This question pertains to non-Western music and hence a bit of explanation before a question is stated.

I listen to and play Hindustani Sangeet (i.e., North Indian Classical Music) which does not use chords. A singer is accompanied by tabla (two drums used for percussion but also tuned to the singer's tonic) and tanpura drone, harmonium, or sarangi.

Most of our music is without notation (sheet music). There are Ragas, which are like melody standards with certain rules, such as which notes are allowed and which are principal notes and important phrases for each Raga. Based on those rules and using arpeggios we compose on line music.

The tonic (which is called Shadaj, similar to C of a scale) of a singer is fixed by his/her choice in any note of an octave but such that he/she gets a two-octave voice range.

Singing is accompanied by the tanpura, which is a drone tuned to the tonic of a singer, so it helps in identifying the tonic.

However, some singers do not use a drone, but harmonium may accompany. In this case mostly harmonium plays a tune sung by the singer almost in unison.

I have analysed notations of some study songs and observed that the tonic is hardly there. In such case how can I identify the tonic?

One idea what I have is that if I can transcribe a floating phrase of notes, my brain may extrapolate to identify a tonic even when it is physically almost not there. Please reflect.

  • 5
    Namaste! Welcome to Music.SE. I'm not even vaguely qualified to answer this question (I hope someone who is will come by), but curiosity has gotten the best of me, so I ask: when you say that "the tonic is hardly there", what do you mean? Do you mean that it is not marked as such in the notation? Or that you would expect to be able to figure out the tonic by which note is used most, or at endings, or some such, and that is not the case? May 5, 2014 at 4:49
  • I think my response will work if you would post a series of samples I will run an example for you.
    – user10164
    May 10, 2014 at 8:42
  • @Codeswitcher Hopefully he comes back, but I believe the "melody" lines imply a tonic, but may only rarely pass through whatever note that is (albeit an octave or two above the drone). It would be great if some examples could be provided.
    – NReilingh
    May 12, 2014 at 4:37
  • can you post a link to piece of Hindustani Sangeet music?
    – Anthony
    May 14, 2014 at 14:43
  • "I have analysed notations of some study songs and observed that the tonic is hardly there." It would help to know how exactly these study songs are being notated. In Indian classical music, the notation for a song is anyway expressed using the notes Sa-Ri-Ga-Ma-Pa-Dha-Ni (analogous to Do-Re-Mi-Fa-So-La-Ti). So, right off the bat there can be no question of a hidden tonic in a song analyzed from its notation, making this question entirely unclear. I presume the large number of upvotes are due to its age?
    – user77458
    Nov 2, 2021 at 15:18

6 Answers 6


I am going to preface my answer by saying that I know very, very little about North Indian Classical music. However, from your descriptions, I believe that there are parallel techniques used in Western music that may help you define tonic.

Essentially, it appears that you're asking how to define tonic when the harmony / melody is not functioning in the way it is expected.

If the Shadaj is fixed and the singer is adhering to this rule (despite not using the drone), then they will invariably end up returning to a certain pitch over and over again as they are defining that as the tonal center of their scale. By choosing to sing some pitches more often than others, they are creating a tonal hierarchy, and thus, are creating a tonic through pitch-emphasis rather than adhering to the drone.

So, I would listen for notes that the singer keeps returning to and / or emphasizing through neighbor-note motion; which creates (in some aspect at least) a form of tension - release that will emphasize a certain tone more than others.

You have also mentioned that you have studied some scores. In the scores, you can physically see which notes appear more often than others (you can even count the number of times they appear if you'd like, or if the music is particularly complex.) Once you've established a tone-hierarchy, it then becomes a matter of determining the relationships of the most prominent tones -> are some creating / resolving tension somehow? If so, in what way(s)?

If North Indian Classical Music is similar at all to the Maqams of Egyptian / Israeli music, then you may be able to also use modulations to help you determine how the scales are being used and which pitches within those scales seem to be more emphasized than others.

I'd also add that I think you have a great idea about transcribing what you're hearing - not only can it be a great tool for understanding what's going on, but it also makes a great record of the culture's music that can then be shared with others.

Lastly, it may also be a good idea to talk to those musicians and ask them how they are able to find tonic if they are not using the drone. It may very well be that there in fact is no tonic at all, which I think, would represent a fascinating shift in music if it was a new phenomenon.


Key and tonic are highly contextual in all music, so if a melody line is all one has to go on and that line is ambiguous in what it implies (it does not contain enough information to narrow down all possibilities to one answer), it may not always be possible to arrive at one answer with a high degree of certainty without additional information.

Of course, knowing that we are talking about North Indian classical music gives us quite a bit of context. We know that this music is almost always accompanied by a drone on the shadaj, we know that the shadaj forms the basis for a raga, which defines a melodic construct, and we know that instruments like the tabla and tanpura are tuned to the shadaj.

Harmonium, as you mention, can be used to play melody, but can also be used to provide a drone (some harmoniums are provided with a drone stop to make this easier).

So, to the point of figuring out what the shadaj is from a subset of all of this contextual information, the job is easy if any drone is present OR if the tabla can be heard. If, however, all we have is the melody alone (either in audio or transcribed form), then we need to reverse-engineer the sequence described above. Since the shadaj informs the raga and the raga informs the melody, with enough melodic information we should be able to ascertain the raga in use, and in doing so we determine the shadaj.

Ragas are different from "Western" scales in that the set of notes, their tunings, and the melodic embellishments used on those notes are all defined by the raga in use. So, we have all of those factors to go by when narrowing down the possibilities of which raga, and therefore shadaj, is associated with a melody. This also means that having the shadaj missing from the melody that is being sung should not be such a great burden with all of the other information to be used to ascertain the raga.

In terms of actually making this determination of which raga is in use, the process should not be so different for an expert of Indian classical music as it is for a Western classical or jazz musician to do the same for a scale or key. From a cognitive perspective (for either style), the musician that hears a melody associates it with his or her existing knowledge and training so he or she has an expectation of where the melody will go and what notes will be used. After this guess is made, a few more seconds of listening can confirm or refute it, and if correct, the determination has been made within a margin of error provided by the extent of musical context present.

The untrained ear can do the same process logically, of course, though at a much slower pace, and with a higher rate of error. (This applies to any algorithmic or computerized methods as well.) The sheer number of ragas makes this a bit of a tall order for Indian classical music compared to common practice Western classical music, but if you only need to find out the shadaj, you can group your possible ragas by shadaj to make your process of elimination a bit more manageable.

Disclaimer: I am not formally trained in Indian classical music; please let me know if I have gotten something wrong.


You are right, there are songs without the singer actually singing the Shadaj. If there is an accompanying drone, the drone will keep playing the Shadaj. But if there is no drone, and the singer is not physically singing the Shadaj, it still virtually exists and is an integral part of the Sruthi/Pitch the user is trying to sing in. If you want to findout the Shadaj, then you need to find out the notes that are being sung and extrapolate it to find out the Shadaj. I do it all the time and is a principal technique.


Are you familiar with the structure of the key.

In western music, major scale intervals (in half tones) would be: 2 2 1 2 2 2 1.

I am guessing the indian system has a similar structure. If you know it. Just list all the used pitch classes, put them in order. And see where the half tones are. This will help you determine the scale.

Once you know the scale, and if it sounds major or minor, you can identify the tonic. In classic music theory, it also helps for instance to look at starting and end notes.

  • I don't know nearly enough about Indian music to answer the question, but scale structure is different enough that a comparison with the common western scales is difficult.
    – Karen
    May 13, 2014 at 18:25
  • I wasn't suggesting using the same intervals. Could this be of help? raag-hindustani.com/Scales1.html
    – dorien
    May 13, 2014 at 20:10

Once we comprehend the tune as intended by the composer, our brain has already identified the tonic. Without identifying the tonic, the same tune will sound very different. So, the question about identifying the tonic becomes that of locating the tonic inside your brain, which has made you recognize the tune that it actually is. If you are already a musician, it will be easier to recognize the tonic note, else it requires some amount of training in identifying the notes. There are musical structures or parts in which the tonic may be absent. But in classical music, it goes on, on the Tanpura, which continously supplies the tonic. However, in light music, the tanpura may not be used, where it may become a bit harder to identify the tonic in absence of the Tanpura.


In the general case you cant. Actually you cant even if the supposed tonic is in there.

Let's take this example, the notes G, B, D, C in order. Now I can find multiple major candidate keys for just this excerpt alone (for instance both G major and C major fit on top of it, in fact the notes G B D form the G major chord). Depending on my interpretation, the tonic could be either G or C.

It becomes even more complicated even if I use the relative minor scale. For instance the relative minor of C major is A minor, and all these notes fit on top of that too. So just listening to these four notes could also interpret this as a melody with the tonic on A, but being played on minor.

Also note all these interpretations are equally valid. The only way to disambiguate the first conundrum would be if the note F# or F was played (thereby clearly disambiguating it as C major or G major) but that does not disambiguate between A minor.

There is also the added complexity that you could use a church mode instead (for instance G major scale is also the B phrygian mode which makes the tonic B or the A dorian mode). Hell I could invent an arbitrary scale on top of these notes with any tonic.

Now while I used western theory to highlight a point, the argument still holds in ICM. For instance you can replace Major scale with anything from thaat bilawal and Minor scale from anything in thaat Asavari and the central problem is still exactly the same.

To summarize my point, for the general case (i.e. any arbitrary combination of notes) it is not possible to infer the tonic. The attempt to do so often relies on hazy definitions of tonic identification like "I just can, my brain just knows its C". without providing a tangible rubric of understanding how that inference came into being.

  • I think the best "general case" should be at least a piece of music - yes, four notes is usually ambiguous, but usually an entire song has enough contextual information to define a particular tonic (at least in Western music, maybe it's different for North Indian?). No, there's nothing mathematical about it, but that doesn't mean that it's not possible to infer the "correct key".
    – user45266
    Mar 24, 2020 at 19:13
  • As a counter example to your assertion, there is nothing stopping from defining my song as an ostenato that repeats the above 4 notes over and over again. In fact, a lot of modern western pop do in fact does it as their background score.
    – Khalian
    Mar 24, 2020 at 19:27
  • Yes, but those modern western pop songs that do that still make sense in a certain key. As an example, the first part of "Livin' on a Prayer" has the underlying ostinato EEBDEEBD. Yes, those notes can belong to many keys. However, the song's melodies (and the rest of the song in general) make it obvious that E minor is the only sensible interpretation (or G major, but that's another argument).
    – user45266
    Mar 25, 2020 at 20:24
  • "There is nothing stopping from defining my song as an ostenato that repeats the above 4 notes over and over again" Yes, that is valid. It's perfectly fine to do that, and in that case you would be right that the idea of a key is useless. But that's not very common in Western Music. Ostinatos are common, sure, but songs usually don't just consist of a couple of notes in isolation. (Also, you can ping me so that I get notified about a comment by using the @ symbol, like so: @Khalian)
    – user45266
    Mar 25, 2020 at 20:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.