Since Bhatkhande’s documentation of raga grammar in the first quarter of the 20th century, “Vadi” and “Samvadi” swaras have been the pivotal notions describing the melodic personalities of ragas. This set of notions, identifying two dominant swaras, one primary and the other secondary, each located in a different tetrachord, has immense intuitive appeal. But, on close scrutiny, the theory as well as practice of specific ragas appears to warrant re-examination of the “Vadi-Samvadi” notion. Ahir Bhairav, a mature and popular raga, is an interesting example of this phenomenon.
Ahir Bhairav belongs to the Bhairav parent scale, and is a variant of Bhairav, the foundation raga of the scale. It is documented as a compound raga, derived as a blend of a folk/ tribal melody called Ahiri and Bhairav. Texts on raga grammar do not appear to suggest a standardised melodic identity for Ahir Bhairav. In terms of characteristic melodic patterns, Subbarao [Subbarao, B. Raga Nidhi, Vol. I, 4th edition, 1996, Music Academy, Madras] lists two variants of Ahir Bhairav. Thakurdas, [Thakurdas, Manikubuwa, Raga Darshan, Vol.IV, 1st edition, 1997, Lakshminarayan Trust, Rajpipla] presents a third and, substantially different, interpretation of the raga.
For the Ahir Bhairav of his era, Bhatkhande considers Sa as the vadi swara. He avers that because the frequent stand-alone use of Ma, this swara could also be considered the vadi. However, he argues that [komal] Re is also an important swara in the treatment of the raga. Thakurdas suggests Pa and Sa as the vadi and samvadi.
Survey of recordings
Important clues to the raga’s melodic personality may be gleaned from a survey of seven recordings by modern masters: Gangubai Hangal [Inreco:2711-0078], Bhimsen Joshi [TCICL-062C], Kishori Amonkar [Music Today: A91006], Ravi Shankar [STCS:850094], Brij Bhushan Kabra [STCS:850321], Shivkumar Sharma [6TCS:O4B:7175], Pandit Jasraj [STCS:O4B:7402].
The trickiest issue in decoding raga grammar pertains to identification of the “vadi” and “samvadi” swaras in the raga. The raga is consistently treated as a Bhairav variant, with its Bhairav personality being highlighted by the oscillated [komal] Re in the descent [G-M-r-S]. If the vadi [primary dominants] is intended to indicate dominance over the totality of the aural experience, [komal] Re would be a probable choice. Considering the melodic contours of the mukhdas of most bandishes, Ga would also be a candidate for the status.
While Sa and Pa do not enjoy even near-dominant status in the surveyed recordings, Ma appears to have a following, though not categorical enough to qualify it as a dominant. Kishori Amonkar, Pandit Jasraj, Brij Bhushan Kabra, and Shivkumar Sharma have invested a considerable amount of improvisational energy with Ma as a melodic focus. Kabra and Shivkumar have reinforced this emphasis with an appropriate tuning of the acoustic ambience [chords/ drones] on their instruments, exploiting the first-third harmony of Ma and Dh. Kabra’s bandish itself is centred around Ma, and is perhaps the only conscious treatment -- in this sample of recordings -- of the swara as a “Vadi”.
While the “vadi” seems elusive on the surveyed recordings, the “samvadi” does not even appear faintly on the horizon. In the uttaranga, Dh appears to enjoy some significance, but not sufficient to claim samvadi status. The samvadi of the raga remains indeterminate.
The evidence considered here suggests a debatable vadi, and an indeterminate samvadi, with almost all swaras [Ni excluded] usable as terminal points in phrasing [Nyasa swaras]. Ahir Bhairav is thus a raga whose melodic personality is sustained entirely by its distinctive swara material, and a reasonable consistency in its “Chalan”. This makes it a raga of immense melodic tenacity, and improvisational potential. But, this does not make for a comfortable relationship between theory and practice.
The fundamental issue
Ahir Bhairav is supposedly a compound raga, and it has often been argued that compound ragas, by definition, defy the codification of their grammar. But, this is not an isolated case. There are many ragas whose melodic personalities are adequately sustained by their swara-material and their “Chalan” [skeletal phraseology] or what the Dhrupad tradition describes as “Raga Swarup” [melodic form], without reference to “Vadi-Samvadi” swaras.
All this is not terribly original. For many ragas, Bhatkhande himself has shied away from identifying the “Vadi-Samvadi” pair. Later grammarians, also of formidable stature, have often either differed from Bhatkhande’s documentation, or been vague on this matter. A similar divergence is noticed in the practice of several ragas by the leading musicians of our times, including those known for their theoretical soundness. Despite the vagueness of grammar on this dimension long believed to be crucial, several ragas identify themselves beyond reasonable doubt when performed.
What we are looking at is not simply a chasm between theory and practice [Lakshya-Lakshana divergence]. What we are looking at is the inadequacy of analytical rigour used by grammarians, and insufficient commitment to empirical research. The research method required for such enquiries is simple enough, and need not hold our attention here. What is perhaps more crucial is our willingness to make a conceptual leap in understanding the role of raga grammar, and accept its implications.
The purpose of raga grammar is to establish the identity of the raga beyond reasonable doubt, without the risk of confusion with other known ragas. It appears possible to prove that this purpose is served without the categorical identification of “Vadi-Samvadi” pair in musical performance across a large number of ragas. This could mean that this notion of dominant tones is superfluous, though probably not redundant.
In such an event, we may need to acknowledge two levels of raga grammar -- a “mandatory” level which identifies the raga beyond reasonable doubt, and a “prescriptive” level which helps by its observance, but does no damage by its breach. The notion of “Vadi-Samvadi” may, then, belong more appropriately to the “prescriptive” level rather than the “mandatory” level at which it is currently regarded in musicological literature.
After I proposed this two-level notion of raga grammar in my book [Hindustani music: a tradition in transition], experts in linguistics took strong objection. They argued that there is no such thing as “prescriptive grammar”. If anything is grammar, it can only be mandatory. The answer to this is that, the notions of grammar must take into account the nature of the language to which they pertain. Notions relevant to a spoken and written language need not be binding on a “language” that uses melody as its vehicle of ideas. Moreover, when we consider the uniquely Indian problem of codifying the “raga-ness” of ragas, we are persuaded that Indian musicology will have to develop its own “grammar” with its own conceptual and analytical tools.
© Deepak S. Raja 2007