Why show Carnatic music in staff notation?

Mar 16, 2014  

Recently, I added support for viewing Carnatic notation on Patantara as western staff notation, albeit with some known deficiencies that I’m working to address. In this post, I talk about the reasons behind the support for staff notation and their influence on the format and extent of support for the notation on the site.


Our main reason for supporting viewing Carnatic compositions as staff notation is so that musicians who are familiar with it but not with the textual “sa ri ga ma” notation need not go through the learning curve just to access the musical material. It also helps Carnatic musicians who don’t know staff notation to reach a wider musical audience by speaking their language. I hope this would encourage musicians to collaborate across cultural boundaries.

Taking this perspective places some constraints on the extent to which the resultant notation conforms to the conventions of publication of western classical music or Jazz. Two important aspects where the goal has a say are how to indicate the tala on the staves, and how to indicate the raga using a “key signature” next to the clef.

Indicating the tala using “bar lines”

The tala is, as of this writing, not visible in the staff notation since I’m still considering a few approaches and haven’t settled on one yet.

Consider the “ata talam” which has a pattern of 5+5+2+2. The following approaches are possible -

  1. Treat the entire tala cycle as “one measure” and use 144 as the time signature.

  2. Indicate the angas of the tala as separate measures, adding the relevant time signatures to it. In this case, that would make four bars with time signatures 54, 54, 24 and 24 respectively for each cycle of the tala.

  3. Use a “compound time signature” like [5+5+2+2]/4 and indicate the anga divisions using dotted lines and the sama using a bar line. (Thanks to Michael Schachter for this suggestion.)

Now, the first approach is alright, but subdivisions of a tala are important to Carnatic music in their relationship to the text and melodies. This importance would be lost in the staff notation.

The second approach would convey the subdivisions perfectly. I believe most sight readers would be comfortable with this. However, the larger time structure of the tala is now lost, though any musician with modest competency can see that pattern. It also looks unneessarily repetitive. Michael also suggested that this would permit western musicians to not be aware of the tala concept at all, which is not necessarily a good thing.

The third approach looks ideal and is what I’m currently considering implementing. It shows both the internal structure of the tala as well as the larger cycle itself.

The point to note here is that this choice is dictated by the fact that we do not want to lose the essential aspects of Carnatic music in translating it into staff notation, even if it might make it easier to read in the target musical language.

Indicating the raga using “key signatures”

When a Carnatic melody is presented without a leading “key signature”, many ragas would need ♯ and ♭ signs for several notes. This can become tedious to read and visually redundant. Using a “key signature” to improve the readability is therefore useful.

Now, should we map ragas to one of the known key signatures so that musicians know how to play them? I believe doing so would result in some of the melodic aspects lost or distorted in translation.

To address this, I’m now considering an approach that reaches out to (what I see as) the basic principle behind the key signature notation in western music, which is that ♯ and ♭ signs are placed against those lines, notes on which should be rendered with an accidental were it not for the signature. Calling on this principle, it appears sufficient to compute the notes on a line that are always to be rendered flat or sharp and place ♯ and ♭ signs on those lines near the leading clef.

As a Carnatic musician with some modest familiarity with staff notation, this works for me, though it still distorts ragas whose arohana and avarohana using different svaras and might need different ♯ and ♭ indicators. I believe this notation is usable by western musicians. You can see for yourself on Patantara whether this works, and please do comment on this point if you don’t find this acceptable or have a better suggestion.