An x-ray of two raga-s: Darbari Kanada and Jaunpuri

Colonial musicology and its language

The colonial heritage of musicological thinking starts from language. English became the dominant world language, American its neocolonial pendant. The choice of language inevitably brings along a mindset that derives from the culture to which that language belongs. The English musicological vocabulary is both inadequate and confusing when speaking of other musics. Speaking of Indian music for instance, there simply is no word in English for raga. On the other hand the concepts of harmony, chord, symphony—to mention just a few—have no equivalent in Hindi. The problem of Indian music can be partially solved by writing in Hindi. Partially: when writing about Hindustani music, for writing about Karnatak music is another, more complex issue. All that doesn’t solve the problem of writing about Chinese music. Indeed, then we would have to write in Chinese. OK, so for convenience we use one language, the lingua franca of this era: not Esperanto or Klingon, but English. Bummer. The problem with musicological English is its deep relation with European classical music. My colleagues in the field of WAM studies would argue that German would be a more suitable language. And then come French and Italian. Some of them don’t even know English! Ah, you say, what is WAM? It stands for Western Art Music, more or less the same as European classical music. These terminologies are intricate. Why Western, why European? Western sometimes means the western hemisphere, but most of Europe is not there. Sometimes it is used to refer to the region where the European culture dominates, i.e. Europe + the Americas. Some people may not include the Latin American countries because they are ‘underdeveloped’. And since Russia also stays out, western could be equated with NATO. I even used the expression North Atlantic music. My colleague Nicholas Cook suggested we speak of western music, with a lowercase “w”, to make clear that it is something else than Western. Great idea, especially when I’m lecturing. Anyway, many colleagues prefer to speak of ‘the European musical tradition’. This, by the way, includes both Russia and the Latin American countries. Whether it includes Greece and Southern Spain is another question. Of course Greece is the most important fountainhead of European culture, but some of its musical traditions, such as Rebetika, are non-European. Same for Flamenco. And of course the music of first nations / indigenous Americans and Afro-Americans.

The other terms “classical” and “art” are perhaps even more confusing. Classical refers to ancient Greek or Latin culture. Which, in the case of music, is rather ridiculous. Sculpture, architecture, literature and perhaps even painting in Europe were influenced by ancient Greece and Rome, but for music this is nonsense. This reference to the european ancient cultures is called the renaissance (rebirth), which took place in the fifteenth and sixteenth centuries. And yes, the “classical” music of Europe also flourished in that period. But the heyday of classical music is later, while its early developments starts much before, certainly from the twelfth century onwards. Many scholars limit the term classical to the music of the 18th and 19th century, and what comes before they call “early music”. The word ‘art’ is really the worst of the whole problem. I recently visited the KMSKA, Koninklijk Museum voor Schone Kunsten van Antwerpen (Royal Museum for Fine Arts). One of the meanings of ‘art’ is “a skill at doing a specified thing, typically one acquired through practice” (Apple Dictionary 2.3.0). A very broad definition indeed. But what we see in before-mentioned museum are (mainly) paintings of high class and high value. Top quality, bags of money. And the so-called ‘works’ of European music are also dependent on high skill and represent high quality. It’s really an elite business. I’m using elite here in a general sense, persons or a group that excel in a particular field, political, economic, religious, creative, sportive, intellectual. The “great” works of art are made or performed by people who possess a rare talent and have worked hard through training and practice to polish and develop their talent. They are the creative elite, they are specialists in their field. Elite derived from elect, though select would come a little closer. Most of the time however it’s a matter of a little in-group, either hereditary or through band forming. That’s why we don’t favour elitism.

This digression is to localize the language of my musicology. And since English is the colonial language par excellence, the English of musicology is also as colonial as it gets. I learned some european language in school and later tried to study Swahili, Amharic and Chinese but abandoned that before I could get anywhere. In Hindi I became reasonably proficient, enough to read books on Indian music and conduct a simple conversation. I also did a year of Sanskrit, but that wasn’t enough for anything useful. So, I am doomed to write in English or Dutch. And even though English isn’t my native tongue I will use it, especially because my Indian readers often know English better than Dutch. As I mentioned above, Germany may be the epicentre of the European musical tradition, and hence German the most suitable language if we are talking about that music. But I’m not. I’m talking about Indian music, more precisely about Hindustani music. And in India, or preferably the South Asian subcontinent, we meet with similar problems as in Europe. I already mentioned the two main forms of Indian classical music, Hindustani and Karnatak. Roughly speaking they are practiced in the north and the south. But there are other traditions that do not belong to either: Odissi (in the east) and Sufiana Kalam (in Kashmir). And the traditional music of Kathakali, Sopanam, is not really Karnatak music, although it has similarities. The language of Indian musicology is possibly more diverse and confusing than the European situation. Early musicological works have been written in Sanskrit and Tamil, in the Middle Ages Persian studies were added and in more recent times there are texts in some 20 different languages. The language of musicology depends on the language in which the musicologist writes. But the terminology is related to the music at hand. So it happens that when we write in English about Indian music we use lots of terms from the Sanskritic tradition. Raga, tala, svara, matra, etc. But sometimes we use an approximate equivalent: thus matra can be called a beat and a svara a note. Or should it be a tone? Actually, the English note and tone are not quite the same as the Dutch noot en toon. Sometimes we don’t really know the meaning of a certain term, like the sruti of ancient Indian musicology. When English researchers tried to grasp it, they thought it was a unit in a system of tuning schemes. At the time, proponents of harmonic tunings were fighting equal temperament and they suggested the 3-sruti interval was a minor whole-tone whereas the 4-sruti interval was a major whole-tone. Thus, the 7-sruti interval would be the harmonic major third, the 9-sruti interval the perfect fourth and the 13-sruti interval the perfect fifth. Mind you, these theories emerged in the very beginning of the nineteenth century, well before Hertz and Helmholz.

The words “microtone” and “microtonal” were coined before 1912 by Maud MacCarthy Mann in order to avoid the misnomer “quarter tone” when speaking of the srutis of Indian music. Prior to this time the term “quarter tone” was used, confusingly, not only for an interval actually half the size of a semitone, but also for all intervals (considerably) smaller than a semitone. It may have been even slightly earlier, perhaps as early as 1895, that the Mexican composer Julián Carrillo, writing in Spanish or French, coined the terms microtono/micro-ton and microtonalismo/micro-tonalitĂ©.

(https://en.wikipedia.org/wiki/Microtonal_music)

However, contemporary Indian musicians refer to sruti-s as a special way of rendering a note in a raga. In western musicological terminology it would be called ornamentation, but that again creates a misunderstanding. For the sruti is not the icing on the cake of the note, it is the rendering of the note as a gestalt, the shape of the tonal space is its very essence, not something added to the discrete pitch. I have explained this on my page about gandhara in Darbari. I remember having a discussion with the Dutch composer Ton de Leeuw (1926-1996) who was into microtonality at a conference about Indian music, a major source of inspiration for him. He explained that to produce a microtone, the pitch had to be extremely steady, there should be no vibrato whatsoever. I objected that the microtones of Indian music were never steady, but rather moved about in tonal space. He angrily professed this was impossible. However, the highly respected Indian musician present at the conference, sitar player Imrat Khan (1935-2018), brother of Vilayat Khan and son of Inayat Khan, confirmed my position, much to the disappointment of our composer friend. Of course, in reality we were talking about two completely different things: sruti as a pitch and sruti as a gestalt. In which the former meaning is an insubstantiable interpretation of sruti in ancient texts, whereas the latter is a term from contemporary practice. The word sruti in Sanskrit refers primarily to the revelation of the sacred texts, in particular the Veda’s. Literally it means “that which is heard”, or perhaps “that which was heard” or “that which has been heard”. In musicology it has been suggested that this should be interpreted as the smallest interval we can distinguish. Or, when can we hear two tones being just ever so slightly different. But considering the way the term is used nowadays would require a different understanding of “what was heard”. By the way, the term ‘gestalt’ is not quite suitable for describing the phenomenon because it suggests a fixed shape, whereas in actual practice the shape develops and transforms over time.

Tonal structure of raga-s

I felt this somewhat tedious introduction about musicology, terminology and language was necessary before I broach the subject of this post: the tonal structure of raga performance. My colleague, a pioneer in computerized analysis of music, invented the tonagram. The computer program may be measuring the pitch 30 times per second, and each measurement is counted and put in a table. The table is turned into a graph. Below is the tonagram of Uday Bhawalkar’s rendition of Darbari Kanada on AUTRIM.

In this graph 0 is Sa, 1 Re, 2 Re, 3 Ga, 4 Ga, 5 Ma, 6 Ma etc. There are thin lines at 100, 200, 300, 400 etc. cents (if you don’t see them open or download the graph and enlarge it). I use underlining for komal svaras, but in the case of Ma the underlined version (Ma) is shuddh, whereas tivra Ma is written without underlining.

The tonagram is really an x-ray of the performance. What we see could be called the skeleton of the raga. There are some concepts in Indic musicology that have bearing on the tonal structure of a raga. I am quoting from my Hindustani Music in the Twentieth Century (1980, 19-20):

Allowed and forbidden phrases depend on the context and can only be derived from the outline of the raga. The same is true for duration, stress and accent. It is a mistake to think that these factors can be described in terms of nyasa, vadi and amsa or even bahutva and alpatva of tones. The concept of nyasa is still the easiest to handle; those notes on which one can stop are nyasa, e.g. sa, re and pa in Miyan ki Malhara. Vadi and samvadi should be understood in relation to musical phrases and elaboration. Each piece of elaboration makes a particular note ‘shine’ or ‘sound’. To this note there may be a corresponding note which sustains the main note and has a perfect fifth relation with it. Therefore a raga usually contains many vadis and samvadis. Bahutva and alpatva have a more scientific meaning as they indicate the frequency with which a tone occurs. This can be worked out graphically.

I think I was struggling with those concepts at the time I wrote my PhD (the book was the published version of the PhD from 1978). I couldn’t really make sense of the concept of amsa, as it was not in use among musicians, it belonged to ancient musicology. Nyasa was also an ancient concept but it had a very clear-cut and useful meaning in raga phrasing. Vadi and samvadi were concepts propagated in some musicological literature and in music schools, but they were controversial among musicians. Samvada could be best explained as ‘harmonic relation’. Nothing to do with the European concept of ‘harmony’. Thus in the Bilaval scale SP, PR, RD, DG, GN could be samvada, but NM is not. Hence the tendency would be to add tivra ma or komal ni, or both, as we see in Yamankalyan, Khamaj or Kedar to mention just a few. Nazir Jairazbhoy (1971) has considered this the driving force behind the evolution of scales and raga-s, although he formulated the process in terms of parallel tetrachords rather than samvada relations. It’s the same, Winnie the Pooh would say. Of course, we also know that DG aren’t samvada if G is a harmonic major third, pace Pythagoras. However, in quotidian parlance vadi and samvadi are used to denote the “most prominent” notes of a raga, whatever that means. The tonagram as devised by Bernard Bel was used to correctly measure the pitch in performances of a raga. The ISTAR team, with Bernard, Jim Arnold, Joep Bor and myself felt the methods used by Jairazbhoy and his student Levy were unreliable, and moreover we felt that the outcome of their result that sruti-s don’t exist was an insult to Indian music, Indic musicology and the subaltern in general. Later I came to the conclusion that Jairazbhoy’s theory was more practical than the artificial squeezing of contemporary musical practice into fantasized reconstructions of ancient theory. Ancient theory that was devised by Europeans longing for a pristine untempered music to be found among primitives and subalterns. And the sacrosanct subalterns themselves were more catholic than the pope and championed the ideology of the superalterns. My guru, Pt. Dilip Chandra Vedi, was a different kind of scholar-musician. He was blessed with a scientific mind and could separate ideologically based theories from practice. He possessed critical scrutiny, he took very little for granted, he was very independent, he could weigh the advantages and disadvantages of opposing theories and he tested theory against practice. No-one is perfect, not even Vediji. I won’t go into detail but those who have known him are aware he, being human, also had his defaults and prejudices. I don’t of course, but that’s another matter. What about you, dear reader?

We soon understood that the tonagram of a full performance doesn’t say much about intonation. The whole idea was utterly silly, because Indian music is not focussed on discrete pitches. If you make a tonagram of a piano, an organ, even a guitar or an oboe you may get discrete pitches, but in India only with a jaltarang. During a performance pitches can shift, and moreover the pitch may not be steady at all. Part of the debate between Jairazbhoy-Levy and the ISTAR team was about the gandhara of Darbari. But as explained in my page on that subject, as well as our article in the ISTAR Newsletter 3-4 (On measuring notes, p.50), that gandhara is never a steady pitch. Levy took the everage on the pitch positions he had measured, we suggested one should perhaps look at the lower passing of the andol (oscillation). But even that suggestion is moot, the lower position isn’t fixed. In the course of a single performance it changes, between performances it changes and from one artist to another it changes. Levy’s suggestion that the perceived pitch would be the average is of course untenable. In the case of vibrato this is true if the frequency of the vibratoes is more than five per second. But andol-s are slow oscillations, and we do hear the pitch going up and down very clearly. In short, there is no fixed pitch there and consequently there is no sruti. By the way, if the Ga of Darbari would be atikomal, it would have to be five sruti-s above Sa. But obviously Re is the standard four sruti-s above Sa, which means Re-Ga would be one sruti which is not a possible configuration in the tradition of Bharata.

When we understood that the full tonagram of a performance wasn’t going to solve the question of intonation Bernard devised a method of creating a selective tonagram. This meant that the software isolated pitches that were held steady for a certain amount of time, e.g. one second, and between certain limits of pitch variance (note that in vocal music especially pitch is never completely steady, even if we perceive it as such). This resulted in an overview of more or less discrete pitches. Perhaps this could serve as a theoretical backbone of the scale being used. We could also have resorted to measuring the pitch of the sympathetic strings on a sarangi or sitar. Interestingly Bernard had created a sruti harmonium, in which every pitch was fully tunable with a read-out of the pitch (like some Yamaha keyboards). The result was baffling: musicians were totally inconsistent in tuning it, both between musicians and between sessions of one musician. Bernard concluded that this approach was useless and that pitch had to be measured in performance, not in an experimental environment. I wasn’t present at this phase of the research and I might surmise that the timbre of the sruti harmonium was such that the pitch couldn’t be estimated very well. The lack of partials in such timbres makes pitch perception very difficult. In fact, a pure sinus tone always seems to be in tune somehow. In tune? Funny concept. In tune with what? What this means is that the produced tone is consonant with an external tone. And consonance is established by matching partials. Since pure sinus tones don’t have partials we can establish neither consonance nor dissonance. Certain flutes and even some high-pitched female voices also have that quality. Which brings me to another point I have to mention: the pitch extraction system that Bernard had developed in the nineteen eighties (and all other pitch trackers before and around that time) used hardware filters to isolate the fundamental frequency. Modern programs such as PRAAT use an algorithm to estimate the perceived pitch. The whole process is therefore performed in the software. That simplifies the job but the result, especially with vocal music, is not shockingly different. Of course there is ample evidence that perceived pitch is not necessarily the fundamental o f a complex wave form. Therefore software that takes into account the state-of-the-art models of pitch perception should be preferred over the older methods.

Back to the tonagram of Uday’s Darbari: what strikes immediately is the prominence of Sa, Re and Pa. What also strikes is the shape of Ga: it is stocky. That’s because of the andol-s that make the tonal space broader. And it seems to be quite glued to Re, which, in the case of Uday’s interpretation is explained because he lets the Ga emerge slowly from Re. You may notice the average pitch of Ga seems very low and some people might think “see, the sruti of Ga in Darbari is ati-komal after all”. But I will retort “No no no, it is not a pitch that is lower than “normal”, it is a moving pitch, you cannot define it in terms of the nineteenth century interpretation of Bharata’s tuning scheme”. Moreover, in that interpretation ati-komal Ga would be 292 cents, whereas the peak in the graph is much lower, 260 cents.

We may also note that Dha and Ni are quite broad, whereas Sa, Re, Ma and Pa are more pointed. That’s because the tonal space around them is used more intensively, though not as much as Ga.

Now let’s have a look at Uday’s Jaunpuri:

Basically, Darbari and Jaunpuri use the same scale but the tonagrams are quite different. In the image below the two are compared more easily:

To start with, Sa is the most prominent note in Darbari, while Pa is in Jaunpuri. In general the whole configuration of Jaunpuri is higher. Although Jaunpuri is not generally considered to be uttaranga pradhan (focussed on the second tetrachord, from Pa to high Sa), we can say that it is more so than Darbari. Darbari is a very solemn raga, and that requires elaboration in the lower regions. The central note, aptly called Ma (Madhyama, the middle), is almost identical in both ragas. In Jaunpuri however Dha is much more present than in Darbari, and the space between Ni and Dha is much more “filled”. Curiously the mind from Ma to Ga is much less prominent than the mind from Ni to Dha.

I will never forget the first time I heard Kishori Amonkar. It was 1970 on AIR (All India Radio), and she sang Jaunpuri. She had been announced but I didn’t know her name or background. She immediately reminded me of Mogubai Kurdikar, my favourite singer. And as I found out later she was Mogubai’s daughter. My wife Marilene got hold of her LP and we felt deeply moved by her spine-chilling Ga that seemed to come out of the deepest layers of the universal soul. The difference between the Ga of Darbari and the Ga of Jaunpuri is like the difference between the slowly undulating and swelling waves of the ocean and the outpouring of power from Iguaçu waterfalls (sorry, but I haven’t been to Niagara or Victoria). If you have read my Hindustani Music in the Twentieth Century you will know that phrasing is the main character of raga. The order of notes, the notes that are prolonged, the notes that end a phrase, the inflections of the notes. All these bring out the face of the raga (the raga murti). The scale says very little about the raga, and yet, as I have tried to demonstrate here, the x-ray of the skeleton of the raga can help to understand some salient details.

As has been so often the case, when I started learning Jaunpuri with Vediji a whole new world opened up. In many ways his interpretation and that of Kishoritai were similar (Vediji loved her music, which was rather unusual as he was generally critical of the younger generation), but there was one very big difference: the treatment of Dha. Most musicians approach Dha in a mind from Ni, but Vediji preferred to start from somewhere between shuddh Dha and Ni. It is so haunting. Perhaps it is a rare occurrence of the seventh harmonic, although to prove that would be impossible as the intonation is never steady but part of a mind. On the AUTRIM website there is a Jaunpuri by Aslam Khan, who belongs to Agra gharana, like Vediji. Below is the tonagram of his interpretation and I invite you to see if it makes sense.

Are the selected recordings representative?

To finish, a note about the representativeness of the recordings and as such of the tonagrams. In the much touted Raga Guide we asked four renowned artists to make a very short recording of a total of 74 ragas. This was inspired by the 78RPM recordings of the great masters of the second quarter of the 19th century. We felt that since those stalwarts were able to summarize a raga in less than three minutes we could ask our contemporary masters to do the same. Thus, with a booklet and four CDs we presented a short compendium of Hindustani music. When internet became a viable option, Suvarnalata Rao and myself thought we could go beyond the Raga Guide. First, we gave artists more leeway, we required them to make a brief recording but to use as much time as would be necessary to bring out all the essential characteristics of the raga. This varies by raga. Hamir is obviously more limited than Miyan ki Malhar. Second, instead of the staff/sargam combined notation of the Raga Guide we opted for the live graphic representation (a subject I have discussed elsewhere). Third, we had a different approach to de accompanying description. In the Raga Guide we had a fixed scheme of aroha-avaroha and calan. The description in the AUTRIM website is more open-ended and includes also extensive references from other experts in the field. In a sense the AUTRIM website is an enlarged, extended version of the Raga Guide but it also involves different approaches to description and representation. Moreover the number of raga-s has increased and the number of recordings even far more as we often have several recordings of the same raga by different artists.

Still, because of the very idea that the artists would provide a full picture of a raga without entering into superfluous elaborations, we feel that the recordings do represent a reasonably complete insight into the raga. Many musicians take pride in performing a single raga for hours on end. They suggest they never repeat a single phrase identically, which I have no reason to doubt. But other musicians, and the famous sarod player Hafiz Ali Khan springs to mind, never performed a raga for more than some 20 minutes. Vediji was also not in favour of excessively long-drawn performances of a single raga and I guess Bach would agree, if he had been born in India.

Bibliography

Arnold, W. J., Joep Bor, Wim van der Meer, “ON MEASURING NOTES, A Response to N.A. Jairazbhoy”, ISTAR Newsletter 3-4, p. 46-59, New Delhi 1985

Bor, Joep, editor, author, Suvarnalata Rao, Wim van der Meer, Jane Harvey co-authors, Henri Tournier, music transcriptions Lalita du Perron, song text translations Robin Broadbank, recordings, Hariprasad Chaurasia, flute, Buddhadev DasGupta, sarod Shruti Sadolikar-Katkar, vocal, Vidyadhar Vyas, vocal, The Raga Guide, A Survey of 74 Hindustani Ragas, Nimbus Records with Rotterdam Conservatory of Music, 1999.

Deva, B.C. and P.S. Nair, “Forms in Music”, in Sangeet Natak 2, Apr. 1966, 106-117.

Jairazbhoy, N.A., The Rags of North Indian Music, Their Structure and Evolution, Faber and Faber, London, 1971.

Levy, M. Intonation in North Indian Music – A Select Comparison of Theories with Contemporary Practice, Bib!ia Impex, New Delhi 1982.

Meer, W. van der, Hindustani Music in the Twentieth Century, Martinus Nijhof, The Hague, 1980.

Other references have direct links in the text.