pallavi alone, we see how the sangati phrases ally an ornamental presentation of the third that is not directly induce a sense of climax and release. The first three lines simply heard in the pallavi phrase until line 5, after the first sangati on repeat the pallavi phrase, which is confined to the fourth from the line 4. There, the second note of the phrase is Eb rather than D. tonic C to F. This phrase has an upward contour, aptly setting the However, the raga's third degree is still represented by the oscilla- words: "Is salvation attainable?"The first sangati phrase on line 4 tion ornament on beats 5 and 6. After the second sangati phrase expands the range of the initial fourth by adding a phrase that rises has reached the high Eb in line 8, the pallavi phrase is articulated adjacently beyond the F to Bb before falling back at beat 6 via two to show its rising third without the oscillation. This occurs on line parallel two-beat phrases to the eduppu on beat 2. The next san- 9, beat 6. The clear rising pattern turns out to elide with the scalar gati begins on line 8, beat 2, and promptly climbs up to high Eb. rise to high C mentioned above. After this point in the music, the The descent from the 5th beat of the tala to the next line can be di- pallavi phrase retains the oscillation on beats 5 and 6, but its sec- vided into an initial upper decent from Eb through C to Bb, fol- ond note is always EB. lowed by a beat's move from Bb to F, and concluding with a two- This unfolding of the pallavi phrase throughout the pallavi sec- beat prefix of running sixteenths to the eduppu on line 9. Thus, tion is a special process in this composition, but one which is be- from lines 4 to 9 we see a double transposition of the pallavi's ini- lied by at least one South Indian source that presents "Mokshamu- tial tetrachord range: up a fourth twice followed by a retraction. galadha" in Indian notation: "GanamruthaKeerthana Malika, part Lines 9 and 10 summarize this up-then-down gesture with a rise IV," a series of Caratic compositions compiled and notated by though the scale from madhya C to tara C. This rise is simply A. S. Panchapakesa Iyer "for music students," as the author states adorned with upper neighbor inflections and is followed by a in his brief preface.23 I have notated Panchapakesa's version of highly ornamented motion through the descending pentatonic "Mokshamugaladha"on the first line of Example 6(b). Below this scale. One more presentation of the pallavi phrase in line 11 leads are the series of pallavi-phrase presentations discussed above, to a long oscillation on the third, ending with three tonic quarter plus the one on line 26, which occurs after the charanam in Bala- notes. chander's performance. As we can see, there are many remarkable The anupallavi, like the pallavi, tends not to stay in its charac- teristic register. From line 13, it gradually creeps down in line 15 23PanchapakesaIyer 1995. Variation and Process in South Indian Music 87

Example 6(b). Comparisonof pallavi of "Mokshamugaladha"as transcribedby A. S. PanchapakesaIyer with performancesby S. Balachanderand S. Ramanathan

line 1.

' :1 iC, 7 Tjj n nj3 X 113 line 5,

lalachander fromEx. 6) line 19. * (1 74 J L 7 j I; J

line 26

S. Ramanathan

1.- I I i 1 -2. -bA, r

I1L , k I I 3. ., I x differences between Balachander's performance and Panchapa- process of thematic transformation I have discussed and yet also kesa's notation. Only the last line of Example 6(b) gets very near ask whether his performance would be traditionally sanctioned- to the Indian source on the top line. If we were to regard Pancha- or even be deemed incorrect. There is some justification for con- pakesa's notation as definitive, we might assume that Bala- cern here since Balachander places the eduppu on beat 2, not half chander's performance was at least stretching the boundaries of a beat later, as in Panchapakesa. Further,he never plays anything tradition by delaying the definitive version of the phrase until line like what Panchapakesa places on beat 5. 26 in his performance. Since many professional musicians and To address this concern, we must first recall that, in India, scholars (but not all) consider Balachander to be a great master music is transmitted aurally. The status of the Indian notation as musician, we might both admire his ingenuity in improvising the a "text" is therefore not at issue. But we might regard the Indian _

88 Music Theory Spectrum

Example 6(c). Comparisonof anupallaviof "Mokshamugaladha"as transcribedby A. S. PanchapakesaIyer with performanceby S. Balachander Anupallavi(as writtenin "GanamruthaKeerthana Malika")

-- JiS.. or >r-pr ?, . I ^ _

line 14. A I ml-" "-I ' lin, 1

- r- 5 1 Ezit I linpe...._ 29 _. c pz-. Ia IrrIZJ~7 -f r -r t - V i L i I I I I I I .1 L L I] I |I -l _

line 8. tAb ,i j ,rr'T I

notation as a rough transcription of what is learned by rote in We might conclude that beats 5 and 6 of the Panchapakesa nota- Indian music lessons; nuances and other precise compositional de- tion represent a simplification for beginning students of what tails, which are always followed to the letter in musical practice, Ramanathansings and Balachander plays. are thus not notated. In an attempt to determine the status of But when we consider the rest of Ramanathan's and Balachan- Panchapakesa's notation, I offer a brief transcription of another der's performances of "Mokshamugaladha,"Ramanathan agrees performance of the opening of "Mokshamugaladha,"given at the to a much greater extent with Panchapakesa than does Balachan- bottom of Example 6(b), in which S. Ramanathan was the singer. der. Example 6(c) contrasts Balachander's performance of the The transcription shows that Ramanathan'sversion is much nearer anupallavi with Panchapakesa's notation.24However, yet another to Panchapakesa's than is Balachander's; the metric positions of performance of the piece, sung by the eminent vocalist K. V. the eduppu correspond, as do the Eb-C figures on beat 4. How- Narayanaswamy,25agrees to a substantial degree with Balachan- ever, Ramanathan's performance is almost the same as Balachan- der's version. Both master musicians place the eduppu on beat 2 der's on the last four beats of the tala. Still, on the third line of the transcription, Ramanathanintroduces a phrase, marked with X on the example, that does not occur in Balachander's performance 24Example6(b) also shows how Balachander'ssecond playing of the anu- but does to the 6th beat of Since correspond Panchapakesa's. pallavi (not given in Example 6) preserves the contour in his first playing on Ramanathan, like Panchapakesa, is a scholar and pedagogue, line 18 but also makes a cross referenceto a passage on line 8 of the pallavi. Ramanathan's performance corroborates Panchapakesa's notation. 25Guru Padam: K. V Narayanaswamy,Sangeetha Compact Disk, KDV 063. Variation and Process in South Indian Music 89

and perform identically the transformations of the pallavi phrase Catlin, Amy Ruth. 1980. "Variability and Change in Three Kar- listed in Example 6(b).26 nataka Kriti-s: A Study of South ." Of course, there is much more to say about the identity of Ph.D. diss., Brown University. Carnatic compositions that involve transmission and lineage. Nelson, David Paul. 1991. " 'Mrdangam' Mind: The 'Tani Avar- Suffice it to say that many versions of a piece can and do exist tanam' in KarnatakMusic." Ph.D. diss., Wesleyan University. side by side. On one hand, multiple versions influence each other Kaufmann, Walter. 1976. The Ragas of South India: A Catalogue and evolve as musicians listen to each others' performances and of Scalar Material. Bloomington: Indiana University Press. introduce new sangatis into a performance. On the other hand, Kumar, K. and Stackhouse, J. 1987. Classical Music of South each version resists change, due to the accuracy of transmission India: Karnatic Tradition in Western Notation. Stuyvesant, inherent in a teacher-student relation that hands down composi- N.Y.: Pendragon. tions aurally by rote. Panchapakesa Iyer, A. S. 1995. Ganamrutha Keerthana Malika In this study of melodic and rhythmic transformation in two (4 vols.) Madras:A. R. K. Printers. kritis, I have tried both to show the inner workings of the music Powers, Harold S. 1958. "The Background of the South Indian and to go far beyond the usual descriptive discourse found in most Raga-System." Ph.D. diss., Princeton University. Indian and Western accounts of Caratic music. Within the syn- Reck, David B. 1990. "India/South India." In Worlds of Music. tactic constraints of raga and tala, we have witnessed a number of 2nd edition, Jeff Titon, ed. New York: Schirmer. intricate processes of variation involving prefix, suffix, insertions, Sambamurthy,P. 1964. South Indian Music. 6 vols. Madras: The rhythmic permutation and reconfiguration, contour transformation Indian Music Publishing House. and affiliation, and the like. I hope that this study, besides provid- Subba Rao B. 1996. Raganidhi. 4 vols. Madras: The Music ing an in-depth look at compositional process in Caratic music, Academy. might reflect back on Western music, helping us to develop better Wade, Bonnie. 1988. : The Classical Traditions. ways of understanding and describing music of architectonic Englewood Cliffs: Prentice-Hall. complexity-whether traditional or not-in which melody and Visvanathan, Tanjor. 1974. "Raga Alapana in South Indian rhythm are the dominant dimensions instead of harmony or coun- Music." Ph.D. diss., Wesleyan University. terpoint.27 ABSTRACT This paper analyzes two contrastingcompositions of Carnatic(South LIST OF WORKS CITED Indianclassical) music. I focus on one of Caratic music's most wide- musicalforms: the a devotional in three Bhagyalakshmy, S. 1990. Ragas in Carnatic Music. Trivandrum: spread kriti, usually major C. B. H. Publications. sections. Kritisoften employ extensive use of certainmelodic/rhythmic variationscalled sangatis.With the use of sangatis,the kriti becomes a long, dynamiccomposition, often containing a climaxreplete with intricate motivic The focus of this is to the in which 26S.Balachander died in 1990 at the age of 63. The fact that Balachander development. paper study way the work and Narayanaswamyplay the piece similarly is interestingsince Balachander sangatiphrases to providelong-range continuity and a sense of was an innovatorwhile Narayanaswamyis a traditionalist. autogeneticprocess within a kriticomposition. The analysesare based on 27I am grateful to Lewis Rowell, who offered suggestions and advice for transcriptionsmade by the authorof recordingsof the compositions"Kar- improvingthe firstdraft of this paper. pagame,"by PapanasamSivan, and of "Mokshamugaladha"by Tyagaraja.