Dhrupad and khyāl are the two forms of classical singing that exist today in North India. Dhrupad, the older form, enjoyed wide popularity till the seventeenth or early eighteenth century, after which it gradually declined with the emergence of khyāl, a more entertaining style. The decline of dhrupad accelerated during the last two centuries, with many of its practitioners switching over to the new form, which progressively increased in popularity and attracted greater patronage. Dhrupad however remained the favoured style in a few imperial courts, mainly in Rajasthan and Bihar, where some dhrupad traditions continued till the late 1940’s, when these states were assimilated into the Indian republic. The period after Indian independence till the present times was a difficult one for this art, for it called into question its very survival in a society in which it was not popular, but needed the patronage of a new ruling class of bureaucrats and politicians who unlike its previous aristocratic patrons, had no connection with it at all. There are at present very few practitioners of dhrupad left in India, and as can be expected in such a situation, there is an enormous fragmentation and erosion of knowledge about the art, even among its few remaining practitioners.

The whole body of composed work of the tradition has been practically decimated. Lack of knowledge about it has reached a point where it is difficult, even in the literature of music and in musical circles, to find a proper definition of what dhrupad is, and what sets it apart from its modern derivative - the khyāl. It is common in India now to find dhrupad described in terms of the language and the grammar of khyāl. Most descriptions list the obvious structural differences between khyāl and dhrupad and emphasize that in dhrupad, ornaments and melodic devices like murkī, khatkā, phirat, and particularly fast passages called tāns, which are characteristic of khyāl singing, are strictly avoided.

The decline of dhrupad during the last two centuries coincides in my opinion with a paradigm shift in Indian music, in which it came to be accepted that music must primarily entertain. This is a concept that reigns supreme in India today, and therefore precludes any attempts to revive or even initiate a serious study of dhrupad. But the sophistication of the musical concepts underlying dhrupad, and its objective of creating a music that uplifts, but does not necessarily entertain, and that embodies the essence of Indian spiritual thought, has found for it a growing acceptance and admiration in the West. Since the visit of the elder Dāgar brothers to the West in the 1960’s and the efforts of Alain Danielou kindled interest in dhrupad, many singers have given performances there, and the number of concerts, workshops and seminars of dhrupad in the West now significantly exceed those in India. It would not be an exaggeration to say that the interest of people in the West has made dhrupad singing, financially a more viable profession for its few remaining practitioners. From the knowledge that I gained during my apprenticeship under various maestros of the Dāgar family, and particularly under Rahīm Fahīmuddīn Dāgar, the eldest surviving member of this family of musicians that has practiced dhrupad for many generations, I would say that the fundamental feature of dhrupad singing which characterizes it and sets it apart from other kinds of singing, is the use of the three vedic accents of udātta, anudātta and svarita, and the complex dynamics of sound and the musical possibilities that this creates. The fundamental role played by these accents in creating this music actually underlies its grammar, for these accents define the patterns in time and the internal dynamics of the melodic phrases. The dynamics of sound employed in dhrupad go back to the researches into sound embodied in the vedic recitations and Sanskrit grammar, and also lead to the origin of the concept of a rāg. The use of these accents is also the clearest evidence that dhrupad has evolved from the vedic chantings.

The importance of the vedic accents in dhrupad singing is evidenced by the numerous references to them in texts of traditional dhrupad compositions. For example a composition in rāg Rupāvatī from the Dāgar family repertoire, attributed to the eighteenth century dhrupad singer Adāran۠g and quoted by Sanyal and Widdess [1] lists the three vedic accents along with ten of the basic ornaments used in singing. Udātta, anudātta and svarita are defined as follows by Panini, who is credited with creating the grammar of classical Sanskrit.

That which is pronounced higher at its place of articulation is termed udātta. That which is pronounced lower at its place of articulation is termed anudātta. That which is pronounced with a combination of features of udātta and anudātta is termed svarita. [2] Similar definitions can be found in the various prātiśākhyas of the Vedas for example the Śukla Yajurveda Prātiśākhya [3] Udātta, anudātta and svarita here refer to the dynamics of sound that accompanies Sanskrit pronunciation. Pronunciation is accompanied by movements of energy. An upward movement of position or energy in the body is udātta, a downward movement anudātta and a horizontal movement svarita. In dhrupad singing the positions used for producing sound are: the navel, heart, throat, palate, lips, tongue, teeth, head, the two nostrils and the region between the eyebrows and sound constantly moves through these positions. There is also a major difference between the way changes of position are used in the vedic recitations and in dhrupad, which is that the processes of udātta, anudātta and svarita in dhrupad are made independent of the syllables themselves, but subservient to the requirements of the melody, while in the recitation of the vedas, the dynamics of sound embodied in udātta, anudātta and svarita are dependent on the words and syllables. This separation, in dhrupad, of the dynamics of sound from the pronunciation of syllables, is done for the purpose of producing a complex music, in which the melodic devices and ornaments become independent of the syllables or the text, but obey the logic of the music. Making the processes of udātta, anudātta and svarita independent of the syllables requires knowledge of the distinction between two kinds of sound or nād: ādhār nād and nirādhār nād. Ādhār and nirādhar literally mean with support and without support. Ādhār nād is a frontal throaty sound anchored to the vocal chords, while nirādhār nād is sound that fills the body like a fluid entity and can freely change position. It is only with nirādhār nād and a special kind of pronunciation (nād kā uccāran or the pronunciation of nād) in which the syllables are made so soft that they don’t influence the dynamics of sound, that the processes of udātta, anudātta and svarita can be made independent of the syllables and dependent only on the logic of the melodic development.

Another important possibility that emerges from the complex dynamics of sound with the vedic accents is the use of a note that is completely or partly silent. It is not articulated clearly as sound, but is merely an internal change of position that prepares for the next note that is to be produced. The vedic accents and this concept of gupt and prakat, or concealed and revealed notes, enables a dhrupad singer to create hidden relations or consonances (samvād) between notes. It is this concept of samvād or the consonance of notes that is the origin of the concept of a rāg.

The recitation of mantras of the vedas is accompanied by formalized gestures and movements of the arm, the hand, and the fingers, corresponding to the three accents [4]. These gestures and movements have a profound effect on the sound. In the melodically more complex dhrupad, these formalized movements are replaced by more free and fluid movements that however, still have a direct relation to the internal dynamics of sound in the body of the singer.

The processes of udātta, anudātta and svarita and the use of concealed notes leads to the microtonal conception of a note as a fluid entity, for udātta or anudātta within a note and silent changes in internal position actually create a play on the microtonal shades of the note. The dynamics of sound within a note makes it a fluid entity that is not fixed, but is a part of the infinite spectrum of notes created by overtones. Through the dynamics of sound, a note within a rāg constantly establishes samvād with other notes of the rāg. The background of a spectrum of overtones, to which a dhrupad singer sings, is provided by the drone instrument the tānpurā, whose curved bridge with its shifting point of contact with the strings passing over it, embodies in its design the concept of a note as a fluid entity. An important consequence of this concept of a fluid note with infinite microtonal shades is that a note within a rāg also captures in it the entire information of samvād or consonance within the rāg, and thus also characterizes the rāg itself. A rāg can be thus characterized or identified by the characteristic microtonal shades of its notes. The samvād within a rāg produces a microtonal shift in the tonic, and consequently also in the other notes which are related to the tonic through overtones. This concept is also reflected in the tuning of the drone instrument. A skilled dhrupad singer would tune the tānpurā to reflect the character of the rāg being performed. Of the four strings of the tānpurā, the two middle strings tuned to the tonic form the reference pitch, and the last string is tuned to the microtonally shifted tonic that characterizes the samvād of the rāg to be performed. A rāg is seen as a personality or a mood or a state of being, and not as a sequence of notes sung in a certain way. The character of the rāg is established by the information of samvād contained in the microtonal shifts in its notes.

The concept of a shifted tonic for a rāg in relation to the reference pitch becomes more understandable when one considers the process of the generation of notes by natural overtones. The successive generation of six fifths above, and five fifths below a reference pitch yields twelve notes. If we carry this process further then the generation of eleven fifths above and ten fifths below the reference would yield twenty two pitches. These twenty two pitches are schematically represented on a logarithmic spiral in diagram1 while the second diagram shows 53 such pitches generated by perfect fifths, of which the twenty two represented in diagram1 are a subset. A repetition of this process would eventually yield an infinite continuum of notes and infinite microtonal shades of the reference pitch. From these infinite variations of the reference pitch or the sā and of the other notes, the rāg chooses the ones it needs, depending on the mutual consonances of its notes. Another way of expressing this would be to say that the consonances and relations of the notes within a rāg push and pull the notes into their correct places. For example although the rāgs Puriā and Mārwā use the same notes, the vastly different internal relations and consonances within these rāgs results in a lowered or komalān۠g sā for Puriā and a higher or tīvrān۠g sā for Mārwā. This concept of the essence of a rāg being contained in its sā is called the bīja rupa or the seed form of the rāg.

The internal consonances within a rāg do not merely change the pitch of its sā and the other notes, but also require changes in the resonance and quality of the sound produced by the voice. For example the term lajjit or shy is used to characterize the quality of the notes of Puriā, while those of Mārwā are termed tejasvī or radiant. In rāgs with lowered notes, a special technique called sakārī which produces a soft diffused resonance in the voice, is employed to make these extremely shifted notes sound natural to the ear. If the low notes of Ābhogī or Puriā were to be sung with the same quality of the voice as the raised notes of Mārwā, they would sound very discordant. In an actual rendition of a rāg, depending on the ornaments and relations that are being explored, a singer has to constantly adjust the pitches by minute amounts and also change the resonance and quality of the voice using various internal processes described in the preceding paragraphs. An important aspect of vocal Dhrupad is the possibility of making continual changes in sakārī or resonance in the voice along with changes of microtone.

Among the rāgs in the audio samples accompanying this article Ābhogī and Jaijaivantī are examples of komalān۠g rāgs, with Ābhogī being an extreme example with drastically lowered notes, while Mīyā Kī Malhār is sung in the tīvrān۠g with raised notes. In rāgs like Ābhogī or Puriā, the low pitches would sound very odd and discordant if the background drone were to be suddenly made silent. This is something I have sometimes seen some of my teachers doing by abruptly placing the hand on the tanpura strings to cut off the drone to show how shifted the pitches of the rāg being performed are. However the presence of the drone with its overtones and the skilful use of the resonance of the voice with sakārī, not only makes these shifted notes sound natural, but also greatly enhances the feeling of the rāg by reinforcing the samvād or the entire internal consonance of the rāg.

Dhrupad is for the major part an abstract singing without a text using abstract sequences of syllables like ā, ra, ra , nā , ra, nā , noom, na …. that have no literal meaning. These monosyllabic sounds are traditionally said to be derived from a mantric text. The singing is mostly improvised, but the improvisation follows a grammar and structure. A dhrupad performance begins with this abstract improvised singing called the ālāp, with slow free flowing melodic phrases that are not initially differentiated into beats. The phrases of the ālāp take up the different facets or an۠gs of the rāg one by one, and their gradual succession creates an impression of the rāg slowly unfolding itself. Using sequences of these abstract syllables, the singer improvises phrases within the framework of the rāg and the grammar of dhrupad. A phrase of dhrupad ālāp is constructed much like a sentence of speech, with a beginning, one or more intermediate parts, and a final part which brings it to a conclusion. Sometimes a phrase could have just one part. Even one note can sometimes constitute a phrase, carrying within itself the germ of a beginning, a middle, and an end, through the use of the dynamics of sound described in the preceding paragraphs. The phrases are constructed in sequence with each phrase slightly varying the musical idea presented in the preceding phrase, and anticipating and preparing the ground for the idea to be expressed in the next one. Often a whole group of phrases illustrates a particular an۠g of the rāg. A group of phrases expressing a major aspect of the rāga is ended by a characteristic phrase like the ending of a paragraph or a chapter in a long essay.

Typically, though not always, the regions or facets being explored start around the tonic: the sā of the middle octave, gradually descend into the lower octave and again rise to the tonic, expand into the middle register, progressively rise to the highest register and finally descend again to the tonic. The objective is an orderly progression through the an۠gs of the rāg, though expert singers can also produce exceptional developments in a rāg, where the an۠gs being explored do not strictly follow this sequential progression. Indeed in certain rāgs a sequential progression is not possible because of the very nature of the rāga.

After this exploration of the rāg through free flowing patterns comes a phase where a rhythmical pulse is introduced into the ālāp, and the facets or an۠gs of the rāg are explored within this pulse, which is quickened in stages. The introduction of the pulse and the progressively accelerating tempo leads to the introduction of new melodic elements and ornaments, that are nearly absent in the slower preceding part. This part concludes the ālāp and is followed by the singing of a composed melody with lyrics, set to a tāl or a cycle of beats accompanied by a barrel drum - the pakhāwaj.

Strictly speaking it is this composed melody, which has four parts, that bears the name dhrupad, because this composed melody actually embodies within it all the principles of the music, and is actually a model of the exposition of the rāg, composed and set to lyrics by a master dhrupad singer and taught to his students for subsequent oral transmission to succeeding generations. After singing this composed melody, the singer explores the rāg again with improvised phrases, using the lyrics of the composition with accompanying rhythmical improvisations on the drum. These phrases obey the same grammar as the alāp, but are bound now by the cycle of beats - the tāl. In the audio samples accompanying this article is an excerpt from the slow part of an ālāp in Ābhogī, an excerpt of moderate tempo from an alāp in rāg Mīyā Kī Malhār and and an example of a dhrupad composition in rāg Jaijaivantī showing the first two parts of the composition followed by rhythmical improvisations.

A consequence of the fundamental role of the vedic accents is that phrases of dhrupad cannot be too fast. An overly rapid sequence of notes would make it impossible to execute the udātta and anudātta accents with the accompanying internal changes of position, and to correctly use hidden (gupt) notes and consonances (samvād) and put in the microtonal details that come from the concept of a note as a fluid entity. That is why in khyāl this technique of the voice is abandoned for a more frontal throaty sound or ādhār nad as opposed to the nirādhār nād of dhrupad, for this is conducive to the production of the very quick ornaments and fast passages and permutations of notes that are characteristic of khyāl.

However, though this also leads to a music that is in some sense more entertaining and exciting, it also fundamentally alters the very concept of rāg, for devoid of the vedic accents, the music cannot adequately explore the concept of samvād and microtones, but must start regarding rāgs as sequences of notes executed in a certain way.


  1. Dhrupad Tradition and Performance in Indian MusicRitwik Sanyal, Richard Widdess, Ashgate Publishing Limited, page161
  2. The Ashtadhyayi of Panini – Rama Nath Sharma, Munshiram Manoharlal Publishers Pvt. Ltd. sutra numbers 1/2/29, 1/2/30, 1/2/31 pages 99,100
  3. Shukla Yajur Veda – Pratishakhya, with commentary by Dr. Virendra Kumar Verma, Choukhamba Sanskrit Pratishthan, Delhi, ch. 1 sutra no. 108, pages 59,60
  4. Hidden Face of Ancient Indian Song- Solveig McIntosh, Ashgate Publishing Limited, chapter 4


The author wishes to express his gratitude to Skye Lofvander and Knud Brant for their numerous stimulating discussions and insights on the subject of microtones, and to Skye Lofvander for the two diagrams accompanying this article. He also expresses his gratitude to Solveig McIntosh for reading and correcting a major part of the article. The author is however solely responsible for any inadvertent errors that may still be present. Thanks are also due to Martin Redfern and Resul Pookutty for the work they did on the audio recordings, excerpts from which accompany this article and to Dalchand Sharma for his pakhawaj accompaniment in one of the audio samples. The author dedicates this article to the maestros of the Dagar family - Ustad Nasir Aminuddin Dagar, Ustad Rahim Fahimuddin Dagar, Ustad Zia Mohiuddin Dagar, Ustad Zia Fariduddin Dagar and Ustad Hussain Sayeeduddin Dagar, who gave him the knowledge that made the writing of this article possible.

The full article here.

Download PDF