Caryāgītikoṣa - Fascimile edition
1. Description of the manuscript
1.1. The discovery of the caryagitikosa manuscript by Mahamahopadhyaya Haraprasad Sastri is one of the most significant events in the history of the Eastern New Indo Aryan languages. In 1907 A.D., during his third visit to Nepal in search of Buddhist scriptures, Sastri discovered the Cgk. text in the Raj Darbar Library of Nepal1 and subsequently published it in Bengali, with three other Mss. , in 1916 A.D., from Bangiya Sahitya Parisat, Calcutta, under the title, hajar bacharer purana bahgala bhasay bauddha gan o doha (the Buddhist songs and dohas written in thousand years old Bengali language). Sastri published the four Mss. in one volume under the wrong impression that all of them were written in the earliest form of the Bengali language. This is evident from the following comment made by him in the introduction of this book.
In 1907, again I went to Nepal and discovered a few Mss. One of them entitled, caryacaryabiniscaya, contains a few kirttana songs, and the Sanskrit commentaries thereof. The songs are similar to the kirttana songs of the Vaisnavas, and are known as caryapada. I got another Ms., a dohakosa (anthology of doha songs). The composer’s name is Sarohavajra. The commentary is in Sanskit and the commentator’s name is Advayavajra. I saw another Ms., which also is a dohakosa. The composer is Krsnacarya. This text also bears a Sanskrit commentary.. .1 believe, the writers of this language belong to Bengal or its neighbouring areas. There are evidences to prove that some of the poets are Bengali. In spite of slight grammatical differences all these Mss. seem to be written in Bengali.
1.2. Suniti Kumar Chatterji, while analysing the characteristics of Old Bengali in his ODBL. (1926), has clearly pointed out that only the carya songs available in the Cgk. Ms., are written in Bengali, and the doha songs available in the three other Mss. are composed in Eastern Apabhrarrria. In spite of some superficial affinity between the languages of these three Mss. and Cgk., they are basically different. Chatterji’s considered opinion has been accepted by other scholars as authentic.
1.3. According to Sastri the title-of the c. MS. is caryacaryabiniScaya. But such a title is not mentioned anywhere in the text. In the invocation sloka there is a reference to the c. verses as ascrya caryacaye, meaning literally ‘in the wonderful caryds,\ but there is no mention of the word biniscaya, as used by Sastri. When the Ms. was discovered by Sastri the title-page and the colophon page were already lost. The present title carydcaryatika, written on the recto side of the folio No. 1, is a later insertion, probably made during the time of its entry into the Raj Darbar Library of Nepal.1 But in any case, the word biniscaya (ascertaining), as used by Sastri, has not been mentioned eithei- in the invocation Sloka or in the title given by the cataloguers of the Raj Darbar Library of Nepal.
1.4. It is now known to us that more than one Tibetan translation of this work was in vogue in the Buddhist society of that time. Suniti Kumar Chatterji made a mention of this in his ODBL. (pp.199) and also published one such translated song (No. 29) in the Indian Historical Quarterly, Calcutta, 1927, II, (pp. 676-682). Prabodh Chandra Bagchi subsequently discovered a complete Tibetan translation of this text. Both the folios, bearing the title and colophon, being available there, it is now confirmed that the Sanskrit comm¬entary incorporated in Cgk. was written by Munidatta and that the Tibetan translation was made by Kirtticandra or Chandrakirtti. The title of the com. is carydgitikosavrtti. The relevant portion available in the colophon of the Tibettan text may be cited here:1
tatrahrtanam ca vicaritanam caryasatenahrtagitikanam.
.A. sattvaistu sambodhivicaranartham kosam budah samracayamvabhuvuh. .
ardhasya tasmanmunidattakena risyavabodhapratipadanaya.
jnanaya sarvasya tathaiva so’yam kosasya carthah prakatikrto’atra.. satpurusavabodhaya tikam krtva hyuparjitam.
munidattena yatpunyam tenastu sajjano jinah. .
The commentary of Munidatta is known as carydgitikosavrtti, which obviously presupposes that the title of the anthology was cryd- gitikosa. In the absence of any other specific title mentioned in the original Bengali MS., the name caryagitikosa may reasonably be accepted as the title of the anthology, and carydgitikosavrtti that of the commentary by Munidatta.
1.5. Carydgitikosavrtti is a palmleaf manuscript now preserved in the National Archives of Nepal (known as rdstriya abhilekhalaya'). Originally the MS. constituted more than 69 folios (probably 71 folios, including the pages containing title and colophon). Before the discovery of the MS. by Sastri, even before its accession to the Raj Darbar Library, at least six numbered folios (Nos. 35, 36, 37, 38, 66, 70) and the folio containing the title, were already lost. Only
sixty-four folios containing forty-six full songs and the first six lines of another ten-line song, along with the corresponding Sanskrit commentaries, are now available to us. The scribe used the old Bengali script both for the Bengali verses and the corresponding Sanskrit commentaries. It has much affinity with the
scripts of the neighbouring Eastern NIA. languages also. It may be mentioned here that all the sister NIA. languages of this region viz, Assamese, Oriya, and Maithili, are now claiming that this earliest vernacular text was written in their respective languages. Even Hindi has registered her claim on it.
1.6. The condition of the palmleaf MS., as recently examined by me, is quite good. The size of the folio is 121' x li". Both the sides, recto and verso were used for writing. There are five lines on each page; the first and the fifth lines are run-on, the medial three lines have a gap of about one inch in the middle. All the leaves are tied with a central string and a square like blank space is left in the
middle for it. In spite of so much care taken by the scribe, a few portions of the writings have already been damaged due to the friction of the string. The handwriting is generally distinct, neat and well-shaped. Written in bright black ink, the characters are slightly slanted towards the right-side. Certain portions have faded beyond legibility. In all such places I have followed the readings rendered by Sastri.
1.7. From a careful examination of the handwriting it appears to me that the whole manuscript, except in the cases of a few correc¬tions, was copied by a single scribe. In some of the pages the scribe was more careful; but so often, two or three different characters of the same letters have been used. Still from the mould of the characters it is evident that the entire Ms. has been copied in one handwriting.
1.8. The internal evidences of the Ms. show that the scribe copied this text from two different sources: one having only the Bengali verses, and another having only the Sanskrit commentaries of them. In more than fifty cases the language of the songs as cited in the commentary is different from that of the songs quoted in full at the beginning.»
Another important evidence has been given by the scribe himself. On folio No. 18-A, at the end of the commentary of song No. 10, he has made a small comment as, nadidombipadanam sunetyadi caryaya vydkhya nastr. the commentary of the carya of Nadidombipada, beginning with sane etc., is not available. It shows that as the commentary was not available the scribe did not reproduce this song. The third evidence has been given by the Tibetan translator Kirtticandra. In the colophon it is stated that from an anthology of hundred c. songs Munidatta selected half the number and wrote commentaries on them.
Another indirect evidence is available in the language of the commentary. The commentaries of most of the songs start with words like, tamevartham prakathayanti\ to explain it as referred to. The question of such reference arises only when the original text is not supplied along with the commentary. Such evidences clearly prove that the scribe brought together the Bengali c. songs and the Sanskrit commentaries of them from two different sources.
2. Script and handwriting
2.1. It has been mentioned earlier that the Cgk. text was copied in one hand, and that the handwriting is generally neat, distinct and legible. Certain portions of some of the folios have faded beyond legibility perhaps due to exposure to light? On pages 1-B and 3-B there are evidences that some over enthusiastic readers attempted to over-write a few indistinct letters in modern Nagri. Some of the characters, as written by the scribe, are confusing. It is
difficult for the readers to distinguish between tu and da/da,ta and dhafdha, ba and ca, cha and conjunct ccha, kr/ku/kya and kca, na and la, nu and nna, na and ia, nasal conjuncts and non-conjunct nasalized letters, etc. This is perhaps one of the main reasons why the editors have considerably differed in their reading of the Cgk. text. Two such significant instances, as experienced by me, may be mentioned here. All the previous editors, including Sastri, Bagchi and Sahidullah, have read the following words with initial cluster like, cchadi (15), cchupai (6) and cchinali (18). But not a single instance of such initial cluster of ca and
ch is available in any of the E.NIA. texts of the Middle period. When the scribe himself had made no distinction between the characters of cha and ccha, and when those three words are still in use in Bengali and Assamese as chadi, chupai and chinali, it seems to me more logical to read them as such. In another case, the name of the composer of song No. 33 has been read by Sastri and other editors as dhendhana pa(da). But when the scribe had made no distinction between the characters of la and dha, and in the Tibetan text the name has been read as tentanapada, I find no reason to discard the Tibetan reading.
The absence of spacing between words is another obstacle faced by the editors in their correct reading of the text. Where Sastri read a word as pucchatu (41), I preferred to read it as puccha tu; his suname heri (13) has been read by me as suna meheli; in song No.50, he read a line as, tahi toli sabaro hakaela kandasa saguna siali, while I have read it as, tahi toli sabaro ddha kaela kandasa saguna sidli. Many similar instances may be cited where the editors have differed from one other in their reading of the text.
2.2. The script used in the Cgk. Ms. has been identified as Old Bengali by Sastri, Chatterji, Bagchi, Sahidullah, Sen and many other linguists. Paleographist R. D. Banerji also had endorsed their view; but it is interesting to note that he was inclined to fix the date of its copying even later to that of Skk. During my recent visit to the National Archives of Nepal I had an opportunity to consult some tradi¬tional pundits working there as professional readers and scribes of
old manuscripts. They identified its script as Old Newari. In their opinion the language of the songs is Maithili, and that of the commentary is Sanskrit. One of them read some of the pages of the MS. to me in the same reading, as given by Sastri. However, they admitted that in such an old period Maithili had little
difference with Bengali or Assamese. Haraprasad Sastri prepared a copy of this Ms., now preserved in the Library of the Asiatic Society of Calcutta, the script of which is modern Newari, a close associate of Nagri. So it is quite likely that the Cgk. was written in a common script which was in vogue in Bengali-Assamese and Maithili and some other E.NIA. vernaculars of that period.
2.3. According to Tarapada Mukherji the script-character of Cgk. has a close affinity with that of a dated Ms., panca- kara (1199 A.D.).8 The songs collected in Cgk. are presumed to have been composed in a period roughly covering the 9th to 12th centuries A.D., and all the twenty-three composers hailed from the eastern region of India. The Sanskrit annotation might have been written by Munidatta in the 13th century A.D., for the understanding of the Buddhist sahajiya
group of Nepal and Tibet. A hybrid style of Sanskrit was developed through the Sanskritization of various Buddhist scriptures from Prakrit and vernaculars. The language of this commentary of the c. verse, has much similarity with that style.3 The present Cgk. Ms. might have been copied in the late thirteenth or early fourteenth century A.D. Probably no appreciable change of the script-face was made in that early period, between Bengali-Assamese and Maithili.
The second oldest Bengali Ms. available so far is srikrsnakirttana, a long narrative verse based on the popular love story of Krsna and Radha. The story was written by Badu Candidasa in the late fourteenth or early fifteenth century, and was copied, approximately, in the early sixteenth century. A comparative study of the script-character of these two Mss. may give us a clearer picture of the development of Bengali script in OB. and MB. periods. This has been attempted in this chapter; and a chart has been annexed showing the three stages of Bengali-Assemese, along with the Modern Maithili, Oriya, and Newari, and Nagri scripts (pp. XXIII-XXIV).
2.4. Vowels: initial.
a, a: In Cgk. these two initial vowels are written as^.^or ■?, ^. The first has similarities with the scripts ofSkk., written as , <p. The second pair resembles the Maithili characters, , '5*. From the characters of Cgk. it appears that the scribe was accustomed to write in both the styles, and made no distinction in their use. In Oriya these two letters are written as ’SI, Sl| . In modern Bengali-Assamese typography they are printed as, «r, ^1