Raga-Ness of Raga II
by Deepak S. Raja

The melodic grammar of raga-s Melodic grammar begins with the permissible swara-s of a raga. The vast majority of raga-s utilise a minimum of five swara-s. The few raga-s, which use four swara-s are performed only rarely -- obviously due to their limited melodic potential. On this basis, raga-s are classified into chatuswara [quadratonic], audava [pentatonic], shadava [hexatonic] and sampoorna [heptatonic]. Raga-s which use a different number of swara-s in the ascent and the descent are described by a dual description such as audava-sampoorna or a shadav-sampoorna.

In this classification, Indian theory does not habitually double-count the komal [flat] and tivra [sharp] shrutis [microtones] of the same swara, when both are used. For communicating this feature to foreign readers, I have found it useful to use terms like “hyper-hexatonic” or “hyper-heptatonic” for such raga-s. Although the swara-s permitted in a raga constitute the basis of its melodic grammar, and are therefore mandatory, they are subject to an occasional influence of aesthetic grammar. This dimension arises from the practice, in some raga-s, of non-standard intonation, deploying pitches [relative to the fundamental / scale-base] other than the standard pitches for the swara-s as defined in Hindustani music. Many of these prescriptions are based on sound principles of correspondence between swara-s even if musicians advocating the practice cannot always articulate them. But, some of them might well be purely aesthetic. In raga Puriya, for instance, some stylistic traditions prescribe the intonation of komal [flat] Re [2^nd] in the microtonal region below the standard frequency of komal [flat] Re. The komal Re of raga Shree, too, is aesthetically sensitive. Some musicians prescribe its intonation below the standard komal Re pitch, while others commend a pitch above the standard pitch. In a similar deviation, the komal [flat] Ni [7^th] of raga Adana is occasionally prescribed at a pitch between the standard komal and shuddha [natural] Ni frequencies. 

Such instances abound. Non-standard intonation of swara-s, though controversial, does not belong to the territory of folklore. It is a very explicit part of the grammar of some raga-s. Its practice is documented in several texts on raga-grammar, and perpetuated by the personalized system of aesthetic indoctrination prevalent in Hindustani music. It is also practiced with explicit intent by its advocates. However, it is unclear whether this feature of raga-ness should be classified as melodic grammar, which is mandatory, or aesthetic grammar which is prescriptive. Only a substantial survey of performances of relevant raga-s, using contemporary pitch measurement equipment, can determine this issue.

The second dimension of melodic grammar is its scale, consisting of the aroha [ascent] and the avaroha [descent]. Although a simple, sequential, listing of the permissibleswara-s in ascending and descending order is often encountered, it is considered incorrect and misleading because it fails to account for the manner in which the raga permits their deployment. On these grounds, there exists a strong, and valid, opposition to the very notion of aroha-avaroha as representative of a raga’s melodic personality. This objection is particularly potent because the same scale can deliver a multiplicity of raga-s, as in the case of Marwa, Puriya and Sohini. It seems legitimate, therefore, to accept the “chalan” [skeletal phraseology] as a superior alternative.

The chalan consists of a set of phrases, generally arranged in ascending and descending formation. Most raga-s can be adequately represented by a chalan of eight or fewer phrases. This device is indeed superior to the simple, sequential, scale because it provides the essential building blocks of the melody. It always includes the pakad [signature/ catch phrase], which identifies the raga beyond reasonable doubt. Collectively, the chalan phrases represent the identifiable and distinctive melodic personality of the raga. They achieve this status by also incorporating several elements fundamental to defining the raga’s melodic grammar.

The first such element is the sequence in which the permissible swara-s can be used for phrasing. The phrasing structure also reflects the permissible alpatva / bahutva [suppressed / enhanced intonation] of the swara-s within a phrase and also the manner of their treatment or intonation. The second important element is the identification of the “nyasa” swara-s of the raga – swara-s permitted for use as terminal or resting points in the phrasing. The nyasa swara-s are extremely important because raga-s, using an identical set of swara-s, are crucially differentiated by terminal points of phrasing. The distinctive aural experience of Marwa, for instance, is considered safe with only two permissible resting/ terminal points: Dh and Re. Any other terminal point in phrasing can cast an unintended shadow of either Puriya or Sohini to fall on the raga.

Two of the nyasa swara-s, the most frequently used, and the second most frequently used, are the vadi [primary dominant] and the samvadi [secondary dominant] swara-s of the raga respectively. The vadi and the samvadi are generally located in different halves of the scale. The vadi-samvadi notions, memorized unquestioningly by all students of music theory, are far trickier than most imagine. Raga-s do not reveal their melodic personalities in as simplistic terms as musicologists would like them to do. Far too often for my comfort, I have found even Bhatkhande circumspect and tentative about the identity of the vadi-samvadi swara-s of raga-s. In addition, authorities often differ on this subject. In many cases, authorities are also found to have identified nyasa swara-s, but shied away from identifying either the vadi or samvadi swara-s, or both. 

Approaching this problem from an empirical and analytical perspective is not always more helpful.  My attempt at verifying the melodic personality of the popular, and seemingly non-controversial raga, Ahir Bhairav, was an eye-opener. I surveyed eight recordings of the raga by the biggest names in contemporary Hindustani music. I found that only four of the eight maestros treated Ma, [4^th ] the documented vadi swara of the raga, as even a nyasa swara. Of these, only two made a special effort to give it the status of a probable vadi. Only one musician had squarely, and explicitly, treated Ma as the vadi  of the raga.

This need not mean that the texts are wrong or confused. Nor need it imply that leading musicians perform the raga incorrectly. It could merely suggest that the constant dialogue between theory and practice required of a vibrant art music tradition has ceased to function, and that the vadi-samvadi notions, now require fresh thinking, probably along with several other notions of raga grammar. If the vadi-samvadi are indeterminate for some raga-s, we should accept this reality. If this is found to occur on a large scale, we might even shift this facet of raga-ness out of mandatory / melodic grammar, treating it as a facet of prescriptive / aesthetic grammar. The intuitive and conceptual appeal of the vadi-samvadi notions is, however, so strong, that music theory will probably persist with their mandatory status, and performance will continue to handle music as it has always done.  

Most modern authorities on raga-s have documented the above aspects of raga grammar. Selectively, rather than systematically, some texts have ventured beyond them into what I have described as the “aesthetic grammar”. The aesthetic grammar of a raga is related partially to its melodic grammar, and partially to the “rasa” [emotional content] associated with the raga. This is why I have described it as lying on the borderline between grammar and literature.  

The region defining a raga’s centre of gravity is clearly related to the vadi swara of the raga, the most frequent terminal point for the phrasing. It is important to define such an epicentre as a region, in addition to its focal point [vadi], in order to avert the risk of its dilution by a disproportionate amount of improvisational energy being invested outside this region. If, for instance, a musician tarries substantially in the lower tetrachord in Sohini, he exposes the raga to a risk of confusion with Marwa, since both raga-s have an identical set of swara-s, and they differ in their respective centres of gravity. Even for raga-s which do not have so fragile a raga-ness, the systematic development of a melodic centre of gravity is important.  When I come across a piece of music about which there is nothing obviously wrong, nor anything remarkably right, I discover that, in most cases, the musician has been indecisive about the raga’s melodic centre of gravity.  On a bad day, even great musicians can fall prey to such a dilatory tendency.  

A magnified version of the epicentre/ centre of gravity notion, is the notional scale-base. Fundamentally, all raga-s make melodic sense only in relation to the scale-base at the Madhya Saptak [middle octave] Sa and a scale of Sa to Sa. Yet, some raga-s are most effectively rendered assuming a different scale base. Again, Marwa, Puriya, and Sohini are good examples. The contemporary Marwa is most effectively performed assuming the scale-base at the lower-octave Dh [Scale: Dh to Dh]. It is differentiated from Puriya with the scale base at the lower-octave Ni [Scale: Ni to Ni], and Sohini with its scale base at the middle-octave Ga [Scale: Ga to Ga].

In this context, I recall a remarkable recording of Sohini by Shivkumar Sharma. Throughout the performance, he totally avoids melodic execution in the region below the middle-octave Ga, using the scale-base at Sa merely to provide the scalar reference point for the raga to make melodic sense. This is, of course, a rare instance of fastidiously sound music. But, it demonstrates the logic of a notional scale-base as an important operating principle for the aesthetic integrity of some raga-s. 

Related to these prescriptions are biases in favour of specified melodic directions. Certain raga-s are described as “aroha-pradhan” [ascent-dominated] while others are described as “avaroha-pradhan” [descent-dominated]. Learned texts frequently refer to the relationship between the melodic centre of gravity of a raga and its directional bias. Bhatkhande, for instance, virtually equates raga-s with their epicentre in the uttaranga with those that are descent-dominant. Contemporary practice does not support so sweeping a generalization. But, there is ample justification for regarding the directional mdominance of melodic movements as an independent, and additional, discriminator between raga-s.

The operation of this discriminant can be seen clearly, once again, in the treatment of Marwa, Puriya and Sohini by great musicians. Marwa renditions tend to be  dominated by descending phrases; Puriya phrasing tends to be almost equally split between ascending and descending phrases; while Sohini renditions tend to be dominated by ascending phrases. This prescription is neither far fetched, nor imaginary. I have verified its validity through a complete phrase-by-phrase notation of these three raga-s, recorded in succession by Ustad Vilayat Khan on a CD due for publication by India Archive Music, New York. On this recording, the maestro has demonstrated, in quantifiable terms, the creation of the desired directional biases by the use of ascents structured as strings of descending phrases, and descents structured as strings of ascending phrases.

This sequential rendition of Marwa, Puriya, and Sohini by Ustad Vilayat Khan is significant also for its demonstration of the other two aspects of aesthetic grammar discussed here.  A melodic plot of the renditions of the three raga-s also demonstrates the use of different notional scale-bases, and different melodic centres of gravity/ epicentres as differentiators between raga-s using the same set of permissible swara-s.

The Vilayat Khan rendering suggests that the deployment of these devices is probabilistic in nature. No raga presentation can entirely forego any region of the melodic canvas, or entirely ignore the universal scale-base at Sa, or entirely neglect ascending or descending melodic phrasing. But, the great musician, being sensitive to these facets of raga grammar, allows the dominant emotional content of the raga to express itself by a multi dimensional reinforcement of the dominant melodic idea.

This brings us back to where we started – that a raga is a melodic representation of an emotional statement, and a melodic vehicle for its communication. An understanding of its raga-ness helps the musician to create a stimulus, which has a high probability of attaining a well-defined emotional state for himself, and of inducing a similar emotional state amongst his audiences.  Truly great music, like great literature, tends to transcend grammar, often liberating it from itself. When this happens, grammar gets re-written. But, for lesser mortals, grammar – as understood in each era – remains the stepping stone to literature.”

© Deepak S. Raja, March 2001

Raga Sangeet Menu    Raga Sangeet Home    WUWF Home