Note: A post after a long time. Perhaps I am rusty as I had a real hard time penning this one. I had broach this subject (a deeper look into the rhythmic maze woven in kalpanaswaras) for a while, but I could not get my thoughts ordered correctly, and thus it seemed quite hard to prevent it from getting out of hand and too expansive. Also, this is certainly an area where I am shaky at best, and so in spite of the effort I may have gone off the beam. Advance apologies if that turns out to be the case!
The term kaNakku in Tamil means calculation, and is used to refer to many things, the most common of which is a colloquial reference to the subject of math in general.
In Carnatic Music also, it finds use albeit mainly by the tamil speaking cross section of the Carnatic Music populace. Here, like most unofficial, colloquial terms, it does not have a precise, specific meaning. I think it refers to “complexity”, in particular “rhythmic complexity” – an obviously nebulous reference. When used to describe the music of an artist, it is used to describe the artistic expression of the musician’s kalpanaswaras. When someone says His/Her music has a lot of kaNakku, it means that during kalpanaswaras, the musician typically weaves seemingly complicated rhythmic patterns presumably involving elaborate calculations i.e. kaNakku. This may involve swaras/pauses of varied lengths, and also employing complex techniques like switches in gati/gait. This to most of us would seem complicated, and that is perhaps so because it is “non intuitive” in the sense that a steady, smooth flowing rhythm is not easily apparent in the constructed melodic patterns.
The other end of the spectrum to this type of kalpanaswaras is one which is referred by another nebulous term – sarvalaghu. I do not know what exactly the word means, or again, whether it has a precise meaning that every one agrees upon. At best, I can think we can describe it as weaving kalpanaswara patterns where a somewhat steady, smooth flowing rhythm natural rhythm is obvious to the listener. Of course, I realize that the phrase steady, smooth flowing rhythm is perhaps as nebulous as the term it is trying to shed light upon on 🙂 !
Hence, perhaps the difference would be better demonstrated via examples:
Here is an example of the sarvalaghu type of kalpanaswaras:
The musician here is Semmangudi Srinivasa Iyer (SSI) who almost always did kalpanaswaras in this fashion, and he always does a superb job. In fact, this is the more common type of kalpanaswaras which almost all musicians follow. The great Madurai Mani Iyer who was famous for his kalpanaswara renditions excelled in sarvalaghu style. Today, for example, T.M. Krishna amongst the current day vocalists employs a style very similar to SSI above. In fact, even those who employ “kanakku” usually start (or end) with this type of kalpanaswaras.
Contrast the above with this by T.N. Seshagopalan (TNS), particularly the iteration starting with him:
The piece starts with the violinist’s response which is more or less a straightforward sarvalaghu type style – you can clearly make out swaras in groups of 4. However, compare that with TNS’s next iteration? Can you see that starting with it, the rhythm of the melody seems somewhat complicated? Of course the pattern is certainly catchy but to me, it seems it isn’t as “easy flowing” as the SSI sample.
But why is this so? .
Sarvalaghu style – a deeper look:
Tala, gati, and mAtra: The song that SSI is singing swaras to is mArubalka (Sriranjani, Thyagaraja), which is set to Adi Tala, in catusra gati. What does this mean? This implies that the “beat” of the song, which has the same duration as an akshara of the tala (but need not align with it and hence is not necessarily synonymous with it) can be further divided into 2 or 4 or even 8 smaller units of time. This can be in theory be 16, 32 etc. in theory, but let us consider simpler cases.
If we assume that it is 4 divisions (most common), then within the duration of a whole cycle of Adi tala (which has 8 aksharas) you have 32 such time units as follows:
1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | ||||||||||||||||||||||||
1 | 2 | 3 | 4 | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 4 | 1 | 2 | 3 | 4 |
This smaller time division is called a mAtra or mAttirai (tamil). So if a musician is singing swaras for one tala cycle as above, and each swara‘s duration is equal to this smaller time unit, then he/she would be able to fit 32 swaras in a cycle of Adi tala. This is the typical way things are done in most Adi tala songs (in catusra gati).
Now in sarvalaghu singing, for catusra gati, I believe that the swaras (and pauses) are almost always 1, 2, and 4 mAtras, (i.e. 1/4th 1/2, or 1-whole akshara durations), and you will also find them grouped such thatas per the tala “beats” (either on the “beat” or half-way in between). Not all swaras need be of same duration (e.g. like srgm rgmd gmdn mdns), and thus it can be a mixture of these. If you mostly have 1, 2, or 4 mAtras for swaras (and pauses), then you can see in general you will find that they align “to the beat”.
Here is the notations for the swaras the vocalist (SSI) sings in that piece:
Note that the swaras shown above are grouped in 4s (not necessarily aligned to tala), and a set of 4 have the same duration as the akshara, which is the span of the natural “beat of the song”. As you can see, the overwhelming # of swaras are 1/4th akshara in duration, and there are almost no pauses. This is typical of sarvalaghu type of swaras singing and it helps establish a riveting, easy to grasp rhythm.
“Kanakku” style: In the case of that T.N. Seshagopalan piece, as we will see, the swaras do not follow this style. In fact, it took me quite a bit of analysis and very very careful listening to arrive at the correct rhythmic pattern being used here (a reflection of my weakness w.r.t laya 🙂 ). Here is the notation for the first iteration that TNS sings:
Again swaras are grouped in 4 which span an akshara. The blue [ ] markers are used to group the swaras as per the musical pattern. Just from the look of the above, it should be clear that the piece is more complex in its composition that the previous case. If you look at it carefully, you can see that TNS is weaving a 5-count (i.e. khaNDa) based melodic pattern within the catusra gati framework (i.e. where the mAtras are spaced apart in groups of 4). The base melodic pattern is 10 mAtras i.e. 2 and 1/2 aksharas long, and is mostly made up of 5 swaras divided as 1, 1 (pause), 2, 1, 5 (e.g m , d , m g , , , , and m , d , m d , , , , ). There is one case in the middle where I think he intentionally or unintentionally leaves a pattern as 9 mAtras, but stretches the following one to 11 mAtras, and thus maintain the 20 mAtra count for two successive patterns. But in general it is a 10-mAtra pattern. Although, I have not analyzed all iterations it seems the violinist’s response to this iteration, and the entire next iteration is based on such a pattern.
As I mention above, it took quite a bit of effort for me to arrive at the pattern, and be reasonably sure it is indeed the case! Let us say I am still not 100% sure! I actually I had to slow down the song a lot for me to be able to count the mAtras for the swaras and thus determine that it is a khaNDa pattern.
What was interesting is that I got the rhythmic sense of the pattern quite easily registered in my brain, could relate to it well, and even able to reproduce it from memory almost immediately. However to find out how many mAtras it took etc. was still a difficult challenge. Even now, I have to sing it extremely slowly to correctly count the 5-akshara swara as well as the 1+1(pause)+2+1 pattern right :-).
So given that, some of you may also be wondering if this is really correct. So I present another way to look at this – a “video” that demonstrates that the notation above is correct. Note however that the tempo slowed down by about 30% so that each mAtra is stretched out to be long enough to be more discernible to the eye:
In that video, I have placed markers at regular intervals spanning the duration of a mAtra. Then I placed labels for the swara pattern for the first iteration. As you can see the swara pattern matches the audio (mostly – there is some jitter as humans are not robots). Of course it might seem silly to do this but I also think that this is a nice instructional method for illustrating laya aspects. What do you think?
In any case, why is this piece “more complex” and “less intuitive” (at least to me)? Based on above, I am thinking that perhaps the odd numbered 5-mAtra long pause-laden swara is something “atypical” that my mind does not grasp easily as compared to a 2-mAtra or 4-mAtra swara (which is ubiquitous). This is especially so when done within a catusra gati song which till that point had laid out the more ubiquitous 4-based i.e. ta-ka-dhi-mi rhythm clearly (this is why I chose to start the sample with the violinist’s portion). On top of that we have 10-base pattern for an entire iteration, done within a catusra gati which seems atypical.
So I wonder if this is perhaps the reason why the rhythm while attractive has a complicated feel to it. When musicians employ “kANakku” it is typical for them to employ notes and pauses like these of “atypical” length. When we listen, we do know something complicated is going on, but perhaps most of us are not able to tell why.
Only when you peer inside you see the sophistication involved, and you begin to understand why many feel that Carnatic Music has one of the most sophisticated rhythmic components amongst all music systems of the world!
But you ain’t seen nothing yet !!!
Carnatic Musicians go even further in terms of rhythmic sophistication. I will post about this with one mind-boggling kalpanaswara performance from a genius of a musician in the next part
July 29, 2010 at 3:58 pm
Thanks Arun for this topic! I never understood the meaning of sarvalaghu myself nor even the word derivation. I can follow your explanations lucidly. It can be catusram/thisra/… but always constant. Is it not? Now the ‘kaNakku’ part is still hazy! How is 20 matras accommodated within the 32 matra sarvalaghu Adi taaLam? Is it done by changing the ‘akshara kaalam’ ?
July 29, 2010 at 4:21 pm
No cml, instead think of whole iteration as one. It is 60 mAtras long i.e. start of the m , , , , till g r s r g (leading to the final ma). This would fit 15 aksharas of the song, where an akshara’s duration is the same as that of the ta-ka-dhi-mi (from vocalist part) leading to this point. This akshara spacing does not change in the song – it is a constant – TNS and others would put tala with this spacing throughout.
Also, this pattern that spans the duration of 15 aksharas need not be aligned with the tala since that depends on the “take-off” point i.e. where the first m , , , , starts. One can figure this out by keeping tala from the beginning or better yet note down the take off point for the kalpanaswara line. I didnt do that).
However, if we assume it does align with tala, then you if imagine TNS starting on second beat of tala (i.e. little finger) – then the pattern would take that entire tala cycle (7 aksharas), and one other tala cycle aksharas (8 more making it 15) such that the final ma is at samam for the 3rd cycle. This ma as I have marked is probably 4 more mAtras adding up 64 i.e. exactly 2 cycles (at which point the violinst starts his/her response).
July 29, 2010 at 4:24 pm
Oh also it isnt that swaras are of constant duration in sarvalaghu (in fact it isnt even in ssi’s case as he does r g m , where m is twice as long) – but they are mostly so, and longer swaras are mostly twice as long etc. And yes if it is for tisra gati, you would have mostly 1/3 and some 2/3 etc.
July 29, 2010 at 4:46 pm
cmlover, say that a passage ends 555 (ThaDhiGhiNaThom three times). This is 15 matras: if we are to fit it into a half of adi talam, we must leave one matra gap (or fill).
|,Tha dhi Ghi Na Thom
Tha dhi Ghi Na Thom
Tha dhi Ghi Na Thom || Tha
This is the first and simplest such a pattern that mridangam students learn, and yet it occurs in countless korvais, both in percussion and kalpanaswara. Other calculations grow in complexity, but the principle will remain the same. 20 matras in 32 leaves 12 that must have something done with them!
July 29, 2010 at 4:48 pm
This piece of Kanakku, by the way, is one of my on-line personas 🙂
…Nick
Arun: That is a cool one! Particularly the Thad part :)!!
July 29, 2010 at 4:56 pm
Nick – I would be wrong, IMHO while what you say makes sense, I would wager that to non-laya folks (like me), it is just confusing because it does not directly address the piece in question :)! We are not trying to fit 15 matras in 16 nor 20 in 32 in this particular example (even if in principle it would be similar).
Arun
July 29, 2010 at 6:28 pm
Thx Arun/Nick:
For me 15×4 +4 =64 settles the account. But complexities can be introduced through eduppu, kuraippu etc., to settle the final account. At any rate the akshara kaalam is inviolate throughout the concert. It can be doubled/halved/quadrupled etc., but only in powers of 2 and not of any other number (say 3); Am I right?
Arun: Hmm. You are perhaps confusing Akshara kaalam with something else. Let is say that for this song, an Adi tala song, tala is kept such that 1 cycle comes to 8 seconds. So your akshara spacing (i.e. the kriyas) 1 second apart. This 1 second between the kriyas of the tala remains constant throughout – that us inviolate and that is refers to keeping tala. This is so even during the above 64 mAtras (2 cycles i.e. in 16 seconds) done as 4 + 60 where 60 is a khanDa pattern. But WITHIN that khanDa pattern each 5 mAtras is 5/4 i.e. 1.125 seconds – but since tala is still kept at 1 second apart, this 1.125 isnt akshara duration.
The double,halve,quadruple I was referring to length of swaras in sarvalaghu – predominantly they would 1, 2, 4 matras – atleast that is what I think/em>
July 29, 2010 at 7:37 pm
So happy you are back to blogging 🙂 Great post and for a laya challenged person like me, the lucid explanations helped a great deal!
Arun: Thanks. I am actually laya challenged too. If any I understand things behind it, but in practice I find it quite hard to grasp.
July 30, 2010 at 9:54 am
Is it permitted to mix different gatis within a taaLa cycle, say khanda and catusra in different variations and come out even at the end of n cycles where n is an integer ?
Arun: Certainly but you are getting a little ahead 😉
This is where complicated kaNakku are needed to keep the account balanced; right?
Arun: Yes – but not necessarily i.e. you can have quite complex stuff even without gati switches – I think. and not all gati switches are necessarily complex in terms of counting etc. (but still complex/challenging execution-wise)
August 9, 2010 at 9:24 am
[…] Carnatic, Carnatic in Western world Leave a Comment A short detour before the 2nd of part of Kanakku vs. Sarvalaghu in Carnatic Music. But this one is related to […]
August 21, 2010 at 3:49 pm
[…] August 21, 2010 Kanakku vs. Sarvalaghu in Carnatic Music – Part 2 Posted by Arunk under Carnatic, Carnatic Music Leave a Comment This is a Part 2 of a post that tries to highlight some of the rhythmic wizardry employed by Carnatic musicians during kalpanaswaras.1 Click here for Part 1. […]
August 24, 2010 at 5:59 am
The first swaram in the SSI clip should be notated as
r,gr,snd-n,sn,dmd
(In carnatic music, a long note can occur by not only elongation but also silence and other assimilation effects)
Since the take-off point is 2 inner-beats, it is better to notate as (starting from the first drutam):
| r,gr,snd | n,sn,dmg || md (dArinerigi)
More later
August 24, 2010 at 6:00 am
(Edited post)
The first swaram in the SSI clip should be notated as
r,gr,snd-n,sn,dmgmd — this line edited
(In carnatic music, a long note can occur by not only elongation but also silence and other assimilation effects)
Since the take-off point is 2 inner-beats, it is better to notate as (starting from the first drutam):
| r,gr,snd | n,sn,dmg || md (dArinerigi)
More later
August 24, 2010 at 8:01 am
Thanks. You are right. It should have been r,gr (and n,sn,). I will make the correction when I get a chance (not at the right computer now)
I thought about showing it as per tala but then I thought it may “hide” the (relative) simplicity of the intrinsic rhythmic aspect of the melody of the sarvalaghu here. This is why I havent shown any tala anga markers. (I do realize this makes the notation technically not correct).
Arun
October 19, 2010 at 9:07 pm
I think the accompanists are Sri Lalgudi G Jayaraman for Semmangudi, and Sri T N Krishnan for TNS. Interestingly, Lalgudi is known to use more kanakku (in a different manner) when he gets the centre stage, where as TNK adopts elements from the SSI style.
October 30, 2010 at 3:06 pm
The violinist in the TNS clip is Shri Nagai Muralidharan. This is a commercial release, I guess with Shri Krishnam Bhaja Manasa as the main.
Sathej
October 24, 2012 at 6:40 pm
Can sombody here teach me solkattu thanks.Contact me AT sukhforyou@yahoo.com
August 18, 2013 at 4:30 pm
Dear Sri Arun,
Very good exposure to Tala nuances in Carnatic music. By the way, as per my wife who is a knowledgeable carnatic vocalist, 10-10-…-10-11-9 pattern adopted in the swaraprasthara presented by the great TNS , is definitely intentional for a surprising break of monotony.
As an initial guide to carnatic music listners, I have written a blog on Rhythmic aspect of Carnatic music as below.
http://lvnaga.wordpress.com/2012/11/16/talas-or-rhyth…carnatic-music/
I request your comments on the same. You may also see my other attempts on music in
http://lvnaga.wordpress.com/wp-admin/edit.php?category_name=music
August 20, 2013 at 2:11 pm
Hi, I am Nagarajan. I forgot to give my Email ID.
lvnaga@yahoo.com
November 15, 2013 at 10:02 am
Very nice article. You have really put ina lot of thought and effort into explaining a very complex musical term for readers like me. Awesome effort!
Arun: Thanks!
January 31, 2015 at 8:51 am
It would have been better if simple English was used in first few paragraphs. Not all musicians are experts in English. We should not end up in referring to dictionary while reading music related articles.
January 31, 2015 at 8:52 am
Also more explanation could be given for Korvai/Koraipu.
December 4, 2016 at 1:50 am
Tks, THOROUGHLY confused!! My rasika-ship is confined to ragalapana & shloka-singing/playing as ragamalika. Thus, I am happily RID of the entire TAALAM imbroglio! Great \, isn’t it? – tho I still like Madurai Mani Iyer’s swaras, ignoring the ‘kaala’ aspect COMPLETELY. I recommend these options to all UIP rasikas(u=most UN-important or most UN-informed yet happy rasikas like me – from age 05 to 84 now).
March 12, 2019 at 8:34 am
Where is the video? I couldn’t see any link either! – Ignorant violinist of decades with an audience of only two!! now nearing 86 but g/son of a “renowned”(within his village, accompanying a deity’s procession + teaching kids @ 8 annas pm!!)
March 12, 2019 at 9:10 am
Respected ArunKm,
At my age, I only wish to continue my v.v.happy association with Karnataka sangeetham(only on my violin) as long as I live. Seeing/hearing your video on sarvalaghu will greatly help. Hence my comment on your write-up Part 1.
I hope you will guide me to do just that.
Rgds. ramani
________________________________
July 24, 2020 at 6:47 pm
Hi there! I have not read it in its entirety yet, but I just wanted to point out that your audio samples are not playing. It would be great if you could look into that and fix the issue!
Thanks,
Sridhar