Devanagari Source of the Ujjain Sarada Manuscript

Devanagari Source of the Ujjain Sarada Manuscript of the Matsya Purana / Shri Anand Swarup Gupta 163-174

[ matsyapuranasyaikah sarada lipikosa ujjayinisthayam 'simdhiya-oriyantala imstityuta ' samsthayam samraksito'sti | sa kosastatra kasmiradesatpraptah, matsya- puranasya kasmirapathaparamparam canusarati | suddhascayam kosah, tathapi lekhakapramada- janyah kascidasuddhayastatropalabhyante, tasvaneka asuddhayah kesamcit saradaksaranam nagaraksaranam ca parasparam samanakrtitvaddha torbhrantena lipikarena janitah | devanagaraksaranameva sarada lipyam pratilipikarane eta asuddhayah sambhavitumarhanti nanyathetyatra pradarsya kasmirapathanusari kascitpracino devanagarikosa evasya saradakosasyadhara asidityasminnibandhe pracina lipivijnanamasritya pratipaditam ] The Sarada Ms. of the Matsya Purana depositedin the Scindia Oriental Institute, Ujjain, and bearing No. 4481 contains important textual peculiarities, some of which have already been discussed. It also contains certain peculiar scribal errors and textual corruptions which can be explained palaeographically only if we suppose that some Devanagari original was the direct or indirect source of this Sarada Ms. And as this Ms. comes from Kashmir and represents a Kashmirian version of the Matsya Purana, its supposed Devanagari original also must have belonged to the same version, and also must have been written in Kashmir or the North-western parts of India. From this it may also be inferred that the Devanagari script was in use in Kashmir prior to or side by side with the use of the Sarada script. The Devanagari script has been used in India since about the 8th century A. D. For a long time it has generally been writing Sanskrit works and recognised as an all-India script for Vide my article 'A study of the Textual Peculiarities of a Sarada Ms. of the Matsya Purana' published in 'Purana' Vol. I, pp. 58-71. 7

164 puranam - PURANA inscriptions. Devanagari and Sarada are [Vol. I, No. 2 two sister scripts, both being evolved from the Kutila lipi which was in use in Northern India from Nepal to Kashmir between the 6th and the 9th centuries A. D. But Devanagari seems to be the older of the two. The Sarada script is found used in an inscription of the Chamba state of the 10th century A. D.1, when Devanagari was already in use both in Northern and Southern India. The Devanagari script has been used in some of the inscriptions and grants of South India about the 10th century A. D. About this time Devanagari had its full sway in Gujarat, Malwa, Rajasthan, Madhyadesa (including the modern U.P.) and Berar.Even in Bengal and Nepal Devanagari was the script which was used for writing Sanskrit upto the 10th or 11th century A.D.3, when the Bengali and Nevari scripts were evolved. regards the use of the Devanagari script in Kashmir about this time, no documentary evidence is perhaps available; but in order to explain the above-mentioned textual corruptions of the Ujjain Sarada Ms. we have to start with the hypothesis that there was a time when in Kashmir the Devanagari script was in use prior to or side by side with the Sarada script, and that the Ujjain Sarada Ms. is either a direct or an indirect transcript from some non-extant Devanagari original. In the present article an effort is made to show that some of the corrupt readings of the Ujjain Sarada Ms. are really the corrupt forms of the corresponding correct readings of the supposed Devanagari original. As There are certain Devanagari letters or symbols which, through a little carelessness of the scribe, may be written as resembling each other; e. g. 7 and a, and c, c and c, etc. va, pa ya, frequently interchage in Devanagari manuscripts. Confusion between such letters may easily arise in the case of a scribe who, though being less familiar with the Devanagri script, is transcribing from a Devanagari original in his own script or in some other more familiar script. 1. Cf. G. H. Ojha, Bharatiya Prachina Lipimala (2nd ed.), p. 73. 2. Ibid., pp. 68 ff. 3. Ibid., pp. 47, 77; also cf. Plate XXIV for a Nepal Ms. of the 10th centuryA. D.

Feb., 1960] SOURCE OF UJJAIN SARADA MANUSCRIPT 165 Below are given some of the cases of such confusion which have been found in the Ujjain Sarada Ms. of the Matsya Purana (hereafter referred to as $)- 1. The Letter a confused as ta (a) An., 152. 36- atha samjnamavapyasu garudo'pi sakesavah | paranmukho ranattasmat palayata mahanavah || S- palayana "" "" rara "" [Here, the reading '' (palayata) is the past imperfect tense () in the atmanepada, of the root to ayu go with c as an upasarga, and agrees with its subject Garudah. Obviously it is the correct reading and is found in all other Mss., both Northern and Southern. The S-reading c(palayana) is clearly the result of the confusion between the Devanagari letters 7 and 7] (b) An., 154. 526 cd - kamarupa mahotsaha maharupagunanvitah (ganesah ) | S- mahansaha fura ww [Here, it may be noticed that the scribe first confused the letter of the original reading HET: (mahotsahah) tu as, by which the correct original reading became corrupt as : (mahonsahah); he then might have emended the portion (mahon) as (mahan), and thus the corrupt reading : (mahansahah) was the result. Thus mahotsahah > mahonsahah > mahansahah might have been the process.] (c) An., 220. 6 cd - avinitakumaram hi kulamasu visiryate || S--avinina 0 dw dw ww 1. Anandasrama edition of the Matsya Purana, Poona, 1907. T

166 puranam - PURANA [Vol. I, No. 2 2. The letter confused as a na (a) An., 127. 19 ab - evam dhavaniyukto'sau bhramate jyotisam ganah | The Devanagari Ms. (No. 4646) of the Scindia Oriental Institute, Ujjain,' which is closely allied to S, reads as follows:- evam vanibaddho'pi bhramati jyotisam ganah | S - evam dhruvatibaddho'pi (b) An., 191. 118 ab- "" wr ankolasya samipe tu natidure tu tasya vai | 99 " tatidure " In all these instances the confusion occurs between the two letters 't' (a) and 'n' (7). This confusion is possible only in the case of the Devanagari letters- and -and never in the case of the Sarada letters; for in the Sarada script these two letters do not resemble at all, and so they can never be confused with each other. Hence, it may be inferred that S is a copy of some Devanagari original, or of some intermediate Sarada exemplar which was itself a direct or an indirect descendant of the supposed Devanagari original. This inference may further be strengthened by the following instances 3. The conjunct (tpa) confused as (tya) An., 153.24 ab- tasyaraksatpadam savyam maruto'mitavikramah | D reads- tasyaraksatpadam saumyam • But S reads- tasyaraksatyadam saumyam • 1. This Ujjain Devanagari Ms. will be referred to as D.

Feb., 1960] SOURCE OF UJJAIN SARADA MANUSCRIPT 167 [Here, it may be seen that the readings of both the Ujjain Mss. -D and S-represent the same version, but while the reading of D is correct, that of $ is quite defective and corrupt and carries no sense at all. The reading (tasyarakshat padam) is given by all other Mss. The defect in the S-reading (tasyarakshatyadam) is clearly the result of confusion arising between the Devanagari conjuncts T (tpa) and (tya) owing to their deceptive similarity. The Sarada conjuncts tpa and tya are so distinct from each other that they can never be confused. Hence the creeping of this textual error in $ can be explained only if we assume some Devanagari original as its source.] 4. The letter (t) confused as (v) An., 191, 29 a b- tato gacchettu rajendra turasangamamuttamam | S- " 99 " vurikacchamanuttamam || [Here, some of the Mss. read nadisamgama for turasamgama, and the three Devanagari Mss.-No. E. 3550 of the India Office Library, London, No. 2288 of the Deccan College Manuscripts Library, Poona, No. 10443 of the Sarasvati Mahal Library, Tanjore-read (turikachchha), but the Ujjain Sarada Ms. reads g (vurikachchha), and the Ujjain Devanagari Ms. (mentioned above) also gives this reading in an emended form as a (varikachchha). The S-reading gis clearly the corrupt form of the Devanagari reading g which must have been the reading of its hypothetical Devanagari original also, and which 1. In the Colophon of its Adhyaya 198 (corresponding to the An, Adh. 191) which deals with the mahatmya of the various tirthas in connection with the the Ujjain Sarada Ms (5) again gives the same reading g and the Ujjain Devanagari Ms. (D) which closely follows S in giving the same detailed Colophon of this Adh. gives the reading 3 (urikachchha) instead of its previous reading See my article referred to in footnote I

168 puranam - PURANA [Vol. I, No. 2 has fortunately been preserved in the above-mentioned three Devanagari Mss. There is no possibility of confusion between the Sarada symbols for the letters a and a; but the Devanagari letter a (t) if written hurriedly and carelessly may look like (n) or (v) and may easily be confused with these letters.] There is also some similarity between certain letters of the Devanagari and Sarada scripts. Thus the Sarada symbol for the letters (4) resembles the Devanagari letter (bh), and similarly the Sarada symbol for the letter s () resembles the Devanagari letter (m). Now a scribe who is more familiar with the Sarada script, and is more accustomed to use it, may easily confuse the Devanagari letters (bh) and H (m) with the Sarada symbols for s and respectively, for these Sarada symbols with which he is more familiar may intrude into his memory owing to their deceptive resemblance with the Devanagari symbols and which he actually wants to copy, and then he may copy these Sarada symbols of his memory instead of the Devanagari letters. bha and ma of the original. The opposite may also be true; that is, a scribe who is transcribing in Devanagari from a Sarada original may also be subject to similar psychological or visual confusion. But then we have to decide whether a particular reading in a manuscript is a correct or probable one. If it is not, then we shall have to assume that it is a corrupt form of some original correct reading of the scribe's exemplar which might have been in a script different from the one in which he has transcribed, and then we shall have to show that the corruption was also palaeographically possible. Below are given a few instances of such psychological confusion, which will further corroborate the

igno Feb., 1960] SOURCE OF UJJAIN SARADA MANUSCRIPT 169 inference that S is a descendant of some Devanagari original :- 5. The Devanagari (bh) confused with the Sarada symbol for sa ( 8 ). An., 153.22cd, 23- airavate caturddante matange'calasamsthite || mahamadajalasrave kamarupe satakratuh | tasthau himagireh srnge bhanumaniva diptiman || S --sanumaniva for bhanumaniva [ Here, only S reads sanumaniva for bhanumaniva, while all other Northern and Southern Mss. give the same reading as the An. But in this context ' seems to be the original and correct reading, and the reading of S may be taken as a corrupt one resulting from the confusion between the Devanagari letter c (bh) and the resembling Sarada symbol for the letter s] 6. Devanagari (m) confused with the Sarada symbol for (s) An., 206. 41- krsnepsitam krsnamrgasya carma dattva dvijendraya samahitatma | yathoktametanmaranam na socetprapnotyabhistam manasah phalam tat || [Here, S and some other Mss. read' for 'krsnepsitam ', but 'maranam ' is the reading of all the Mss. except S which reads Saranam in place of maranam. Here also the reading of S does not seem to be a correct one, and may be the result of the confusion between the two letters Devanagari H (m) and Sarada (s), which resemble each other, for the Sarada letters 'm' and 's' have little chance of being confused with each other.] These are some of the instances which are sufficient to prove that S is a descendant of some Kash-

170 puranam - PURANA [Vol. I, No. 2 mirian Devanagari manuscript of the Matsya Purana. Now it remains to be ascertained whether a is a direct transcript from the hypothetical Devanagari original, or is a copy of some intermediate Sarada exemplar which might itself be a direct or an indirect descendant of the Devanagari original. There are certain corrupt readings in S which may throw some light on this problem also. The following instances may be studied :- 1. An., 154. 188 cd, 189 ab- sada || yathahamuktavanasya hyuttanakaratam uttano varadah paniresa devyah sadaiva tu | D - yattavaduktavanasya uttanakarata maya || uttano varadah paniresa devyah sadaiva hi | S- yattavaduktavanasya uttanakarana maya uttano varadah paniresa devyah sadaiva hi [Here, the reading 'uttanakarata ' must be the original correct reading, for it is supported by the next line 'uttano varadah pani0 '; it is also corroborated by the following previous reading: uttanahastata prokta yacatameva nityada | an., 154. 170 c-d. Here also, both D and S give this same reading 'uttanahastata ' which is the same in sense as 'uttanakarata '. So it is clear that the Sarada scribe might have confused the Devanagari letter with the Sarada symbol for the letter (), and the corrupt reading uttanakaraja ' was the result.] 2. an., 271. 16- iksvakunamayam vamsah sumitranto bhavisyati | | sumitram prapya rajanam samstham prapsyati vai kalau ||

Feb., 1960] SOURCE OF UJJAIN SARADA MANUSCRIPT 171 The reading of all the Mss. except D and S is the same as the above An. reading. The reading of the first line is uniform in all the Mss. including D and S; but the reading of the second line in D and $ differs from each other as follows:D S yatastam prapya rajanam samstham prapsyati vai kalau || yajnastam prapya rajanam samstham prapsyati vai kalau || [Here the S-reading c (yajnas) in place of the Dreading (yatas) seems to be corrupt in this context. It may be a transcriptional error of the Sarada scribe who seems to have confused the Devanagari letter (t) of the original with the Sarada symbol for the letter), and then to have given play to his intelligence in emending the reading 'yajas' (thus formed) as 'yajnas'. Thus the process might have been- yatas (yatas) > yajas (yajas) > yajnas (yajnas).] Now in both these instances the corrupt reading of S must be the outcome of the confusion between the Devanagari letter a and the Sarada symbol for . The symbols of these two letters of the two scripts, Devanagari and Sarada, though having divergent values, must have had some similarity, so that the confusion between them might have been possible. In this Sarada Ms. the letter is written somewhat like the Devanagari and has no resemblance with the Devanagari letter J. So the confusion between this Sarada symbol of letter and the Devanagari symbol is not possible. Therefore it may be inferred that S is not the direct descendant of the Devanagari original. Moreover, we find this same form of the Sarada symbol for in a manuscript of the Atharva-Veda and also in another manuscript of the Sakuntala, both Mss. belonging to the 16th century A. D. So $ might have been written near 8

172 puranam - PURANA [Vol. I, No. 2 about that date or possibly much later as it appears from its general appearance. But, in some inscriptions of the Chamba State belonging to the 10th century A.D. we find the Sarada symbol for inscribed as resembling the Devanagari letter of that time, which was not much different from the modern Devanagari , as is clear from a Nepal Ms. of about the 10th century A.D.,³ and also from an Ujjain inscription of the 11th century A.D.4 Anyhow, it is probable that the a letter of the Devanagari original might have been mistaken as the Sarda symbol for , by the first Sarada transcriber, and the corrupt readings such as ' and might have resulted. Then these corrupt readings might have been continued in the subsequent Sarada Mss. until they appeared in S. Thus all such corrupt readings of $ can be palaeographically explained only if we assume that S is the descendant of some remote Devanagari original, between which and S several successive intermediate Sarada manuscripts might have passed. The Ujjain Devanagari Ms. (referred to as D) is closely allied to the Ujjain Sarada Ms. (referred to as S). It seems to be the Devanagari transcript of some Sarada Ms., for in one place it confuses the letter (s) as (ch), which in the Sarada script closely resemble each other, and so gives the corrupt reading ccc (prapachyati) for the correct reading cc34fa (prapasyati). Morever, both these manuscripts (D and S) come a big collection purchased by the Scindia Oriental Institute in 1936 from Srinagar (Kashmir)5. So both these Mss. represent the same version of the Matsya Purana, and as already mentioned they are closely allied too. But this Devanagari Ms. generally contains correct readings in place of from 1. Cf. G.H. Ojha, Bharatiya Prachina Lipimala (2nd ed.) Plate XXXI. 2. Ibid., Plate XXVIII. 3. Ibid., Plate XXIV. 4. Ibid., Plate XXV. 5. This information was kindly supplied to me by Sri S. L. Katre, Curator, Scindia Oriental Institute, Ujjain, in his letter dated 18-7.58,

Feb., 1960] SOURCE OF UJJAIN SARADA MANUSCRIPT 173 the above-quoted corrupt readings of the S Ms. So one of the two things is possible either the scribe of this Devanagari Ms. emended all the above-mentioned corrupt readings of its Sarada exemplar, or there might have been two streams of the Sarada Mss. of the Matsya Purana issued from the same Devanagari original, the one confusing the letters of the two scripts and the other not making such confusion; and D might be the Devanagari transcript of a Sarada exemplar belonging to the non-confusing stream. The points discussed in this article may briefly be recapitulated as follows:- 1. There are certain corrupt readings in the Ujjain Sarada Ms. of the Matsya Purana, which can be explained only if we suppose that it had some Devanagari original as its direct or indirect source. 2. The general appearance and the forms of some of the letters of this Sarada Ms. show that it was written after the 16th century A.D., when Sarada symbol for the letter resembled the Devanagari letter . 3. Certain corrupt readings of the Sarada Ms. show that the hypothetical Devanagari original might have belonged to the 10th or the 11th century A.D., when the Sarada symbol for the letter somewhat resembled the Devanagari letter a: 4. The Ujjain Sarada Ms. is not the direct or immediate transcript of the Devanagari original, but one or more intermediate Sarada Mss. might have passed between the Devanagari original and this Ujjain Sarada Ms. 5. There might have been two streams of the Sarada Mss. of the Matsya Purana issuing from the same Devanagari source, the one confusing certain letters of the two sister scripts-Devanagari and Sarada- and the other not subject to such confusion and hence preserving the correct readings of of the Devanagari original.

174 puranam - PURANA [Vol. I, No. 2 6. The Ujjain Devanagari Ms. (No. 4646) is the Devanagari transcript of a Sarada Ms. belonging to the non-confusing group, hence this Devanagari Ms. of the Ujjain Oriental Institute generally gives the correct readings in place of certain corrupt readings of the Ujjain Sarada Ms. These points require further study based on sufficient manuscript evidence and epigraphic data. Until some more Sarada Mss. of the Matsya Purana are available and studied thoroughly, these points may be taken as tentative only. akrodhanah krodhanebhyo visista- statha titiksuratitiksorvisistah | amanusebhyo manusasca pradhano vidvamstathaivavidusah pradhanah || ( matsya 0, 37.6 )