Sri Srinivasa Yajvan’s Svarasiddhāntacandrikā – Introduction in English

Sri K. A. Sivaramakrishna Sastri, after careful research of several manuscripts, brought out an edition of Svarasiddhāntacandrikā by Srinivasa Yajvan. This book is an exhaustive commentary on Pāṇini’s Svarasūtras.

This very well written introduction Sri K. A. Sivaramakrishna Sastri gives an overview of the origin and nature of tones / accents or svaras, their transformation and uses. He describes at length the utility and application of svaras in the Vedas. He goes on to provide an insight into ekaśruti, and its place in both Vedas and Classical Sanskrit. The reader’s attention is also drawn to an interesting aspect of this article, where he elucidates the author’s argument of the validity of applying svaras to texts in Classical Sanskrit!

This book was originally published in 1936 as part of “Annamalai University Sanskrit Series – 4”, the book can be downloaded from




Śabda (tone) is the origin of all sounds.”[1] “Of all Varṇās (sounds) dhvani (tone) is the root-cause“[2]

Dhvani the original means of communication of ideas

Man is endowed with the two powers of producing – (1) the articulate sound and (2) the tone. By mere birth he is not able to produce the articulate sound; only after closely observing the conversations of his elders does he try to imitate them. Even in this attempt he does not all on a sudden become an adept but only after exerting himself for long in that direction. The case of tone, however, is different. A newborn child afflicted by hunger cries out; as its suffering is much or slight, it cries aloud or in a low tone. This capacity of the child to produce different tones is, therefore, innate, while the capacity to produce articulate sound is acquired by constant practice. The former power being independent manifests itself before the acquired power begins to operate. Hence man’s tone is anterior to articulate sound.[3]

Tone is the means of communication for those who have no power of speech. We observe that animals and birds convey their thoughts by gestures and tones. In some cases tone alone is employed, since gestures serve no purpose in the absence of company. Men also make use of tones in their dealings with animals and birds; so do children and dumb people.[4]4 Hence tones are the means of communication when language and gestures fail. To this effect Hari says — “Tones, it is observed, influence the actions of animals, etc.” Hence modern Philologists accept the view that language has developed from tones, as may be inferred from the attempts of animals and children to convey their ideas.[5]

Tone ever associated with and accessory to speech

Even after the advent of language, tones have persisted in it since they are  natural to men. Angry, happy or terrified people, unable to express themselves fully, are seen producing only tones. All languages, therefore, abound in indeclinables like ha, hi without any derivative sense.[6] Not only are they convenient in language but also necessary. When three or four men are suffering from an ague and when they are asked about the nature of their suffering, all of them may reply ‘unbearable’. Does this reply convey accurately the particular degree of suffering that each man undergoes? Since all of them use the same expression, it is not possible to have a comparative knowledge of their suffering except from their tones, indistinct, soft, harsh etc. Similarly emotions like pleasure, sorrow and terror cannot be so well known in their different degrees from their respective articulate expressions as they could be from their respective tones. Hence tones are indispensable to language.

Language superior to tone

Better and greater however are the uses of language for mankind than those of tone. To illustrate, a mother hears her child crying, thinks it is hungry, and tries to give it some food, but it does not take it in and cries aloud. Then the mother understands that the child is having some pain. But until she knows that the cause of suffering is fever, stomach ache or flea bite, she cannot think of any remedy. How can she know the disease from the tones of the child unless it is given expression to by articulate speech? On the other hand, from articulate speech the child’s suffering and its cause can be known clearly and precisely. Moreover, tones are expressive of emotions like pleasure, agony and terror for the time being in as much as they are the very expressions of those emotions. When these emotions disappear, they too disappear, and so they cannot explain things that are of the past. Hence, since man will have no power of knowing things of the past and future, mankind, nay, the whole world, will sink bank into the animal state. Danḍin has expressed the same idea when he praised speech thus. The entire three fold world would have been enveloped in darkness, had it not been for the illumination called Śabda (speech) all over the world.[7] So the communication of one’s thoughts to others, and ones understanding of others from their expression, etc., are made possible best by language, yet language serves its purpose better when it is attended by tones suited to varied emotions.

Transformation of tones

Philologists hold that language had originally been unitary and that it became diversified according to the changes wrought by time and space. Ordinary people use certain words which are not generally used by the learned though they are as expressive as the words of the latter. Similarly even in cases of words used by both sets of people there is a perceptible difference in their utterance, so also in the way of conveying ideas by words. From these differences a careful observer can know the learning and habitat of persons. The reasons for these differences are the varied contacts of people with learned men, their different capacities for learning and their divergent ideals of conduct. Hence divergences in their expressions and in the control of their emotions like anger, joy and terror and consequently in their tones.

Definite Tones the cause of beauty

In the development of language and tone, man, not satisfied with the mere communication of his thoughts to others, introduced into language particular arrangements of words in prose and poetry and methods of expressing beautiful ideas. Similarly he regulated tones for the sake of beauty and convenience in utterance. All people feel happy in hearing words that are uttered in a particular tone, even when they do not understand their meanings. Learned people are doubly happy, because they understand the meaning of those words and appreciate their appeal to the ear. Among public speakers some are able to move the audience to tears, to make them indignant and to make them dance with joy, there are others who do quite the reverse. Such regulated tones admit of a two-fold division, some subordinate to the meaning conveyed and the others, independent. When they are independent of the meaning conveyed they become beautiful songs under restrictions of order and time in various degrees. When they are subordinate to ideas, this restriction of order and time may delay and sometimes obstruct the comprehension of ideas. Hence tones have different rules in language.[8] Tones differ according to the attitudes and circumstances of speakers. When many speak in different tones, at once there is disharmony. Such occasions unavoidably arise in festivities. So it is desirable to have some rules to regulate the tones, so that no disharmony might arise.

Tones or accents manifold

The rule regulating tones could not be the same in all languages, since different races possess different tastes. What pleases one need not please another. Beauty is more subjective than objective even when individuals are considered. The elements of one language are not found another – for example, in Sanskrit there exist the dual number, svarabhakti, the sound, etc., which are not found in many other languages; the sound found in Tamil is absent from Non-Dravidian languages. A similar lack of uniformity exists in the case of tones also. These tones are employed in letters, words and sentences with a view to producing manifold effects in language (as explained above) and therefore become an integral part of language.

Tones are otherwise called svaras (accents). Emotions like anger, pleasure, terror, hatred and envy are known from different tones. To illustrate the word subhāṣitam, when uttered in a particular tone, indicates contempt and therefore means apabhāṣitam. In other tones it indicates a query, an answer, praise or envy. So also the reduplicated form tiṣṭha tiṣṭha indicates extreme regard or threat. In a similar manner words with different accents or tones like ādyudātta, madhyodātta and antodātta convey different ideas. This is explained by the accent sutras of Pāṇini. “Satyam praśne, etc.”[9] These sutras record the existing varieties of accents on the basis of natural emotions. And it does not stand to reason to say that these accents were introduced into language at a particular time, as they occur naturally in daily conversation.

Uses of accents

Accents are primarily intended for the expression of emotions and our sense of beauty. There are other uses, which have, as has been said, an important part to play language:

(1) Economy may be mentioned first. Man, particularly a learned man, wishes to speak as briefly as possible. For example, in the statements – ‘gacchayati kim? gacchatyeva na tiṣṭhati’ (does he go? he goes only and does not stand), the senses of interrogation and assertion are conveyed by the words kim and eva respectively; but when it is possible to convey those ideas by particular accents even without the use of those expressions, no purpose is served by the latter. So also ideas of contempt and praise can be made known by accents without the use of their respective expression. Hence economy is an important use of accents.

(2) In all languages numerous homonyms exist. They cannot convey their meanings without reference to the intention of the speaker by means of the context and the like. Even the context cannot sometimes bring out the intention since it may be common to more than one sense. But the accent in a word which forms an integral part of the expression fixes the sense. In accordance with this says the Mahābhāṣya (Paspaśāhnīka) – “Yājñīkāh paṭhantı sthūlaprasatim agnivārunim ālabheta”.[10]  The sacrificers prescribe that a sacrifice is to be performed with a cow possessing wide spots for the twin-gods, Agni and Varuṇa. There arises a doubt in the compounds – sthūlaprasatīm in the passage referred to, whether it is to be split into “sthūla cāsau prasati ca”. sthūlaprasati or “sthūlāni prasanti yasyāḥ sā sthūlaprasati”.  This doubt cannot be cleared without the help of accent. If it has the accent of the pūrvapada, it is a bahuvrīhi, or if it has the antodātta of a compound, it is a tatpuruṣa.[11]

Similarly in the Upaniṣad passages explained in support of dualism, absolute monism, qualified monism, etc., there is no means of knowing their intention and no conclusion can, therefore, be drawn from them. In such cases accent would be very helpful. For example, in the Mahāvākya – (ātma) tat tvam asi (That thou art) – some split it as atmā-tat while others as atmā-atat. If the sound a after t is udātta the splitting tat is correct; if it is svarita, it is atat. Similar explanations on the basis of accent are largely found in the Vedas. As now-a-days accent has fallen into disuse in many Upaniṣads, their intention has to be ascertained in other ways.

(3) Mimāmsakas are of opinion that spiritual merit accrues from the recital of the Vedas with accent.[12]

(4) Another use of accent may be mentioned. In ancient times owing to the absence of books and other modern requisites of learning, there must have been emphasis on the traditional method of reciting the Vedas, and for the unsullied perpetuation of the traditional method, the employment of accent would be very useful, if not necessary. Further, learning by rote would be facilitated by the employment of accent.

(5) Lastly as sentence split is of vital importance in the interpretation of the Vedic texts, the help rendered to it by accent cannot be exaggerated

The advantages of accent detailed above may be common to all languages, but its special characteristics differ in different languages. It is a laborious task to make a comparative study of accent.[13]

We can fully explain the varieties of Vedic accent as the chanting of the Vedic hymns with the proper accent is in full swing at the present day.

Origin of Accent

The Prānavāyu starting from the lungs and coming into contact with the organs above the lungs, produces dhvani. It is audible or inaudible as it is produced in the throat or below the throat. When the throat is of sub normal elasticity, the breath has not full scope to manifest itself, and the abhighāta increases. When the throat is of more than normal elasticity and functions, the prāṇa comes out fully and the abhighāta decreases. When the elasticity of the throat is normal, a dhvani is produced different from that described in the other cases and akin to the utterance of h. Thus dhvani is of three kinds – nāda, śvāsa and hakāra. With the help of these three dhvanis all articulate sounds are uttered – so says the Taittirīya Prātisakhya.[14] The four sutras of this Prātisakhya explain the three dhvanis along with their significant names. Though the three dhvanis are produced by the Prāṇa, the particular dhvani called is so called because it requires for its production more breath than do the others.

In the languages of the world one and the same sound is produced sometimes by śvāsa dhvani and sometimes by nāda dhvani. These dhvanis applied to vowels are called svaras. Though they are applicable to consonants as well, they are not regarded as svaras. So two varieties of svaras exist – nāda and śvāsa.

Some Philologists[15] opine that there are many indications in Sanskrit and kindred languages of the fact that Śvāsa was once prominent in those languages, though they are now considered to have nāda predominantly. The Taittirīya Prātiśākhya[16] clearly says that the vowels in Sanskrit have nāda. But on the authority of Uvvaṭa[17], the commentator on the R̥gVeda Prātiśākhya, and of Nāgeśabhatta[18], is some say that the other variety of accent, śvāsa, exists in the Vedas. The second view need not be taken as going against the first, in view of the fact that those authors have most probably referred to the contact of air quite indispensable to the production of nāda. The Taittirīya Prātiśākhya sutraNādo’nupradānam – expressly explains the importance of nāda. So it is indisputable that nāda is the only chief svara in Sanskrit.[19]

Varieties of Nada

There are three accents in the Vedas – udātta, anudātta and svarita. Both udātta and anudātta are independent while svarita is only a combination of the other two. So say Pāṇini and the Prātiśākhyakāras. The three sutras[20]Uccairudāttaḥ, Nīcairanudāttaḥ and Samāhāraḥ svaritaḥ – are found in the Aṣṭādhyāyī, in the Taittirīya Prātiśākhya,[21] and in Kātyāyana’s Prātiśākhya[22] with the verbal change of the sūtraubhayavān svaritaḥ. In the R̥gVeda Prātiśākhya different definitions of svaras are given: (1) The three svaras are pronounced with ayāma, viśrambha and ākṣepa[23] respectively, 2) Svarita is that svara in which there is the combination of udātta and anudātta in one syllable.[24] It is explained by Paṇini and Kātyāyana as the combination of two svaras, udātta being the first and anudātta the second. cf (a) tasyādita udāttam arddhahrasvam (P. S. I. 2. 32) (b) tasyādita udāttam svarārddhamātram (K. P. S I 111.) Thus udātta has high pitch, anudātta, low pitch and svarıta, a combination of both. The term udātta is explained as that which is uttered in a high tone and the term anudātta, as that which is uttered in a low tone.

Character of svarita discussed

Many scholars[25] have observed that the Paṇiniyan system of grammar is subsequent to the Prātiśākhya literature. Though there are many portions of the Prātiśākhyas proving their posteriority to Pāṇini, yet those portions are regarded as interpolations. It must, however, be admitted that the nucleus of the Prātiśākhyas is pre-Pāṇiniyan. That this view is acceptable to Patañjali, the author of the Mahābhāṣya, is clear from his bhāṣya on the sutra 1 2 32. From this bhāṣya it is evident that Pāṇini has in the main dealt with those aspects of the subject not included in the texts dealing with the places, means and modes of articulation. Similarly Patañjali regards the three sutras giving the definitions of the three svaras as superfluous in that they can well be known from other sources. Moreover, Pāṇini while enjoining ekaśruti, by the sutra – ekaśruti dūrāt sambuddhau[26] has not defined it, since it can be known from other sources as in the case of prathama. The sutras – nityam mantre,[27] vibhasa chandası[28]  yajuṣyeke sam[29] – are rules applicable to mantras, Brāhmaṇas etc. Even in those cases, sometimes, no strict rule can be framed. Hence the sutras like vyayatyayo bahulam,[30] justifying the Vedic forms that have been transmitted by oral tradition. The Bhāṣyakāra clearly explains in the sutra – tatpuruṣe krti bahulam[31] the nature and purpose of Pāṇini’s Vyākaraṇa as distinct from those of the Prātiśākhyas. We should necessarily employ terms like – bahulam, anyatarasyam, vıbhasa, ekesam since the Vyākaraṇa Sastra is applicable to all the Vedas, so a single rule (covering all cases) cannot be framed[32]. Thus it is clear that each Prātiśākhya deals with one recension of the Vedas while the Vyākaraṇa Śāstra treats the Vedas as a whole. Hence the impossibility of the latter making common rules for the multifarious changes in the Vedic forms. To explain these changes a mere reference to Pāṇini’s sūtras is therefore, hardly sufficient, unless they are fully borne out by the Prātiśākhyas, etc.

Even in the Prātiśākhyas, if the rules enjoin grammatical operations common to many, their acceptability is open to question, unless they are supported by some specific authority. Grammar too cannot render any help in that direction, since it is more comprehensive in character. To say that the Prātiśākhyas are unreliable serves no purpose. So in those instances decision should be arrived at only with reference to the tradition of Vedic recitation, which however cannot be set aside as unreliable, as the chief aim of the Vyākarana Sastra and of the Prātiśākhyas is the safe preservation of the time-honoured method of Vedic recitation

Tradition is valuable not only in the study of the Vedic language, but also in that of other languages. All the peculiarities of a language cannot be known from its grammar: for example, the roots and śav are given the senses of cutting and going; but on their restricted usages grammar is silent. That these usages can be known from local practices is well attested by a passage in the Mahābhāṣya[33] “The root śav is used in the senses of ‘to go’ and ‘to change’ by the Kāmbojas and the Āryas”

The differences in the utterance of tones and Varṇās cannot be learnt from grammar but only from experience. The Prātiśākhya sūtra[34]jihvāgreṇa prativeṣṭya mūrdhani ṭavarge [ जिह्वाग्रेण प्रतिवेष्ट्य मूर्धनि टवर्गे ] – explains the place and means of articulation of the avarga. One who has not heard that sound cannot exactly understand its nature this sūtra. Unless a man hears the sound of , a knowledge of its definition cannot help him in producing that sound. Only persons born in England or those who come into contact with Englishmen can pronounce English correctly. The knowledge of correct pronunciation and usage is therefore indispensable to the understanding not only of sounds but also of their definitions given in the various Prātiśākhyas.

Since the Vyākaraṇa Śāstra, as has been said, does not in the main deal with the nature of the three accents, we must turn our attention to the Prātiśākhyas and tradition. According to Indian tradition the three accents are pronounced thus – udātta in a middle tone, svarita in a high tone and anudātta in a low tone. This practice is prevalent today throughout India in the recitation of the Vedas with three accents. The peculiarities of each recension of a Veda are known; one syllable is uttered one recension in the udātta accent while the same syllable is uttered in another in the anudātta accent, as in the Vājasaneya Brāhmaṇa. This change cannot be effected by the student of the Vedas of his own accord, but only with the authority of the relevant Prātiśākhya and other works on accent. A similar practice is found regarding anudātta and svarita; but never is udātta pronounced in a low tone, or svarita in a middle tone. If it is proved that these accents are pronounced in some recensions differently from that indicated above, the question naturally arises – which of these accents is the oldest? What the science of grammar explains regarding the position of accent in a word is closely followed in practice even at the present day. Hence in this case there is no contradiction between theory and practice.

As regards the nature of udātta and anudātta there arises no controversy, but not so in the case of svarita, since the Śāstra explains it as the combination of the two basic svaras – udātta and anudātta. It is, therefore, natural that in the pronunciation of svarita both udātta and anudātta should be heard. But the traditional practice is that svarita is pronounced in a high tone different from that of udātta and anudātta. This is well borne out by many Sūtras in the Prātiśākhyas. One sūtra from the R̥gVeda Prātiśākhya (III-4) says that the first half mātra or first half of svarita is higher than udātta; the rest is anudātta but heard as udātta. Many sutras[35] taken at random from the other Prātiśākhyas elucidate the nature of svarita in a slightly modified way. According to some, the first part of svarita is udāttasama and the second part, anudāttasama, anudāttatara and udāttasama. The whole svarita has the high tone, which is considered to be higher than the middle tone – udātta. Thus the difference between udātta and svarita well establishes the latter as a separate accent.

Similarly the Vaidikābharaṇa[36], a commentary on the Taittirīya Prātiśākhya, explains the middle tone while commenting on the sūtra – dhr̥taḥ pracayaḥ Kaundinyasa – that the pracayasvara is called dhr̥ta which stands exactly n the middle of kruṣṭa, prathama and dvıtīya on one side and of caturtha, mandra and atisvarya on the other, and that it is therefore an unchanging middle tone. The definitions of the three accents in the R̥gVeda Prātiśākhya[37] and the Kātyayana Prātiśākhya[38] also favour the above view. Uvvaṭa while commenting on these two definitions says[39] clearly that “the svarita accent is different from udātta and anudātta in that it has a distinct sound. It is produced by the combination of the two accents just as bronze is produced when copper and tin are alloyed.”

The term udātta does not signify the sense of increase of tone. This is explained by the bhāṣya[40] on the sūtra – uccairudāttaḥ “the words ucca and nīca are of relative significance, one thing may be ucca (high) to a person while it is nīca (low) to another”. Hence their scope is relative. Kaiyaṭa[41] and Nageśabhaṭṭa[42], while commenting on that sūtra, hold the same view. Kaiyaṭa[43] further remarks that the particular accent is to be known by constant practice, like the musical tones, ṣadja, etc. This shows clearly that the accents like udātta stand on a par with the musical tones. Patañjali while commenting on the sūtra – tasyādita udāttam arddhahrasvam – interprets[44] udātta, the first part of svarita, as high tone and indirectly emphasizes that the name udātta in the sūtra – uccairudāttaḥ – is not to be regarded as significant. Thus he agrees with the Prātiśākhya view that the tone of svarita is higher than that of udātta[45].

It may also be said that Pāṇini accepts the view that the pitch of svarita is higher than that of udātta. Though he allows two udāttas in a word, he does not countenance the combination of two svaritas owing to the high pitch of svarita. Further, he inferred that the svarita and udātta are of different pitch. The utterance of many sounds of middle and low tones is perhaps possible, but it is absolutely impossible in a sound of high tone because an incessant flow of air current exhausts the speaker easily. On this ground the Prātiśākhyas[46] introduce anudātta between two svaritas. This intervening tone is called kampa, which is of two kinds, the tone preceding anudātta is sometimes udātta and sometimes svarita. The Taittirīya Prātiśākhya defnitely says[47] that in the Taittirīya Samhita the preceding tone is always svarita. This is fully borne out by the commentary, Vaidikābharaṇa[48], which restricts the other variety of Kampa to the Kuśmanda mantras and other recensions of the Yajurveda. The Kampa in the Kuśmanda mantras is in practice to day. It is highly probable that Pāṇını’s sutra – tasyādita udāttam arddha hrasvam – founded upon those mantras and the recensions mentioned above. So the high pitch of svarita characteristic of modern practice is nothing but the continuance of the tradition, the persistence of which in ancient India is indubitably testified to by the Prātiśākhyas, their commentators, Pāṇını, Patañjali and his commentators[49].

Whitney’s position examined

Though the nature and validity of the traditional point of view are well understood by Max Muller[50], Whitney regards[51] anudātta as the normal tone, udātta as a high tone and svarita as intermediate between the two and gives the following illustration.

This is based primarily on the misinterpretation of the Pāṇinīya sutra – uccairudātiaḥ. The word uccaiḥ is taken by Whitney to mean high, but its real sense is the uccaiḥ sthāna of the throat, etc., as understood by all Sanskrit grammarians. Further, the term udātta is regarded by him as significant, but the original authorities already quoted disallow the interpretation of udātta as high. Thus his basic assertion is contradicted by Indian authors. Further, if uccaiḥ means high nīcaiḥ should mean low, but Whitney gives it the sense of normal on the basis of his interpretation of the term anudātta as non-high. Moreover, the intermediate place assigned by him to svarita is hard to understand as he himself says that svarita is a term of doubtful meaning. It seems that the main cause of his confusion is his neglect of the value of Indian practice at the present day for the interpretation of ancient Indian texts.

Variations of accents

These three accents undergo certain variations when they occur in the Samhitāpāṭha. Generally the word has one udātta or one svarita; the other accents are anudāttas. Hence Pāṇini’s sūltra – “…anudāttam padamekavarjam[52]”. So in dissyllabic words there are many anudāttas followed by udātta or svarita. The last of those anudāttas cannot be pronounced like the preceding anudāttas because the former is inevitably influenced by the succeeding udātta or svarita. Hence it is uttered in a lower tone than in the ordinary case. Such an anudātta is called anudāttatara, sannatara, a term used by Pāṇini ın I 2 40, or Vikrama in the Taittirīya Prātiśakhya[53]. Similarly arises the term udāttatara (Pāṇini, I. 2. 35). Svarita is of two kinds, nitya and anitya. Nitya is independent of other accents whereas anitya is dependent on them. Another division of svarita is into prākr̥ta and aprākr̥ta; the former comes within the scope of the niṣedhasūtra (viii. 4 67) but not the latter[54]. In the Prātiśākhyas many other divisions and definitions, even an eight-fold division, are found[55]kṣipra, nitya, abhinihita, Praśliṣṭa, Tairovyañjana, pādavrtta, Tairovirāma and Tathābhavya. Consequently shades of difference in pronunciation are indicated.

A fourth variety of accent is called ekaśruti, the nature of which is not clear from the diverse accounts of grammarians[56]. Patañjali says that this accent is a combination of udātta and anudātta like svarita, but Kaiyata says that udātta and anudātta are clearly heard only in svarita. According to the Kāśikāvr̥tti, ekāśruti obliterates the distinction of the three accents. The Āśvalāyanasūtra explains it as the continuance of one of the three accents without the intervention of the others. According to the Taittirīya Prātiśākhya it is more or less udātta. Our author attempts without success to harmonise these conflicting accounts. If the name of the accent in question is regarded as significant, the view of the Āśvalāyanasūtra is the most satisfactory, and practice at the present day conforms to the Prātiśākhya definition. If ekaśruti is treated as a separate accent, six accents are on the whole to be recognized[57]udātta, udāttatara, anudātta, anudāttatara, svarita and ekaśruti. If we separate the udātta of svarita as is done by Patañjali[58], we get seven accents.

Studying the question of accents with reference to the Vedic Śākhās we come across a welter of conflicting opinions which make it impossible to say definitely in a particular case whether the proper accent is udātta, anudātta or svarita, though there is no manner of uncertainty regarding the pronunciation of these accents. According to Katyāyana’s Bhāṣikapariśiṣṭasūtra[59] there are only two accents, udātta and anudātta, in the Śatapathabrāhmana. In the mantra portion of the Vājasaneya samhita three accents are distinguished. The Carakas maintain the traisvarya of the mantras in the Brāhmaṇas; among them the Khānḍikīya and Aukhīya sections employ cātussvarya[60]. Among the Bahvr̥cas[61] some pronounce the last mātrā of pracaya in anudātta tone. In a series of pracaya the last, or more than one from the last syllable, are uttered by some others like anudātta; but the very last syllable all Bahvr̥cas pronounce alike. Owing to these perplexing varieties the accents of śākhās cannot be accurately determined except with the help of their respective Prātiśākhyas.

Identity between traisvarya and saptasvarya

In the recensions of the Sāmaveda there are seven accents with their own varieties described above, and these do not differ from the three main accents and from the ṣadja and other svaras of music. The fact is well borne out by the Taittirīya Prātiśākhya[62], Nāradaśikṣā[63] and Yājñavalkyaśikṣā[64]. But some southern scholars in music and some westerners like Dr A. C. Burnell[65] question the identity between the main accents and the tones of music like ṣaḍja. Dr. Burnell goes even the length of saying that unfounded statements like these identifying dissimilar things are not rare in the works of ancient Indian scholars. Siddheshwar Varma attempts to justify the statements of the Śikṣās. On such controversial matters unbiased judgments are to be pronounced only by those well-versed in music and Vedic lore. It is for them to decide on what basis the Śikṣās referred to identify the traisvarya with the musical sāptasvarya.

Accents sometimes belong to consonants

These accents belong to vowels and not to consonants. Hence they are known as svaras. By their association consonants sometimes possess tones. Those acquiring the tone of svarita are important, and they are anusvāra, svarabhakti, ṅ, ṇ and n, etc., which may be well illustrated by the following passages:

वा॒य॒व्यश्वे॒तमालभेत्, दो॒षावस्तोर् ह॒विष्मती, गर्भमायन्, ब्रह्मण् व॒न्तः॒, न्यङ् ।

These consonantal tones cannot be pronounced in the duration of a half mātrā. This excess of time is given in some Śikṣās recorded in the Vaidikābharaṇa[66]. Similarly they become uddātta, etc., as found in these passages:

अरुणाः काण्डर्‍षयः, ए॒तद॑धिवि॒धाय॒ र्‌षि॒रवो॑चत् |

Of these svarabhakti is independent. Hence consonants also possess at times accents like udātta.

Accents common to Vedic and Classical Sanskrit languages

That these accents are common to Vedic and Classical Sanskrit languages is accepted by the Śāstrakāras since both words and their meanings are the same in those languages. This principle has been accepted not only by the Mīmāmsakas[67] but also by Pāṇini who has not made any adhikāra like chandasi for his svarasūtras. That he has introduced the word chandasi into some of his svarasūtras[68] indicates clearly that other sūtras are common to both. Moreover, most of the Pāṇiniyan indicatory sounds – anubandhas – have the indications of accents as one of their main purposes. If there were no accent in classical Sanskrit, most of the indicatory sounds would be purposeless. Śrīnvasayajvan, author of the Svarasiddhāntacandrika[69], explains the application of svaras to all languages, as follows. “In short, just as all the Vedas are glorified for the chief purpose of explaining the nature of the one Supreme Brahman, the whole domain of Vyākaraṇa Śāstra exists for the explanation of svaras. The Vārttikakāra too has not introduced the chando’dhikāra into the svarasūtras, and so it is clear that he has approved of the traisvarya even in classical Sanskrit. The Bhāṣyakāra’s approval of traisvarya for the Aṣṭādhyayī in his comment on the sūtra VI.4.174 shows the application of traisvarya to both Vedic and classical Sanskrit. Patañjalis very substitution of the words ‘duṣṭaḥ Śabdaḥ[70] for ‘mantro hīnaḥ’ in the Śikṣā supports hıs view of the general application of svaras. Nāgeśabhaṭṭa[71] and others[72] to say that the view denying accent to secular speech is unauthoritative.

Traisvarya in all the Vedas

Now the question arises – whether the recital of the Vedic texts like the Aitareya Brāhmaṇa without traisvarya at present has been the practice from the very beginning, or whether the once existing svara has been given up in course of time. Bhaṭṭojī Dīkṣita gives an answer to this question in his Śabdakaustubha I. 2 36[73]. “Traisvarya does not prevail in all the Vedas. The sūtra – vibhāṣā chandasi – enjoining ekaśruti optionally is to be interpreted not as an ordinary but as a vyavasthitavibhāṣa, with the result that traisvarya is to be practiced in the mantra portions and ekaśruti in the Brāhmaṇa portions of the R̥gVeda, but traisvarya is to be followed in the Taittirīya recension of the Yajurveda; no other decision to be arrived at altering the practice of the Vedas.”

The Vr̥ttikāra[74], Haradatta, Nāgeśabhaṭṭa[75] and others explain the option of traisvarya and of ekaśruti in all the Vedas. In his Sabdenduśekhara[76], Nageśabhaṭṭa criticises the above view of Bhaṭṭojī Dīkṣita thus: “the Bhāṣyakāra has not included the sūtra – vibhasā chandasi – in the list enumerating the Vyavasthitavibhāṣās under his bhāṣya on VII 4.41. In some places the Vedas having three accents are recited in ekaśruti, only. The bhāṣya – ekaśrutiḥ svarasarvanāma – also explains ekaśruti as a fitting substitute for all svaras in the Vedas. This does not support the Vyavasthitavikalpa mentioned above. Even the Vr̥ttikāra accepts the optional aikaśrutya in the Vedas, when he says: “the adoption of traisvarya is with a view to acquiring spiritual merit or adr̥ṣṭa, and aikaśrutya is come by tradition, and people generally following tradition do not chant the Vedas differently. So the Vedic option is not violated.[77]

What Nāgeśa has said is true and appropriate. It cannot be maintained that traisvarya was never prevalent in the Brāhmaṇas on the ground that it is not practiced now. It cannot be argued that if traisvarya were ever practiced in those Vedic texts, it would have come down to us; for, it is hard to believe that all sampradāyas have been transmitted to us without a break though the discontinuity of tradition is not easy to account for. From the present day existence of traisvarya in one Brāhmaṇa we may infer its existence in other Brāhmaṇas as well, even in by-gone ages.

The argument of Bhaṭṭojī Dīkṣita that no change could be imposed on the Vedas against practice is fallacious, in view of the fact that in the Taittirīyāraṇyaka many varieties of accents and many readings of words and passages are found. If it is argued that one of them alone is traditional and correct, the other is to be set down as contrary to tradition, so that the break in tradition at times becomes an established fact. The Sārasvatopākhyāna of the Mahābhārata[78] describes a great famine which caused much confusion and misery, and Brahmins, afflicted by hunger, completely lost their power of memory, with the result that they lost their Vedic lore. Similarly in periods of cataclysmic changes in the country, the Vedas might have suffered modifications sufficient to effect discontinuity in tradition.

But only those well versed in the Vyākaraṇa Śāstra and Śikṣās can save the Vedas from corruption. Others can only learn them by heart, as they have no means of clearing their doubts, apabhramśas and apasvaras that are introduced into the Vedic texts. The Śāstraic scholars at the present day condemn the learning of the Vedas by rote without understanding their meanings. So a healthy combination of the two is essential. To acquire these two requires prolonged study and extraordinary patience. Moreover, Vedic studies are not at present encouraged sufficiently in India, and though the texts are preserved intact to some extent, their continued existence in future cannot be expected if the break in tradition widens.

Ekaśruti in the place of traisvarya

In the Vedic sacrifices, the hymns are chanted in ekaśruti, a substitute for traisvarya which has declined with the passage of time.

(1) As the study of the Brāhmaṇas having traisvarya was considered indispensable to the attainment of the great puruṣārtha and as traisvarya could not be introduced into them, ekaśruti, one of the permitted accents, was substituted.

(2) In classical Sanskrit poets have employed figures of speech like śleṣa. Ekaśruti, but not traisvarya, is suited to the employment of those figures of speech.

(3) Women and Śūdras are prohibited from the study of the Vedas. For their understanding of dharma various Purāṇās have been composed. If the Purāṇās also are to be studied in traisvarya, sufficient distinction could not be made between the two. Hence ekaśruti for the Purāṇās.

(4) Every spoken language possesses svaras. They are practiced in every age without any difficulty. In Sanskrit svaras are to be understood from the treatises on them. Moreover, the inflexional nature of the language adds to the difficulty of its study. So any compulsion in the adoption of svara would discourage its study. Further, Mīmāmsakās hold that spiritual merit is acquired by the proper study of the Vedas, and this is dented to classical Sanskrit. Pūrva and Uttara Mīmāmsakās recite the Vedas and Upaniṣads in ekaśruti because knowledge of their contents is emphasized. Hence the neglect of traisvarya in the Vedas and classical Sanskrit

Most of the overzealous āstikas however recite the Vedas in traisvarya, lest they should sin. But ekaśruti is sanctioned by the sūtra – vibhāṣā chandasi. The Śiksa – mantro hīnaḥ, etc. – does not enjoin anything new, but only condemns those who discontinue traisvarya in the course of their recital. So both in the Vedas and in classical Sanskrit one is not obliged to confine one’s attention to a particular accent.

Some[79] opine that Pāṇini enjoins the option of accent in the Vedas only, so that traisvarya is to be necessarily employed in classical Sanskrit. This view is untenable since it goes against the opinions of Patañjali[80] and Mammaṭabhaṭṭa[81]. Both Bhaṭṭoji and Nageșa split the sūtra vibhāṣā chandasi – into vibhāṣa and acchandasi, so that the sutra can be interpreted as enjoining ekaśruti, optionally in classical Sanskrit. That Patañjali regards the use of incorrect words as bringing in some pratyavāya to the speaker in the Vedic sacrifices only, proves the contrary of such a practice in classical Sanskrit. What is to be applied to the words in language is applicable to svaras also. Hence there is no rule for the regulation of traisvarya in Sanskrit. If the yogavibhāga of that sūtra is accepted, traisvarya also can be maintained as sanctioned by the Śāstras, particularly by the Bhāṣya, the Kāvyaprakāśa and time honoured practice. So the option of ekaśruti must be acceptable to all.

It is hard to appreciate the view of some scholars who support the usage udgātāraḥ on the mere authority of the Bhāṣya but reject the validity of ekaśruti in Bhāṣya though this svara is accepted by the Bhāṣyakāra and justified by the Sūtra.

Author of the work

Our author Śrīnivāsa Dīkṣita or Yajvan, the son of Kr̥ṣṇa Vipaścit and Anantāmba, of the Sankr̥ti gotra, was an expert in the Vyākaraṇa, Mīmāṁsā and Nyāya. He was also well versed in the Yajurveda and its aṅgas. He was the disciple of Rāmabhadra Dīkṣita, the author of the Jānakīpariṇaya and many other works. That he had more than one teacher is known from the introductory verses to his work. He must have, therefore, flourished in South India towards the close of the seventeenth century[82]. It is a well-known fact that Shahaji, the great Maratha king at Tanjore, famous for his gift of the whole village Thiruvisalur to renowned Pandits, made Rāmabhadra Dīkṣita reside in that village. As our author was the disciple of Rāmabhadra Dīkṣita he might have lived there along with his great Guru.

Nāgeśabhaṭṭa was probably his contemporary. But neither quotes the other. So it is highly doubtful if they ever came together. Nageśa had been at Benares when he was invited by Jayasimha, the king of Jayapuri for the Aśvamedha. This happened in 1714[83].

The Work

The Svarasiddhāntacandrikā is an exhaustive commentary on the Svarasūtras of Pāṇını. It contains also adverse criticisms on the Svaramañjari of Nr̥simha, son of Rudrārya.

Unlike the Bhāṣya and the Kāśikāvr̥tti, this work comments on the sutras on the basis of the logical relation between a general rule and its exception. It, therefore, contributes to an easy grasp of the sutras. Again, unlike the Siddhānta Kaumudi, it gives illustrations as far as possible from the Yajurveda. Hence its popularity with the students of that Veda. As against the five divisions – sādhāraṇa, prakr̥ti, pratyaya, samāsa and tiṅanta of the Siddhānta Kaumudi, it gives twelve divisions – upodghāta, samīna, paribhasa, samhitā svara, dhātu svara, pratyaya svara, samāsa svara, phit svara, subanta-sarvānudātta, tiṅanta-sarvānudātta, pluta svara and ekaśruti. It points out the full defects of the Svaramañjari as follows – Omission of many svarasūtras, inappropriateness of illustrations, wrong conclusions etc., elaborated in its introductory verses. Its distinctive merit is a comparative study of the Pāṇinīya svarasūtras with the Taittirīya Prātiśākhya and Śikṣā. It discusses the various views before it arrives at a considered judgment. Though it has not departed much from the path chalked out by the previous works, it covers a vast field and is remarkably successful as a compendium. Its language is simple, lucid and attractive

Grounding himself on the principle of Yogavibhāga, Bhaṭṭoji Dīkṣita explains ekaśruti with reference to classical Sanskrit, on the strength of the ekaśrutipāṭha of the word aikṣvāka in the danḍināyana sūtra and usages like śveto dhāvati in the Bhāṣya. This interpretation is rejected by our author[84] because of the inappropriateness of the illustrations. The ekaśruti mentioned may be explained by the rule – chandovat sūtrāṇi bhavanti. The sentence śveto dhāvati conveys only one idea but reminds the hearer of a similar sentence conveying the other idea. But this repudiation of Bhaṭṭojī’s interpretation is untenable. In the danḍināyana sūtra, ekaśruti is treated as a substitute for any other svara, just as in doubtful cases of gender, napumsaka is taken for granted. This simile and the context here make it sufficiently clear that ekaśruti may be employed in all cases. This rule ‘chandovat sūtrāṇi bhavanti’ applies only to extreme cases. In the instance of ‘śveto dhāvati’ the mediate suggestion of the second idea is impossible, because the accent of that sentence cannot suggest a different accent. If, however, another sentence is mediately suggested somehow or other, the difference of accent will be a hindrance to the attainment of the idea to be conveyed by that sentence.

On the whole, our author’s work places him in the forefront of commentators on the Svarasūtras of Pāṇini as he has handled a difficult subject with admirable mastery and ease.








[1] Taittīrīya Prātiśākhya, XXII, 1

[2] Tribhāṣyaratna, Mysore Edn, p 491

[3] Prapañcasāra, I 43

[4] Taittirīya Brāhmaṇa, II, 884, Bhaṭṭa Bhaskara Bhāṣya on the above, Vākyapadīya, II, 119, 148 and 149, Puṇyarāja’s commentary on the above, De Laguna, Speech, its Functions and Development, p 78

[5] O Jesperson, Language its Nature, Development and Origin, p 27, De Laguna, op cit, pp 75 and 81

[6] Benfey quoted by O Jesperson op cit p 415 Jesperson op cit p 433

[7] Kāvyadarśa

[8] S B Dumville The Science of Speech p 20.

[9] Vide the Sanskrit Introduction p vii

[10] N S Edn 1, p 24

[11] Mahābhāṣya I, p 27, Vide Taittirīya Samhita, II, 5. 2

[12] Tantravārttika (Poona Edn), I, 3.9 p. 268

[13] Brugmann Comparative Grammar I p 528 and Whitney J.A.O.S Vol V p 196

[14] Taittirīya Prātiśākhya, II 4-7 pp 68-70

[15] Brugman op cit I, p 528 and B. Dunville, op cit, p 104

[16] II 8 p 70

[17] Rk Prātiśākhyabhāṣya Benares Edn, (1903) p 98.

[18] Uddyota N S Edn, Vol II, p 26

[19] Siddheshwar Varma’s Critical Studies in the Phonetic Observations of Indian Grammarians p 162

[20] I 2 29 31

[21] I 1 38 40

[22] I 108-110

[23] III 1

[24] III 3

[25] Keith A History of Sanskrit Literature p 423 and Siddheshwar Varma op cit Introduction pp 21 96

[26] I 2 33

[27] VI 1 290

[28] VI 2 164

[29] VIII 3 104

[30] III 1 85

[31] VI 3 14

[32] Kielhorn’s Edn III p 146

[33] N S Edn, Vol I p 65

[34] Tai Prā II 37.

[35] Vide the Skt Intro, p XVI

[36] Mysore Edn., p 452

[37] III 1

[38] I 31

[39] R̥k Prā bhāṣya p 98 Kā Prā Benares Edn p 12

[40] N S Edn Vol II

[41] Pradīpa Vol II p 26

[42] Uddyota  do

[43] Pradīpa do

[44] N S Edn Vol II p 29

[45] Svara Siddhāntacandrika (below) p 372

[46] Vide Tai Prā XIX 3 and Rk Prā III 34

[47] XIX 4

[48] p459

[49] Vide the Śabdakaustubha Benares Edn Vol I p 400

[50] Sanskrit Grammar p 290-3

[51] J A O S Vol V pp 196 ff and Sanskrit Grammar p 448 and A Bergaigne and V Henry Manual pour etudier Le Sanskrit Vedique pp 6 and 7

[52] VI, 1 158

[53] XIX I

[54] Vide Śabdakaustubha Vol I, p 397

[55] See Skt Intro p XXIII

[56] See Svara Siddhāntacandrika (below) pp 371 and 372, G

[57] Ibid p 14

[58] Mahābhāṣya, N S Edn, Vol II, p 31

[59] I 14 and 15, Benares Edn, p 441 and 442

[60] Vide Bhāṣika Pariśiṣṭa sūtra, III, 25 and 26

[61] R̥k Prā III 20 and 21, 27 and 28

[62] XXIII, 17

[63] Vide Śikṣāsangraha, Benares Edn, p 410

[64] Ibid, p 1 and 2

[65] See Skt Intro f-n p XXVII

[66] P 41

[67] Vide Tantravārtika, 1, 3, 9 Poona Edn. P 292

[68] Vide VI, 2 164, VI. 2 119, VIII 1 35, etc.

[69] p. 3

[70] Mahābhāṣya, N S Edn, Vol I p 27

[71] Śabdenduśekhara, Benares Edn, p 936

[72] Svarasiddhāntacandrikā p 34

[73] Benares Edn Vol II, p 17

[74] Kāśikā I, 2 36

[75] Padamañjari, do

[76] Benares Edn, p 396

[77] Vide Vaidikābharaṇa p 411

[78] Satyaparva, 52 42 Bombay Edn, p 103

[79] Vide C V Vaidya’s History of Sanskrit Literature  – Section III p 133 and Svarasiddhāntacandrika (below) pp. 382-3

[80] Mahābhāṣya N S Edn Vol I p 68

[81] Kāvyaprakāśa II Trivandrum Sanskrit Series 88 p 75.

[82] Vide Indian Antiquary 1904 p 126 ff

[83] Vide Belvalkar’s Systems of Sanskrit Grammar

[84] Vide Svarasiddhāntacandrika (below) pp. 382

Leave a comment


  1. Absolutely wonderful. Eternally grateful to you for these posts which are so true to Vedic tradition. i know i am being greedy but just wish you could post more often 🙂


  2. Many thanks!


  3. Absolutely good work kindly find the book link which will be helpful for others.



Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: