Articles on Carnatic Music

Two Faces Of Music
by Ganapathi Narsimhan

There is more to music than meets the ear. There are two ways in which any piece of music or sequence of intonated phrases can be discussed. On the one hand, we can talk about its 'physical' characteristics such as sound patterns which can be measured as the progression of frequencies, mediated by amplitudes(loudness) in decibels. On the other hand, we can talk about "meaning." These two ways are independent of each other. We can talk about the duration, loudness and sequence of swaras without reference to what that particular sequence means. We do not even need to know the art and science of music. But if we want to talk about "meaning" it is essential that we know music science, the language of music. We might agree that chatusruti rishabam and suddha gandharam have a common swara and frequency representation and that the phrases: pa,da,sa and pa ni sa are different without having to debate whether the phrases are written down or intonated loud or soft, and in tempo, slow or fast. To borrow two terms from psycho-linguistics, the underlying discipline in comprehension and learning, we have ‘surface structure’ of music representing the physical characteristics and the 'imbedded' or 'deep' structure, the meaning. In short these represent the totality of music, both the form and content. The term “imbedded” is appropriately chosen because meaning lies deeper than mere sequence of sounds or of written syllables. Here there is cause for much theoretical difficulty and therefore confusion. Meaning does not lie in the realm of musical language, written or intonated, but only in the underlying thought processes of the user or the listener. Swara sequences are not swaras intonated in isolation. The way in which swaras are intonated and are combined in sequences forming musical phrases, determines the meaning The bridge between the surface and the deep structure is called 'syntax' or the rules of musical grammar. It is the performer’s latent knowledge that enables him to interpret and deliver musically correct phrases. A listener acquires meaning from the surface structure presented by the performer only if he/she is aware of some rules of grammar. A critical listener comprehends scientific music by imposing his own limited imbedded structure. While this provides him aesthetic sensations, by virtue of his familiarity of musical grammar, he is also able to pass judgment on the performance. A critic need not be a performer because of the lack of essential attributes like a well trained voice and training. An amateur and an aesthete can be both.

Meaning does not travel from the message to the listener until the listener brings meaning to the message. Music comprehension is also a matter of anticipation or prediction of what the performer might deliver at a given moment of time. Cognitive prediction means the prior elimination of unlikely alternatives in the flow of musical phrases, in raga alapana for example. Such predictions are neither too narrow nor too broad. A probability measure is brought to bear on the termination of a given phrase at a time t to qualify the range of predictions. A well structured raga alapana, unlike what a nadaswara vidhwan undertakes, of Anandabairavi, for example, may be considered. This delectable raga is derived from the melakartas, Kharaharapriya. The arohana-avarohana format of this janya raga is:

Arohana: s g R g m P D P S ; Avarohana: S n D P m g R s

with an occasional prayoga of antaragandharam, G. This raga acquires a rare and short-lived flavor of Reetigowlai from the rendering by intellectual stalwarts like GNB or Mahalingam when a prayoa like n D m and the occasional phrase, g m P m are intonated subtly. A well structured raga alapana signifies an ordered musical exercise which might commence at the adhara swara like sa and which builds up an edifice by moving successively to higher notes in the arohana and frequently looping backwards in the domain of avarohana.

Take the phrase that is delivered at time, t :

Pa, Da Pa ma Pa ma ga Ri ga ...

The listener endowed with sufficient raga gnana can legitimately anticipate the following sequence as a follow up with a probability of 0.8 at time ( t + δt ):

ga ma Da pa ma ga Ri, ni sa ga Ri ga ma...

but might not anticipate the phrase:

ga ma pa Da pa Sa, ni Da ma , ga ma pa ma ga Ri ga....

Intonated with the liberty taken to use the segment ni Da ma, deleting pa, the phrase would be a ‘special’ prayoga, giving a subtle flavor of Reetigowlai. This is a privilege of the mighty and not of the ordinary mortals.

It is rarely necessary to predict precisely the structure of an evolving series, that a raga alapana traverses. The more knowledgeable we become and the more familiar we become with the virtuosity of the artist, our predictions come closer to near certainty. But the excitement resides only in the unpredictable.>

In summing up, we could say that we are usually able to anticipate what phrases the artist is about to intonate because we have the abiliy to make such predictions based on our musical knowledge and critical listening experience. We thus sample the remaining set of alternatives for limited matching. An accompanist on the violin, for example, keeps these alternatives in his mind while supporting the vocalist. I have often heard Lalgudi Jayaraman amongst others, accompanying Madurai Mani Iyer and GNB. While accompanying Madurai Mani Iyer on a raga alapana of Bairavi or Kambodhi, the superposition of the sounds of the vocal chords and the violin would be on phase. But repeating the exercise with GNB, there would be a perceptible lag in the follow up. Herein is then the example of the predictable as against the unpredictable. Often the violinist like Chowdiah would be carried away by his own exuberance and would go on the lead. This is often risky because of the existence of similarities of approach of many ragas. For example, Ariyakkudi Ramanuja Iyengar is prone to keep the pattern hanging to confuse the violinist like Chowdiah. In one concert Ariyakkudi was rendering a raga alapana starting with madhyamam and progressing towards the upper shadjamam only without extension to the rishaba-gandhara domain. The raga flavor could have been that of Pantuvarali, Subhapantuvarali or Varali. Chowdiah took the lead and intonated the suddha rishabham and antara gandharam, thinking that Iyengarval was going to elaborate Pantuvarali. Iyengar , looked at Chowdiah and said softly: " I am going to sing : 'Karuna Yelagante' . This is a composition of Thyagaraja in the raga Varali ! Because music is embedded in meaning (order) and meaning is limited by the purpose and understanding of the performer and the listener, the occurrence of ambiguity in practice is a very rare event; and when it occurs and is detected, it is a cause for amusement.

Syntax is concerned with order and semantics with meaning. Musicologists do not assert that our facility in music appreciation is not rule-governed and that the rules are not generative in the sense illustrated in the miniature grammar indicated below:>

Generative Grammar

(Symmetric and Asymmetric Pentatonic Scales)


r m d

s P S ..1



{ A= s; B=(r,R), C=(m,M), D=P, E=(d,D), F=S } ..2


X 1 = A + B + C ; Y1 = D + E + F ; X 2 = C + B + A; Y 2 = F + E + D ..3

If S 1 and S 2 are the arohana and the avarohana then they are given by:

S 1 = ( X 1 + Y 1 ) and ( X 1 + Y 1 ) = / ( Y 1 + X 1 ) ..4

S 2 = ( Y 2 + X 2 ) and ( Y 2 + X 2 ) = / ( X 2 + Y 2 ) ..5

There are three bivariants (r,R; m,M; and d,D) which lead to a total of 2 3 = 8

arohana formats and for each of these 8 arohana formats there are 8 avarohana formats which are the exact reverse of the 8 arohana formats. Thus we have a total of 64 musical scales of which 8 are symmetric and 56 are asymmetric. The music-scientist would argue that rules of music must be rooted in meaning and related to meaning in all their operations. Consider a miniature generative grammar of English:

Syntax Lexicon

S = X + Y A => A(1)= a; A(2)= the

X = A + B and B => B(1)= ball; B(2)=bicycle; B(3)= bird;

Y = C + X and thus: B(4) = cat; B(5)= dog; B(6)= girl; ..6

S = A + B + C + A + B B(7) = man; B(8) = teacher

C =>C(1)= alarms; C(2) =annoys;

C(3)= avoids

C(4) = bites; C(5) = chases.

The number of perfectly grammatical sentences that can be formed (S's) are: 1280 given by the product: 2x 8 x5 x2 x 8; the numbers correspond to the degrees of

freedom of the components of S . Let us write down some sentences that make sense and some that do not.


S(1) = A(2) + B(8) + C(5) + A(1) + B(5) => The teacher chases a dog

S(2) = A(1) + B(4) + C(1) + A(2) + B(3) => A cat alarms the bird ..7


S(3) = A(2 )+ B(2) + C(4) + A(1) + B(7) => The bicycle bites a man

S(4) = A(1) + B(6) + C(2) + A(1) + B(1) => A girl annoys a ball ..8

So grammar is not concerned with making sense out of a sentence. Who then is responsible for the sensible content of an utterance? The answer is course, the speaker. When the speaker cranks up his grammar to produce an S he wants one that will represent a meaning he has in his mind. He begins with an imbedded structure (thoughts) and the grammar enables him to generate a surface structure for the meaning. The listener reverses the flow by translating the surface structure to his own characteristic imbedded structure. Thus the meaning that the speaker intended and the meaning that the listener abstracted may not be identical.

The syntax that determines the structure of a raga must be based not on how swaras can be put together with grammatical rules at the level of surface structure(arohana-avarohana) but how musical expressions are related at the deeper level of aesthetic appeal. Under such conditions and constraints, most of the 64 janya ragas would be devoid of musical appeal.

Communication in music, a complex term, may be defined as producing a surface structure (sequence of swara phrases) that someone can make sense of in the way that was intended by the communicator.

Ganapathi Narsimhan is a Professor (Article contributed by Srini Ramachandran) and composer of Tamil and Sanskrit kritis. He lives in Melbourne.

back to Articles

Contact us.

updated on