కవిజనాశ్రయము/Preface
PREFACE.
Kavijanasrayamu or the Refuge of the Poets is now being published for the first time in a complete form, although it is, perhaps, the oldest book extant on Telugu Prosody. The text is the result of the collation of ten manuscripts of which nine are written on palm-leaves and one on paper. The palm-leaf books are all in the library of the Andhra Sahitya Parishat (Telugu Academy) and they have been mostly found in the Tamil Districts of this Presidency. The paper copy is in the Brown Collection of the Government Oriental Manuscripts Library, Madras. Appakavi's remark that “perverse poets have interpolated many curious things into the Kavijanasrayamu" is fully borne out by the state of the manuscripts. I have tried my best to restore the original text, although there is possibly room for further pruning.
The book purports to have been written by Malliya-Recha (Recha son of Malliya or Mallaya) and dedicated to himself. One of the manuscripts "జ" contains a short preface of seven verses, which says that “Bhima Kavi, the beloved of the (God) Bhimesvara of Vemulavada" composed the book and dedicated it to Recha, making it appear, at the same time, that Recha composed it himself. This Statement bears out the universal tradition which ascribes the book to Vemulavada Bhima Kavi. It has been recently suggested that the author of the book was Gokarna, brother of Bhima who was king of Vardhamanapura during the reign of Pratapa Rudra of Warangal; but this suggestion is entirely unsupported. I hold, therefore, subject to future correction that Vemulavada Bhima Kavi was the author of the book and I also hold, for the reasons detailed in the Telugu preface, that Bhima was a native of Vemulavada (alias Lemulavada) in the Karimnagar District of H. H. the Nizam's dominions and not, as is supposed by some, of a village of that name near Draksharama in the Godavery District.
The only material available for fixing the age of Bhima Kavi is furnished by two stray verses which tradition ascribes to that poet. It is said that the poet went to see king Kalinga Gangu, that he was told that the king was too busy to see him, that the poet got enraged and uttered a curse in the form of an extempore verse and that, as a result of this curse, the king was soon afterwards deprived of his throne and kingdom by his enemies and reduced to beggary. The poet happened to meet the king again while in this plight and, pitying his condition, he delivered himself of another extempore verse which led to the king's early restoration to his throne. Ordinarily, stray verses are not the best kind of evidence; but they are the only evidence available in this case and I am aware of nothing which militates against that evidence. I, therefore, accept it subject to future correction.
Now, we have to see who this Kalinga Gangu was and when he flourished. Among the kings of Kalinga, there was only one king who was surnamed Ganga. That was king Anantavarma surnamed Chodakalinga Ganga, who was crowned in the Saka year 999 and reigned for 60 years[1]. About this time, there was another king who bore the name Choda Ganga-viz. Raja Raja Choda Ganga of Vengi who was crowned in Saka year 1006[2]. He was the grandson of the great Chalukya king Raja Raja and governed the Vengi country on behalf of his father Kulotthunga Choda I. For our present purposes it is immaterial which of these two kings was the patron of Bhima Kavi but the appellation Kalinga Gangu appropriately belongs to Anantavarma.
This leads to the conclusion that the poet Bhima must have flourished about the close of the 11th century or the beginning of tho 12th century A. D., and, therefore, subsequent to tho time of Nannaya Bhatta who was the poet Laureate of Raja Raja. It was held by the late Mr. Gurazada Sriramamurthi that Bhima Kavi was a contemporary of Raja Raja's father Vimaladitya[3]. This view was based on an incorrect reading of another verse which is also attributed to this poet and, in the Telugu preface, I have exposed the untenability of this reading. On the other hand, Rao Bahadur K. Viresalingam Pantulu thinks that our poet must be referred to the 14th century A.D.[4]. This view is sought to be supported by four verses which are all attributed to Bhima Kavi and three of which are assumed to belong to the present work. These latter verses really belong to a later work, Kavyalankara Chudamani of Vinnakota Peddanna and there is nothing reliable to show that the fourth verse was composed by Bhima Kavi. I have no hesitation, therefore, in rejecting both these theories[5].
We do not know much about Recha. His description as an ornament of the Sravakas, a disciple of Vadindra Chudamani and a friend of Jina Mata, shows that he was a Jain by faith and in two verses he is said to be a Komati or Vaisya by caste. He was evidently a patron of learning and like many of his class aspired to authorship.
This is the only book so far discovered, which throws any light on the connection between Jainism and Telugu literature. The oldest works in Canarese were composed by Jains and I believe there are also some Jain books in Tamil. It is very probable that there was a Jain period in Telugu literature also, a supposition which derives additional support from the fact that some of the great Canarese Jain poets such as Pampa and Nagavarma emigrated from the Telugu Country, but, for some reasons which have yet to be ascertained, that literature has not come to light. Let us hope that it is not lost for ever.
A perusal of the text will show that the book is modelled after some of the Canarese classics such as Kavirajamarga, Chandombudhi and Sabdamanidarpana. Some of the points of similarity are noted in footnotes in their appropriate places.
I will now endeavour to give a short sketch of the general principles of Telugu Prosody:—
TELUGU PROSODY.
Syllables:—Telugu prosody, like the Canarese prosody, is based on the Sanskrit model except in regard to the purely vernacular metres. All these systems are founded on the quantitative values of syllables. The unit of quantity is a Matra (Mora) which is the measure of a Laghu (Short). In terms of letters a laghu may be defined as a short vowel either standing by itself or superimposed on a consonant. Syllables longer than a laghu are guru. In this book, they are defined as syllables which end in long vowels or are succeeded by compound syllables or anusvara and are, at the same time, “stressed". By virtue of the last mentioned qualification, a short syllable occuring at the end of a word remains short though followed by a word beginning with a compound syllable. In other words, the rule applies only to syllables occuring in the same word. Compounds must, of course, be taken as single words. The same qualification also accounts for a syallable preceding a compound syllable of which the sound r forms the second and penultimate element being sometimes counted as short. This definition of guru is not, however, exhaustive. It does not, for instance, cover cases of short syllables followed by consonants as ral in varal which is guru. The guru is said to contain two matras and the laghu one. This, again, is not strictly accurate. For, while a laghu can be neither more nor less than a matra, a guru may be anything above a matra. The syllable tan in vanitan contains 1½ matras (taking a consonant as equivalent to half a matra), while the syllable gan in anagan counts for 2½ matras. Dirgha is, of course, two matras but pluta is three. So, the correct way of stating the difference, in terms of matras, between a guru and a laghu would be that a laghu contains a single matra and a guru more than one.
There is a rule in Sanskrit by which a short syllable occuring at the end of a pada or verse line is counted as a guru. This rule is also adopted in Canarese but rejected in Telugu.
Ganas:—The Sanskrit term gana corresponds to the English term foot, but, in Sanskrit prosody, the idea of a foot (pada or charana) is appropriated to a verseline on the analogy of a four footed animal, Sanskrit verses being, as a rule, quatrains. Ganas are composed either of syllables (aksharas) or of matras. The former consist of three syllables each. If all the three syllables are guru, the gana is denoted by the syllable ma and if all the syllables are lagnu, it is na. If the first, second or third sylable only is guru, the gana is called bha, ja or sa. Similarly, if the first, second or third syllable only is laghu, the gana is either ya, ra or ta. This gives us eight akshara ganas which are exclusively used in Vritha metres. None of these ganas can be substituted for another though of the same quantity.
There are twenty eight matra ganas which are used only in the jati metres. These ganas are composed of matras and may be substituted one for another of the same group. They are derived by the process of prastara or permutation from two, three or four gurus as illustrated below.[6]
The first row gives four ganas, the second eight and the third sixteen. In Canarese, all these are adopted under the names of Brahma, Vishnu and Rudra ganas respectively. In Telugu, the first two ganas of each group are eliminated and the rest adopted and are called the Surya, Indra and Chandra ganas respectively.
Vrittas:—Metres are either Vrittas or Jatis. The former are composed entirely of akshara ganas. They may be divided into two main groups, ordinary and uddharamala. Ordinary vrittas are grouped into twenty six groups called Chhandas beginning with Ukta and ending with Utkriti. There are two vrittas in Ukta, each line of which consists of a single syllabie guru or laghu. A verse of each succeeding group has one syllable to the line more than a verse of the immediately praceding group, so that a verse of the Utkriti group has twenty six syllables to the line. The number of metres in each group is determined by the permutation of the syllables in a verse line in that group. This gives a total of 134217726 metres for the twenty six groups. And this is only the number of the Sama or regular Vrittas—i.e.—metres whose lines are uniform in regard to the number and order of ganas. There are also arthasama or semi regular and vishama or irregular vrittas. A semi regular vritta is that two lines of which belong to one metre and the other two to another metre. Where the two metres appertain to the same group or Chhanda, the Vritta is called Svasthanarthasama vritta and where the two metres belong to two different groups, the vritta is called Parasthanarthasama vritta. Where each line of a vritta is of a different metre from the rest, that verse is called an irregular vritta. It seems permissible to make the distinction of Svasthana and Parasthana in regard to irregular vrittas also. It will thus be seen that the total number of ordinary metres in Sanskrit (and hence also in Canarese and Telugu) is well nigh incalculable. On the top of these, come the Uddharamala vrittas i.e. metres which contain more than twenty six syllables in a line. A wellknown instance of these is the Dandaka metre.
Jatis:-Jatis which are made up of matra ganas seem to have been, in their origin, popular or folk metres, while vrittas were learned metres. They are either Sanskritic such as Kanda or purely Dravidian such as Sisa.
Yati or Caesura:—Yati in Sanskrit is a synonym for Visrama or Virama and means a pause. It indicates the place in a verse-line where the reader needs to stop for breath. As it is hardly convenient to stop in the middle of a word, Yati has come to be defined as the breaking off of a word. (యతిర్విచ్ఛేదసంజ్ఞకః). Pingala allows a deviation from this rule in rare cases in the interests of free flow of versification. There must always be yati at the end of a line. That is to say, every line must end in a complete word. In addition to this, there is internal yati in the case of all metres whose lines contain ten or more syllables each. Some long metres contain more than one internal yati. No internal yati is prescribed for metres of the first nine groups for the evident reason that the lines are too short to necessitate a pause in the middle in reading. But it can, by no means, be said that the internal yati always marks the place where an average reader would need to stop for breath. These places seem more or less arbitrary.
In Telugu, there is no final yati and the internal yati indicates the recurrence of the initial syllable of the line or its equivalent rather than the breaking off of a word. In its essence, Telugu yati is more akin to alliteration than to Caesura. Sometimes, the two ideas are combined; but, as the combination is calculated, except in the case of masters, to impede the course of versification besides leading to the use of expletives, the expedient of using cognate syllables (Savarnaksharas) seems to have been hit upon. For purposes of yati, a syllable means the entire syllable i. e., consonant as well as vowel but where the alliterating syllables contain more than one consonant each, it is enough if any one of them in the one agrees with any one in the other. The subject of yati has always been a fruitful source of discussion with the Telugu grammarians. The present work treats of ten kinds of yati. Vinnakota Peddanna describes twelve and Ananta twenty four sorts of yati, while Appakavi treats of no less than forty one kinds. A good many of these can be grouped under Akhanda yati which the purism of Appakavi rejects as barbarous. Akhanda yati may be described as yati with a syllable which is the result of Sandhi, as bha in Vrishabhadhipa (Vrishabha+adhipa). Appakavi lays down that, for purposes of yati, you can only take the a in bha and not the whole syllable bha. This would be yati in the Sanskrit sense. Some writers on prosody recognize Akhanda yati but it must be said that it is very sparingly used even by those who defend it.
Prasa or Rhyme:-Prasa corresponds to the English rhyme but occurs in the second syllable of each line of the verse. When it occurs also at the end of the line, it is called antya prasa but this is used only as an embellishment and very oocasionally. It is the consonant that counts for prasa and the vowel does not matter. The same consonant or set of consonants must recur in all the lines. Canarese prosody allows the first four consonants óf each varga or class to rhyme with one another, but this is allowed only to a limited extent in Telugu. Even in Canarese, the latitude allowed by the rules is not availed of to any large extent in practice. If the rhyming syllable is preceded by a long syllable or Anusvara in one line, it must be so in all the other lines.
The double restriction of yati and prasa, no doubt, renders versification in Telugu more difficult than in most other languages, but it is not without its uses. Besides rendering verses more melodious than they would otherwise be, yati and prasa afford an invaluable aid in investigating the morphology of words. Their disadvantage lies, as has already been stated, in limiting the scope for the choice of expression and leading to the use of inappropriate or unnecessary words. Both these evils are greatly minimized in standard classics not only by their authors possessing unlimited command of language-both Telugu and Sanskrit but also by the adoption of two expedients already referred to viz, the use of substitutes (Savarnas) and the elimination of the idea of pause in yati. That these restrictions are not felt as a great hardship is evident from the fact that there is perhaps more versification going on at the present time in Telugu than in any other language—at least in Southern India. It must be noted, moreover, that the full force of this restriction is felt only in vrittas and that much latitude is allowed in the composition of jatis.
J. RAMAYYA.
- ↑ (a) Epigraphia Indica. Vol. 18.
- ↑ (b) Epigraphia Indica Vol. ?
- ↑ (a) Mr. Sriramamurthi's Lives of the Telugu poets.
- ↑ (b) Lives of the poets.
- ↑ (c) This argument is elaborated in the Telugu Preface.
- ↑ (a) The process of prastara may be briefly described as follows:--Write down as many gurus as you want to permutate. This will be the first line, put a laghu below the first guru of the previous line and reproduce the syllables to the right. Repeat this in the third and subsequent lines but reverse the syllable to the left of the first laghu that you write in these lines. Continue this process till you get all laghus and prefix a laghu to each line which commences with a laghu.